A full year of college Calculus I (limits, derivatives, and integrals) taught unit by unit to the College Board course framework, one lesson for each required topic (closely related topics share a lesson). Every lesson quotes the required content, works examples with the numbers checked, and adds figures where a picture makes a problem clear. Work through the lessons at your own pace, and book a tutor the moment a topic doesn't click. Exam format from 2027: this is a hybrid digital exam: multiple choice in Bluebook, free response handwritten in a paper booklet. Work the practice problems on paper and check your steps against the worked answer, the way you will sit the free-response section in May.
3H 10M EXAM42 MCQ6 FRQ72 LESSONS80 PRACTICE PROBLEMSPREREQ: PRECALCULUS
What this course covers, and how the exam weights it.
AP Calculus AB follows the eight units of the College Board course framework, in 72 lessons. Each unit below shows its share of the multiple-choice section, so you can see where to spend your time: Units 5 and 6 together are roughly a third of the exam.
U1Limits and Continuity10–15%
U2Differentiation: Definition and Fundamental Properties10–15%
U3Differentiation: Composite, Implicit, and Inverse Functions5–10%
U4Contextual Applications of Differentiation10–15%
U5Analytical Applications of Differentiation15–20%
U6Integration and Accumulation of Change15–20%
U7Differential Equations5–10%
U8Applications of Integration10–15%
All eight units are open, 72 topics in all. Every topic opens with the method, worked examples, and a set of practice problems. Every problem has a full worked solution, so you can find the step where yours went wrong. Each unit closes with a ten-problem mixed review.
Free preview: all of Unit 1, its lessons and its review set,
is open to everyone with no account. Past that, a few more lessons open
for free, and a free account can start a free month of this course: no
card, every lesson and practice set open for 30 days. Grading of
written work and teacher review come with a paid plan.
Introducing calculus: can change occur at an instant?
A speedometer reads 62 miles per hour. But speed is distance divided by time, and at a single instant no time has passed and no distance has been covered. How can a car have a speed at an instant at all? That puzzle is the reason calculus exists. Algebra can only describe change over an interval. Calculus uses the idea of a limit to describe change at an instant, and everything in this course, the derivative, the integral, series, grows from that one idea.
CHA-1.A. Interpret the rate of change at an instant in terms of average rates of change over intervals containing that instant.
Essential knowledge.
CHA-1.A.1. Calculus uses limits to understand and model dynamic change.
CHA-1.A.2. Because an average rate of change divides the change in one variable by the change in another, the average rate of change is undefined at a point where the change in the independent variable would be zero.
CHA-1.A.3. The limit concept allows us to define instantaneous rate of change in terms of average rates of change.
Average rate of change
For a function \(f\), the average rate of change over the interval from \(x=a\) to \(x=b\) is the change in the output divided by the change in the input:
\[ \frac{f(b)-f(a)}{b-a} \]
Geometrically it is the slope of the secant line through the two points \((a,f(a))\) and \((b,f(b))\) on the graph. Its units are the output units per input unit: if \(s(t)\) is distance in meters and \(t\) is time in seconds, the average rate of change is in meters per second. The same expression is often written with a step size \(h\): the average rate of change from \(x=a\) to \(x=a+h\) is \(\dfrac{f(a+h)-f(a)}{h}\). The two forms are equivalent (put \(b=a+h\)), and you will meet both throughout the course.
Notice what the formula needs: a denominator that is not zero. If the two input values are the same, \(b-a=0\), the quotient is \(\frac{0}{0}\) and the average rate of change is undefined. An average needs an interval of positive length to average over. A single point has no length, so on its face the formula says nothing about the rate of change at a point.
From average rates to an instantaneous rate
Here is the key move. We cannot divide by a zero-length interval, but we can use shorter and shorter intervals and watch what the averages do. If the averages settle toward a single number as the interval shrinks, that number is the best possible meaning of "the rate of change at the instant." The instantaneous rate of change at \(x=a\) is defined in terms of average rates of change: it is the number that the average rates over smaller and smaller intervals containing \(a\) approach.
Figure 1.1. Secant lines through \(P(2,\,1.5)\) and points \(Q\) that slide toward \(P\) along \(f(x)=\tfrac14x^2+\tfrac12\). As \(Q\) closes in, the secant slopes (the average rates of change) approach the slope of the tangent line, the instantaneous rate of change.
In the figure, the point \(P\) is fixed and \(Q\) slides toward it. Each secant line through \(P\) and \(Q\) has a slope equal to an average rate of change. As \(Q\) approaches \(P\) the secant lines pivot toward one limiting line, the tangent line at \(P\), and the slopes approach the slope of that line. In the next lessons, the idea of "approaches" gets a precise name (a limit) and a precise notation. For now the point is the logic: the instantaneous rate of change is not a quotient with a zero denominator; it is the value that a family of ordinary quotients approaches.
Worked example · A falling object
A ball dropped from a bridge has fallen \(s(t)=5t^2\) meters after \(t\) seconds (using \(g=10\) m/s\(^2\)). Estimate its speed at the instant \(t=1\) by finding average velocities over shorter and shorter intervals starting at \(t=1\).
Over \([1,2]\). \(s(1)=5\) and \(s(2)=20\), so the average velocity is \(\dfrac{20-5}{2-1}=15\) m/s.
Over \([1,1.1]\). \(s(1.1)=5(1.21)=6.05\), so \(\dfrac{6.05-5}{0.1}=10.5\) m/s.
Over \([1,1.01]\). \(s(1.01)=5(1.0201)=5.1005\), so \(\dfrac{0.1005}{0.01}=10.05\) m/s.
Over \([1,1.001]\). \(s(1.001)=5.010005\), so \(\dfrac{0.010005}{0.001}=10.005\) m/s.
The averages 15, 10.5, 10.05, 10.005 are closing in on 10. The instantaneous velocity at \(t=1\) is 10 m/s. Algebra confirms it: over \([1,1+h]\) the average is \(\dfrac{5(1+h)^2-5}{h}=\dfrac{10h+5h^2}{h}=10+5h\), which approaches 10 as \(h\) shrinks.
Worked example · Reading a table
The table gives the distance \(d(t)\), in feet, that a cyclist has traveled after \(t\) seconds.
\(t\) (s)
0
2
5
8
\(d(t)\) (ft)
0
30
90
192
Find the average rate of change over \([2,5]\) and over \([5,8]\) and say what each means.
Over the first interval the cyclist averaged 20 feet per second, and over the second, 34 feet per second. The cyclist is speeding up. These are averages; the table alone cannot tell us the speed at \(t=5\) exactly, though it suggests a speed somewhere between 20 and 34 ft/s there.
Why the limit idea is needed
Every method in this course, finding slopes of curves, the speed of a particle at an instant, the area under a curve, the sum of an infinite list of numbers, runs into the same obstacle: the naive computation would divide by zero or add infinitely many things. Each time, calculus resolves the obstacle by replacing the impossible single computation with a sequence of possible ones and asking what they approach. That is why Unit 1 is about limits before any derivative is mentioned. When you reach the formal definition of the derivative in Unit 2, you will see that it is exactly the expression \(\dfrac{f(a+h)-f(a)}{h}\) with the instruction "let \(h\) approach \(0\)."
Calculator
On the AP exam a graphing calculator can build a table of average rates for you. Enter \(Y_1=5X^2\), then compute \(\dfrac{Y_1(1+h)-Y_1(1)}{h}\) for \(h=0.1,\,0.01,\,0.001\). A table of values is a quick way to see what the averages are approaching, but a pattern in a table is evidence, not proof. Later lessons show how to find limits exactly.
Exam focus
Average rate of change over \([a,b]\) is \(\dfrac{f(b)-f(a)}{b-a}\), the slope of the secant line; its units are output units per input unit.
It is undefined when the two input values are equal, because the denominator is zero.
The instantaneous rate of change at \(x=a\) is defined through the limit of average rates of change over shorter and shorter intervals containing \(a\).
Tables and graphs give average rates; a value read from a table over a short interval is an estimate of the instantaneous rate, not its exact value.
On free-response questions, "include units" is part of the answer: write them.
Practice
Find the average rate of change of \(f(x)=x^2-3x\) over \([1,4]\).
Show answer
\(f(1)=-2\) and \(f(4)=4\), so the average rate is \(\dfrac{4-(-2)}{4-1}=\dfrac{6}{3}=2\).
A student tries to find the average rate of change of \(g\) over the "interval" \([3,3]\). What goes wrong?
Show answer
The change in input is \(3-3=0\), so the quotient is \(\dfrac{g(3)-g(3)}{0}=\dfrac{0}{0}\), which is undefined. An average rate of change needs an interval of positive length.
The temperature of a cup of tea is \(T(0)=95\) and \(T(10)=71\), where \(T\) is in degrees Fahrenheit and \(t\) is in minutes. Find the average rate of change of \(T\) over \([0,10]\) and include units.
Show answer
\(\dfrac{71-95}{10-0}=-2.4\) degrees Fahrenheit per minute. The negative sign means the tea is cooling.
For \(f(x)=\sqrt{x}\), simplify \(\dfrac{\sqrt{4+h}-2}{h}\) by multiplying by the conjugate, then evaluate it for \(h=0.01\). What number do the averages seem to approach as \(h\) shrinks?
Show answer
\(\dfrac{\sqrt{4+h}-2}{h}\cdot\dfrac{\sqrt{4+h}+2}{\sqrt{4+h}+2}=\dfrac{h}{h(\sqrt{4+h}+2)}=\dfrac{1}{\sqrt{4+h}+2}\). For \(h=0.01\) this is about \(0.2498\). As \(h\) shrinks the values approach \(\dfrac{1}{2+2}=0.25\).
The position of a runner is \(p(t)=2t^2\) meters at \(t\) seconds. Find the average velocity over \([3,3.01]\) and use it to estimate the velocity at \(t=3\).
Show answer
\(p(3)=18\) and \(p(3.01)=2(9.0601)=18.1202\), so the average velocity is \(\dfrac{0.1202}{0.01}=12.02\) m/s. The velocity at \(t=3\) is approximately 12 m/s (the exact value is 12).
Explain in a sentence why a speedometer reading cannot be computed as distance divided by elapsed time at a single instant, and how calculus fixes this.
Show answer
At one instant the elapsed time is zero, so the quotient would be \(\frac{0}{0}\). Calculus instead takes average speeds over shorter and shorter time intervals around the instant and defines the instantaneous speed as the limit of those averages.
Lesson 1.1 ended with a promise: instantaneous rates of change are defined by what average rates "approach." Now we make approach precise. A limit describes the value a function gets close to as its input gets close to some number. Limits are the language the rest of calculus is written in, so it is worth learning the notation until it reads like a sentence.
LIM-1.A. Represent limits analytically using correct notation.
LIM-1.B. Interpret limits expressed in analytic notation.
LIM-1.C. Estimate limits of functions.
Essential knowledge.
LIM-1.A.1. Given a function \(f\), the limit of \(f(x)\) as \(x\) approaches \(c\) is a real number \(R\) if \(f(x)\) can be made arbitrarily close to \(R\) by taking \(x\) sufficiently close to \(c\) (but not equal to \(c\)). If the limit exists and is a real number, then the common notation is \(\lim_{x\to c}f(x)=R\).
LIM-1.B.1. A limit can be expressed in multiple ways, including graphically, numerically, and analytically.
LIM-1.C.1. The concept of a limit includes one-sided limits.
The idea of a limit
Suppose \(f\) is a function and \(c\) is a number. We say that the limit of \(f(x)\) as \(x\) approaches \(c\) is \(L\), and write
\[ \lim_{x\to c} f(x)=L, \]
when the values \(f(x)\) can be made as close to \(L\) as we like by taking \(x\) close enough to \(c\), but not equal to \(c\). The words "but not equal to \(c\)" matter. A limit is about what happens near \(c\), not at \(c\). That is exactly what we needed in Lesson 1.1: a quotient that is undefined at \(h=0\) can still have a perfectly good limit as \(h\) approaches \(0\).
You read the notation left to right: "the limit, as \(x\) approaches \(c\), of \(f\) of \(x\), equals \(L\)." A common shorthand is \(f(x)\to L\) as \(x\to c\). The number \(L\) is a real number if the limit exists; later (Lesson 1.14) we will also use the symbol \(\infty\) to describe certain behavior, but \(\infty\) is not a number and a limit that "equals infinity" is a limit that does not exist as a real number.
Figure 1.2. The graph of \(f\) follows \(y=x+1\) except at \(x=1\), where the function value is \(f(1)=0.5\). The limit \(\lim_{x\to1}f(x)=2\) depends on the values near \(x=1\) and ignores the value at \(x=1\).
The figure shows why the value at the point and the limit at the point are separate things. The graph follows the line \(y=x+1\) everywhere except \(x=1\). Near \(x=1\) the outputs head toward \(2\), so \(\lim_{x\to1}f(x)=2\). The function value happens to be \(f(1)=0.5\), but the limit does not care. The function could even be undefined at \(x=1\) and the limit would still be \(2\).
One-sided limits
Sometimes the left and right sides of \(c\) behave differently, so we need notation for each side. The left-hand limit \(\displaystyle\lim_{x\to c^-}f(x)\) uses only inputs \(x\lt c\), and the right-hand limit \(\displaystyle\lim_{x\to c^+}f(x)\) uses only inputs \(x\gt c\). The small superscript sign tells you which side you approach from: minus for smaller values (the left on a number line), plus for larger values (the right).
The two-sided limit and the one-sided limits are related by one rule that you will use constantly:
In words: the two-sided limit exists only when both one-sided limits exist and agree. If they exist but differ, the two-sided limit does not exist.
Figure 1.3. At \(x=2\) the graph approaches \(1\) from the left and \(3\) from the right, so \(\lim_{x\to2}f(x)\) does not exist even though \(f(2)=2\) is defined.
Worked example · Reading limits from a piecewise function
Let \(f(x)=\begin{cases}x+2, & x\lt1\\ 1, & x=1\\ 5-x, & x\gt1.\end{cases}\) Find \(\displaystyle\lim_{x\to1^-}f(x)\), \(\displaystyle\lim_{x\to1^+}f(x)\), \(\displaystyle\lim_{x\to1}f(x)\), and \(f(1)\).
Left. For \(x\lt1\), \(f(x)=x+2\), and \(x+2\to3\) as \(x\to1\), so the left-hand limit is \(3\).
Right. For \(x\gt1\), \(f(x)=5-x\), and \(5-x\to4\), so the right-hand limit is \(4\).
Two-sided. The one-sided limits are different (\(3\ne4\)), so \(\displaystyle\lim_{x\to1}f(x)\) does not exist.
Value. \(f(1)=1\), which agrees with neither limit. The function value and the limits are three separate pieces of information.
The same limit in three representations
A limit can be shown graphically (follow the curve toward \(x=c\) from each side and read the height), numerically (a table of values for \(x\) near \(c\)), and analytically (with a formula and algebra). All three describe the same number. In this lesson you read limits from graphs and tables and write them correctly; algebraic evaluation comes in Lessons 1.5 and 1.6.
Worked example · A table
Estimate \(\displaystyle\lim_{x\to2}\frac{x^2-4}{x-2}\) from a table.
\(x\)
1.9
1.99
1.999
2.001
2.01
2.1
\(\dfrac{x^2-4}{x-2}\)
3.9
3.99
3.999
4.001
4.01
4.1
The function is undefined at \(x=2\) (it gives \(\frac{0}{0}\)), yet the values from both sides close in on \(4\). So \(\displaystyle\lim_{x\to2}\frac{x^2-4}{x-2}=4\). The table is evidence, not a proof, but here the pattern is clear and an algebraic check agrees: \(\dfrac{x^2-4}{x-2}=\dfrac{(x-2)(x+2)}{x-2}=x+2\) for \(x\ne2\).
Exam focus
\(\displaystyle\lim_{x\to c}f(x)=L\) means \(f(x)\) is close to \(L\) when \(x\) is close to, but not equal to, \(c\).
The limit does not depend on \(f(c)\). \(f(c)\) may equal the limit, differ from it, or not exist.
\(c^-\) means approach from the left (smaller \(x\)); \(c^+\) means approach from the right (larger \(x\)).
The two-sided limit exists only if both one-sided limits exist and are equal. If they differ, write "does not exist" (not "0" and not "undefined").
Write limit notation completely, with the variable and the value approached, every line you use it. Dropping the "\(\lim\)" before the limit is reached is a classic lost point.
Practice
Read aloud and rewrite in words: \(\displaystyle\lim_{x\to3^-}g(x)=5\).
Show answer
"The limit of \(g\) of \(x\) as \(x\) approaches 3 from the left is 5." The outputs of \(g\) get close to 5 as \(x\) approaches 3 through values less than 3.
Use limit notation for this statement: as \(x\) approaches 4 from the right, \(f(x)\) gets close to 6.
Show answer
\(\displaystyle\lim_{x\to4^+}f(x)=6\).
A graph shows an open circle at \((2,3)\) and a closed circle at \((2,7)\), with the curve approaching height 3 from both sides of \(x=2\). Find \(\displaystyle\lim_{x\to2}f(x)\) and \(f(2)\).
Show answer
The curve approaches height 3 from both sides, so the limit is 3. The closed circle gives the function value, \(f(2)=7\). The limit and the function value differ.
For \(h(x)=\dfrac{|x|}{x}\), find \(\displaystyle\lim_{x\to0^-}h(x)\), \(\displaystyle\lim_{x\to0^+}h(x)\), and \(\displaystyle\lim_{x\to0}h(x)\).
Show answer
For \(x\lt0\), \(|x|=-x\), so \(h(x)=-1\); for \(x\gt0\), \(h(x)=1\). The left limit is \(-1\), the right limit is \(1\), and since they differ, \(\displaystyle\lim_{x\to0}h(x)\) does not exist.
The values of \(f(x)=\dfrac{x^3-1}{x-1}\) are \(f(0.9)=2.71\), \(f(0.99)=2.9701\), \(f(0.999)=2.997001\), and \(f(1.001)=3.003001\). Estimate \(\displaystyle\lim_{x\to1}f(x)\). Is \(f(1)\) defined?
Show answer
The values from both sides approach 3, so the limit is approximately 3. \(f(1)\) is not defined, because the formula gives \(\frac{0}{0}\). The limit exists anyway.
True or false: if \(f(c)\) is undefined, then \(\displaystyle\lim_{x\to c}f(x)\) does not exist. Explain.
Show answer
False. A limit depends only on values near \(c\), not at \(c\). For example \(\dfrac{x^2-4}{x-2}\) is undefined at \(x=2\) but has limit 4 there.
You will often be handed a limit with no formula to work with: a graph in a free-response question, or a table of measured data. This lesson is about reading limits from those two representations, and just as important, about knowing when they can fool you. Algebraic techniques come next; first you should be able to look at a picture or a column of numbers and say what a limit is, or why it does not exist.
LIM-1.C.2. Graphical information about a function can be used to estimate limits.
LIM-1.C.3. Because of issues of scale, graphical representations of functions may miss important function behavior.
LIM-1.C.4. A limit might not exist for some functions at particular values of \(x\). Some ways that the limit might not exist are if the function is unbounded, if the function is oscillating near this value, or if the limit from the left does not equal the limit from the right.
LIM-1.C.5. Numerical information can be used to estimate limits.
Estimating a limit from a graph
To find \(\displaystyle\lim_{x\to c}f(x)\) from a graph, put one finger on the curve far to the left of \(x=c\) and slide it toward \(x=c\); note the height it approaches. Then do the same from the right. If both fingers head for the same height \(L\), then \(\displaystyle\lim_{x\to c}f(x)=L\). Three things to remember:
An open circle marks a point the graph approaches but does not include; a closed circle marks the actual function value. The limit is the height the curve approaches, whatever the circles say.
If the two sides approach different heights, the limit does not exist, even though each one-sided limit does.
The function value \(f(c)\) is read separately from the filled-in point. It may differ from the limit.
Worked example · Reading limits from a graph
The graph of \(g\) has these features. Near \(x=-1\) the curve approaches height 2 from both sides, with a closed circle at \((-1,4)\). At \(x=3\) the curve approaches height 1 from the left and height \(-1\) from the right, with a closed circle at \((3,0)\). Find the requested values.
Quantity
Value
Reason
\(\displaystyle\lim_{x\to-1}g(x)\)
2
both sides approach 2
\(g(-1)\)
4
the closed circle
\(\displaystyle\lim_{x\to3^-}g(x)\)
1
left side approaches 1
\(\displaystyle\lim_{x\to3^+}g(x)\)
\(-1\)
right side approaches \(-1\)
\(\displaystyle\lim_{x\to3}g(x)\)
does not exist
the one-sided limits differ
\(g(3)\)
0
the closed circle
Three ways a limit can fail to exist
A limit might not exist for some functions at particular values of \(x\). There are three common reasons, and you should be able to name them from a graph or a formula.
Figure 1.4. Three ways a limit can fail to exist as \(x\) approaches a point: the one-sided limits disagree (left), the function grows without bound (center), or the function oscillates without settling (right, \(y=\sin(1/x)\) near \(0\)).
The one-sided limits disagree (a jump). Example: the function in the figure's left panel, or \(\dfrac{|x|}{x}\) at \(0\).
The function is unbounded near \(c\) (it grows without bound). Example: \(\dfrac{1}{x^2}\) as \(x\to0\). The outputs do not settle on any real number.
The function oscillates near \(c\) without settling. Example: \(\sin\dfrac{1}{x}\) as \(x\to0\), which swings between \(-1\) and \(1\) infinitely often in every interval around \(0\).
(Notice that a limit that does not exist is a perfectly good mathematical answer. On the exam, "does not exist" with a reason is often the whole point of the question.)
Estimating a limit from a table
A table of values gives numerical information about a limit. Choose \(x\)-values that approach \(c\) from both sides, closer and closer, and watch the outputs. If the outputs from both sides approach the same number, that number is your estimate of the limit.
Worked example · A limit from a table
Estimate \(\displaystyle\lim_{x\to0}\frac{\sin x}{x}\) (with \(x\) in radians).
\(x\)
\(-0.1\)
\(-0.01\)
\(-0.001\)
0.001
0.01
0.1
\(\dfrac{\sin x}{x}\)
0.998334
0.999983
0.9999998
0.9999998
0.999983
0.998334
The function is not defined at \(x=0\), but the values from both sides climb toward \(1\). We estimate \(\displaystyle\lim_{x\to0}\dfrac{\sin x}{x}=1\). (Lesson 1.6 and Lesson 4.7 give ways to prove it.)
Worked example · Different one-sided behavior in a table
For a function \(f\), a table gives \(f(1.9)=4.9\), \(f(1.99)=4.99\), \(f(1.999)=4.999\), and \(f(2.001)=1.001\), \(f(2.01)=1.01\), \(f(2.1)=1.1\). What is \(\displaystyle\lim_{x\to2}f(x)\)?
From the left the values approach \(5\); from the right they approach \(1\). The one-sided limits exist but differ, so \(\displaystyle\lim_{x\to2}f(x)\) does not exist. A table that shows only one side of \(c\) cannot tell you this, which is why a good table always includes both sides.
When graphs and tables mislead
Because limits describe what happens arbitrarily close to \(c\), pictures and finite tables can mislead:
Scale. A graphing window shows only finite detail. A tiny hole or a very narrow spike may not be visible, and a graph can appear to touch an asymptote. When a limit decision matters, zoom in or use algebra.
Unlucky inputs. For \(f(x)=\sin\dfrac{\pi}{x}\), the inputs \(x=0.5,\,0.25,\,0.125,\dots\) all give \(\sin(2\pi),\sin(4\pi),\sin(8\pi)=0\), so that table suggests a limit of \(0\). But at \(x=0.4\), \(0.2222\dots\), \(0.1538\dots\) (that is, \(x=\frac{2}{4n+1}\)) the value is \(1\). The function oscillates and the limit at \(0\) does not exist.
Too few values. Two or three nearby inputs are a suggestion. Use inputs that get steadily closer, from both sides.
A table or graph can estimate a limit; to justify one on the free-response section you will use algebra, limit properties, or a named theorem.
Calculator
A graphing calculator is valuable for limit work: graph the function and zoom in at \(x=c\), or use a table (TABLE or TBLSET) with a small step. Remember that the calculator evaluates at specific numbers only. If your table or window disagrees with an algebraic argument, trust the algebra and then look for why the picture misled.
Exam focus
Read a limit from a graph by following the curve from both sides; circles show values, not limits.
A limit fails to exist when the sides disagree, when the function is unbounded, or when it oscillates.
For a table, use inputs approaching \(c\) from both sides; agreement suggests a limit, disagreement shows it does not exist.
Graphs and tables are estimates: scale and unlucky inputs can mislead, so a justification should use algebra or a theorem.
Practice
A table shows \(f(2.9)=7.41\), \(f(2.99)=7.9401\), \(f(2.999)=7.994001\), \(f(3.1)=8.61\), \(f(3.01)=8.0601\), \(f(3.001)=8.006001\). Estimate \(\displaystyle\lim_{x\to3}f(x)\).
Show answer
The values from both sides close in on 8, so \(\displaystyle\lim_{x\to3}f(x)\approx8\). (These are the values of \(x^2-1\).)
For \(f(x)=\dfrac{1-\cos x}{x^2}\), the table gives \(f(\pm0.1)=0.4995835\) and \(f(\pm0.01)=0.4999958\). Estimate \(\displaystyle\lim_{x\to0}f(x)\).
Show answer
The values approach \(0.5\) from both sides, so the limit is approximately \(0.5\), or \(\frac{1}{2}\).
A graph shows a curve climbing without bound on both sides of \(x=2\). Does \(\displaystyle\lim_{x\to2}f(x)\) exist? Why?
Show answer
It does not exist as a real number, because the function values are unbounded near \(x=2\) and do not approach any single number.
Explain why the table \(x=0.5,\,0.25,\,0.125\) for \(f(x)=\sin\dfrac{\pi}{x}\) does not prove that the limit as \(x\to0\) is 0. Find an input near 0 where \(f\) equals 1.
Show answer
All three inputs give \(\sin\) of a multiple of \(\pi\), which is 0, but the function oscillates. At \(x=0.4\), \(f(0.4)=\sin(2.5\pi)=1\). Inputs of the form \(\dfrac{2}{4n+1}\) give 1 for every \(n\), however close to 0, so the limit does not exist.
On a graph, \(f\) approaches 1 as \(x\to4\) from the left and approaches 3 as \(x\to4\) from the right, and \(f(4)=2\). Find \(\displaystyle\lim_{x\to4^-}f(x)\), \(\displaystyle\lim_{x\to4^+}f(x)\), \(\displaystyle\lim_{x\to4}f(x)\), and \(f(4)\).
Show answer
The left limit is 1, the right limit is 3, the two-sided limit does not exist (the sides disagree), and \(f(4)=2\).
A calculator graph of \(y=\dfrac{x^2-9}{x-3}\) in a standard window looks like the straight line \(y=x+3\) with no break. What does the graph hide, and what is \(\displaystyle\lim_{x\to3}\dfrac{x^2-9}{x-3}\)?
Show answer
The function is undefined at \(x=3\), so the true graph has a hole at \((3,6)\) that the pixels do not show. Since \(\dfrac{x^2-9}{x-3}=x+3\) for \(x\ne3\), the limit is 6.
Determining limits using algebraic properties of limits
Estimating limits from graphs and tables is useful, but it is slow and it is not a proof. Most limits you meet can be found exactly, in a line or two, by using a short list of limit properties. These properties let you take a complicated expression apart, find the limit of each piece, and put the pieces back together.
Required content · LIM-1.D.1, LIM-1.D.2
Learning objective.
LIM-1.D. Determine the limits of functions using limit theorems.
Essential knowledge.
LIM-1.D.1. One-sided limits can be determined analytically or graphically.
LIM-1.D.2. Limits of sums, differences, products, quotients, and composite functions can be found using limit theorems.
The limit properties
Suppose \(\displaystyle\lim_{x\to c}f(x)=A\) and \(\displaystyle\lim_{x\to c}g(x)=B\), where \(A\) and \(B\) are real numbers, and let \(k\) be a constant. Then:
\(\displaystyle\lim_{x\to c}\,[f(x)]^n=A^n\) (and roots, when they are defined)
Composite
\(\displaystyle\lim_{x\to c}F(f(x))=F(A)\) if \(F\) is continuous at \(A\)
In words: limits of sums, differences, products, quotients, and composite functions can be found using limit theorems, as long as the limits of the pieces exist (and, for a quotient, the bottom limit is not zero). The same properties hold for one-sided limits, where \(x\to c^-\) or \(x\to c^+\) throughout.
Direct substitution
Apply the constant and identity properties repeatedly and you get a powerful shortcut. For any polynomial \(p\), \(\displaystyle\lim_{x\to c}p(x)=p(c)\). For a rational function \(\dfrac{p(x)}{q(x)}\), \(\displaystyle\lim_{x\to c}\dfrac{p(x)}{q(x)}=\dfrac{p(c)}{q(c)}\) as long as \(q(c)\ne0\). The same is true for the other familiar functions (roots, exponentials, logarithms, sine, cosine) wherever they are defined. This is direct substitution, and it is the first thing to try on any limit. When it yields a real number, you are done. If it gives \(\frac{0}{0}\), the property of the quotient does not apply and you need another method (Lesson 1.6). If it gives a nonzero number divided by zero, the limit is unbounded (Lesson 1.14).
Worked example · Building a limit from the properties
Find \(\displaystyle\lim_{x\to2}(3x^2-2x+1)\) and justify each step.
The constant multiple, power, difference, and sum properties allowed us to evaluate each piece separately. This agrees with direct substitution, \(p(2)=9\).
You never need a formula for \(f\) and \(g\). The properties work from the limit values alone.
Worked example · One-sided limits of a piecewise function
Let \(f(x)=\begin{cases}x^2-1,&x\lt2\\ 7-x,&x\ge2.\end{cases}\) Find \(\displaystyle\lim_{x\to2}f(x)\).
Left. For \(x\lt2\), \(f(x)=x^2-1\), so \(\displaystyle\lim_{x\to2^-}f(x)=2^2-1=3\).
Right. For \(x\ge2\) (and near 2 on the right), \(f(x)=7-x\), so \(\displaystyle\lim_{x\to2^+}f(x)=7-2=5\).
The one-sided limits differ, so \(\displaystyle\lim_{x\to2}f(x)\) does not exist. Here \(f(2)=5\) is defined, but that does not repair the disagreement.
Composite functions
The composite property says you may "push the limit inside" a function that is continuous at the limit value. For example, \(\displaystyle\lim_{x\to0}\cos(x^2+\pi)=\cos\!\Big(\lim_{x\to0}(x^2+\pi)\Big)=\cos\pi=-1\). Be careful: if the outside function is not continuous at the inner limit, pushing the limit inside can give a wrong answer. Continuity gets its own treatment in Lessons 1.10 through 1.13.
What to write on the exam
Show the substitution and the value: \(\displaystyle\lim_{x\to3}\frac{x^2-1}{x+2}=\frac{3^2-1}{3+2}=\frac{8}{5}\). If an algebraic property requires a condition (nonzero denominator, existence of the pieces), check it. A statement such as "the limit of the quotient is the quotient of the limits" is only valid when the bottom limit is not zero.
Exam focus
Try direct substitution first for polynomials, rational functions (denominator nonzero), and the familiar functions on their domains.
The sum, difference, constant multiple, product, quotient, and power properties require the limits of the pieces to exist; the quotient also requires a nonzero bottom limit.
For a piecewise function, find each one-sided limit with the piece that applies on that side, then compare.
\(\frac{0}{0}\) from substitution means "not finished," not "does not exist."
Practice
Evaluate \(\displaystyle\lim_{x\to-1}(x^3-4x+7)\).
Show answer
By direct substitution, \((-1)^3-4(-1)+7=-1+4+7=10\).
Evaluate \(\displaystyle\lim_{x\to3}\dfrac{x^2-1}{x+2}\).
Show answer
The denominator is \(3+2=5\ne0\), so by substitution the limit is \(\dfrac{9-1}{5}=\dfrac{8}{5}\).
Evaluate \(\displaystyle\lim_{x\to4}\sqrt{x+5}\cdot(x-1)\).
Show answer
\(\sqrt{9}\cdot3=3\cdot3=9\).
Suppose \(\displaystyle\lim_{x\to a}f(x)=-4\) and \(\displaystyle\lim_{x\to a}g(x)=5\). Find \(\displaystyle\lim_{x\to a}[f(x)g(x)+2f(x)]\).
Show answer
By the product, constant multiple, and sum properties: \((-4)(5)+2(-4)=-20-8=-28\).
Let \(g(x)=\begin{cases}2x+1,&x\le3\\ x^2-2,&x\gt3.\end{cases}\) Find \(\displaystyle\lim_{x\to3}g(x)\).
Show answer
Left: \(2(3)+1=7\). Right: \(3^2-2=7\). The one-sided limits agree, so \(\displaystyle\lim_{x\to3}g(x)=7\).
Explain why the quotient property cannot be used to evaluate \(\displaystyle\lim_{x\to1}\dfrac{x-1}{x^2-1}\), and then evaluate the limit another way.
Show answer
The limit of the denominator is \(1^2-1=0\), and the quotient property requires a nonzero bottom limit (substitution gives \(\frac{0}{0}\)). For \(x\ne1\), \(\dfrac{x-1}{x^2-1}=\dfrac{x-1}{(x-1)(x+1)}=\dfrac{1}{x+1}\), whose limit as \(x\to1\) is \(\dfrac{1}{2}\).
Direct substitution is the first move, but it sometimes returns \(\frac{0}{0}\), as it did for \(\dfrac{x^2-4}{x-2}\) at \(x=2\). That form is called indeterminate: it says nothing about the limit by itself, because the limit could be any number, or not exist. The cure is algebra. The graph of the function is the same as the graph of a simpler function except at the one troublesome point, so we rearrange the expression into an equivalent form that substitution can handle. Limits depend only on values near \(c\), and for \(x\ne c\) the two forms are identical.
Required content · LIM-1.E.1
Learning objective.
LIM-1.E. Determine the limits of functions using equivalent expressions for the function or the squeeze theorem.
Essential knowledge.
LIM-1.E.1. It may be necessary or helpful to rearrange expressions into equivalent forms before evaluating limits.
Technique 1: factor and cancel
When substitution into a rational function gives \(\frac{0}{0}\), the numerator and denominator share a factor \((x-c)\). Factor, divide it out (legitimate because \(x\ne c\) in a limit), and substitute again.
Direct substitution gives \(\frac{0}{0}\). Factor: \(\dfrac{(x-3)(x+3)}{(x-3)(x-2)}=\dfrac{x+3}{x-2}\) for \(x\ne3\). Now substitute: \(\dfrac{3+3}{3-2}=6\).
When the \(\frac{0}{0}\) involves a square root, multiply the numerator and denominator by the conjugate of the radical expression, so that the difference of squares removes the root. Keep the other factor unmultiplied so that you can cancel.
The numerator is \(\dfrac{2-(x+2)}{2(x+2)}=\dfrac{-x}{2(x+2)}\). Dividing by \(x\) gives \(\dfrac{-x}{2(x+2)\,x}=\dfrac{-1}{2(x+2)}\) for \(x\ne0\). Then
The first was suggested by the table in Lesson 1.4 and is justified with the squeeze theorem in Lesson 1.8; the second follows from it. Use them together with alternate forms of trigonometric functions, such as \(\tan x=\dfrac{\sin x}{\cos x}\) or the identity \(1-\cos x=\dfrac{\sin^2x}{1+\cos x}\), to rewrite an expression in a form where these limits apply. A frequent move is to scale the argument: \(\dfrac{\sin(kx)}{x}=k\cdot\dfrac{\sin(kx)}{kx}\to k\cdot1\).
Worked example · Trig forms
Find (a) \(\displaystyle\lim_{x\to0}\frac{\sin5x}{3x}\) and (b) \(\displaystyle\lim_{x\to0}\frac{\tan x}{x}\).
(a) \(\dfrac{\sin5x}{3x}=\dfrac{5}{3}\cdot\dfrac{\sin5x}{5x}\). As \(x\to0\), \(5x\to0\) too, so \(\dfrac{\sin5x}{5x}\to1\), and the limit is \(\dfrac53\).
Look at the form of the expression. A rational function with a shared factor: factor. A radical difference: conjugate. A fraction inside a fraction: common denominator. A trigonometric function near \(0\): use \(\frac{\sin x}{x}\to1\), or rewrite with identities. Lesson 1.7 collects these choices into one decision guide.
Exam focus
\(\frac{0}{0}\) is indeterminate: it calls for rearranging the expression, not for answering "does not exist."
Factor and cancel the shared \((x-c)\); multiply by the conjugate for radicals; use a common denominator for complex fractions.
Know \(\displaystyle\lim_{x\to0}\frac{\sin x}{x}=1\) and \(\displaystyle\lim_{x\to0}\frac{1-\cos x}{x}=0\) (radians), and the scaling trick \(\dfrac{\sin kx}{x}\to k\).
After simplifying, write the equivalent expression and then substitute; do not substitute into the original and write \(\frac{0}{0}=\) a number.
Practice
Evaluate \(\displaystyle\lim_{x\to2}\dfrac{x^2-4}{x^2-3x+2}\).
Show answer
Some functions refuse every algebraic trick. Try factoring \(x^2\sin\dfrac1x\) or multiplying it by a conjugate and you get nowhere, and direct substitution at \(x=0\) is not even possible. But such a function has a feature we can exploit: it is trapped between two simpler functions that are heading to the same place. If a function is pinned between two others that both approach \(L\), it has no choice but to approach \(L\) as well. That is the squeeze theorem.
Required content · LIM-1.E.2
Learning objective.
LIM-1.E. Determine the limits of functions using equivalent expressions for the function or the squeeze theorem.
Essential knowledge.
LIM-1.E.2. The limit of a function may be found by using the squeeze theorem.
The theorem
Squeeze theorem. Suppose that \(g(x)\le f(x)\le h(x)\) for all \(x\) near \(c\) (except possibly at \(x=c\) itself), and that \[\lim_{x\to c}g(x)=\lim_{x\to c}h(x)=L.\] Then \(\displaystyle\lim_{x\to c}f(x)=L\).
This is a theorem, so on the exam you must check its hypotheses before you use it: the inequality really holds near \(c\), and the two bounding limits exist and are equal. Then name the theorem and state the conclusion.
Figure 1.5. The oscillating function \(f(x)=x^2\sin(1/x)\) is trapped between \(-x^2\) and \(x^2\). Both bounds approach \(0\) as \(x\to0\), so \(f(x)\) is squeezed to \(0\).
In the figure, \(f(x)=x^2\sin\dfrac1x\) oscillates wildly near \(0\), but because \(-1\le\sin\dfrac1x\le1\) we have \(-x^2\le f(x)\le x^2\). The outer curves both pass through the origin, and the oscillating curve is forced through it too.
Building the bounds
Almost every squeeze problem starts with a bounded function such as \(\sin\), \(\cos\), or \(\sin\dfrac1x\), whose values always lie between \(-1\) and \(1\). Start from the bound on the bounded part, then multiply through by the other factor. If that factor can be negative, multiply by its absolute value or square (which is nonnegative) so the inequality direction is preserved.
Bound. For all \(x\ne0\), \(-1\le\cos\dfrac2x\le1\). Since \(x^4\ge0\), multiplying keeps the direction: \(-x^4\le x^4\cos\dfrac2x\le x^4\).
Limits of the bounds. \(\displaystyle\lim_{x\to0}(-x^4)=0\) and \(\displaystyle\lim_{x\to0}x^4=0\).
Conclusion. By the squeeze theorem, \(\displaystyle\lim_{x\to0}x^4\cos\frac2x=0\).
Worked example · Given inequalities
For all \(x\) near 2 (with \(x\ne2\)), a function \(f\) satisfies \(4x-5\le f(x)\le x^2-1\). Find \(\displaystyle\lim_{x\to2}f(x)\).
\(\displaystyle\lim_{x\to2}(4x-5)=3\) and \(\displaystyle\lim_{x\to2}(x^2-1)=3\). The bounds agree, so by the squeeze theorem \(\displaystyle\lim_{x\to2}f(x)=3\). We learned this without knowing a formula for \(f\).
The limit \(\dfrac{\sin x}{x}\)
The squeeze theorem is the reason \(\displaystyle\lim_{x\to0}\dfrac{\sin x}{x}=1\), the limit you used in Lesson 1.6. Compare areas in the unit circle for \(0\lt x\lt\dfrac\pi2\): the triangle with height \(\sin x\) is inside the sector of angle \(x\), which is inside the triangle with height \(\tan x\). This gives \(\sin x\lt x\lt\tan x\). Dividing by \(\sin x\) and taking reciprocals:
\[\cos x\lt\frac{\sin x}{x}\lt1.\]
Because \(\dfrac{\sin x}{x}\) is an even function, the same inequality holds for negative \(x\) near \(0\). As \(x\to0\), \(\cos x\to1\) and the upper bound is \(1\), so \(\displaystyle\lim_{x\to0}\dfrac{\sin x}{x}=1\) by the squeeze theorem.
Worked example · Why "no limit" for the inner function is not enough
A student argues: "\(\sin\dfrac1x\) has no limit as \(x\to0\), so \(x\sin\dfrac1x\) has no limit either." Is that right?
No. The product of a function with no limit and a function that approaches \(0\) can have a limit. Here \(-|x|\le x\sin\dfrac1x\le|x|\), and both bounds approach \(0\), so \(\displaystyle\lim_{x\to0}x\sin\dfrac1x=0\) by the squeeze theorem. Compare \(\sin\dfrac1x\) alone, which does not settle at all.
Exam focus
State the hypotheses: \(g(x)\le f(x)\le h(x)\) near \(c\), and \(\lim g=\lim h=L\).
Bound a trig factor with \(-1\le\sin\le1\) or \(-1\le\cos\le1\), then multiply by a nonnegative factor (use \(x^2\), \(x^4\), or \(|x|\)).
If the two bounding limits are different, the squeeze theorem says nothing.
\(\displaystyle\lim_{x\to0}\dfrac{\sin x}{x}=1\) is itself a squeeze result.
Practice
Use the squeeze theorem to evaluate \(\displaystyle\lim_{x\to0}x^2\cos\dfrac{5}{x}\).
Show answer
Since \(-1\le\cos\dfrac5x\le1\) and \(x^2\ge0\), \(-x^2\le x^2\cos\dfrac5x\le x^2\). Both bounds approach 0, so the limit is 0.
For all \(x\) near 1, \(2x\le f(x)\le x^2+1\). Find \(\displaystyle\lim_{x\to1}f(x)\), and check that the bounds are valid.
Show answer
\(\displaystyle\lim_{x\to1}2x=2\) and \(\displaystyle\lim_{x\to1}(x^2+1)=2\), so \(\displaystyle\lim_{x\to1}f(x)=2\) by the squeeze theorem. The inequality \(2x\le x^2+1\) holds because \(x^2-2x+1=(x-1)^2\ge0\).
Evaluate \(\displaystyle\lim_{x\to0}x^2\left(2+\sin\dfrac1x\right)\).
Show answer
\(1\le2+\sin\dfrac1x\le3\), so \(x^2\le x^2\left(2+\sin\dfrac1x\right)\le3x^2\). Both bounds approach 0, so the limit is 0.
A student says: "\(g(x)\le f(x)\le h(x)\), \(\lim g=1\) and \(\lim h=3\), so by the squeeze theorem \(\lim f=2\)." What is wrong?
Show answer
The squeeze theorem needs the two bounding limits to be equal. Here \(1\ne3\), so the theorem gives no information; \(f\) could approach any number between 1 and 3 (or fail to have a limit at all).
Show that \(\displaystyle\lim_{x\to0}x\sin\dfrac1x=0\) even though \(\displaystyle\lim_{x\to0}\sin\dfrac1x\) does not exist.
Show answer
\(-1\le\sin\dfrac1x\le1\) gives \(-|x|\le x\sin\dfrac1x\le|x|\); both \(\pm|x|\) approach 0, so the limit is 0. The function \(\sin\dfrac1x\) alone oscillates between \(-1\) and \(1\) near 0 and has no limit.
Using \(\cos x\lt\dfrac{\sin x}{x}\lt1\) for small positive \(x\), explain why \(\displaystyle\lim_{x\to0^+}\dfrac{\sin x}{x}=1\).
Show answer
As \(x\to0^+\), \(\cos x\to1\) and the upper bound is the constant 1. Both bounds approach 1, so \(\dfrac{\sin x}{x}\) is squeezed to 1.
Selecting procedures and connecting representations
You now own a toolbox: direct substitution, factoring, conjugates, common denominators, trigonometric limits, the squeeze theorem, and estimates from graphs and tables. The College Board treats these two topics (1.7 and 1.9) as skills topics rather than as new content. There is nothing new to memorize here; the goal is to practice choosing the right tool by looking at the structure of a problem, and to move comfortably between the graphical, numerical, and analytical forms of the same limit. On the exam, mixed limit questions are common, and the first mark often goes to the student who picks the right method quickly.
A decision guide
Always try direct substitution first. Then match what you see to a method.
What you see
What to try
A polynomial, root, exponential, logarithm, sine or cosine, with substitution giving a real number
Direct substitution (the function is continuous at \(c\))
A limit is one number, but you can see it in three forms. Skills questions ask you to connect them, for example to identify which graph, table, or formula matches a given limit statement, or to re-express a limit given in one form in another.
Numerical. At \(x=0.01\) the quotient is \(\dfrac{\sqrt{4.01}-2}{0.01}\approx0.2498\); at \(x=-0.01\) it is \(\dfrac{\sqrt{3.99}-2}{-0.01}\approx0.2502\). Both sides are near \(0.25\).
Graphical. The graph is a smooth curve through height \(0.25\) with a hole at \(x=0\) (the function is undefined there).
Analytical. Multiply by the conjugate: \(\dfrac{(x+4)-4}{x(\sqrt{x+4}+2)}=\dfrac{1}{\sqrt{x+4}+2}\to\dfrac14\).
All three agree: the limit is \(\dfrac14\). The analytic form is the one you would present as a justification; the other two confirm it and help you spot mistakes.
Worked example · A limit hidden in a graph and a statement
A graph of \(f\) has a hole at \((3,6)\), with the curve approaching height \(6\) from both sides, and a closed circle at \((3,2)\). Write the limit statement and the value statement that describe this.
\(\displaystyle\lim_{x\to3}f(x)=6\) and \(f(3)=2\). The graph shows both facts; they are different statements and neither implies the other.
Exam focus
Try direct substitution first; if it gives a real number, you are done.
For \(\frac{0}{0}\), match the structure: factor, conjugate, common denominator, trig limit.
A bounded factor times something approaching \(0\) suggests the squeeze theorem.
Move easily between a graph, a table, and a formula for the same limit; use algebra for the justification.
Name the method in your work and write the equivalent expression before you substitute.
Practice
Evaluate \(\displaystyle\lim_{x\to9}\dfrac{x-9}{\sqrt{x}-3}\) and name the method.
Show answer
Conjugate (or factor \(x-9=(\sqrt x-3)(\sqrt x+3)\)): the expression equals \(\sqrt x+3\) for \(x\ne9\), so the limit is \(6\).
Evaluate \(\displaystyle\lim_{x\to0}\dfrac{\frac{1}{x+5}-\frac15}{x}\) and name the method.
Show answer
Common denominator: the numerator is \(\dfrac{5-(x+5)}{5(x+5)}=\dfrac{-x}{5(x+5)}\), so the quotient is \(\dfrac{-1}{5(x+5)}\to-\dfrac{1}{25}\).
Evaluate \(\displaystyle\lim_{x\to0}\dfrac{\sin2x}{\sin5x}\).
Show answer
Evaluate \(\displaystyle\lim_{x\to-3}\dfrac{x^2+x-6}{x+3}\).
Show answer
Factor: \(x^2+x-6=(x+3)(x-2)\), so the expression is \(x-2\) for \(x\ne-3\), and the limit is \(-5\).
Evaluate \(\displaystyle\lim_{x\to0}x^3\sin\dfrac{1}{x^2}\).
Show answer
\(-1\le\sin\dfrac1{x^2}\le1\) and \(|x^3|\) bounds the product: \(-|x|^3\le x^3\sin\dfrac1{x^2}\le|x|^3\). Both bounds approach 0, so the limit is 0 (squeeze theorem).
The table of \(g(x)=\dfrac{x^2-9}{x-3}\) gives \(5.999\) at \(x=2.999\) and \(6.001\) at \(x=3.001\); the graph shows a hole at \((3,6)\). Which representation would you use to justify \(\displaystyle\lim_{x\to3}g(x)=6\) on a free-response question, and why?
Show answer
The algebraic one: \(\dfrac{(x-3)(x+3)}{x-3}=x+3\) for \(x\ne3\), so the limit is \(6\). The table and graph support the result, but a justification needs a mathematical argument.
Types of discontinuities and continuity at a point
Informally, a function is continuous if you can draw its graph without lifting your pencil. That picture is a good start, but calculus needs a definition precise enough to prove things with. Limits provide it. In this lesson you will state exactly what continuity at a point means, test it step by step, and name the three types of break in a graph.
Required content · LIM-2.A.1, LIM-2.A.2
Learning objective.
LIM-2.A. Justify conclusions about continuity at a point using the definition.
Essential knowledge.
LIM-2.A.1. Types of discontinuities include removable discontinuities, jump discontinuities, and discontinuities due to vertical asymptotes.
LIM-2.A.2. A function \(f\) is continuous at \(x=c\) provided that \(f(c)\) exists, \(\lim_{x\to c}f(x)\) exists, and \(\lim_{x\to c}f(x)=f(c)\).
Continuity at a point
A function \(f\) is continuous at \(x=c\) if all three of these conditions hold:
\(f(c)\) exists (the function is defined at \(c\));
\(\displaystyle\lim_{x\to c}f(x)\) exists (the left and right limits exist and agree); and
\(\displaystyle\lim_{x\to c}f(x)=f(c)\) (the limit equals the function value).
In one line: \(f\) is continuous at \(c\) provided that \(f(c)\) exists, the limit as \(x\to c\) exists, and \(\displaystyle\lim_{x\to c}f(x)=f(c)\). If any condition fails, \(f\) is discontinuous at \(c\). When you are asked to justify continuity or discontinuity, check the conditions in order and say which one fails, if one does. "The graph has a hole" is a description; "\(\lim_{x\to2}f(x)=4\) but \(f(2)=3\)" is a justification.
Types of discontinuity
Figure 1.6. Three types of discontinuity: a removable discontinuity (a hole, where the limit exists), a jump discontinuity (one-sided limits differ), and an infinite discontinuity (a vertical asymptote).
Discontinuities come in three kinds, according to which condition fails and how.
Type
What happens
Which part fails
Removable
The limit exists, but \(f(c)\) is missing or has a different value (a hole)
\(f(c)\) undefined, or \(\lim\ne f(c)\)
Jump
The one-sided limits exist but are not equal
the two-sided limit does not exist
Infinite (vertical asymptote)
The function is unbounded near \(c\)
the limit does not exist (infinite)
The word removable is a promise: since the limit exists, you could repair the graph by defining or redefining \(f(c)\) to equal the limit. Lesson 1.13 does exactly that. Jump and infinite discontinuities cannot be repaired by changing one value.
Worked example · The three-condition test
Let \(f(x)=\begin{cases}\dfrac{x^2-4}{x-2},&x\ne2\\ 3,&x=2.\end{cases}\) Is \(f\) continuous at \(x=2\)?
1. Does \(f(2)\) exist? Yes, \(f(2)=3\).
2. Does the limit exist? For \(x\ne2\), \(f(x)=\dfrac{(x-2)(x+2)}{x-2}=x+2\), so \(\displaystyle\lim_{x\to2}f(x)=4\). It exists.
3. Does the limit equal \(f(2)\)? \(4\ne3\). Condition 3 fails.
So \(f\) is not continuous at \(x=2\). Because the limit exists, the discontinuity is removable: redefining \(f(2)=4\) would make the function continuous.
Worked example · Classifying discontinuities
Find and classify the discontinuities of \(h(x)=\dfrac{x+3}{x^2+x-6}\).
Factor the denominator: \(x^2+x-6=(x+3)(x-2)\), so \(h(x)=\dfrac{x+3}{(x+3)(x-2)}\). The function is undefined at \(x=-3\) and \(x=2\).
At \(x=-3\). For \(x\ne-3\), \(h(x)=\dfrac{1}{x-2}\), so \(\displaystyle\lim_{x\to-3}h(x)=\dfrac{1}{-5}=-\dfrac15\). The limit exists, so this is a removable discontinuity (a hole at \(\left(-3,-\frac15\right)\)).
At \(x=2\). The numerator is \(5\) and the denominator approaches \(0\), so \(|h(x)|\) grows without bound. This is an infinite discontinuity (a vertical asymptote at \(x=2\)).
Worked example · A jump
Let \(p(x)=\begin{cases}x^2,&x\lt2\\ 8-x,&x\ge2.\end{cases}\) Show that \(p\) is discontinuous at \(x=2\) and classify the discontinuity.
\(p(2)=8-2=6\). Left limit: \(\displaystyle\lim_{x\to2^-}x^2=4\). Right limit: \(\displaystyle\lim_{x\to2^+}(8-x)=6\). The one-sided limits are not equal, so \(\displaystyle\lim_{x\to2}p(x)\) does not exist: condition 2 fails. This is a jump discontinuity, with a jump of \(6-4=2\).
Reading the conditions from a graph
On a graph, check by eye: Is there a filled-in point at \(x=c\)? (Condition 1.) Do the left and right sides of the curve arrive at the same height? (Condition 2.) Is the filled-in point at that height? (Condition 3.) If you can answer yes three times, the function is continuous there.
Exam focus
Continuous at \(c\) means three things: \(f(c)\) exists, \(\lim_{x\to c}f(x)\) exists, and the two are equal. Say which condition fails when it is not.
Removable: the limit exists but does not match \(f(c)\) or \(f(c)\) is undefined. Jump: one-sided limits differ. Infinite: vertical asymptote.
A rational function has a removable discontinuity where a factor cancels and an infinite one where it does not.
A limit existing at \(c\) does not make \(f\) continuous at \(c\): the function value must agree.
Practice
Which condition for continuity fails for \(g(x)=\dfrac{1}{x-1}\) at \(x=1\), and what type of discontinuity is it?
Show answer
\(g(1)\) is undefined, and the limit does not exist (the function is unbounded near 1). It is an infinite discontinuity (vertical asymptote).
Is \(f(x)=\dfrac{x^2-1}{x-1}\) continuous at \(x=1\)? Classify the discontinuity.
Show answer
\(f(1)\) is undefined, so \(f\) is discontinuous at 1. Because \(\dfrac{x^2-1}{x-1}=x+1\) for \(x\ne1\), the limit is 2. The discontinuity is removable.
Let \(h(x)=\dfrac{|x|}{x}\) for \(x\ne0\). Classify the discontinuity at \(x=0\).
Show answer
The left limit is \(-1\) and the right limit is \(1\). They differ, so the two-sided limit does not exist: a jump discontinuity.
Is \(f(x)=\begin{cases}x^2+1,&x\le2\\ 7-x,&x\gt2\end{cases}\) continuous at \(x=2\)? Justify with the three conditions.
Show answer
\(f(2)=5\). Left limit: \(2^2+1=5\). Right limit: \(7-2=5\). The limit exists and equals 5, which equals \(f(2)\). All three conditions hold, so \(f\) is continuous at 2.
A graph has a closed circle at \((4,1)\), and the curve approaches height 3 from both sides of \(x=4\). Which condition fails? What type of discontinuity is it?
Show answer
\(f(4)=1\) exists and the limit is 3, but \(3\ne1\), so condition 3 fails. The discontinuity is removable.
True or false: if \(\displaystyle\lim_{x\to c}f(x)\) exists, then \(f\) is continuous at \(c\). Explain.
Show answer
False. The limit may exist while \(f(c)\) is undefined or has a different value, as with a hole in the graph. Continuity needs the limit to exist and equal \(f(c)\).
Continuity at a single point is a local property. Most of the time we care about continuity on a whole interval, because the big theorems of calculus (the Intermediate Value Theorem, the Extreme Value Theorem, and eventually the Fundamental Theorem of Calculus) all begin with the hypothesis "\(f\) is continuous on \([a,b]\)." This lesson shows how to confirm that hypothesis quickly, using the shape of a function's formula instead of testing every point.
Required content · LIM-2.B.1, LIM-2.B.2
Learning objective.
LIM-2.B. Determine intervals over which a function is continuous.
Essential knowledge.
LIM-2.B.1. A function is continuous on an interval if the function is continuous at each point in the interval.
LIM-2.B.2. Polynomial, rational, power, exponential, logarithmic, and trigonometric functions are continuous on all points in their domains.
Continuity on an interval
A function is continuous on an interval if it is continuous at each point in the interval. For an open interval \((a,b)\), that means continuity at every \(x\) with \(a\lt x\lt b\). For a closed interval \([a,b]\), we need continuity at each interior point andone-sided continuity at the endpoints:
At an endpoint only one side of the function exists inside the interval, so only one side can be checked. The graph of a function that is continuous on \([a,b]\) is one unbroken curve from the point \((a,f(a))\) to the point \((b,f(b))\).
Continuity of the familiar functions
You do not have to verify the three conditions point by point for every function. These families are continuous at every point of their domains:
polynomial functions (continuous on all real numbers);
rational functions (continuous wherever the denominator is not zero);
power and root functions, \(x^r\);
exponential and logarithmic functions (\(e^x\) everywhere, \(\ln x\) for \(x\gt0\));
Sums, differences, products, and composites of continuous functions are continuous where they are defined, and a quotient of continuous functions is continuous wherever the denominator is not zero. These properties follow directly from the limit properties of Lesson 1.5. The practical message: a function built from the familiar pieces is continuous on every interval inside its domain. To find where it might fail, find where it is not defined, and check any places where the formula changes.
Worked example · Intervals of continuity from the domain
Find the intervals on which \(f(x)=\dfrac{\sqrt{x-1}}{x^2-9}\) is continuous.
The square root needs \(x-1\ge0\), so \(x\ge1\). The denominator is zero when \(x^2=9\), that is \(x=3\) or \(x=-3\); only \(x=3\) lies in \(x\ge1\). The domain is \([1,3)\cup(3,\infty)\).
Since \(f\) is built from continuous pieces, it is continuous on its domain: on \([1,3)\) and on \((3,\infty)\). (At \(x=1\) the function is continuous from the right, which is all that is possible there.) At \(x=3\) there is an infinite discontinuity.
Worked example · A logarithm
On what intervals is \(g(x)=\ln(x^2-4)\) continuous?
\(\ln\) requires a positive input: \(x^2-4\gt0\), so \(x\lt-2\) or \(x\gt2\). Thus \(g\) is continuous on \((-\infty,-2)\) and on \((2,\infty)\). The function is not defined (so not continuous) at \(x=\pm2\) or anywhere between.
Piecewise functions
For a piecewise function, each piece is continuous on its own interval because it is built from continuous parts. The only places to check are the boundaries where the formula changes, using the three-condition test of Lesson 1.11.
Worked example · Checking the boundary
Is \(f(x)=\begin{cases}x^2+1,&x\lt1\\ 3x-1,&x\ge1\end{cases}\) continuous on \((-\infty,\infty)\)?
Each piece is a polynomial, so \(f\) is continuous for \(x\lt1\) and for \(x\gt1\). At the boundary \(x=1\): \(f(1)=3(1)-1=2\); the left limit is \(1^2+1=2\); the right limit is \(3(1)-1=2\). All three conditions hold, so \(f\) is continuous at \(1\), and therefore continuous on all real numbers.
Why this matters
When a theorem requires "\(f\) is continuous on \([a,b]\)," you verify it by pointing to the type of function (or to a boundary check) and say so: "\(f\) is a rational function whose denominator is nonzero on \([a,b]\), so it is continuous there." Watch for a closed interval that contains a point outside the domain. For example, \(\tan x\) is not continuous on \([0,\pi]\) because it is undefined at \(x=\dfrac\pi2\), so the Intermediate Value Theorem cannot be applied there.
Exam focus
Continuous on a closed interval \([a,b]\) means continuous inside \((a,b)\), right-continuous at \(a\), and left-continuous at \(b\).
Polynomial, rational, power, exponential, logarithmic, and trigonometric functions are continuous at every point of their domains.
Find intervals of continuity by finding the domain (denominators not zero, even roots of nonnegative numbers, logarithms of positive numbers).
For piecewise functions, check each boundary with the three-condition test.
Justify continuity by naming the function type and the domain, not by saying "the graph has no breaks."
Practice
On what intervals is \(f(x)=\dfrac{x+2}{x^2-x-6}\) continuous?
Show answer
\(x^2-x-6=(x-3)(x+2)\) is zero at \(x=3\) and \(x=-2\). The function is continuous on \((-\infty,-2)\), \((-2,3)\), and \((3,\infty)\).
Is \(g(x)=\sqrt{4-x^2}\) continuous on \([-2,2]\)? Explain what happens at the endpoints.
Show answer
Yes. The domain is \(-2\le x\le2\), and \(g\) is continuous at every point of its domain. At \(x=-2\) it is continuous from the right and at \(x=2\) from the left, which is what continuity on a closed interval requires.
Is \(f(x)=\begin{cases}x^2+1,&x\lt1\\ 3x-1,&x\ge1\end{cases}\) continuous at \(x=1\)? Is it continuous everywhere?
Show answer
\(f(1)=2\), and both one-sided limits equal 2, so \(f\) is continuous at 1. Each piece is a polynomial, so it is continuous elsewhere; \(f\) is continuous on all real numbers.
Can the Intermediate Value Theorem be applied to \(f(x)=\tan x\) on \([0,\pi]\)? Why or why not?
Show answer
No. \(\tan x\) is undefined at \(x=\dfrac\pi2\) in the interval, so \(f\) is not continuous on \([0,\pi]\), and the hypothesis of the theorem is not satisfied.
Where is \(h(x)=\sin\dfrac1x\) continuous?
Show answer
\(h\) is a composite of continuous functions wherever \(\dfrac1x\) is defined, so it is continuous on \((-\infty,0)\) and \((0,\infty)\). It is not defined at \(0\), and (as Lesson 1.3 showed) it oscillates there, so it cannot be repaired.
Find the value of \(k\) that makes \(f(x)=\begin{cases}kx+1,&x\le3\\ x^2-2,&x\gt3\end{cases}\) continuous at \(x=3\). (This is the question type of Lesson 1.13; try it now.)
Show answer
Left side: \(f(3)=3k+1\). Right limit: \(3^2-2=7\). Set \(3k+1=7\), so \(k=2\).
If a graph has a hole, we can imagine filling it in. If it has a jump or an asymptote, no single dot can repair it. This lesson turns that picture into two skills the exam asks for again and again: redefining a function value to make a function continuous, and solving for an unknown parameter so that a piecewise function is continuous.
Required content · LIM-2.C.1, LIM-2.C.2
Learning objective.
LIM-2.C. Determine values of x or solve for parameters that make discontinuous functions continuous, if possible.
Essential knowledge.
LIM-2.C.1. If the limit of a function exists at a discontinuity in its graph, then it is possible to remove the discontinuity by defining or redefining the value of the function at that point, so it equals the value of the limit of the function as \(x\) approaches that point.
LIM-2.C.2. In order for a piecewise-defined function to be continuous at a boundary to the partition of its domain, the value of the expression defining the function on one side of the boundary must equal the value of the expression defining the other side of the boundary, as well as the value of the function at the boundary.
Removing a removable discontinuity
If the limit of a function exists at a discontinuity, then it is possible to remove the discontinuity by defining or redefining the value of the function at that point so that it equals the limit. In symbols: if \(\displaystyle\lim_{x\to c}f(x)=L\) exists but \(f(c)\ne L\) (or \(f(c)\) is undefined), the function \[F(x)=\begin{cases}f(x),&x\ne c\\ L,&x=c\end{cases}\] is continuous at \(c\). Only one value changes. The graph of \(F\) is the graph of \(f\) with the hole filled in (or with a misplaced point moved to the hole).
Worked example · Filling a hole
Redefine \(f(x)=\dfrac{x^2-x-6}{x-3}\) at \(x=3\) so that it becomes continuous there.
Factor: \(x^2-x-6=(x-3)(x+2)\), so \(f(x)=x+2\) for \(x\ne3\) and \(\displaystyle\lim_{x\to3}f(x)=5\). Define \(f(3)=5\). With that value, \(f(x)=x+2\) for all \(x\), a polynomial, which is continuous everywhere.
Jump and infinite discontinuities cannot be removed this way. At a jump the two one-sided limits are different numbers, so no single value of \(f(c)\) can equal both; at a vertical asymptote there is no real number to equal the limit.
Continuity of piecewise functions
A piecewise function can be forced to be continuous at a boundary by choosing parameters. The recipe: at each boundary \(x=b\), set the value of the left expression equal to the value of the right expression (and to the function value, if it is given there separately). Each boundary produces one equation.
Worked example · One parameter
Find \(k\) so that \(g(x)=\begin{cases}\dfrac{x^2-9}{x-3},&x\ne3\\ k,&x=3\end{cases}\) is continuous at \(x=3\).
\(\displaystyle\lim_{x\to3}\dfrac{(x-3)(x+3)}{x-3}=6\). For continuity, \(g(3)=\displaystyle\lim_{x\to3}g(x)\), so \(k=6\).
Worked example · A trigonometric limit
For \(h(x)=\begin{cases}\dfrac{\sin2x}{x},&x\lt0\\ 3x+k,&x\ge0\end{cases}\), find \(k\) so that \(h\) is continuous at \(0\).
Left limit: \(\displaystyle\lim_{x\to0^-}\dfrac{\sin2x}{x}=2\cdot\lim\dfrac{\sin2x}{2x}=2\). Right side: \(h(0)=3(0)+k=k\), and the right limit is also \(k\). Continuity requires \(k=2\).
Worked example · Two parameters
Find \(a\) and \(b\) so that \(f(x)=\begin{cases}2x,&x\lt1\\ ax+b,&1\le x\le3\\ x^2-1,&x\gt3\end{cases}\) is continuous everywhere.
Each piece is a polynomial, so the only trouble is at \(x=1\) and \(x=3\).
At \(x=1\). The left limit is \(2(1)=2\) and the middle expression gives \(a+b\). So \(a+b=2\).
At \(x=3\). The middle expression gives \(3a+b\); the right limit is \(3^2-1=8\). So \(3a+b=8\).
Subtract the first equation from the second: \(2a=6\), so \(a=3\), and then \(b=2-3=-1\). Check: at \(x=1\), \(3(1)-1=2\) matches; at \(x=3\), \(3(3)-1=8\) matches.
Notice that each of these is the three-condition test of Lesson 1.11 with an unknown in it. When you write a solution, show the one-sided limits (or the value from each side), set them equal, and solve.
Exam focus
A removable discontinuity can be removed by (re)defining \(f(c)=\lim_{x\to c}f(x)\). Jump and infinite discontinuities cannot.
To make a piecewise function continuous at a boundary, set the two sides (and the function value) equal and solve for the parameter; use one equation per boundary.
Evaluate the limit that involves the unknown-free piece first; trigonometric and rational pieces may need Lesson 1.6 techniques.
Check your parameter values by substituting them back at each boundary.
Practice
What value should be assigned to \(f(4)\) so that \(f(x)=\dfrac{x^2-16}{x-4}\) is continuous at \(x=4\)?
Show answer
\(\dfrac{(x-4)(x+4)}{x-4}=x+4\) for \(x\ne4\), so the limit is 8. Define \(f(4)=8\).
Find \(k\) so that \(f(x)=\begin{cases}x^2-k,&x\lt2\\ 3x+1,&x\ge2\end{cases}\) is continuous at \(x=2\).
Show answer
Left limit: \(4-k\). Right side: \(3(2)+1=7\). Set \(4-k=7\), so \(k=-3\).
For each, say whether the discontinuity can be removed: (a) a hole at \(x=2\); (b) a jump at \(x=0\); (c) a vertical asymptote at \(x=1\).
Show answer
(a) Yes, define the value at \(x=2\) to be the limit. (b) No, the one-sided limits differ. (c) No, the limit is not a real number.
Find \(a\) and \(b\) so that \(f(x)=\begin{cases}x+2,&x\lt-1\\ ax+b,&-1\le x\le2\\ 3x+1,&x\gt2\end{cases}\) is continuous everywhere.
Show answer
At \(x=-1\): \(-a+b=1\). At \(x=2\): \(2a+b=7\). Subtract: \(3a=6\), so \(a=2\) and \(b=3\).
How should \(f(0)\) be defined so that \(f(x)=\dfrac{\sin3x}{x}\) is continuous at \(x=0\)?
Show answer
\(\displaystyle\lim_{x\to0}\dfrac{\sin3x}{x}=3\), so define \(f(0)=3\).
A student claims the discontinuity of \(f(x)=\dfrac{1}{x^2}\) at \(x=0\) can be removed by defining \(f(0)=0\). Explain the error.
Show answer
\(\dfrac1{x^2}\) is unbounded near 0, so there is no real limit for \(f(0)\) to equal. Defining \(f(0)=0\) does not make the function continuous: the discontinuity is infinite, not removable.
Connecting infinite limits and vertical asymptotes
So far, a limit that did not exist was simply a failure. Some of those failures are very informative. When a function grows without bound near a point, the direction it runs off in (up or down) tells you a great deal about the graph. In this lesson we extend limit notation to describe unbounded behavior, and connect it to vertical asymptotes.
Required content · LIM-2.D.1, LIM-2.D.2
Learning objective.
LIM-2.D. Interpret the behavior of functions using limits involving infinity.
Essential knowledge.
LIM-2.D.1. The concept of a limit can be extended to include infinite limits.
LIM-2.D.2. Asymptotic and unbounded behavior of functions can be described and explained using limits.
Infinite limits
We write \(\displaystyle\lim_{x\to c}f(x)=\infty\) when the values of \(f(x)\) grow larger without bound as \(x\) approaches \(c\), and \(\displaystyle\lim_{x\to c}f(x)=-\infty\) when they decrease without bound. The same notation works for one-sided limits. Be careful with the meaning: \(\infty\) is not a real number, so a limit that "equals \(\infty\)" is a limit that does not exist as a real number. Writing \(=\infty\) is a way of describing why it does not exist, and it is the preferred way to answer on the exam when the function is unbounded in one direction.
If the two one-sided limits are \(+\infty\) and \(-\infty\), the two-sided limit does not exist and we do not use \(\infty\) for it; we describe each side separately.
Vertical asymptotes
The line \(x=c\) is a vertical asymptote of \(f\) if at least one of the one-sided limits at \(c\) is \(+\infty\) or \(-\infty\). In words, the graph climbs or dives along the line \(x=c\). Vertical asymptotes are an example of how asymptotic and unbounded behavior of functions can be described and explained using limits.
Figure 1.7. For \(f(x)=\dfrac{1}{x-2}\), the one-sided limits at \(x=2\) are infinite and have opposite signs, so the two-sided limit does not exist. The line \(x=2\) is a vertical asymptote.
For \(f(x)=\dfrac{1}{x-2}\), as \(x\to2^+\) the denominator is a small positive number, so \(f\to+\infty\); as \(x\to2^-\) the denominator is a small negative number, so \(f\to-\infty\). The line \(x=2\) is a vertical asymptote even though the two-sided limit does not exist.
Finding vertical asymptotes of rational functions
For a rational function in lowest terms, the vertical asymptotes are at the zeros of the denominator. The steps:
Factor and cancel common factors. A canceled factor leaves a hole (removable discontinuity), not an asymptote.
Find the remaining zeros of the denominator. At each, the numerator is not zero, so the quotient is (nonzero)/(small), which is unbounded.
Decide the sign from each side by checking the sign of each factor just to the left and just to the right of \(c\).
Worked example · One-sided infinite limits
Find \(\displaystyle\lim_{x\to2^+}\dfrac{x+1}{x^2-4}\) and \(\displaystyle\lim_{x\to2^-}\dfrac{x+1}{x^2-4}\).
Near \(x=2\) the numerator is close to \(3\gt0\). The denominator \(x^2-4=(x-2)(x+2)\): for \(x\) slightly greater than \(2\), \((x-2)\gt0\) and \((x+2)\gt0\), so the denominator is a small positive number; for \(x\) slightly less than \(2\), it is a small negative number.
The line \(x=2\) is a vertical asymptote. (The same analysis at \(x=-2\) gives \(\displaystyle\lim_{x\to-2^-}=-\infty\) and \(\displaystyle\lim_{x\to-2^+}=+\infty\), because the numerator is near \(-1\).)
Worked example · A hole and an asymptote
Find the discontinuities of \(f(x)=\dfrac{x^2-1}{x^2-3x+2}\) and describe each with limits.
Factor: \(\dfrac{(x-1)(x+1)}{(x-1)(x-2)}\). The factor \(x-1\) cancels, so for \(x\ne1\), \(f(x)=\dfrac{x+1}{x-2}\).
At \(x=1\). \(\displaystyle\lim_{x\to1}f(x)=\dfrac{2}{-1}=-2\): a hole at \((1,-2)\), a removable discontinuity.
At \(x=2\). The numerator is near \(3\), the denominator \(x-2\to0\): \(\displaystyle\lim_{x\to2^+}f(x)=+\infty\) and \(\displaystyle\lim_{x\to2^-}f(x)=-\infty\). The line \(x=2\) is a vertical asymptote.
The square is positive on both sides of \(3\) and approaches \(0\), so the quotient is a large positive number from both sides. Both one-sided limits are \(+\infty\), so \(\displaystyle\lim_{x\to3}\dfrac{1}{(x-3)^2}=\infty\) (the limit does not exist as a real number, but the unbounded behavior is the same from each side).
Other functions with vertical asymptotes
Rational functions are not the only source. \(\displaystyle\lim_{x\to0^+}\ln x=-\infty\), so \(x=0\) is a vertical asymptote of \(y=\ln x\). And \(\tan x\) has vertical asymptotes at \(x=\dfrac\pi2+k\pi\): \(\displaystyle\lim_{x\to(\pi/2)^-}\tan x=+\infty\) and \(\displaystyle\lim_{x\to(\pi/2)^+}\tan x=-\infty\), because \(\tan x=\dfrac{\sin x}{\cos x}\) with \(\sin\frac\pi2=1\) and \(\cos x\) a small positive then small negative number.
Exam focus
\(\displaystyle\lim_{x\to c}f(x)=\infty\) describes unbounded growth; the limit does not exist as a real number.
\(x=c\) is a vertical asymptote if at least one one-sided limit at \(c\) is \(\pm\infty\).
For rational functions, cancel common factors first: canceled factors are holes; the remaining zeros of the denominator are asymptotes.
Decide \(+\infty\) or \(-\infty\) from the signs of the factors on each side of \(c\).
If one side is \(+\infty\) and the other \(-\infty\), describe each side; do not write a two-sided \(\infty\).
Practice
Find \(\displaystyle\lim_{x\to3^-}\dfrac{2}{x-3}\).
Show answer
For \(x\lt3\) the denominator is a small negative number and the numerator is 2, so the limit is \(-\infty\).
Find \(\displaystyle\lim_{x\to0}\dfrac{1}{x^2}\).
Show answer
The denominator is a small positive number from both sides, so both one-sided limits are \(+\infty\): the limit is \(\infty\) (it does not exist as a real number).
Find the vertical asymptotes and holes of \(f(x)=\dfrac{x-5}{x^2-25}\).
Show answer
\(f(x)=\dfrac{x-5}{(x-5)(x+5)}=\dfrac1{x+5}\) for \(x\ne5\). There is a hole at \(x=5\) (limit \(\frac1{10}\)) and a vertical asymptote at \(x=-5\).
Find \(\displaystyle\lim_{x\to1^+}\dfrac{x+2}{x-1}\).
Show answer
The numerator approaches 3 and the denominator is a small positive number, so the limit is \(+\infty\).
Find \(\displaystyle\lim_{x\to-1^-}\dfrac{x}{x+1}\).
Show answer
The numerator is near \(-1\). For \(x\lt-1\), \(x+1\) is a small negative number, so \(\dfrac{-1}{\text{small negative}}\) is a large positive number: the limit is \(+\infty\).
Describe the behavior of \(y=\ln x\) near \(x=0\) using a limit, and state the asymptote.
Show answer
\(\displaystyle\lim_{x\to0^+}\ln x=-\infty\), so the line \(x=0\) (the \(y\)-axis) is a vertical asymptote. The function is not defined for \(x\le0\).
Connecting limits at infinity and horizontal asymptotes
Lesson 1.14 asked what happens when the output grows without bound. This lesson asks a different question: what happens to the output when the input grows without bound? If you let \(x\) become huge, positive or negative, does \(f(x)\) settle down? A limit at infinity describes the end behavior of a function, and the number it settles on gives a horizontal asymptote.
LIM-2.D. Interpret the behavior of functions using limits involving infinity.
Essential knowledge.
LIM-2.D.3. The concept of a limit can be extended to include limits at infinity.
LIM-2.D.4. Limits at infinity describe end behavior.
LIM-2.D.5. Relative magnitudes of functions and their rates of change can be compared using limits.
Limits at infinity
We write \(\displaystyle\lim_{x\to\infty}f(x)=L\) if the values \(f(x)\) can be made as close to \(L\) as we like by taking \(x\) sufficiently large, and \(\displaystyle\lim_{x\to-\infty}f(x)=L\) if the same is true when \(x\) is large and negative. If such a limit is a real number \(L\), the line \(y=L\) is a horizontal asymptote of the graph. A function can have a different horizontal asymptote on each end, or one, or none.
Figure 1.8. For \(f(x)=\dfrac{2x+1}{x-1}\), the values approach \(2\) as \(x\to\infty\) and as \(x\to-\infty\) (end behavior), so \(y=2\) is a horizontal asymptote. There is also a vertical asymptote at \(x=1\).
The basic facts: \(\displaystyle\lim_{x\to\pm\infty}\frac{1}{x^n}=0\) for any positive \(n\), and \(\displaystyle\lim_{x\to\infty}e^{-x}=0\) (as does \(\displaystyle\lim_{x\to-\infty}e^{x}=0\)). Everything else comes from rewriting an expression so these facts apply.
Rational functions
To find \(\displaystyle\lim_{x\to\pm\infty}\frac{p(x)}{q(x)}\), divide the numerator and denominator by the highest power of \(x\) in the denominator. The result depends on the degrees:
Divide top and bottom by \(x^2\): \(\dfrac{5-\frac{3}{x}+\frac{1}{x^2}}{2+\frac{7}{x^2}}\). As \(x\to\infty\), \(\frac3x\to0\), \(\frac1{x^2}\to0\), \(\frac7{x^2}\to0\), so the limit is \(\dfrac{5}{2}\). The line \(y=\dfrac52\) is a horizontal asymptote.
Worked example · Different degrees
Find \(\displaystyle\lim_{x\to\infty}\frac{4x+1}{x^2+3}\) and \(\displaystyle\lim_{x\to\infty}\frac{x^3+1}{2x^2-5}\).
The first has a lower degree on top. Dividing by \(x^2\): \(\dfrac{\frac4x+\frac1{x^2}}{1+\frac3{x^2}}\to\dfrac{0}{1}=0\). The second has a higher degree on top; dividing by \(x^2\) gives \(\dfrac{x+\frac1{x^2}}{2-\frac5{x^2}}\), whose numerator grows without bound while the denominator approaches \(2\), so the limit is \(+\infty\) (the function is unbounded, with no horizontal asymptote).
Square roots and the sign of \(x\)
When a square root of \(x^2\) appears, remember that \(\sqrt{x^2}=|x|\), which equals \(x\) for positive \(x\) and \(-x\) for negative \(x\). The two ends can then give different answers.
Worked example · Different limits at the two ends
Find \(\displaystyle\lim_{x\to\infty}\frac{3x+2}{\sqrt{x^2+1}}\) and \(\displaystyle\lim_{x\to-\infty}\frac{3x+2}{\sqrt{x^2+1}}\).
Divide numerator and denominator by \(|x|\), since \(\sqrt{x^2+1}\approx|x|\) for large \(|x|\).
As \(x\to\infty\), \(|x|=x\): \(\dfrac{3+\frac2x}{\sqrt{1+\frac{1}{x^2}}}\to\dfrac{3}{1}=3\).
As \(x\to-\infty\), \(|x|=-x\), so dividing by \(|x|\) means dividing the numerator by \(-x\): \(\dfrac{-3-\frac2x}{\sqrt{1+\frac1{x^2}}}\to-3\).
The function has two different horizontal asymptotes, \(y=3\) on the right and \(y=-3\) on the left.
Comparing growth rates
Limits at infinity let us compare how fast functions grow: the relative magnitudes of functions and their rates of change. A short ranking, from slowest to fastest growth as \(x\to\infty\), is: logarithms, then powers (\(x^p\) with \(p\gt0\)), then exponentials (\(b^x\) with \(b\gt1\)). In limit form, for any positive \(n\) and \(p\),
So \(e^x\) eventually dwarfs every power of \(x\), and every power of \(x\) dwarfs \(\ln x\). When one function in a quotient dominates the other, the limit is \(0\) or \(\pm\infty\) accordingly.
Worked example · A growth comparison
Evaluate \(\displaystyle\lim_{x\to\infty}\frac{x^5}{e^x}\) and \(\displaystyle\lim_{x\to\infty}\frac{e^x}{x^5}\).
An exponential grows faster than any polynomial, so the denominator dominates in the first quotient: the limit is \(0\). The second is the reciprocal and is unbounded: the limit is \(+\infty\). (L'Hospital's Rule in Lesson 4.7 can prove these.)
Exam focus
\(\displaystyle\lim_{x\to\infty}f(x)=L\) means the end behavior of \(f\) approaches \(L\); \(y=L\) is then a horizontal asymptote. Check both ends.
Rational functions: compare degrees (lower: \(0\); equal: ratio of leading coefficients; higher: unbounded).
Treat \(\sqrt{x^2}\) as \(|x|\); the left and right ends may give opposite signs.
Growth order: \(\ln x\ll x^p\ll b^x\). Use it to decide limits of quotients at infinity.
A function can cross a horizontal asymptote; the asymptote describes the far ends, not every point.
Practice
Evaluate \(\displaystyle\lim_{x\to\infty}\dfrac{7x-2}{3x+5}\).
Show answer
The degrees are equal; the ratio of leading coefficients is \(\dfrac73\).
Evaluate \(\displaystyle\lim_{x\to-\infty}\dfrac{2x^3-x}{x^3+4x^2}\).
Show answer
Divide by \(x^3\): \(\dfrac{2-\frac1{x^2}}{1+\frac4x}\to\dfrac21=2\).
Evaluate \(\displaystyle\lim_{x\to\infty}\dfrac{x^2+1}{3x-1}\). Does the graph have a horizontal asymptote on the right?
Show answer
The degree on top is higher, so the quotient is unbounded: the limit is \(+\infty\). There is no horizontal asymptote on the right.
Evaluate \(\displaystyle\lim_{x\to\infty}\left(3+e^{-2x}\right)\) and state the horizontal asymptote.
Show answer
\(e^{-2x}\to0\), so the limit is 3. The line \(y=3\) is a horizontal asymptote.
Evaluate \(\displaystyle\lim_{x\to\infty}\dfrac{\ln x}{\sqrt x}\) and explain in a sentence.
Show answer
The limit is 0. A power function \(x^{1/2}\) grows faster than \(\ln x\), so the denominator dominates.
Find all horizontal asymptotes of \(f(x)=\dfrac{5x}{\sqrt{x^2+4}}\).
Show answer
As \(x\to\infty\), \(\sqrt{x^2+4}\approx x\) and the limit is 5. As \(x\to-\infty\), \(\sqrt{x^2+4}\approx|x|=-x\) and the limit is \(-5\). The asymptotes are \(y=5\) and \(y=-5\).
If you start a hike at 3,000 feet and end at 7,000 feet, you must have passed through exactly 5,000 feet at some moment, because you did not teleport. That everyday certainty is the Intermediate Value Theorem. It turns continuity, a property of a graph having no breaks, into a way to prove that an equation has a solution even when you cannot find the solution by algebra.
Required content · FUN-1.A.1
Learning objective.
FUN-1.A. Explain the behavior of a function on an interval using the Intermediate Value Theorem.
Essential knowledge.
FUN-1.A.1. If \(f\) is a continuous function on the closed interval \([a,b]\) and \(d\) is a number between \(f(a)\) and \(f(b)\), then the Intermediate Value Theorem guarantees that there is at least one number \(c\) between \(a\) and \(b\), such that \(f(c)=d\).
The theorem
Intermediate Value Theorem (IVT). If \(f\) is a continuous function on the closed interval \([a,b]\) and \(d\) is a number between \(f(a)\) and \(f(b)\), then there is at least one number \(c\) between \(a\) and \(b\) such that \(f(c)=d\).
Figure 1.9. A continuous function on \([a,b]\) takes every value \(d\) between \(f(a)\) and \(f(b)\). Here the horizontal line \(y=d\) meets the graph at least once (at the open circles). The theorem guarantees at least one such \(c\); it does not say how many or where.
Read the statement as a checklist. A correct use of the theorem names three things:
\(f\) is continuous on \([a,b]\) (say why: the type of function, or a boundary check).
\(d\) is between \(f(a)\) and \(f(b)\) (give the numbers and compare).
Conclusion: there exists \(c\) in \((a,b)\) with \(f(c)=d\) (an existence statement).
The theorem guarantees that a solution exists. It does not say how many there are, where they are, or what the value of \(c\) is.
Showing that an equation has a solution
The most common use is a root-finding argument: to show \(f(x)=0\) has a solution in an interval, take \(d=0\) and show that \(f(a)\) and \(f(b)\) have opposite signs.
Worked example · A root of a polynomial
Show that \(x^3-x-1=0\) has a solution in \((1,2)\).
Let \(f(x)=x^3-x-1\). It is a polynomial, hence continuous on \([1,2]\). Compute \(f(1)=1-1-1=-1\lt0\) and \(f(2)=8-2-1=5\gt0\). Since \(0\) is between \(f(1)=-1\) and \(f(2)=5\), the Intermediate Value Theorem guarantees a number \(c\) in \((1,2)\) with \(f(c)=0\). So the equation has a solution in \((1,2)\).
Worked example · Rewriting an equation
Show that \(\cos x=x\) has a solution in \((0,1)\).
Rewrite as \(g(x)=\cos x-x=0\). Both \(\cos x\) and \(x\) are continuous, so \(g\) is continuous on \([0,1]\). \(g(0)=1-0=1\gt0\) and \(g(1)=\cos1-1\approx-0.4597\lt0\) (with \(x\) in radians). Zero lies between \(g(1)\) and \(g(0)\), so by the IVT there is a \(c\) in \((0,1)\) with \(g(c)=0\), that is, \(\cos c=c\).
Using a table
IVT arguments also work from a table of values, when you are told that the function is continuous.
Worked example · How many solutions are guaranteed?
A continuous function \(f\) on \([0,7]\) has \(f(0)=-4\), \(f(3)=2\), and \(f(7)=-1\). What does the IVT guarantee about the equation \(f(x)=0\)?
\(f(0)=-4\lt0\lt2=f(3)\), so there is at least one \(c_1\) in \((0,3)\) with \(f(c_1)=0\). And \(f(3)=2\gt0\gt-1=f(7)\), so there is at least one \(c_2\) in \((3,7)\) with \(f(c_2)=0\). These intervals do not overlap, so there are at least two solutions in \((0,7)\). The theorem does not rule out more.
When the hypotheses fail
The continuity requirement is essential. Consider \(f(x)=\dfrac1x\) on \([-1,1]\): \(f(-1)=-1\) and \(f(1)=1\), so \(0\) is between them, but \(f(x)=0\) has no solution. The IVT does not apply because \(f\) is not continuous on \([-1,1]\) (it is undefined at \(0\)). Similarly, if \(d\) is not between \(f(a)\) and \(f(b)\), the theorem says nothing: a continuous function with \(f(0)=-3\) and \(f(4)=5\) need not ever equal \(6\).
A related remark: because the theorem is only an existence result, you cannot use it to say a value is unique. Later, when you know a function is also increasing, you can combine the two ideas to show exactly one solution.
Exam focus
IVT: \(f\) continuous on \([a,b]\) and \(d\) between \(f(a)\) and \(f(b)\) implies \(f(c)=d\) for some \(c\) in \((a,b)\).
A complete justification states: continuity on the closed interval, the values \(f(a)\) and \(f(b)\) with \(d\) between them, and the conclusion.
To show \(f(x)=0\) has a solution, find a sign change; to solve \(g(x)=h(x)\), apply the theorem to \(g-h\).
It proves existence only; it does not locate \(c\) or count the solutions.
If \(f\) is not continuous on the whole interval, or \(d\) is outside the range between the endpoint values, the theorem does not apply.
Practice
Show that \(x^3+x-3=0\) has a solution in \((1,2)\).
Show answer
\(f(x)=x^3+x-3\) is continuous (a polynomial). \(f(1)=-1\lt0\) and \(f(2)=7\gt0\). By the IVT there is a \(c\) in \((1,2)\) with \(f(c)=0\).
A function \(f\) is continuous on \([2,6]\) with \(f(2)=5\) and \(f(6)=1\). Must \(f(x)=3\) have a solution in \((2,6)\)? Justify.
Show answer
Yes. \(f\) is continuous on \([2,6]\) and \(3\) is between \(f(6)=1\) and \(f(2)=5\), so the IVT guarantees a \(c\) in \((2,6)\) with \(f(c)=3\).
Explain why the IVT does not show that \(f(x)=\dfrac1x\) has a zero in \([-1,1]\), even though \(f(-1)\lt0\lt f(1)\).
Show answer
\(f\) is not continuous on \([-1,1]\) because it is not defined at \(x=0\). The hypothesis fails, and indeed \(\dfrac1x=0\) has no solution.
Show that \(e^x+x-3=0\) has a solution in \((0,1)\).
Show answer
\(h(x)=e^x+x-3\) is continuous. \(h(0)=1+0-3=-2\lt0\) and \(h(1)=e-2\approx0.718\gt0\). By the IVT there is a \(c\) in \((0,1)\) with \(h(c)=0\).
A continuous function has \(f(0)=-3\) and \(f(4)=5\). Does the IVT guarantee a solution of \(f(x)=6\) in \((0,4)\)?
Show answer
No. The number 6 is not between \(f(0)=-3\) and \(f(4)=5\), so the theorem does not apply. (The function might never reach 6.)
At 8 a.m. the temperature is \(54^\circ\)F; at 2 p.m. it is \(71^\circ\)F. Assuming the temperature varies continuously, explain why it was exactly \(60^\circ\)F at some time between.
Show answer
The temperature function is continuous on \([8,14]\) (hours), and \(60\) is between \(54\) and \(71\). By the IVT there is a time \(t\) in \((8,14)\) when the temperature was \(60^\circ\)F.
Unit 1 practice · 10 problems
Unit 1 practice: Limits and Continuity
Ten problems covering the whole unit, in roughly exam order. Work each one
on paper before revealing the answer: the reveal shows the key steps, not
just the number. No calculator needed.
Multiply by the conjugate: \(\dfrac{(x + 9) - 9}{x\left(\sqrt{x+9} + 3\right)} = \dfrac{1}{\sqrt{x+9} + 3} \to \dfrac{1}{6}\).
Let \(g(x) = \dfrac{x^2 - 5x + 6}{x - 2}\). Classify the discontinuity at \(x = 2\), and give the value of \(g(2)\) that would remove it.
Show answer
\(g(x) = \dfrac{(x-2)(x-3)}{x-2} = x - 3\) for \(x \ne 2\), so the limit at 2 is \(-1\). The discontinuity is removable; defining \(g(2) = -1\) removes it.
For large negative \(x\), \(\sqrt{x^2 + 1} \approx |x| = -x\), so the expression behaves like \(\dfrac{3x}{-x} = -3\). The limit is \(-3\).
\(f\) is continuous on \([1, 5]\) with \(f(1) = -2\) and \(f(5) = 7\). Must there be a \(c\) in \((1, 5)\) with \(f(c) = 3\)? Justify.
Show answer
Yes. \(f\) is continuous on \([1, 5]\) and \(f(1) = -2 < 3 < 7 = f(5)\), so by the Intermediate Value Theorem there is a \(c\) in \((1, 5)\) with \(f(c) = 3\).
Use the Squeeze Theorem to evaluate \(\displaystyle\lim_{x \to 0} x^4\cos\!\left(\frac{2}{x}\right)\).
Show answer
\(-x^4 \le x^4\cos\!\left(\tfrac{2}{x}\right) \le x^4\), and both bounds approach 0. The limit is \(0\).
Average and instantaneous rates of change at a point
Lesson 1.1 introduced the central problem of calculus: the rate of change at a single instant. Lesson 1.2 gave us limits. Now we put the two together. The result, the derivative at a point, is the most important single idea in the course. Everything in Units 2 through 5 is about computing it, interpreting it, and using it.
Required content · CHA-2.A.1, CHA-2.B.1
Learning objectives.
CHA-2.A. Determine average rates of change using difference quotients.
CHA-2.B. Represent the derivative of a function as the limit of a difference quotient.
Essential knowledge.
CHA-2.A.1. The difference quotients \(\dfrac{f(a+h)-f(a)}{h}\) and \(\dfrac{f(x)-f(a)}{x-a}\) express the average rate of change of a function over an interval.
CHA-2.B.1. The instantaneous rate of change of a function at \(x=a\) can be expressed by \(\lim_{h\to 0}\dfrac{f(a+h)-f(a)}{h}\) or \(\lim_{x\to a}\dfrac{f(x)-f(a)}{x-a}\), provided the limit exists. These are equivalent forms of the definition of the derivative and are denoted \(f'(a)\).
Difference quotients
For a function \(f\) and a number \(a\) in its domain, there are two equivalent ways to write the average rate of change of \(f\) over an interval starting or ending at \(a\):
Both are called difference quotients. In the first, \(h\) is the width of the interval (it may be positive or negative); the second names the other endpoint \(x\) and the width is \(x-a\). They are the same quotient with \(x=a+h\). Each is the slope of a secant line through the points \((a,f(a))\) and another point on the graph, and each expresses the average rate of change of \(f\) over an interval.
Figure 2.1. The difference quotient \(\dfrac{f(a+h)-f(a)}{h}\) is the slope of the secant line through \(P\) and \(Q\). As \(h\to0\), \(Q\) slides toward \(P\) and the secant slope approaches the slope of the tangent line, which is \(f'(a)\).
The derivative at a point
If we let the second point slide toward the first, that is, let \(h\to0\) or \(x\to a\), the secant lines approach the tangent line at \(x=a\). When the limit exists, it is the instantaneous rate of change of \(f\) at \(a\):
provided the limit exists. These are two equivalent forms of the definition of the derivative, and the number is read "\(f\) prime of \(a\)." If the limit does not exist, \(f\) has no derivative at \(a\).
Use the form that suits the problem. The \(h\) form is usually easier when \(f\) is a polynomial or a root (expand \((a+h)^n\)); the \(x\to a\) form is convenient when you can factor \(x-a\) out of the numerator, as with \(x^3-a^3\).
Worked example · The derivative of \(x^2\) at \(a=3\)
With \(h\). \(\dfrac{f(3+h)-f(3)}{h}=\dfrac{(9+6h+h^2)-9}{h}=\dfrac{6h+h^2}{h}=6+h\), and \(6+h\to6\). So \(f'(3)=6\).
With \(x\to a\). \(\dfrac{x^2-9}{x-3}=\dfrac{(x-3)(x+3)}{x-3}=x+3\to6\). Same answer, \(f'(3)=6\).
The tangent to \(y=x^2\) at the point \((3,9)\) has slope \(6\).
Worked example · A reciprocal
Use the definition to find \(f'(2)\) for \(f(x)=\dfrac1x\).
As \(h\to0\) this approaches \(\dfrac{-1}{4}\). So \(f'(2)=-\dfrac14\): the graph of \(y=\frac1x\) is falling at \(x=2\), with slope \(-\frac14\).
Worked example · Recognizing a derivative as a limit
Identify \(f\) and \(a\), then evaluate: \(\displaystyle\lim_{h\to0}\frac{(5+h)^3-125}{h}\).
The numerator has the form \(f(a+h)-f(a)\) with \(f(x)=x^3\) and \(a=5\), since \(f(5)=125\). So the limit equals \(f'(5)\). Expanding, \(\dfrac{125+75h+15h^2+h^3-125}{h}=75+15h+h^2\to75\). Hence the limit is \(75\).
Meaning and units
The average rate of change gives the slope of a secant; the derivative gives the slope of the tangent. As in Lesson 1.1, the units of \(f'(a)\) are the units of \(f\) per unit of \(x\). If \(s(t)\) is position in meters and \(t\) is in seconds, \(s'(a)\) is the instantaneous velocity at time \(a\) in meters per second. A derivative at a point is a single number; in Lesson 2.2 we will let the point vary and get a function.
Exam focus
Difference quotients \(\dfrac{f(a+h)-f(a)}{h}\) and \(\dfrac{f(x)-f(a)}{x-a}\) express the average rate of change.
\(f'(a)=\displaystyle\lim_{h\to0}\dfrac{f(a+h)-f(a)}{h}=\lim_{x\to a}\dfrac{f(x)-f(a)}{x-a}\), if the limit exists. It is the slope of the tangent line at \((a,f(a))\).
Many limits on the exam are derivatives in disguise: find the function \(f\), the point \(a\), and evaluate \(f'(a)\).
Write the limit notation on each line until the limit is taken.
Give units with interpretations: units of \(f\) per unit of \(x\).
Practice
Use the limit definition to find \(f'(2)\) for \(f(x)=x^2+1\).
Show answer
\(\dfrac{(2+h)^2+1-5}{h}=\dfrac{4h+h^2}{h}=4+h\to4\). So \(f'(2)=4\).
Use the definition to show that for \(f(x)=3x-2\), \(f'(a)=3\) for every \(a\).
Show answer
\(\dfrac{f(a+h)-f(a)}{h}=\dfrac{3(a+h)-2-(3a-2)}{h}=\dfrac{3h}{h}=3\), so \(f'(a)=3\). (A line has the same slope everywhere.)
Find \(f'(1)\) for \(f(x)=\dfrac1x\) using the definition.
Show answer
\(\dfrac{\frac{1}{1+h}-1}{h}=\dfrac{-h}{(1+h)h}=\dfrac{-1}{1+h}\to-1\). So \(f'(1)=-1\).
Evaluate \(\displaystyle\lim_{h\to0}\dfrac{\sqrt{16+h}-4}{h}\) by recognizing it as a derivative.
Show answer
It is \(f'(16)\) for \(f(x)=\sqrt x\). Multiply by the conjugate: \(\dfrac{h}{h(\sqrt{16+h}+4)}=\dfrac1{\sqrt{16+h}+4}\to\dfrac18\). So the value is \(\dfrac18\).
A particle's position is \(s(t)=t^2+t\) meters. Find the average velocity over \([2,2.5]\) and the instantaneous velocity at \(t=2\).
Show answer
\(s(2)=6\) and \(s(2.5)=8.75\), so the average velocity is \(\dfrac{2.75}{0.5}=5.5\) m/s. Instantaneous: \(\dfrac{(2+h)^2+(2+h)-6}{h}=\dfrac{5h+h^2}{h}=5+h\to5\) m/s.
Use \(\displaystyle\lim_{x\to a}\dfrac{f(x)-f(a)}{x-a}\) to find \(f'(1)\) for \(f(x)=x^3\).
Show answer
\(\dfrac{x^3-1}{x-1}=x^2+x+1\to3\). So \(f'(1)=3\).
The derivative as a function and derivative notation
In Lesson 2.1 we computed a derivative at one chosen point. But the slope of a curve changes from point to point, so it makes sense to ask for the slope at every point at once. Doing the same limit calculation with \(x\) left as a variable produces a new function, the derivative function, and a family of notations for it that you will meet every day for the rest of the course.
CHA-2.B. Represent the derivative of a function as the limit of a difference quotient.
CHA-2.C. Determine the equation of a line tangent to a curve at a given point.
Essential knowledge.
CHA-2.B.2. The derivative of \(f\) is the function whose value at \(x\) is \(\lim_{h\to 0}\dfrac{f(x+h)-f(x)}{h}\), provided this limit exists.
CHA-2.B.3. For \(y=f(x)\), notations for the derivative include \(\dfrac{dy}{dx}\), \(f'(x)\), and \(y'\).
CHA-2.B.4. The derivative can be represented graphically, numerically, analytically, and verbally.
CHA-2.C.1. The derivative of a function at a point is the slope of the line tangent to a graph of the function at that point.
The derivative as a function
For a function \(f\), the derivative of \(f\) is the function whose value at \(x\) is
\[f'(x)=\lim_{h\to0}\frac{f(x+h)-f(x)}{h},\]
provided this limit exists. The domain of \(f'\) is the set of \(x\)-values where the limit exists; it can be smaller than the domain of \(f\). The process of finding \(f'\) is differentiation, and we say \(f\) is differentiable at every point where \(f'\) exists.
Worked example · Differentiating from the definition
As \(h\to0\), this approaches \(2x-3\). So \(f'(x)=2x-3\).
Worked example · A tangent line
Write an equation of the line tangent to \(f(x)=x^2-3x\) at \(x=1\).
\(f(1)=1-3=-2\), so the point is \((1,-2)\). \(f'(1)=2(1)-3=-1\). The tangent line is \[y-(-2)=-1(x-1)\quad\Longrightarrow\quad y=-x-1.\]
The derivative at a point gives the slope of the tangent line at that point: the equation of the line tangent to \(y=f(x)\) at \(x=a\) is \(y-f(a)=f'(a)(x-a)\). Writing this equation is a standard task on the exam: get the point \((a,f(a))\) and the slope \(f'(a)\), and use point-slope form.
Notation
The same derivative has several names, and you must be fluent in all of them. For \(y=f(x)\), notations for the derivative include
The Leibniz symbol \(\dfrac{dy}{dx}\) is read "the derivative of \(y\) with respect to \(x\)." It is not a fraction yet (you cannot cancel the \(d\)'s), but it behaves like one in many calculations and it reminds you of the units: \(\dfrac{dy}{dx}\) has units of \(y\) per unit of \(x\). To evaluate at a point write \(f'(2)\), or \(\left.\dfrac{dy}{dx}\right|_{x=2}\). The prime notation is compact and the Leibniz notation makes clear which variable is changing, which matters when there are several (Unit 4).
Four representations of the derivative
The derivative can be represented graphically, numerically, analytically, and verbally:
Graphically: the slope of the tangent line at each point; where the graph of \(f\) rises, \(f'\) is positive; where it falls, \(f'\) is negative; where it has a flat tangent, \(f'=0\).
Numerically: from a table, by the slopes between nearby points (Lesson 2.3).
Analytically: by a formula, such as \(f'(x)=2x-3\).
Verbally: "the instantaneous rate of change of \(f\) with respect to \(x\)" or "the slope of the tangent line."
Figure 2.2. Left: the graph of \(f\) with tangent lines at \(x=-1,0,1\) (slopes \(0,-1,0\)). Right: the graph of \(f'\), whose height at each \(x\) is the slope of the tangent line to \(f\) there. Where \(f'\gt0\), \(f\) rises; where \(f'\lt0\), \(f\) falls.
Worked example · Reading a derivative graph
In the figure, \(f(x)=\dfrac{x^3}{3}-x\) and \(f'(x)=x^2-1\). Where does \(f\) have horizontal tangents? Where is \(f\) decreasing?
\(f'(x)=x^2-1=0\) at \(x=-1\) and \(x=1\): horizontal tangents there. \(f'\lt0\) for \(-1\lt x\lt1\) (the parabola is below the axis), so \(f\) is decreasing on \((-1,1)\) and increasing for \(x\lt-1\) and \(x\gt1\). The graph of \(f'\) tells the story of \(f\) without looking at \(f\) at all.
Exam focus
\(f'(x)=\displaystyle\lim_{h\to0}\dfrac{f(x+h)-f(x)}{h}\), where the limit exists. This defines a function.
Notation: \(f'(x)\), \(y'\), \(\dfrac{dy}{dx}\), \(\dfrac{d}{dx}[f(x)]\). Evaluate with \(f'(a)\) or \(\left.\dfrac{dy}{dx}\right|_{x=a}\).
Tangent line at \(x=a\): \(y-f(a)=f'(a)(x-a)\). Find the point and the slope.
\(f'>0\) means \(f\) is increasing, \(f'<0\) means decreasing, \(f'=0\) means a horizontal tangent.
Do not treat \(\dfrac{dy}{dx}\) as \(dy\) divided by \(dx\) until the course says so.
Practice
Use the limit definition to find \(f'(x)\) for \(f(x)=x^2+2x\).
Show answer
\(\dfrac{(x+h)^2+2(x+h)-x^2-2x}{h}=\dfrac{2xh+h^2+2h}{h}=2x+2+h\to2x+2\). So \(f'(x)=2x+2\).
Use the definition to find \(f'(x)\) for \(f(x)=\dfrac{1}{x+1}\).
Show answer
\(\dfrac{\frac1{x+h+1}-\frac1{x+1}}{h}=\dfrac{-h}{(x+h+1)(x+1)h}=\dfrac{-1}{(x+h+1)(x+1)}\to\dfrac{-1}{(x+1)^2}\). So \(f'(x)=-\dfrac1{(x+1)^2}\).
Write an equation of the tangent line to \(f(x)=x^2-4x\) at \(x=3\).
Show answer
\(f(3)=-3\) and \(f'(x)=2x-4\), so \(f'(3)=2\). The line is \(y+3=2(x-3)\), or \(y=2x-9\).
For \(f(x)=x^2-6x+5\) with \(f'(x)=2x-6\), at what \(x\) is the tangent line horizontal? State the point.
Show answer
Set \(2x-6=0\), so \(x=3\). \(f(3)=9-18+5=-4\), so the point is \((3,-4)\).
Write the derivative of \(y=f(x)\) in three notations, and explain what the number \(\left.\dfrac{dy}{dx}\right|_{x=2}\) means graphically.
Show answer
\(f'(x)\), \(y'\), and \(\dfrac{dy}{dx}\) (also \(\dfrac{d}{dx}[f(x)]\)). The number \(\left.\dfrac{dy}{dx}\right|_{x=2}\) is the slope of the tangent line to the graph at \(x=2\).
If \(f'(x)=3x^2-2\), what is the slope of the tangent line to \(f\) at \(x=-1\)? Is \(f\) increasing or decreasing there?
Show answer
\(f'(-1)=3-2=1\). The slope is 1, which is positive, so \(f\) is increasing at \(x=-1\).
Real data does not come with formulas. A weather station records the temperature each hour, a car's computer logs its distance every few seconds, and a graph in an exam question may have no equation at all. You can still estimate a derivative from that information, because the derivative is a limit of average rates of change, and average rates of change can be computed from any two data points. This lesson shows the standard estimates, and the justifications the exam expects.
Required content · CHA-2.D.1, CHA-2.D.2
Learning objective.
CHA-2.D. Estimate derivatives.
Essential knowledge.
CHA-2.D.1. The derivative at a point can be estimated from information given in tables or graphs.
CHA-2.D.2. Technology can be used to calculate or estimate the value of a derivative of a function at a point.
From a table
Given values of \(f\) at several inputs, the derivative at a point \(x=a\) can be estimated with an average rate of change over a small interval containing or touching \(a\). The better the estimate, the smaller the interval, and (for smooth data) the more symmetric the interval is about \(a\).
One-sided estimates. \(f'(a)\approx\dfrac{f(b)-f(a)}{b-a}\) with \(b\) the next input to the right, or with \(b\) the previous input to the left.
Symmetric (two-sided) estimate. If \(a\) lies between data points \(a-p\) and \(a+q\), use \(f'(a)\approx\dfrac{f(a+q)-f(a-p)}{q+p}\). For equally spaced data this is the average of the left and right slopes, and it is usually the most accurate.
On the free-response section, when a question says "use the data in the table to estimate," write the quotient with the numbers in it and the units, then evaluate it: \(\dfrac{f(7)-f(3)}{7-3}\), not just a decimal.
Worked example · Equally spaced data
The table shows values of a function \(f\).
\(x\)
1.0
1.5
2.0
2.5
3.0
\(f(x)\)
3.2
4.1
5.5
7.6
10.4
Estimate \(f'(2)\).
Left slope. \(\dfrac{5.5-4.1}{2.0-1.5}=2.8\). Right slope. \(\dfrac{7.6-5.5}{2.5-2.0}=4.2\). Symmetric estimate. \(\dfrac{7.6-4.1}{2.5-1.5}=\dfrac{3.5}{1}=3.5\), the average of the two. We estimate \(f'(2)\approx3.5\). The one-sided slopes straddle this value because \(f\) is getting steeper.
Worked example · Unequal spacing and units
A runner's distance \(s(t)\), in meters, is recorded at some times \(t\), in seconds:
\(t\) (s)
0
2
5
9
12
\(s(t)\) (m)
0
14
60
170
262
Estimate the runner's velocity at \(t=5\) and interpret.
Because the times are unequally spaced, use the endpoints of the two neighboring intervals: \(s'(5)\approx\dfrac{s(9)-s(2)}{9-2}=\dfrac{170-14}{7}\approx22.3\) meters per second. (The one-sided slopes are \(\dfrac{60-14}{3}\approx15.3\) m/s on the left and \(\dfrac{170-60}{4}=27.5\) m/s on the right.) At \(t=5\) seconds the runner's velocity is approximately 22.3 meters per second, and the runner is speeding up.
From a graph
To estimate \(f'(a)\) from a graph, find the point \((a,f(a))\), picture the line that just touches the curve there, choose two points on that line, and compute its slope by rise over run. A good estimate uses two points on the tangent line that are far apart, so that reading error matters less. Say which points you used.
Worked example · Tangent line on a graph
A tangent line to the graph of \(g\) at \(x=2\) passes through \((0,1)\) and \((4,9)\). Estimate \(g'(2)\).
Slope \(=\dfrac{9-1}{4-0}=2\). So \(g'(2)\approx2\).
With technology
A graphing calculator can compute a numerical derivative at a point (on many calculators, nDeriv or a "d/dx" feature). It is one of the four calculator capabilities listed for the AP exam, so on a free-response question you must still write the setup, for example "\(f'(2)\)" or "\(\dfrac{d}{dx}(x^2\ln x)\)" at \(x=2\), along with the value the calculator gives. The calculator returns a number, not a formula, and it can be inaccurate at points where the derivative does not exist (such as a corner), where it may return a plausible but meaningless value.
Calculator
To find \(f'(a)\) for \(f(x)=x\ln x\), enter \(Y_1=X\ln(X)\) and compute the numerical derivative at \(X=3\). You should get \(\ln3+1\approx2.099\). Always write the quantity you are computing, such as \(f'(3)\approx2.099\), and report decimals to three places unless told otherwise.
Exam focus
The derivative is estimated by an average rate of change over a small interval: \(\dfrac{f(b)-f(a)}{b-a}\).
Symmetric intervals about the point usually give the best estimate; for unequal spacing use the nearest data points on either side.
From a graph, use two points on the tangent line far apart and compute rise over run.
Show the quotient with its numbers and give units; an estimate is "approximately," not "equals."
A calculator derivative is a numerical estimate; write the setup, and do not trust it at a corner or cusp.
Practice
Use the table \(f(1)=3.2,\ f(1.5)=4.1,\ f(2)=5.5,\ f(2.5)=7.6,\ f(3)=10.4\) to estimate \(f'(2.5)\) with a symmetric interval.
Show answer
\(\dfrac{f(3)-f(2)}{3-2}=\dfrac{10.4-5.5}{1}=4.9\). So \(f'(2.5)\approx4.9\).
The temperature \(T(t)\), in \(^\circ\)F, at \(t\) hours after noon is \(T(1)=70\), \(T(2)=75\), \(T(4)=79\). Estimate \(T'(2)\) and interpret with units.
Show answer
The symmetric estimate is \(\dfrac{T(4)-T(1)}{4-1}=\dfrac{79-70}{3}=3\). At 2 p.m. the temperature is rising at about 3 degrees Fahrenheit per hour.
A tangent line to \(h\) at \(x=3\) passes through \((1,7)\) and \((5,-1)\). Estimate \(h'(3)\).
Show answer
\(\dfrac{-1-7}{5-1}=-2\). So \(h'(3)\approx-2\).
A table gives \(v(3)=12\) m/s and \(v(3.1)=12.4\) m/s. Estimate the acceleration at \(t=3\). Include units.
Show answer
\(\dfrac{12.4-12}{3.1-3}=4\). The acceleration is approximately 4 meters per second per second (m/s\(^2\)).
Why is the symmetric estimate for \(f'(a)\) usually better than a one-sided estimate with the same spacing?
Show answer
The one-sided slope is tied to a point on one side of \(a\) and is biased by the way the slope changes there. The symmetric estimate averages the left and right behavior, so that error mostly cancels.
A student estimates \(f'(2)\) from the table \(f(2)=5.5\), \(f(2.5)=7.6\) and writes "\(f'(2)=4.2\)." What is wrong with the statement?
Show answer
The quotient \(\dfrac{7.6-5.5}{0.5}=4.2\) is only an estimate, so it should read \(f'(2)\approx4.2\), and the quotient with the numbers should be shown. It also uses only one side; the actual derivative could differ.
A derivative is a limit, and limits can fail to exist. So not every function has a derivative at every point. This lesson answers two questions the exam asks over and over: where can a derivative fail to exist, and what is the relationship between being differentiable and being continuous?
Required content · FUN-2.A.1, FUN-2.A.2
Learning objective.
FUN-2.A. Explain the relationship between differentiability and continuity.
Essential knowledge.
FUN-2.A.1. If a function is differentiable at a point, then it is continuous at that point. In particular, if a point is not in the domain of \(f\), then it is not in the domain of \(f'\).
FUN-2.A.2. A continuous function may fail to be differentiable at a point in its domain.
Differentiability implies continuity
A function \(f\) is differentiable at \(x=a\) if \(f'(a)\) exists, that is, if \(\displaystyle\lim_{h\to0}\dfrac{f(a+h)-f(a)}{h}\) exists as a real number. Differentiable functions are always continuous:
Theorem
If \(f\) is differentiable at \(x=a\), then \(f\) is continuous at \(x=a\).
Here is why. If \(f'(a)\) exists, then \(f(a)\) must be defined (the quotient needs it), and \[f(a+h)-f(a)=\frac{f(a+h)-f(a)}{h}\cdot h\ \longrightarrow\ f'(a)\cdot0=0\quad\text{as }h\to0,\] so \(\displaystyle\lim_{h\to0}f(a+h)=f(a)\), which is the definition of continuity at \(a\). In particular, if a point is not in the domain of \(f\), then it is not in the domain of \(f'\).
The converse is false. A continuous function may fail to be differentiable at a point in its domain. Continuity says there is no break; differentiability says there is no break and the graph is smooth enough to have a single tangent line.
Four ways a derivative can fail to exist
Figure 2.3. Four ways differentiability can fail at \(x=0\): a corner (\(y=|x|\)), a cusp (\(y=x^{2/3}\)), a vertical tangent (\(y=x^{1/3}\)), and a discontinuity (a jump). In each case the difference quotient has no finite limit there.
A corner. The one-sided derivatives (left and right slopes) exist but are different. Example: \(f(x)=|x|\) at \(0\).
A cusp. The slopes approach \(+\infty\) from one side and \(-\infty\) from the other. Example: \(f(x)=x^{2/3}\) at \(0\).
A vertical tangent. The slope is unbounded with the same sign on both sides. Example: \(f(x)=x^{1/3}\) at \(0\). The function is continuous there but the tangent line is vertical and has no slope.
A discontinuity. If \(f\) is not continuous at \(a\), it cannot be differentiable there (this is the contrapositive of the theorem).
Worked example · A corner
Show that \(f(x)=|x|\) is continuous but not differentiable at \(x=0\).
Continuous. \(f(0)=0\) and \(\displaystyle\lim_{x\to0}|x|=0\). The three conditions hold.
Not differentiable. \(\dfrac{f(0+h)-f(0)}{h}=\dfrac{|h|}{h}\). For \(h\gt0\) this is \(1\); for \(h\lt0\) it is \(-1\). The one-sided limits are \(1\) and \(-1\), so \(\displaystyle\lim_{h\to0}\dfrac{|h|}{h}\) does not exist and \(f'(0)\) does not exist.
Worked example · A vertical tangent
Show that \(f(x)=x^{1/3}\) is not differentiable at \(x=0\).
\(\dfrac{f(h)-f(0)}{h}=\dfrac{h^{1/3}}{h}=h^{-2/3}=\dfrac{1}{h^{2/3}}\), which is positive and grows without bound as \(h\to0\) from either side. The limit is \(+\infty\), not a real number, so \(f'(0)\) does not exist. The graph has a vertical tangent line at the origin. (Notice that \(f\) is continuous at \(0\).)
Piecewise functions: smooth joins
A piecewise function is differentiable at a boundary only if it is first continuous there, and the slopes of the two pieces match. To find parameters that make a function differentiable at a boundary \(x=c\), solve two equations: equal values (continuity) and equal derivatives (matching slopes).
Worked example · Making a join smooth
Find \(a\) and \(b\) so that \(f(x)=\begin{cases}x^2,&x\le1\\ ax+b,&x\gt1\end{cases}\) is differentiable at \(x=1\).
Continuity at 1. Left value \(1^2=1\); right value \(a+b\). So \(a+b=1\).
Matching slopes. The derivative of \(x^2\) at \(1\) is \(2(1)=2\); the derivative of \(ax+b\) is \(a\). So \(a=2\).
Then \(b=1-2=-1\). Check: with \(a=2\), \(b=-1\) the line \(y=2x-1\) is the tangent to the parabola at \((1,1)\), so the two pieces join without a corner.
What to write
When asked whether a function is differentiable at a point, first check continuity (if it fails, you are done). If it holds, compare one-sided derivatives or use the definition. A statement like "the left derivative is \(-1\) and the right derivative is \(1\), so \(f'(0)\) does not exist" earns the point. "It has a sharp point" does not.
Exam focus
Differentiable at \(a\) \(\Rightarrow\) continuous at \(a\). The converse is false.
A derivative fails to exist at a corner, a cusp, a vertical tangent, or a discontinuity.
For a corner, show the one-sided derivatives are different numbers; for a vertical tangent, the difference quotient is unbounded.
Piecewise, to be differentiable at \(c\): the pieces agree in value and in derivative at \(c\).
Continuity alone is not enough to conclude the derivative exists; a theorem requiring "differentiable on \((a,b)\)" needs a check of this kind.
Practice
Is \(f(x)=|x-2|\) differentiable at \(x=2\)? Justify using one-sided derivatives.
Show answer
For \(h\gt0\), \(\dfrac{|h|}{h}=1\); for \(h\lt0\), \(\dfrac{|h|}{h}=-1\). The one-sided limits differ, so \(f'(2)\) does not exist. (The function is continuous at 2 but has a corner there.)
True or false: if \(f\) is continuous at \(a\), then \(f\) is differentiable at \(a\). Give a counterexample if false.
Show answer
False. \(f(x)=|x|\) is continuous at \(0\) but not differentiable there.
Is \(f(x)=\sqrt{x}\) differentiable at \(x=0\)? Explain.
Show answer
\(\dfrac{f(h)-f(0)}{h}=\dfrac{\sqrt h}{h}=\dfrac1{\sqrt h}\to\infty\) as \(h\to0^+\). The slope is unbounded, so there is a vertical tangent and \(f'(0)\) does not exist. (The function is also not defined for \(x\lt0\).)
Find \(a\) and \(b\) so that \(f(x)=\begin{cases}x^2+1,&x\le2\\ ax+b,&x\gt2\end{cases}\) is differentiable at \(x=2\).
Show answer
Continuity: \(5=2a+b\). Slopes: \(2(2)=4=a\). So \(a=4\) and \(b=5-8=-3\).
A function \(g\) has a jump discontinuity at \(x=3\). Can \(g'(3)\) exist? Explain.
Show answer
No. If \(g\) were differentiable at 3, it would be continuous there. Since it has a jump, it is not continuous, so \(g'(3)\) does not exist.
A graph of \(f\) is smooth except for a sharp point at \(x=1\), where the graph is unbroken. What type of failure is this, and what can you say about \(\displaystyle\lim_{x\to1}f(x)\)?
Show answer
It is a corner (or cusp). The function is continuous at \(x=1\), so \(\displaystyle\lim_{x\to1}f(x)=f(1)\), but \(f'(1)\) does not exist.
The power rule and the sum and constant multiple rules
Finding a derivative from the limit definition works, but each one is a small algebra project. The good news is that derivatives of whole families of functions follow patterns. After this lesson you will differentiate any polynomial, and any sum of powers, in a single line. These first rules are the foundation for all of the rules that follow.
FUN-3.A. Calculate derivatives of familiar functions.
Essential knowledge.
FUN-3.A.1. Direct application of the definition of the derivative and specific rules can be used to calculate the derivative for functions of the form \(f(x)=x^r\).
FUN-3.A.2. Sums, differences, and constant multiples of functions can be differentiated using derivative rules.
FUN-3.A.3. The power rule combined with sum, difference, and constant multiple properties can be used to find the derivatives for polynomial functions.
The power rule
For any real number \(n\), the derivative of \(f(x)=x^n\) is
\[\frac{d}{dx}\big[x^n\big]=n\,x^{n-1}.\]
In words: bring the exponent down as a coefficient, then reduce the exponent by one. Direct application of the definition of the derivative shows why. For \(n=3\), \[\frac{(x+h)^3-x^3}{h}=\frac{3x^2h+3xh^2+h^3}{h}=3x^2+3xh+h^2\ \longrightarrow\ 3x^2,\] and the same pattern with the binomial expansion gives \(nx^{n-1}\) for any positive integer \(n\). The rule is also valid for negative and fractional exponents, which is how you differentiate reciprocals and roots: rewrite them as powers first.
Rewrite
As a power
Derivative
\(\dfrac{1}{x}\)
\(x^{-1}\)
\(-x^{-2}=-\dfrac{1}{x^2}\)
\(\dfrac{1}{x^3}\)
\(x^{-3}\)
\(-3x^{-4}=-\dfrac{3}{x^4}\)
\(\sqrt{x}\)
\(x^{1/2}\)
\(\tfrac12x^{-1/2}=\dfrac{1}{2\sqrt{x}}\)
\(\sqrt[3]{x^2}\)
\(x^{2/3}\)
\(\tfrac23x^{-1/3}\)
Constant, constant multiple, sum, and difference
Three more rules make the power rule work for expressions with several terms. If \(c\) is a constant and \(f\), \(g\) are differentiable:
Constant rule: \(\dfrac{d}{dx}[c]=0\). (The graph is a horizontal line; its slope is zero.)
Sum and difference rule: \(\dfrac{d}{dx}[f(x)\pm g(x)]=f'(x)\pm g'(x)\).
Sums, differences, and constant multiples of functions can be differentiated using these derivative rules, and combined with the power rule they differentiate every polynomial: \(\dfrac{d}{dx}\left[a_nx^n+\dots+a_1x+a_0\right]=na_nx^{n-1}+\dots+a_1\).
Worked example · A polynomial
Differentiate \(f(x)=5x^4-3x^2+7x-9\).
Apply the rules term by term: \(f'(x)=5(4x^3)-3(2x)+7(1)-0=20x^3-6x+7\).
Worked example · Rewrite, then differentiate
Differentiate \(g(x)=\sqrt{x}+\dfrac{4}{x^3}-2\).
Rewrite as powers: \(g(x)=x^{1/2}+4x^{-3}-2\). Then \[g'(x)=\tfrac12x^{-1/2}+4(-3)x^{-4}-0=\frac{1}{2\sqrt{x}}-\frac{12}{x^4}.\] (The derivative of a quotient such as \(\dfrac{4}{x^3}\) can be found with the power rule once you rewrite it; you do not need the quotient rule when the numerator is a constant.)
Applications: horizontal and tangent lines
Worked example · Horizontal tangents
Find the points where \(f(x)=x^3-6x^2+9x+1\) has a horizontal tangent.
\(f'(x)=3x^2-12x+9=3(x^2-4x+3)=3(x-1)(x-3)\). Setting \(f'(x)=0\) gives \(x=1\) and \(x=3\). The points are \((1,\,f(1))=(1,5)\) and \((3,\,f(3))=(3,1)\), since \(f(1)=1-6+9+1=5\) and \(f(3)=27-54+27+1=1\).
Worked example · A tangent line to a fractional power
Write an equation of the line tangent to \(y=x^{3/2}\) at \(x=4\).
\(y(4)=4^{3/2}=8\). \(y'=\tfrac32x^{1/2}\), so \(y'(4)=\tfrac32\cdot2=3\). The tangent line is \(y-8=3(x-4)\), or \(y=3x-4\).
Common mistakes
Differentiating a constant as if it had a variable. \(\dfrac{d}{dx}[\pi^2]=0\), not \(2\pi\). Numbers like \(\pi\), \(e\), and \(k\) (when \(k\) is a constant) are constants.
Forgetting to rewrite. Do not apply the power rule to \(\dfrac1{x^3}\) as written; make it \(x^{-3}\) first.
Dropping the coefficient. The derivative of \(5x^4\) is \(20x^3\), not \(4x^3\) or \(5x^3\).
Power rule: \(\dfrac{d}{dx}[x^n]=nx^{n-1}\) for any real \(n\); rewrite roots and reciprocals as powers first.
Constant: \(\dfrac{d}{dx}[c]=0\). Constant multiple: \(\dfrac{d}{dx}[cf]=cf'\). Sum and difference: differentiate term by term.
Horizontal tangent: solve \(f'(x)=0\), then find the \(y\)-coordinates from \(f\).
Tangent line: slope \(f'(a)\), point \((a,f(a))\).
Simplify the answer only as far as the question requires; on the AP exam an unsimplified correct derivative earns credit.
Practice
Find \(\dfrac{d}{dx}\left(4x^5-x^3+2x\right)\).
Show answer
\(20x^4-3x^2+2\).
Differentiate \(f(x)=\dfrac{1}{x^2}+3\sqrt{x}\).
Show answer
\(f(x)=x^{-2}+3x^{1/2}\), so \(f'(x)=-2x^{-3}+\tfrac32x^{-1/2}=-\dfrac{2}{x^3}+\dfrac{3}{2\sqrt{x}}\).
Find the slope of the graph of \(f(x)=x^3-4x\) at \(x=2\).
Show answer
\(f'(x)=3x^2-4\), so \(f'(2)=12-4=8\).
At what \(x\)-values does \(f(x)=x^4-8x^2\) have a horizontal tangent?
Show answer
\(f'(x)=4x^3-16x=4x(x^2-4)=4x(x-2)(x+2)\), which is zero at \(x=0,\ 2,\ -2\).
Write an equation of the tangent line to \(y=x^2+3x\) at \(x=1\).
Show answer
\(y(1)=4\) and \(y'=2x+3\), so \(y'(1)=5\). The line is \(y-4=5(x-1)\), or \(y=5x-1\).
Find \(\dfrac{d}{dx}\left(\pi^2+e^3\right)\) and explain.
Show answer
Both \(\pi^2\) and \(e^3\) are constants (numbers), so the derivative is \(0\). The constant rule applies; the power rule does not, because the variable is not in the base.
Polynomials are only the beginning. The functions that model waves, growth, and decay are sine, cosine, and the exponential \(e^x\); the function that undoes the exponential is \(\ln x\). Their derivatives are among the formulas you will use most, and you should know all four from memory.
Required content · FUN-3.A.4, LIM-3.A.1
Learning objectives.
FUN-3.A. Calculate derivatives of familiar functions.
LIM-3.A. Interpret a limit as a definition of a derivative.
Essential knowledge.
FUN-3.A.4. Specific rules can be used to find the derivatives for sine, cosine, exponential, and logarithmic functions.
LIM-3.A.1. In some cases, recognizing an expression for the definition of the derivative of a function whose derivative is known offers a strategy for determining a limit.
Specific rules can be used to find the derivatives for sine, cosine, exponential, and logarithmic functions, and you can see where they come from. For sine, with \(x\) in radians, \[\frac{\sin(x+h)-\sin x}{h}=\sin x\cdot\frac{\cos h-1}{h}+\cos x\cdot\frac{\sin h}{h}\ \longrightarrow\ \sin x\cdot0+\cos x\cdot1=\cos x,\] using the two trigonometric limits from Lesson 1.6. The special feature of \(e^x\) is that it is its own derivative: the slope of \(y=e^x\) equals its height at every point, so at \(x=0\) the tangent line has slope \(e^0=1\). All of these rules assume radian measure; with degrees, extra factors of \(\frac{\pi}{180}\) would appear.
Worked example · Combining rules
Differentiate \(f(x)=3\sin x-2\cos x+5e^x\) and \(g(x)=\ln x+3x^2-e^x\).
Using the constant multiple and sum rules with the new derivatives: \[f'(x)=3\cos x-2(-\sin x)+5e^x=3\cos x+2\sin x+5e^x,\] \[g'(x)=\frac1x+6x-e^x.\] Watch the sign: the derivative of \(\cos x\) is \(-\sin x\), so \(-2\cos x\) differentiates to \(+2\sin x\).
Worked example · A tangent line
Write an equation of the tangent line to \(y=\sin x\) at \(x=\dfrac\pi3\).
\(y\left(\tfrac\pi3\right)=\tfrac{\sqrt3}{2}\). \(y'=\cos x\), so \(y'\left(\tfrac\pi3\right)=\tfrac12\). The tangent line is \(y-\tfrac{\sqrt3}{2}=\tfrac12\left(x-\tfrac\pi3\right)\).
Limits that are derivatives
Lesson 2.1 showed that a limit of the form \(\displaystyle\lim_{h\to0}\dfrac{f(a+h)-f(a)}{h}\) is the derivative \(f'(a)\). When \(f\) is a function whose derivative you know, recognizing a limit as the definition of a derivative gives a strategy for evaluating it without any algebra: identify \(f\) and \(a\), differentiate, and substitute.
Worked example · Evaluating limits by recognizing derivatives
Evaluate each limit.
(a) \(\displaystyle\lim_{h\to0}\frac{\cos\left(\frac\pi3+h\right)-\cos\frac\pi3}{h}\). This is \(f'\left(\frac\pi3\right)\) for \(f(x)=\cos x\). \(f'(x)=-\sin x\), so the limit is \(-\sin\frac\pi3=-\dfrac{\sqrt3}{2}\).
(b) \(\displaystyle\lim_{x\to0}\frac{e^x-1}{x}\). Write \(\dfrac{e^x-e^0}{x-0}\); it is \(f'(0)\) for \(f(x)=e^x\), which is \(e^0=1\).
(c) \(\displaystyle\lim_{h\to0}\frac{\ln(2+h)-\ln2}{h}\). It is \(f'(2)\) for \(f(x)=\ln x\); \(f'(x)=\frac1x\), so the limit is \(\dfrac12\).
This skill pays off on the exam: a "limit" question that looks impossible algebraically is often just a derivative in disguise.
Using a calculator
The derivative of \(\sin x\) is \(\cos x\) only for radians. Make sure your calculator is in radian mode before you check a trigonometric derivative or evaluate a trigonometric function in a calculus problem.
Differentiate sums and constant multiples term by term and mind the sign of \(-\sin x\).
A limit \(\displaystyle\lim_{h\to0}\dfrac{f(a+h)-f(a)}{h}\) (or \(\lim_{x\to a}\dfrac{f(x)-f(a)}{x-a}\)) equals \(f'(a)\); identify \(f\) and \(a\), then differentiate.
Learn these four by heart; the derivatives of the other trigonometric functions and of inverse functions come from them.
Practice
Differentiate \(f(x)=2\sin x+e^x-\ln x\).
Show answer
\(f'(x)=2\cos x+e^x-\dfrac1x\).
Find \(\dfrac{d}{dx}\left(\cos x-4x^3\right)\).
Show answer
\(-\sin x-12x^2\).
Find the slope of the tangent line to \(y=e^x\) at \(x=0\), and write its equation.
Show answer
The slope is \(e^0=1\). The point is \((0,1)\), so the tangent line is \(y=x+1\).
Evaluate \(\displaystyle\lim_{h\to0}\dfrac{\ln(e+h)-1}{h}\) by recognizing it as a derivative.
Show answer
It is \(f'(e)\) for \(f(x)=\ln x\), since \(\ln e=1\). \(f'(e)=\dfrac1e\), so the limit is \(\dfrac1e\).
Evaluate \(\displaystyle\lim_{h\to0}\dfrac{\sin\left(\frac\pi2+h\right)-1}{h}\).
Show answer
It is \(f'\left(\frac\pi2\right)\) for \(f(x)=\sin x\), since \(\sin\frac\pi2=1\). \(f'\left(\frac\pi2\right)=\cos\frac\pi2=0\). The limit is 0.
Find the exact \(x\)-coordinate where \(f(x)=e^x-2x\) has a horizontal tangent.
Show answer
\(f'(x)=e^x-2=0\), so \(e^x=2\) and \(x=\ln2\approx0.693\).
It is tempting to guess that the derivative of a product is the product of the derivatives. It is not. Test it on \(x\cdot x=x^2\): the derivatives multiply to \(1\cdot1=1\), but the derivative of \(x^2\) is \(2x\). The correct rule keeps track of how each factor's change contributes to the change in the product.
Required content · FUN-3.B.1
Learning objective.
FUN-3.B. Calculate derivatives of products and quotients of differentiable functions.
Essential knowledge.
FUN-3.B.1. Derivatives of products of differentiable functions can be found using the product rule.
The rule
If \(f\) and \(g\) are differentiable, then \[\frac{d}{dx}\big[f(x)\,g(x)\big]=f'(x)\,g(x)+f(x)\,g'(x).\]
In words: (derivative of the first)(second) plus (first)(derivative of the second). Derivatives of products of differentiable functions can be found using the product rule.
Why it works: picture a rectangle with sides \(f\) and \(g\), whose area is \(fg\). When \(x\) increases a little, side \(f\) grows by about \(f'\,dx\) and side \(g\) by about \(g'\,dx\). The area gains a strip of width \(g\) along one side (area \(g\,f'\,dx\)) and a strip of width \(f\) along the other (area \(f\,g'\,dx\)), plus a tiny corner \(f'g'(dx)^2\) that vanishes in the limit. So \((fg)'=f'g+fg'\).
Worked example · A product with \(e^x\)
Differentiate \(y=x^2e^x\).
Take \(f=x^2\) and \(g=e^x\), so \(f'=2x\) and \(g'=e^x\): \[y'=2x\cdot e^x+x^2\cdot e^x=e^x\,(x^2+2x)=x\,e^x(x+2).\]
Worked example · Checking against expansion
Differentiate \(y=(3x^2-1)(x^3+2x)\) with the product rule, then check by expanding first.
\[y'=6x\,(x^3+2x)+(3x^2-1)(3x^2+2)=6x^4+12x^2+9x^4+6x^2-3x^2-2=15x^4+15x^2-2.\] Expanding: \(y=3x^5+6x^3-x^3-2x=3x^5+5x^3-2x\), so \(y'=15x^4+15x^2-2\). Both methods agree. When you can multiply out easily, either works; the product rule is essential when you cannot, as with \(x^2e^x\).
Worked example · A tangent line
Write an equation of the tangent line to \(h(x)=x\ln x\) at \(x=e\).
\(h'(x)=1\cdot\ln x+x\cdot\dfrac1x=\ln x+1\). At \(x=e\): \(h(e)=e\ln e=e\) and \(h'(e)=2\). The line is \(y-e=2(x-e)\), that is, \(y=2x-e\).
Using given values
A very common exam question gives values of \(f\), \(f'\), \(g\), \(g'\) at a point (often in a table) and asks for the derivative of a product or quotient there.
Given \(f(1)=2\), \(f'(1)=3\), \(g(1)=-4\), \(g'(1)=1\), find \(\dfrac{d}{dx}[f(x)g(x)]\) at \(x=1\).
Show answer
\(f'(1)g(1)+f(1)g'(1)=3(-4)+2(1)=-12+2=-10\).
At what \(x\)-values does \(y=x^2e^x\) have a horizontal tangent?
Show answer
\(y'=e^x(x^2+2x)=xe^x(x+2)\). Since \(e^x\) is never zero, \(y'=0\) at \(x=0\) and \(x=-2\).
A student writes \(\dfrac{d}{dx}\left[x^2\sin x\right]=2x\cos x\). Identify the error and give the correct derivative.
Show answer
The student multiplied the derivatives of the two factors (\(2x\) and \(\cos x\)). That is not the product rule. The correct derivative is \(2x\sin x+x^2\cos x\).
Quotients show up everywhere: average cost, density, a rate per unit, the ratio of two measurements. Their derivatives need a rule of their own, because (as with products) the derivative of a quotient is not the quotient of the derivatives.
Required content · FUN-3.B.2
Learning objective.
FUN-3.B. Calculate derivatives of products and quotients of differentiable functions.
Essential knowledge.
FUN-3.B.2. Derivatives of quotients of differentiable functions can be found using the quotient rule.
The rule
If \(f\) and \(g\) are differentiable and \(g(x)\ne0\), then \[\frac{d}{dx}\left[\frac{f(x)}{g(x)}\right]=\frac{f'(x)\,g(x)-f(x)\,g'(x)}{\big[g(x)\big]^2}.\]
Derivatives of quotients of differentiable functions can be found using the quotient rule. A way to remember it: low d-high minus high d-low, over low squared, where "low" is the denominator and "high" the numerator. The order of the subtraction matters; reversing it flips the sign of the answer.
The rule follows from the product rule. Write \(\dfrac fg=f\cdot g^{-1}\) and differentiate with the product rule and (in the next lesson) the chain rule for \(g^{-1}\). Many students prefer to remember only "\(f'g-fg'\) over \(g^2\)" and write the four pieces first, as in the product rule.
Worked example · A rational function
Differentiate \(y=\dfrac{x^2+1}{x-1}\).
With \(f=x^2+1\), \(g=x-1\), \(f'=2x\), \(g'=1\): \[y'=\frac{2x(x-1)-(x^2+1)(1)}{(x-1)^2}=\frac{2x^2-2x-x^2-1}{(x-1)^2}=\frac{x^2-2x-1}{(x-1)^2}.\]
Worked example · A quotient with \(e^x\)
Differentiate \(y=\dfrac{e^x}{x^2+1}\) and simplify.
\[y'=\frac{e^x(x^2+1)-e^x(2x)}{(x^2+1)^2}=\frac{e^x(x^2-2x+1)}{(x^2+1)^2}=\frac{e^x(x-1)^2}{(x^2+1)^2}.\] The factored form shows that \(y'\ge0\) everywhere and is zero only at \(x=1\).
Worked example · Horizontal tangents
Find the points where \(y=\dfrac{x}{x^2+1}\) has a horizontal tangent.
\(y'=\dfrac{1\cdot(x^2+1)-x(2x)}{(x^2+1)^2}=\dfrac{1-x^2}{(x^2+1)^2}\). The denominator is never zero; \(y'=0\) when \(1-x^2=0\), so \(x=\pm1\). The points are \(\left(1,\tfrac12\right)\) and \(\left(-1,-\tfrac12\right)\).
If the denominator is a constant, rewrite as a constant multiple: \(\dfrac{x^3}{5}=\dfrac15x^3\). If the numerator is a constant, rewrite with a negative exponent: \(\dfrac{4}{x^3}=4x^{-3}\). If a quotient simplifies, simplify first: \(\dfrac{x^2-4}{x-2}=x+2\) for \(x\ne2\). The quotient rule is for genuinely nontrivial quotients.
Exam focus
\(\left(\dfrac fg\right)'=\dfrac{f'g-fg'}{g^2}\); keep the order of subtraction: numerator derivative first.
Write \(f,g,f',g'\) first, then assemble.
Simplify first when you can (constant denominators, cancellable factors, negative exponents).
At a point: \(\left(\dfrac fg\right)'(a)=\dfrac{f'(a)g(a)-f(a)g'(a)}{[g(a)]^2}\).
A quotient's derivative equals zero only where its numerator is zero (and the denominator is not).
Derivatives of tangent, cotangent, secant, and cosecant
You know the derivatives of \(\sin x\) and \(\cos x\). The other four trigonometric functions are quotients or reciprocals of those two, so their derivatives follow from the quotient rule. This is the idea in the course framework: rearranging tangent, cotangent, secant, and cosecant functions using identities allows us to differentiate them using derivative rules. You will want to know the results from memory, but you should also be able to rebuild any of them.
Required content · FUN-3.B.3
Learning objective.
FUN-3.B. Calculate derivatives of products and quotients of differentiable functions.
Essential knowledge.
FUN-3.B.3. Rearranging tangent, cotangent, secant, and cosecant functions using identities allows differentiation of these functions using derivative rules.
A memory aid: the three "co-" functions (\(\cos\), \(\cot\), \(\csc\)) all have negative derivatives.
Where they come from. For tangent, \(\tan x=\dfrac{\sin x}{\cos x}\), and the quotient rule gives \[\frac{\cos x\cdot\cos x-\sin x\cdot(-\sin x)}{\cos^2x}=\frac{\cos^2x+\sin^2x}{\cos^2x}=\frac{1}{\cos^2x}=\sec^2x.\] For secant, \(\sec x=\dfrac1{\cos x}\), so \(\dfrac{0\cdot\cos x-1\cdot(-\sin x)}{\cos^2x}=\dfrac{\sin x}{\cos^2x}=\dfrac{1}{\cos x}\cdot\dfrac{\sin x}{\cos x}=\sec x\tan x\). Cotangent and cosecant work the same way, with \(\sin x\) in the denominator.
Worked example · Combining with the product rule
Differentiate \(y=x\tan x\).
\(y'=1\cdot\tan x+x\sec^2x=\tan x+x\sec^2x\).
Worked example · A sum with secant
Differentiate \(y=\sec x+\tan x\) and factor the result.
Write an equation of the tangent line to \(y=\tan x\) at \(x=\dfrac\pi4\).
\(y\left(\tfrac\pi4\right)=1\). \(y'=\sec^2x\), so \(y'\left(\tfrac\pi4\right)=\left(\sqrt2\right)^2=2\) (because \(\sec\tfrac\pi4=\dfrac{1}{\cos(\pi/4)}=\sqrt2\)). The line is \(y-1=2\left(x-\tfrac\pi4\right)\).
Other forms and recognizing limits
Remember that \(\sec x=\dfrac1{\cos x}\) and \(\csc x=\dfrac1{\sin x}\), and that \(\sec^2x=1+\tan^2x\) and \(\csc^2x=1+\cot^2x\), so a derivative can often be written in more than one equivalent way. As in Lesson 2.7, a limit may be a derivative in disguise: \(\displaystyle\lim_{h\to0}\dfrac{\tan\left(\frac\pi4+h\right)-1}{h}\) is \(f'\left(\frac\pi4\right)\) for \(f(x)=\tan x\), which equals \(2\).
Where these derivatives exist
\(\tan x\) and \(\sec x\) are not defined where \(\cos x=0\), and \(\cot x\) and \(\csc x\) are not defined where \(\sin x=0\); at those \(x\)-values the functions have vertical asymptotes and no derivatives. On any interval inside the domain, the derivative formulas apply.
Is \(f(x) = \begin{cases} x^2 + 2x & x \le 1 \\ 4x - 1 & x > 1 \end{cases}\) differentiable at \(x = 1\)? Justify.
Show answer
Continuity: both pieces equal 3 at \(x = 1\). Slopes: \(2x + 2 = 4\) from the left, \(4\) from the right. Both match, so yes, \(f\) is differentiable at 1 with \(f'(1) = 4\).
Find \(f'\!\left(\dfrac{\pi}{3}\right)\) for \(f(x) = \sec x\tan x\).
Show answer
\(f'(x) = \sec x\tan^2 x + \sec^3 x\). At \(\tfrac{\pi}{3}\), \(\sec = 2\) and \(\tan = \sqrt{3}\): \(2\cdot 3 + 8 = 14\).
Most functions you will meet are not simple powers or trig functions but compositions: \(\sin(x^2)\), \(e^{-3x}\), \(\sqrt{1-x^2}\), \((2x+5)^4\). One function is plugged into another. The power, product, and quotient rules cannot handle these on their own. The chain rule is the rule for differentiating a composite function, and it is the most heavily used rule in all of calculus. Almost every derivative after this lesson uses it somewhere.
Required content · FUN-3.C.1
Learning objective.
FUN-3.C. Calculate derivatives of compositions of differentiable functions.
Essential knowledge.
FUN-3.C.1. The chain rule provides a way to differentiate composite functions.
The idea: rates multiply
Suppose \(y\) depends on \(u\), and \(u\) depends on \(x\). If \(y\) changes 3 times as fast as \(u\), and \(u\) changes 2 times as fast as \(x\), then \(y\) changes \(3\times2=6\) times as fast as \(x\). Rates of change multiply along a chain. In Leibniz notation:
If \(g\) is differentiable at \(x\) and \(f\) is differentiable at \(g(x)\), then the composite \(f(g(x))\) is differentiable at \(x\) and \[\frac{d}{dx}\big[f(g(x))\big]=f'(g(x))\cdot g'(x).\]
In words: the derivative of the outside function, evaluated at the inside, times the derivative of the inside. The chain rule provides a way to differentiate composite functions.
A reliable routine is outside / inside. For \(y=(2x+5)^4\): the outside is "something to the 4th," the inside is \(2x+5\). \[y'=4(2x+5)^3\cdot2=8(2x+5)^3.\] Differentiate the outside as if the inside were just a variable (leaving the inside alone), then multiply by the derivative of the inside.
Patterns you will use constantly
Function
Derivative
\(\big[g(x)\big]^n\)
\(n\big[g(x)\big]^{n-1}g'(x)\)
\(\sin\big(g(x)\big)\)
\(\cos\big(g(x)\big)\,g'(x)\)
\(\cos\big(g(x)\big)\)
\(-\sin\big(g(x)\big)\,g'(x)\)
\(e^{g(x)}\)
\(e^{g(x)}\,g'(x)\)
\(\ln\big(g(x)\big)\)
\(\dfrac{g'(x)}{g(x)}\)
Worked example · Inside functions of several kinds
Combining rules is where most exam questions live. Use the product or quotient rule on the outside structure and the chain rule for any composite factor.
Worked example · Chain inside a product
Differentiate \(y=x^2\sqrt{x+1}\).
Product rule with \(f=x^2\), \(g=(x+1)^{1/2}\); the derivative of \(g\) needs the chain rule: \(g'=\tfrac12(x+1)^{-1/2}\). \[y'=2x\sqrt{x+1}+x^2\cdot\frac{1}{2\sqrt{x+1}}.\] Putting over a common denominator, \(y'=\dfrac{4x(x+1)+x^2}{2\sqrt{x+1}}=\dfrac{5x^2+4x}{2\sqrt{x+1}}\).
Worked example · Values from a table
Let \(h(x)=f(g(x))\). If \(g(2)=1\), \(g'(2)=3\), and \(f'(1)=-2\), find \(h'(2)\).
\(h'(2)=f'(g(2))\cdot g'(2)=f'(1)\cdot3=(-2)(3)=-6\). Notice that the outside derivative is evaluated at \(g(2)=1\), not at \(2\).
Worked example · A tangent line
Write an equation of the tangent line to \(y=\sqrt{3x+1}\) at \(x=8\).
\(y(8)=\sqrt{25}=5\). \(y'=\dfrac{3}{2\sqrt{3x+1}}\), so \(y'(8)=\dfrac{3}{10}\). The line is \(y-5=\dfrac3{10}(x-8)\).
Common mistakes
Forgetting the inside derivative. The derivative of \(\sin(3x)\) is \(3\cos(3x)\), not \(\cos(3x)\).
Evaluating the outside derivative at \(x\) instead of at the inside. With \(f'(g(x))\), plug in \(g(x)\).
Changing the inside. The inside function is left exactly as it was when you differentiate the outside.
Identify the outside and inside; differentiate the outside leaving the inside alone; multiply by the inside's derivative.
Nested compositions need one factor per layer.
Use the product or quotient rule for the overall structure; use the chain rule for composite factors.
With tables: \(\big(f\circ g\big)'(a)=f'(g(a))\,g'(a)\).
Practice
Differentiate \(y=(2x+5)^4\).
Show answer
\(y'=4(2x+5)^3\cdot2=8(2x+5)^3\).
Differentiate \(y=\sin(3x^2)\).
Show answer
\(y'=\cos(3x^2)\cdot6x=6x\cos(3x^2)\).
Differentiate \(y=e^{-x^2}\).
Show answer
\(y'=e^{-x^2}\cdot(-2x)=-2xe^{-x^2}\).
Differentiate \(y=\ln(x^2+4)\).
Show answer
\(y'=\dfrac{2x}{x^2+4}\).
Let \(h(x)=f(g(x))\) with \(g(2)=1\), \(g'(2)=3\), and \(f'(1)=-2\). Find \(h'(2)\).
Show answer
\(h'(2)=f'(g(2))\,g'(2)=f'(1)\cdot3=-6\).
Differentiate \(y=x^2\sqrt{x+1}\) (product rule and chain rule) and give the derivative in the form \(\dfrac{5x^2+4x}{2\sqrt{x+1}}\) or equivalent.
Show answer
Not every curve is the graph of a function \(y=f(x)\). The circle \(x^2+y^2=25\) fails the vertical line test, and a curve like \(x^3+y^3=6xy\) cannot be solved for \(y\) at all by ordinary algebra. But these curves still have tangent lines, and we can find their slopes. The tool is implicit differentiation, which is the chain rule applied to equations.
Required content · FUN-3.D.1
Learning objective.
FUN-3.D. Calculate derivatives of implicitly defined functions.
Essential knowledge.
FUN-3.D.1. The chain rule is the basis for implicit differentiation.
The method
An equation relating \(x\) and \(y\) defines \(y\) implicitly as a function of \(x\) (at least locally, near a point). Treat \(y\) as an unknown function of \(x\), and differentiate both sides of the equation with respect to \(x\). The chain rule is the basis for implicit differentiation: whenever you differentiate an expression containing \(y\), you multiply by \(\dfrac{dy}{dx}\), because \(y\) is a function of \(x\).
For example, \(\dfrac{d}{dx}\big[y^3\big]=3y^2\cdot\dfrac{dy}{dx}\), exactly like \(\dfrac{d}{dx}[(\text{inside})^3]\). And \(\dfrac{d}{dx}[xy]=1\cdot y+x\cdot\dfrac{dy}{dx}\) by the product rule.
The steps:
Differentiate both sides with respect to \(x\), writing \(\dfrac{dy}{dx}\) (or \(y'\)) after every term that contains \(y\).
Collect all terms with \(\dfrac{dy}{dx}\) on one side and all other terms on the other.
Factor out \(\dfrac{dy}{dx}\) and solve.
If a point is given, substitute both coordinates \(x\) and \(y\); the answer usually depends on both.
Figure 3.1. The circle \(x^2+y^2=25\) is not the graph of a single function, but near \((3,4)\) it behaves like one, and implicit differentiation gives its slope there: \(\dfrac{dy}{dx}=-\dfrac{x}{y}=-\dfrac34\). The radius to \((3,4)\) has slope \(\frac43\), and the tangent is perpendicular to it.
Worked example · The circle
Find \(\dfrac{dy}{dx}\) for \(x^2+y^2=25\), then the slope at \((3,4)\) and an equation of the tangent line.
Differentiate both sides: \(2x+2y\dfrac{dy}{dx}=0\). So \(\dfrac{dy}{dx}=-\dfrac{x}{y}\). At \((3,4)\), \(\dfrac{dy}{dx}=-\dfrac34\). The tangent line is \(y-4=-\dfrac34(x-3)\).
Check against geometry: the radius from the origin to \((3,4)\) has slope \(\frac43\), and a tangent to a circle is perpendicular to the radius, with slope \(-\frac34\). ✓
Worked example · A product of \(x\) and \(y\)
Find \(\dfrac{dy}{dx}\) at \((1,2)\) for \(x^2y+y^3=10\).
First confirm the point is on the curve: \(1^2\cdot2+2^3=2+8=10\). ✓ Differentiate: \(2xy+x^2\dfrac{dy}{dx}+3y^2\dfrac{dy}{dx}=0\). Collect: \(\dfrac{dy}{dx}\left(x^2+3y^2\right)=-2xy\), so \(\dfrac{dy}{dx}=\dfrac{-2xy}{x^2+3y^2}\). At \((1,2)\): \(\dfrac{-4}{1+12}=-\dfrac{4}{13}\).
Worked example · A curve that cannot be solved for \(y\)
For \(x^3+y^3=6xy\), find \(\dfrac{dy}{dx}\) and its value at \((3,3)\).
\(3x^2+3y^2\dfrac{dy}{dx}=6y+6x\dfrac{dy}{dx}\). Move the \(\dfrac{dy}{dx}\) terms left: \(\dfrac{dy}{dx}\left(3y^2-6x\right)=6y-3x^2\). So \[\frac{dy}{dx}=\frac{6y-3x^2}{3y^2-6x}=\frac{2y-x^2}{y^2-2x}.\] At \((3,3)\) (which is on the curve, since \(27+27=54=6\cdot9\)): \(\dfrac{6-9}{9-6}=-1\).
Horizontal and vertical tangents
For the circle, \(\dfrac{dy}{dx}=-\dfrac xy\) is zero when \(x=0\): horizontal tangents at \((0,5)\) and \((0,-5)\). The slope is undefined when \(y=0\): vertical tangents at \((5,0)\) and \((-5,0)\). In general, a horizontal tangent occurs where the numerator of \(\dfrac{dy}{dx}\) is zero (and the denominator is not), and a vertical tangent where the denominator is zero (and the numerator is not). Remember to check that any point you find actually lies on the original curve.
Exam focus
Differentiate both sides with respect to \(x\); every \(y\)-term gets a factor of \(\dfrac{dy}{dx}\) (chain rule).
Use the product rule for terms like \(xy\) and \(x^2y\).
Solve for \(\dfrac{dy}{dx}\): collect, factor, divide. The answer may contain both \(x\) and \(y\).
Evaluate at a point with both coordinates; confirm the point is on the curve.
Find \(\dfrac{dy}{dx}\) at \((1,3)\) for \(x^2-xy+y^2=7\).
Show answer
\(2x-y-x\dfrac{dy}{dx}+2y\dfrac{dy}{dx}=0\), so \(\dfrac{dy}{dx}(2y-x)=y-2x\) and \(\dfrac{dy}{dx}=\dfrac{y-2x}{2y-x}\). At \((1,3)\): \(\dfrac{3-2}{6-1}=\dfrac15\).
Find \(\dfrac{dy}{dx}\) for \(e^y=x+y\).
Show answer
\(e^y\dfrac{dy}{dx}=1+\dfrac{dy}{dx}\), so \(\dfrac{dy}{dx}\left(e^y-1\right)=1\) and \(\dfrac{dy}{dx}=\dfrac{1}{e^y-1}\).
Find the slope of the curve \(xy=6\) at \((2,3)\).
Show answer
\(y+x\dfrac{dy}{dx}=0\), so \(\dfrac{dy}{dx}=-\dfrac yx=-\dfrac32\) at \((2,3)\).
Write an equation of the tangent line to \(y^2=x^3\) at \((4,8)\).
Show answer
\(2y\dfrac{dy}{dx}=3x^2\), so \(\dfrac{dy}{dx}=\dfrac{3x^2}{2y}=\dfrac{48}{16}=3\). The line is \(y-8=3(x-4)\).
Why does \(\dfrac{d}{dx}\big[y^3\big]\) equal \(3y^2\dfrac{dy}{dx}\) and not just \(3y^2\)?
Show answer
\(y\) is a function of \(x\), so \(y^3\) is a composite: outside "cube," inside \(y(x)\). The chain rule multiplies by the derivative of the inside, \(\dfrac{dy}{dx}\).
For \(x^2+y^2=25\), identify the points where the tangent line is horizontal and where it is vertical.
Show answer
\(\dfrac{dy}{dx}=-\dfrac xy\). Horizontal when \(x=0\): \((0,\pm5)\). Vertical when \(y=0\): \((\pm5,0)\).
Many important functions come in inverse pairs: squaring and square root, \(e^x\) and \(\ln x\), sine and arcsine. If you know the derivative of one, can you get the derivative of the other without starting over? Yes. The graph of an inverse function is a mirror image of the original, and mirror images have reciprocal slopes. That observation, plus the chain rule, gives a formula.
Required content · FUN-3.E.1
Learning objective.
FUN-3.E. Calculate derivatives of inverse and inverse trigonometric functions.
Essential knowledge.
FUN-3.E.1. The chain rule and definition of an inverse function can be used to find the derivative of an inverse function, provided the derivative exists.
The slope of a mirror image
Two functions \(f\) and \(g=f^{-1}\) are inverses when \(g(f(x))=x\). The graph of \(f^{-1}\) is the reflection of the graph of \(f\) across the line \(y=x\): a point \((a,b)\) on \(f\) corresponds to the point \((b,a)\) on \(f^{-1}\). A reflection swaps horizontal and vertical, so it swaps rise and run. Tangent slopes become reciprocals.
Figure 3.2. The graph of \(f^{-1}\) is the reflection of the graph of \(f\) across the line \(y=x\). A tangent line to \(f\) at \((1,1.5)\) reflects to a tangent line to \(f^{-1}\) at \((1.5,1)\). Reflection swaps rise and run, so the slopes are reciprocals: \(f'(1)=2.5\) and \((f^{-1})'(1.5)=\frac1{2.5}=0.4\).
The formula
If \(f\) is differentiable and one-to-one, and \(f'(f^{-1}(b))\ne0\), then \(f^{-1}\) is differentiable at \(b\) and \[\big(f^{-1}\big)'(b)=\frac{1}{f'\big(f^{-1}(b)\big)}.\]
Where it comes from. Differentiate both sides of \(f\big(f^{-1}(x)\big)=x\) with the chain rule: \(f'\big(f^{-1}(x)\big)\cdot\big(f^{-1}\big)'(x)=1\). Solve for \(\big(f^{-1}\big)'(x)\). The chain rule and the definition of an inverse function can be used to find the derivative of an inverse function, provided the derivative exists.
The formula is usually applied at a specific point. To use it:
You are given \(b\) and need \(\big(f^{-1}\big)'(b)\). Find the input \(a\) with \(f(a)=b\), so that \(f^{-1}(b)=a\). (Often this is guessed or found from a table.)
Compute \(f'(a)\).
Take the reciprocal: \(\big(f^{-1}\big)'(b)=\dfrac{1}{f'(a)}\).
Worked example · Inverse at a point
Let \(f(x)=x^3+2x+1\). Find \(\big(f^{-1}\big)'(4)\).
We need \(a\) with \(f(a)=4\). Try \(a=1\): \(f(1)=1+2+1=4\). ✓ So \(f^{-1}(4)=1\). \(f'(x)=3x^2+2\), so \(f'(1)=5\). Therefore \(\big(f^{-1}\big)'(4)=\dfrac15\).
(We never needed a formula for \(f^{-1}\); finding one would be very hard. The reciprocal formula is the right tool.)
Worked example · From a table
A one-to-one function \(f\) has \(f(1)=3\), \(f(2)=7\), \(f'(1)=4\), \(f'(2)=5\). Find \(\big(f^{-1}\big)'(7)\) and \(\big(f^{-1}\big)'(3)\).
\(f(2)=7\), so \(f^{-1}(7)=2\), and \(\big(f^{-1}\big)'(7)=\dfrac1{f'(2)}=\dfrac15\). Similarly \(f(1)=3\), so \(f^{-1}(3)=1\) and \(\big(f^{-1}\big)'(3)=\dfrac{1}{f'(1)}=\dfrac14\).
Worked example · Rediscovering \(\dfrac{d}{dx}\ln x\)
The functions \(f(x)=e^x\) and \(g(x)=\ln x\) are inverses. Use the formula to find \(g'(x)\).
\(g'(x)=\dfrac{1}{f'(g(x))}=\dfrac{1}{e^{\ln x}}=\dfrac1x\), since \(e^{\ln x}=x\). This is the derivative of \(\ln x\) that you learned in Lesson 2.7, now explained.
When does an inverse exist?
A function has an inverse (on an interval) if it is one-to-one there, for example if it is strictly increasing or strictly decreasing, which is the case when \(f'\) has one sign on the interval. The condition \(f'\ne0\) in the formula is necessary: where \(f\) has a horizontal tangent, the inverse has a vertical tangent, and its derivative does not exist.
Exam focus
\(\big(f^{-1}\big)'(b)=\dfrac{1}{f'\big(f^{-1}(b)\big)}\): reciprocal of the slope at the matching point.
Find \(a\) with \(f(a)=b\) first; then use \(f'(a)\).
The slopes of \(f\) and \(f^{-1}\) at corresponding points \((a,b)\) and \((b,a)\) are reciprocals.
If \(f'(a)=0\), the inverse has a vertical tangent at \((b,a)\).
The formula comes from the chain rule applied to \(f(f^{-1}(x))=x\).
Practice
Let \(f(x)=x^3+x\). Find \(\big(f^{-1}\big)'(10)\).
Show answer
\(f(2)=8+2=10\), so \(f^{-1}(10)=2\). \(f'(x)=3x^2+1\), \(f'(2)=13\). So \(\big(f^{-1}\big)'(10)=\dfrac1{13}\).
A one-to-one function has \(f(1)=3\) and \(f'(1)=4\). Find \(\big(f^{-1}\big)'(3)\).
Show answer
\(\dfrac{1}{f'(1)}=\dfrac14\).
Let \(f(x)=2x+\sin x\). Find \(\big(f^{-1}\big)'(0)\).
Show answer
\(f(0)=0\), so \(f^{-1}(0)=0\). \(f'(x)=2+\cos x\), \(f'(0)=3\). So \(\big(f^{-1}\big)'(0)=\dfrac13\).
The graph of \(f\) has a tangent line with slope \(\dfrac25\) at the point \((3,7)\). What is the slope of the tangent line to \(f^{-1}\) at \((7,3)\)?
Show answer
The slopes are reciprocals: \(\dfrac{5}{2}\).
Use \(f(x)=e^{2x}\) and its inverse \(g(x)=\tfrac12\ln x\) to confirm \(g'(x)\) with the formula.
Show answer
\(f'(x)=2e^{2x}\). \(g'(x)=\dfrac{1}{f'(g(x))}=\dfrac{1}{2e^{2g(x)}}=\dfrac{1}{2e^{\ln x}}=\dfrac1{2x}\), which matches \(\dfrac{d}{dx}\left[\tfrac12\ln x\right]=\dfrac1{2x}\).
Why does the formula fail if \(f'(a)=0\)? What does the graph of \(f^{-1}\) look like there?
Show answer
The reciprocal \(\dfrac10\) is undefined. A horizontal tangent to \(f\) reflects to a vertical tangent on \(f^{-1}\), so \(f^{-1}\) has no (finite) derivative at that point.
The inverse trigonometric functions, \(\arcsin x\), \(\arccos x\), \(\arctan x\), and their relatives, turn a ratio back into an angle. They appear whenever you need an angle from lengths, as in a ladder, a ramp, or a viewing angle. Their derivatives look surprising at first: they are algebraic, with no trigonometry left in them. They come straight from the inverse-function rule and the chain rule.
Required content · FUN-3.E.2
Learning objective.
FUN-3.E. Calculate derivatives of inverse and inverse trigonometric functions.
Essential knowledge.
FUN-3.E.2. The chain rule applied with the definition of an inverse function, or the formula for the derivative of an inverse function, can be used to find the derivatives of inverse trigonometric functions.
For completeness, \(\dfrac{d}{dx}[\operatorname{arccot}x]=\dfrac{-1}{1+x^2}\), \(\dfrac{d}{dx}[\operatorname{arcsec}x]=\dfrac{1}{|x|\sqrt{x^2-1}}\), and \(\dfrac{d}{dx}[\operatorname{arccsc}x]=\dfrac{-1}{|x|\sqrt{x^2-1}}\). As with the other trigonometric functions, the "co-" versions are the negatives of their partners. The first three are the ones you will use most. Note that \(\sin^{-1}x\) is another way to write \(\arcsin x\) (it does not mean \(\frac{1}{\sin x}\)).
Where they come from. Let \(y=\arcsin x\), so \(\sin y=x\) with \(-\tfrac\pi2\le y\le\tfrac\pi2\). Differentiate implicitly: \(\cos y\cdot\dfrac{dy}{dx}=1\), so \(\dfrac{dy}{dx}=\dfrac1{\cos y}\). Since \(\cos y\ge0\) on that interval, \(\cos y=\sqrt{1-\sin^2y}=\sqrt{1-x^2}\). That gives the formula. The chain rule applied with the definition of an inverse function, or the formula for the derivative of an inverse function, can be used to find the derivatives of inverse trigonometric functions. For \(\arctan\): \(\tan y=x\) gives \(\sec^2y\,\dfrac{dy}{dx}=1\), and \(\sec^2y=1+\tan^2y=1+x^2\).
The domain matters. \(\arcsin x\) and \(\arccos x\) are defined for \(-1\le x\le1\) and their derivatives only exist for \(-1\lt x\lt1\) (the graphs have vertical tangents at \(x=\pm1\)). \(\arctan x\) is defined for all real \(x\), with derivative \(\dfrac1{1+x^2}\gt0\) everywhere, so it is always increasing, and flattens out toward horizontal asymptotes \(y=\pm\tfrac\pi2\).
With the chain rule
When the argument is not just \(x\), use the chain rule: replace \(x\) by the inside \(u\) and multiply by \(u'\). \[\frac{d}{dx}\big[\arctan u\big]=\frac{u'}{1+u^2}\qquad \frac{d}{dx}\big[\arcsin u\big]=\frac{u'}{\sqrt{1-u^2}}.\]
\(u=x^2\), \(u'=2x\): \(y'=\dfrac{2x}{\sqrt{1-(x^2)^2}}=\dfrac{2x}{\sqrt{1-x^4}}\), for \(-1\lt x\lt1\).
Worked example · A product and a tangent line
Find the slope of the tangent line to \(y=x\arctan x\) at \(x=1\).
\(y'=1\cdot\arctan x+x\cdot\dfrac{1}{1+x^2}\). At \(x=1\): \(\arctan1=\dfrac\pi4\) and \(\dfrac{1}{1+1}=\dfrac12\), so \(y'(1)=\dfrac\pi4+\dfrac12\approx1.285\).
This is \(f'\left(\tfrac12\right)\) for \(f(x)=\arcsin x\), since \(\arcsin\tfrac12=\tfrac\pi6\). \(f'\left(\tfrac12\right)=\dfrac{1}{\sqrt{1-\frac14}}=\dfrac{1}{\sqrt3/2}=\dfrac{2}{\sqrt3}=\dfrac{2\sqrt3}{3}\).
You now know the full set of differentiation rules for Units 2 and 3: power, constant multiple, sum, product, quotient, chain, implicit differentiation, inverse functions, and the derivatives of the basic and inverse trigonometric functions. Topic 3.5 in the course framework is a skills topic: its purpose is to practice deciding which rule to use, in what order, on a mixed set of problems. That decision is the real difficulty of most derivative questions on the exam.
Read the outermost structure first
The key habit is to identify the last operation performed when you evaluate the expression, because that is the first rule you use.
The last operation is…
First rule
Example
a sum or difference
sum/difference, term by term
\(x^5-4x^{-2}+\sqrt x\)
a product of two functions of \(x\)
product rule
\(x^3\sin x\)
a quotient of two functions of \(x\)
quotient rule (or rewrite)
\(\dfrac{x^2+1}{x-2}\)
applying a function to an inside expression
chain rule (outside, inside)
\(\ln(\sin x)\), \((3x-1)^5\)
an equation in \(x\) and \(y\) that is not solved for \(y\)
implicit differentiation
\(xy+\sin y=2\)
an inverse function whose derivative is not known
inverse-function formula
\(f^{-1}\) at a point
After the first rule, treat each piece as a smaller problem. Often several rules nest: a product whose factors need the chain rule, or a quotient with a composite numerator.
Worked example · Choosing the rule
For each, choose the first rule and find the derivative.
1. \(y=x^5-\dfrac{4}{x^2}+\sqrt x=x^5-4x^{-2}+x^{1/2}\): sum, power rule. \(y'=5x^4+8x^{-3}+\dfrac12x^{-1/2}\).
Algebra done first can save a great deal of work and eliminate mistakes.
Worked example · Rewrite first
Differentiate \(y=\dfrac{x^3-4x}{x}\) and \(y=\dfrac{(x-1)(x+1)}{x-1}\).
Rather than the quotient rule, simplify: \(\dfrac{x^3-4x}{x}=x^2-4\) (for \(x\ne0\)), so \(y'=2x\). Similarly \(\dfrac{(x-1)(x+1)}{x-1}=x+1\) (for \(x\ne1\)), so \(y'=1\). Remember to state the \(x\)-values excluded from the domain.
Worked example · Using values and structure
Let \(k(x)=x\cdot f\big(g(x)\big)\). Suppose \(g(1)=2\), \(g'(1)=3\), \(f(2)=4\), and \(f'(2)=5\). Find \(k'(1)\).
The last operation is a product of \(x\) and the composite \(f(g(x))\). Product rule: \(k'(x)=1\cdot f\big(g(x)\big)+x\cdot f'\big(g(x)\big)\,g'(x)\). At \(x=1\): \(k'(1)=f(g(1))+1\cdot f'(g(1))\,g'(1)=f(2)+f'(2)\cdot3=4+15=19\).
Exam focus
Identify the last operation of the expression; its rule goes first. Then work inward.
Rewrite roots and reciprocals as powers; simplify quotients that cancel, and expand when it is easy, before differentiating.
Combine rules carefully: product or quotient on the outside, chain rule inside any composite factor.
Equations in \(x\) and \(y\) call for implicit differentiation; remember \(\dfrac{dy}{dx}\) after each \(y\)-term.
Check units, domain, and (if a value is requested) substitute at the end.
Practice
Name the first rule and differentiate: \(y=e^{x}\cos x\).
Show answer
Name the first rule and differentiate: \(y=\sqrt{\dfrac{x}{x+1}}\) (write it as \(\left(\dfrac{x}{x+1}\right)^{1/2}\)).
Show answer
Chain rule first (power \(\tfrac12\) on the inside quotient), then the quotient rule inside. \(y'=\tfrac12\left(\dfrac{x}{x+1}\right)^{-1/2}\cdot\dfrac{1}{(x+1)^2}\), since \(\dfrac{d}{dx}\dfrac{x}{x+1}=\dfrac{(x+1)-x}{(x+1)^2}=\dfrac1{(x+1)^2}\).
Differentiate \(y=\ln\left(\dfrac{x^2}{x+1}\right)\) using properties of logarithms first.
Show answer
\(y=2\ln x-\ln(x+1)\), so \(y'=\dfrac2x-\dfrac1{x+1}\).
Find \(\dfrac{dy}{dx}\) if \(x^2+3xy=y^2+4\).
Show answer
\(2x+3y+3x\dfrac{dy}{dx}=2y\dfrac{dy}{dx}\). Collect: \(\dfrac{dy}{dx}(3x-2y)=-2x-3y\), so \(\dfrac{dy}{dx}=\dfrac{-2x-3y}{3x-2y}=\dfrac{2x+3y}{2y-3x}\).
Differentiate \(y=\arctan(e^x)\).
Show answer
Chain rule: \(y'=\dfrac{e^x}{1+e^{2x}}\).
Explain why differentiating \(\dfrac{x^2-9}{x-3}\) with the quotient rule is unnecessary, and give the derivative.
Show answer
For \(x\ne3\), \(\dfrac{x^2-9}{x-3}=x+3\), so the derivative is 1 (for \(x\ne3\)). Simplifying first avoids the quotient rule.
The derivative of a function is itself a function, so we can differentiate it again. The result, the second derivative, measures how the rate of change is changing. It is acceleration when the function is position, and, as Unit 5 will show, it tells us whether a graph is curving up or down. Higher derivatives follow by repeating the process.
Required content · FUN-3.F.1, FUN-3.F.2
Learning objective.
FUN-3.F. Determine higher order derivatives of a function.
Essential knowledge.
FUN-3.F.1. Differentiating \(f'\) produces the second derivative \(f''\), provided the derivative of \(f'\) exists; repeating this process produces higher-order derivatives of \(f\).
FUN-3.F.2. Higher-order derivatives are represented with a variety of notations. For \(y=f(x)\), notations for the second derivative include \(\dfrac{d^2y}{dx^2}\), \(f''(x)\), and \(y''\). Higher-order derivatives can be denoted \(\dfrac{d^ny}{dx^n}\) or \(f^{(n)}(x)\).
Definition and notation
Differentiating \(f'\) produces the second derivative \(f''\), provided the derivative of \(f'\) exists; repeating this process produces higher-order derivatives of \(f\).
Higher-order derivatives are represented with a variety of notations. For \(y=f(x)\), notations for the second derivative include \(f''(x)\), \(y''\), and \(\dfrac{d^2y}{dx^2}\). Higher-order derivatives can be denoted \(f'''(x)\), \(f^{(4)}(x)\), \(\dots\), \(f^{(n)}(x)\), or \(\dfrac{d^ny}{dx^n}\).
Order
Prime notation
Leibniz notation
first
\(f'(x)\)
\(\dfrac{dy}{dx}\)
second
\(f''(x)\)
\(\dfrac{d^2y}{dx^2}\)
third
\(f'''(x)\)
\(\dfrac{d^3y}{dx^3}\)
\(n\)th
\(f^{(n)}(x)\)
\(\dfrac{d^ny}{dx^n}\)
The notation \(\dfrac{d^2y}{dx^2}\) means \(\dfrac{d}{dx}\!\left(\dfrac{dy}{dx}\right)\). It is not the square of \(\dfrac{dy}{dx}\), and \(f^{(n)}\) with parentheses is the \(n\)th derivative, not the \(n\)th power.
Worked example · Repeated differentiation
Find the first four derivatives of \(f(x)=x^4-3x^2+2\).
\(f'(x)=4x^3-6x\). \(f''(x)=12x^2-6\). \(f'''(x)=24x\). \(f^{(4)}(x)=24\). Every derivative beyond the fourth is \(0\): a degree-4 polynomial is eliminated after 5 derivatives.
Worked example · Patterns
Find \(f^{(n)}(x)\) for \(f(x)=e^{2x}\), and the first four derivatives of \(\sin x\).
By the chain rule each derivative of \(e^{2x}\) multiplies by \(2\): \(f'=2e^{2x}\), \(f''=4e^{2x}\), so \(f^{(n)}(x)=2^ne^{2x}\).
For \(\sin x\): \(\cos x\), \(-\sin x\), \(-\cos x\), \(\sin x\), and then the cycle repeats. The fourth derivative of \(\sin x\) is \(\sin x\) again.
Worked example · With the product rule
Find \(f''(x)\) for \(f(x)=xe^x\).
\(f'(x)=1\cdot e^x+x\cdot e^x=e^x(1+x)\). Differentiate again with the product rule: \(f''(x)=e^x(1+x)+e^x\cdot1=e^x(2+x)\).
What the second derivative means
If \(s(t)\) is the position of a particle on a line, then \(s'(t)=v(t)\) is its velocity and \(s''(t)=v'(t)=a(t)\) is its acceleration. The units of the second derivative are the units of \(f\) per unit of \(x\), per unit of \(x\) again: for position in meters and time in seconds, \(\dfrac{\text{m/s}}{\text{s}}=\text{m/s}^2\).
Worked example · Position, velocity, acceleration
A particle moves along a line with position \(s(t)=t^3-6t^2+9t\) feet at time \(t\) seconds. Find its velocity and acceleration, and the acceleration at \(t=3\).
\(v(t)=s'(t)=3t^2-12t+9\). \(a(t)=s''(t)=6t-12\). At \(t=3\): \(a(3)=6\) ft/s\(^2\).
Graphically, the second derivative describes how the slope changes: if \(f''\gt0\) the slopes are increasing and the graph curves upward, and if \(f''\lt0\) the graph curves downward. That is the subject of Lesson 5.6.
Exam focus
\(f''=\dfrac{d}{dx}(f')\); notations \(f''\), \(y''\), \(\dfrac{d^2y}{dx^2}\), and \(f^{(n)}\), \(\dfrac{d^ny}{dx^n}\).
\(\dfrac{d^2y}{dx^2}\) is not \(\left(\dfrac{dy}{dx}\right)^2\).
Differentiate in stages and simplify each stage before the next.
For position \(s(t)\): velocity \(=s'\), acceleration \(=s''\). Units: unit of \(s\) per time per time.
The second derivative is the rate of change of the slope; its sign tells the concavity (Lesson 5.6).
Practice
Find \(f''(x)\) for \(f(x)=x^5-2x^3+x\).
Show answer
\(f'(x)=5x^4-6x^2+1\), so \(f''(x)=20x^3-12x\).
Find \(\dfrac{d^2y}{dx^2}\) for \(y=\ln x\).
Show answer
\(\dfrac{dy}{dx}=\dfrac1x=x^{-1}\), so \(\dfrac{d^2y}{dx^2}=-x^{-2}=-\dfrac1{x^2}\).
Interpreting the meaning of the derivative in context
Up to now a derivative has been a slope or a formula. In applications, a derivative is a rate, and the most valuable skill is saying in plain words what that rate means, with its units. The College Board puts this at the heart of the free-response section: at least two of the six questions are set in a real-world context, and one of the most common tasks is "interpret the meaning of \(f'(a)\) in the context of the problem." This lesson gives you a method for those interpretations.
CHA-3.A. Interpret the meaning of a derivative in context.
Essential knowledge.
CHA-3.A.1. The derivative of a function can be interpreted as the instantaneous rate of change with respect to its independent variable.
CHA-3.A.2. The derivative can be used to express information about rates of change in applied contexts.
CHA-3.A.3. The unit for \(f'(x)\) is the unit for \(f\) divided by the unit for \(x\).
The derivative as an instantaneous rate
The derivative of a function can be interpreted as the instantaneous rate of change of the function with respect to its independent variable. If \(y=f(x)\), then \(f'(x)\) tells how fast \(y\) is changing as \(x\) changes, at the moment (or place) \(x\). The derivative can be used to express information about rates of change in applied contexts.
The unit of \(f'(x)\) is the unit of \(f\) divided by the unit of \(x\):
\[\text{units of }f'(x)=\frac{\text{units of }f(x)}{\text{units of }x}.\]
If \(V(t)\) is volume in gallons and \(t\) is time in minutes, \(V'(t)\) is in gallons per minute. If \(C(q)\) is cost in dollars and \(q\) is the number of items, \(C'(q)\) is in dollars per item. Getting the units right is how you know you have the interpretation right.
A template for interpretation
A complete interpretation of \(f'(a)\) has four parts:
When or where: "At \(x=a\) (give the value with units)."
What quantity: name the function in words, not as "\(f\)."
Direction: "is increasing" if \(f'(a)\gt0\), "is decreasing" if \(f'(a)\lt0\).
How fast: "at a rate of \(|f'(a)|\)" with the units.
So if \(f'(a)=-3\), you do not say "decreasing at a rate of negative three"; you say "decreasing at a rate of 3 units per unit."
Worked example · A tank
The volume of water in a tank at time \(t\) minutes is \(V(t)\) gallons. It is known that \(V'(5)=-3\). Interpret this.
At time \(t=5\) minutes, the volume of water in the tank is decreasing at a rate of 3 gallons per minute. The negative sign means the tank is draining at that moment. (It does not say how much water is in the tank.)
Worked example · From a table
The population of a town, in thousands, is \(P(t)\), where \(t\) is the year. The table gives \(P(2000)=150\), \(P(2010)=165\), \(P(2020)=190\). Estimate \(P'(2010)\) and interpret it.
Use the symmetric estimate: \(P'(2010)\approx\dfrac{P(2020)-P(2000)}{2020-2000}=\dfrac{190-150}{20}=2\). In 2010 the population of the town was increasing at a rate of approximately 2 thousand people per year.
Worked example · A cost function
The cost, in dollars, of producing \(x\) widgets is \(C(x)\), and \(C'(200)=8.50\). Interpret.
When 200 widgets are being produced, the cost is increasing at a rate of \$8.50 per widget. Economists call \(C'\) the marginal cost: it approximates the cost of producing one more widget (the 201st), about \$8.50.
When the variable is not time
The independent variable need not be time. If \(h(x)\) is the elevation, in feet, of a trail at \(x\) miles from the trailhead, then \(h'(x)\) is in feet per mile, the steepness of the trail. If \(T(d)\) is the temperature of the ocean, in degrees Celsius, at depth \(d\) meters, then \(T'(d)\) is in degrees Celsius per meter, the rate at which the water gets colder as you dive deeper. In every case: how fast does the output change per unit of input, right now?
Average rate versus instantaneous rate
An average rate of change over an interval \([a,b]\) is \(\dfrac{f(b)-f(a)}{b-a}\); an instantaneous rate at a point is \(f'(a)\). Exam questions often ask for both and then ask you to compare. For a function that is increasing at a faster and faster rate, for example, the instantaneous rate at the right endpoint is larger than the average over the interval. Always say which one you are describing.
Exam focus
\(f'(a)\) is the instantaneous rate of change of \(f\) with respect to its independent variable at \(x=a\).
Units of \(f'\): units of \(f\) per unit of \(x\).
Interpretation: when (with units), what quantity, increasing or decreasing, at what rate (with units).
A negative derivative means decreasing; give the magnitude as the rate.
Do not confuse \(f'(a)\) with \(f(a)\): the derivative says nothing about the amount, only the rate at which it changes.
Practice
\(W(t)\) is the weight, in pounds, of a puppy at age \(t\) weeks, and \(W'(8)=1.4\). Interpret \(W'(8)\).
Show answer
At age 8 weeks, the puppy's weight is increasing at a rate of 1.4 pounds per week.
\(T(t)\) is the temperature of a cup of coffee, in degrees Fahrenheit, \(t\) minutes after it was poured, and \(T'(10)=-2.5\). Interpret.
Show answer
Ten minutes after it was poured, the temperature of the coffee is decreasing at a rate of 2.5 degrees Fahrenheit per minute.
What are the units of \(R'(x)\) if \(R(x)\) is revenue in thousands of dollars from selling \(x\) hundred tickets?
Show answer
Thousands of dollars per hundred tickets.
A hiker's elevation \(h(x)\) in feet is recorded at \(x\) miles along a trail, with \(h(3)=2400\) and \(h(3.5)=2490\). Estimate \(h'(3)\) and interpret it.
Show answer
\(\dfrac{2490-2400}{0.5}=180\). About 3 miles along the trail the elevation is increasing at approximately 180 feet per mile.
A student says "\(V'(5)=-3\) means there are \(-3\) gallons in the tank at 5 minutes." What is wrong?
Show answer
\(V'(5)\) is a rate, not an amount. It means the volume is decreasing by 3 gallons per minute at \(t=5\); the amount in the tank is \(V(5)\).
The population of bacteria is \(B(t)\) thousand at \(t\) hours. Explain the difference in meaning between \(B(4)=12\), \(B'(4)=3\), and \(\dfrac{B(6)-B(4)}{2}=2.5\).
Show answer
\(B(4)=12\): at 4 hours there are 12 thousand bacteria. \(B'(4)=3\): at 4 hours the population is increasing at 3 thousand per hour (instantaneous). \(\dfrac{B(6)-B(4)}{2}=2.5\): from 4 to 6 hours the population increased by an average of 2.5 thousand per hour.
Straight-line motion: position, velocity, and acceleration
The original application of the derivative, and still the most common one on the exam, is motion along a line: a car on a straight highway, a ball thrown straight up, a particle on a track. Calculus gives exact vocabulary for what the object is doing at each instant: where it is, how fast and in which direction it moves, and whether it is speeding up or slowing down.
Required content · CHA-3.B.1
Learning objective.
CHA-3.B. Calculate rates of change in applied contexts.
Essential knowledge.
CHA-3.B.1. The derivative can be used to solve rectilinear motion problems involving position, speed, velocity, and acceleration.
The vocabulary
For a particle moving on a line with position \(s(t)\) at time \(t\):
Quantity
Definition
Meaning
Velocity
\(v(t)=s'(t)\)
rate of change of position; the sign gives direction
Speed
\(\lvert v(t)\rvert\)
how fast, ignoring direction
Acceleration
\(a(t)=v'(t)=s''(t)\)
rate of change of velocity
The derivative can be used to solve rectilinear motion problems involving position, speed, velocity, and acceleration. By convention, positive velocity means motion in the positive direction (right, or up) and negative velocity means motion in the negative direction (left, or down). The particle is at rest when \(v(t)=0\), and it can change direction only at times when \(v\) changes sign.
Speeding up and slowing down
A particle's speed is increasing when velocity and acceleration have the same sign and decreasing when they have opposite signs. The reason: if \(v\gt0\) and \(a\gt0\), the particle moves right and its velocity is getting larger, so it speeds up. If \(v\lt0\) and \(a\lt0\), the velocity becomes more negative: speed is growing even though the object is moving left. If the signs differ, acceleration pulls against the motion, and the speed shrinks.
Figure 4.1. The graph of \(v(t)=t^2-6t+8=(t-2)(t-4)\). The particle moves right when \(v\gt0\) and left when \(v\lt0\), and its speed increases when \(v\) and \(a=v'\) have the same sign and decreases when they have opposite signs. Here \(a=2t-6\) is zero at \(t=3\), the minimum of \(v\).
A sign chart is the best way to organize this. List the critical times (where \(v=0\) or \(a=0\)), then test the signs of \(v\) and \(a\) on each interval.
Worked example · Reading a velocity graph
The figure shows \(v(t)=t^2-6t+8=(t-2)(t-4)\) for \(0\le t\le5\) (feet per second). When is the particle moving left? When is it speeding up?
Direction. \(v\lt0\) for \(2\lt t\lt4\): the particle moves left. It moves right for \(0\le t\lt2\) and \(4\lt t\le5\). It changes direction at \(t=2\) and \(t=4\), where \(v\) changes sign.
Acceleration. \(a(t)=v'(t)=2t-6\), which is negative for \(t\lt3\) and positive for \(t\gt3\).
Interval
sign of \(v\)
sign of \(a\)
speed
\(0\lt t\lt2\)
\(+\)
\(-\)
slowing down
\(2\lt t\lt3\)
\(-\)
\(-\)
speeding up
\(3\lt t\lt4\)
\(-\)
\(+\)
slowing down
\(4\lt t\lt5\)
\(+\)
\(+\)
speeding up
The speed is increasing on \((2,3)\) and \((4,5)\). Notice that the particle is momentarily at rest at \(t=2\) and \(t=4\), and that at \(t=3\) the acceleration is zero while the velocity is \(-1\) (not at rest).
Worked example · From the position function
A particle moves on a line with \(s(t)=t^3-6t^2+9t\) meters, \(t\ge0\) seconds. Find when it is at rest and describe its motion.
\(v(t)=3t^2-12t+9=3(t-1)(t-3)\) is zero at \(t=1\) and \(t=3\): the particle is at rest at those times. Signs of \(v\): positive on \((0,1)\), negative on \((1,3)\), positive on \((3,\infty)\). So it moves right, then left, then right.
\(a(t)=6t-12\), zero at \(t=2\): negative on \((0,2)\), positive after. Comparing: slowing down on \((0,1)\) (\(v\gt0,a\lt0\)); speeding up on \((1,2)\) (\(v\lt0,a\lt0\)); slowing down on \((2,3)\) (\(v\lt0,a\gt0\)); speeding up on \((3,\infty)\).
Positions at the turning points: \(s(1)=1-6+9=4\) m and \(s(3)=27-54+27=0\) m. The particle goes from \(0\) out to \(4\), back to \(0\), and then out again.
Things to say on the exam
When a question asks "Is the speed increasing or decreasing at \(t=a\)?", compute \(v(a)\) and \(a(a)\), state their signs, and conclude: "\(v(2.5)\lt0\) and \(a(2.5)\lt0\), so the speed is increasing." A single sign statement without the reason does not earn the point. And note that "velocity is increasing" (\(a\gt0\)) is not the same as "speed is increasing."
Exam focus
\(v=s'\), speed \(=|v|\), \(a=v'=s''\).
At rest: \(v=0\). Moving right/up: \(v\gt0\). Moving left/down: \(v\lt0\). Direction changes where \(v\) changes sign.
Speed increases when \(v\) and \(a\) have the same sign; decreases when they differ.
Organize with a sign chart of \(v\) and \(a\); give both signs as the reason.
Units: \(v\) is length per time; \(a\) is length per time per time.
Practice
A particle has \(s(t)=t^2-4t\). Find its velocity and acceleration, and when it is at rest.
Show answer
\(v(t)=2t-4\), \(a(t)=2\). At rest when \(2t-4=0\), at \(t=2\).
For \(s(t)=t^3-6t^2+9t\), is the speed increasing or decreasing at \(t=2.5\)?
Show answer
\(v(2.5)=3(1.5)(-0.5)=-2.25\lt0\) and \(a(2.5)=6(2.5)-12=3\gt0\). The signs differ, so the speed is decreasing.
A particle's velocity is \(v(t)=3-t\) for \(t\ge0\). Describe its direction and when its speed is increasing.
Show answer
\(v\gt0\) for \(t\lt3\) (right) and \(v\lt0\) for \(t\gt3\) (left); \(a=-1\) always. Speed decreases for \(0\le t\lt3\) (\(v\gt0\), \(a\lt0\)) and increases for \(t\gt3\) (\(v\lt0\), \(a\lt0\)).
For \(v(t)=\sin t\) on \([0,2\pi]\), on which intervals is the particle moving left?
Show answer
\(\sin t\lt0\) for \(\pi\lt t\lt2\pi\): moving left on \((\pi,2\pi)\).
A ball thrown upward has height \(h(t)=-16t^2+48t+5\) feet. Find the velocity when \(t=1\) and the time when the ball is at its highest point.
Show answer
\(h'(t)=-32t+48\), so \(h'(1)=16\) ft/s (rising). The ball is momentarily at rest at the top: \(-32t+48=0\), so \(t=1.5\) s.
A student says "\(a(t)\gt0\), so the particle must be speeding up." Is the student right? Explain.
Show answer
Not necessarily. Speed increases only when \(v\) and \(a\) have the same sign. If \(a\gt0\) while \(v\lt0\), the particle is slowing down.
Rates of change in applied contexts other than motion
Motion is only one of the places rates of change appear. The same derivative tells an economist how fast cost rises with output, a biologist how fast a population grows, an engineer how fast a volume changes as a radius changes. This lesson collects the most common applied settings and shows how to calculate and interpret each rate.
Required content · CHA-3.C.1
Learning objective.
CHA-3.C. Interpret rates of change in applied contexts.
Essential knowledge.
CHA-3.C.1. The derivative can be used to solve problems involving rates of change in applied contexts.
Rates with respect to a variable that is not time
The derivative can be used to solve problems involving rates of change in applied contexts. The independent variable may be a length, a quantity produced, a depth, or anything else, and then \(f'\) is the rate of change of \(f\) per unit of that variable.
Geometry. The area of a circle is \(A(r)=\pi r^2\), so \(\dfrac{dA}{dr}=2\pi r\): the area grows \(2\pi r\) square units per unit of radius. (Notice that \(2\pi r\) is the circumference: adding a thin ring of width \(dr\) adds area \(\approx2\pi r\,dr\).) The volume of a sphere is \(V(r)=\tfrac43\pi r^3\), so \(\dfrac{dV}{dr}=4\pi r^2\), the surface area.
Economics. If \(C(x)\) is the cost of producing \(x\) items, \(C'(x)\) is the marginal cost, the approximate cost of one more item. If \(R(x)\) is revenue, \(R'(x)\) is the marginal revenue; for profit \(P(x)=R(x)-C(x)\), \(P'(x)=R'(x)-C'(x)\).
Growth and decay. If \(P(t)\) is a population, \(P'(t)\) is its growth rate in people (or animals, or cells) per unit time. For quantities that grow or decay by a constant factor, the rate is proportional to the amount itself, a theme developed in Unit 7.
Worked example · Marginal revenue
A company's revenue from selling \(x\) items is \(R(x)=50x-0.1x^2\) dollars. Find the marginal revenue at \(x=100\) and interpret.
\(R'(x)=50-0.2x\), so \(R'(100)=50-20=30\). When 100 items are being sold, revenue is increasing at a rate of \$30 per additional item: selling the 101st item brings in about \$30.
Worked example · Population growth
A city's population is modeled by \(P(t)=1200e^{0.03t}\) people, where \(t\) is years after 2020. Find \(P'(10)\) and interpret.
\(P'(t)=1200(0.03)e^{0.03t}=36e^{0.03t}\). Then \(P'(10)=36e^{0.3}\approx48.6\). Ten years after 2020 (in 2030), the population is increasing at a rate of about 48.6 people per year.
Worked example · Geometry with a changing radius
A balloon's volume is \(V(r)=\tfrac43\pi r^3\) cubic centimeters when its radius is \(r\) cm. At what rate does the volume change with respect to the radius when \(r=5\)?
\(\dfrac{dV}{dr}=4\pi r^2\), so at \(r=5\): \(4\pi(25)=100\pi\approx314.2\) cm\(^3\) per cm. (This is per unit of radius, not per second; if the radius itself changes with time you will need the chain rule, in Lessons 4.4 and 4.5.)
Rates from data and from graphs
In practice the formula may be missing. You then estimate the rate from a table or graph as in Lesson 2.3, and interpret it with the same four-part template of Lesson 4.1: when, what quantity, increasing or decreasing, how fast (with units).
Worked example · A table in context
A reservoir's volume \(W(t)\), in million gallons, is measured at several days.
\(t\) (days)
0
3
6
10
\(W(t)\) (million gal)
50
44
41
29
Estimate \(W'(6)\) and interpret.
Use the neighboring data points: \(\dfrac{W(10)-W(3)}{10-3}=\dfrac{29-44}{7}\approx-2.14\). On day 6, the volume of water in the reservoir is decreasing at approximately 2.14 million gallons per day.
Choosing the right interpretation
Read the problem for the independent variable. If it says "per hour," the variable is time in hours; if it says "per item," it is the number produced. The units tell you what the derivative means. Also pay attention to sign: in a context where something is being used up (water draining, a loan being paid), the derivative is negative even though the "rate" is a positive amount.
Exam focus
\(f'(x)\) is a rate of change per unit of \(x\) in any application: geometry, cost, revenue, population, temperature, concentration.
Marginal cost \(C'(x)\) and marginal revenue \(R'(x)\) approximate the cost or revenue of one additional unit.
Always attach units: output units per input unit.
From data, estimate with a difference quotient and say "approximately."
Interpret the sign: positive means increasing, negative means decreasing.
Practice
The area of a square with side \(s\) is \(A(s)=s^2\). Find \(\dfrac{dA}{ds}\) and interpret at \(s=6\) cm.
Show answer
\(\dfrac{dA}{ds}=2s\), so at \(s=6\) the area is increasing at a rate of 12 square centimeters per centimeter of side length.
Cost is \(C(x)=0.02x^2+3x+500\) dollars for \(x\) items. Find the marginal cost at \(x=50\).
Show answer
\(C'(x)=0.04x+3\); \(C'(50)=5\). At a production level of 50 items, the cost is increasing at about \$5 per item.
A population is \(P(t)=2000e^{0.05t}\). Find the growth rate when \(t=4\) and give units if \(t\) is in years and \(P\) in individuals.
Show answer
\(P'(t)=100e^{0.05t}\); \(P'(4)=100e^{0.2}\approx122.1\). At \(t=4\) years the population is growing at about 122 individuals per year.
A table gives the concentration of a drug \(D(t)\) in mg/L at \(t\) hours: \(D(1)=8\), \(D(2)=6\), \(D(4)=3\). Estimate \(D'(2)\) and interpret.
Show answer
\(\dfrac{D(4)-D(1)}{4-1}=\dfrac{3-8}{3}\approx-1.67\). At 2 hours the drug concentration is decreasing at approximately 1.67 mg/L per hour.
The revenue function \(R(x)=80x-0.2x^2\) dollars comes from selling \(x\) units. At what \(x\) is the marginal revenue zero, and what might that mean?
Show answer
\(R'(x)=80-0.4x=0\) at \(x=200\). At 200 units, selling one more unit adds (about) no revenue. Beyond that, revenue decreases with each additional unit.
The surface area of a sphere is \(S(r)=4\pi r^2\) square centimeters when the radius is \(r\) cm. Find the rate of change of \(S\) with respect to \(r\) at \(r=3\), with units.
Show answer
\(\dfrac{dS}{dr}=8\pi r\), so at \(r=3\) it is \(24\pi\approx75.4\) square centimeters per centimeter of radius.
A balloon is inflated, so its volume and radius both change. A ladder slides down a wall, so the height of its top and the distance of its base both change. In situations like these, several quantities change with time, and an equation ties them together. If you know how fast one is changing, you can find how fast another is. Such problems are called related rates, and they are the chain rule at work.
Required content · CHA-3.D.1, CHA-3.D.2
Learning objective.
CHA-3.D. Calculate related rates in applied contexts.
Essential knowledge.
CHA-3.D.1. The chain rule is the basis for differentiating variables in a related rates problem with respect to the same independent variable.
CHA-3.D.2. Other differentiation rules, such as the product rule and the quotient rule, may also be necessary to differentiate all variables with respect to the same independent variable.
The idea
Suppose \(x\) and \(y\) are both functions of time \(t\), and suppose an equation relates them, such as \(x^2+y^2=169\). Differentiate both sides with respect to \(t\). Because \(x\) and \(y\) depend on \(t\), the chain rule applies to each term: \[\frac{d}{dt}\big[x^2\big]=2x\frac{dx}{dt},\qquad \frac{d}{dt}\big[y^2\big]=2y\frac{dy}{dt}.\] The chain rule is the basis for differentiating variables in a related rates problem with respect to the same independent variable. Other differentiation rules, such as the product rule and the quotient rule, may also be necessary to differentiate all variables with respect to \(t\).
This is the same move as implicit differentiation (Lesson 3.2), with \(t\) playing the role of \(x\).
The steps
Draw and label. Sketch the situation. Give a letter to every quantity that changes, and put known constant lengths on the diagram as numbers.
List the given rates and the rate you need, with units and signs. A quantity that decreases has a negative rate.
Write an equation relating the variables that is true at every time (geometry, a formula, similar triangles).
Differentiate with respect to \(t\).
Only now substitute the values at the moment in question, and solve for the unknown rate.
Answer in a sentence with units, saying whether the quantity is increasing or decreasing.
Step 5 is the place students go wrong most often. If you substitute the instantaneous values (like \(x=5\)) before differentiating, you turn variables into constants and your derivative becomes zero.
Figure 4.2. A 13-foot ladder leans against a wall. As the base slides away from the wall (\(x\) increases), the top slides down (\(y\) decreases). The variables are linked by \(x^2+y^2=169\), and differentiating with respect to time gives a relationship between \(\dfrac{dx}{dt}\) and \(\dfrac{dy}{dt}\).
Worked example · The ladder
A 13-foot ladder leans against a wall. The base is pulled away from the wall at 2 ft/s. How fast is the top sliding down when the base is 5 feet from the wall?
Variables: \(x\) = distance from the wall to the base, \(y\) = height of the top; both change with time. Known: \(\dfrac{dx}{dt}=2\) ft/s. Find: \(\dfrac{dy}{dt}\) when \(x=5\).
Relationship. \(x^2+y^2=169\) (Pythagorean theorem; the ladder length is constant).
Differentiate. \(2x\dfrac{dx}{dt}+2y\dfrac{dy}{dt}=0\), so \(\dfrac{dy}{dt}=-\dfrac xy\dfrac{dx}{dt}\).
Substitute. When \(x=5\), \(y=\sqrt{169-25}=12\). So \(\dfrac{dy}{dt}=-\dfrac{5}{12}(2)=-\dfrac56\).
Answer. The top of the ladder is sliding down the wall at \(\dfrac56\) feet per second (about 0.83 ft/s).
Worked example · A growing circle
The radius of a circular oil slick grows at 0.5 cm/s. How fast is the area increasing when the radius is 10 cm?
\(A=\pi r^2\) gives \(\dfrac{dA}{dt}=2\pi r\dfrac{dr}{dt}\). At \(r=10\), \(\dfrac{dr}{dt}=0.5\): \(\dfrac{dA}{dt}=2\pi(10)(0.5)=10\pi\approx31.4\) cm\(^2\)/s.
Worked example · A cube
The edge of a cube increases at 2 cm/s. How fast is the volume increasing when the edge is 5 cm?
\(V=s^3\), so \(\dfrac{dV}{dt}=3s^2\dfrac{ds}{dt}=3(25)(2)=150\) cm\(^3\)/s.
When the product rule appears
If the equation involves a product of two changing quantities, use the product rule.
Worked example · A rectangle
The length \(\ell\) of a rectangle increases at 2 m/s while its width \(w\) decreases at 1 m/s. How fast is the area changing when \(\ell=8\) and \(w=5\)?
\(A=\ell w\), so \(\dfrac{dA}{dt}=\dfrac{d\ell}{dt}\,w+\ell\,\dfrac{dw}{dt}=(2)(5)+(8)(-1)=2\). The area is increasing at 2 square meters per second.
Exam focus
Differentiate an equation relating the variables with respect to time; the chain rule puts a \(\frac{d}{dt}\) factor on every variable term.
Decreasing quantities have negative rates. Signs of the given and the answer matter.
Substitute the instantaneous values after differentiating, never before.
Use the product or quotient rule when the relationship has products or quotients of changing quantities.
State the answer with units and a direction (increasing or decreasing).
Practice
A spherical balloon is inflated so that its radius grows at 2 cm/s. How fast is the volume increasing when \(r=5\) cm? (\(V=\tfrac43\pi r^3\))
Show answer
A rectangle's length is 10 and increasing at 2 per second; its width is 6 and decreasing at 1 per second. Find the rate of change of the area.
Show answer
\(\dfrac{dA}{dt}=2(6)+10(-1)=2\) square units per second; the area is increasing.
In the ladder problem, if the base slides away at 3 ft/s, how fast is the top sliding down when the base is 12 feet from the wall?
Show answer
\(y=\sqrt{169-144}=5\). \(\dfrac{dy}{dt}=-\dfrac{12}{5}(3)=-\dfrac{36}{5}=-7.2\) ft/s: the top slides down at 7.2 ft/s.
Two people start at the same point. One walks east at 4 ft/s and the other north at 3 ft/s. How fast is the distance between them changing after 5 seconds?
Show answer
After 5 s: \(x=20\), \(y=15\), \(z=\sqrt{625}=25\). From \(z^2=x^2+y^2\): \(\dfrac{dz}{dt}=\dfrac{x\,x'+y\,y'}{z}=\dfrac{20(4)+15(3)}{25}=5\) ft/s.
A student substitutes \(x=5\) into \(x^2+y^2=169\) before differentiating, getting \(25+y^2=169\), and then differentiates to find \(2y\dfrac{dy}{dt}=0\). What is the mistake?
Show answer
Substituting \(x=5\) treats \(x\) as a constant, so \(\dfrac{dx}{dt}\) is lost and the derivative of the constant 25 is 0. You must differentiate the general relationship first, then substitute the values for the instant in question.
Lesson 4.4 gave the method. Now you will use it on the problems that appear on the exam: a draining cone, a moving shadow, two vehicles, a rising balloon watched by an observer. The algebra changes from problem to problem; the six steps do not. A related rates problem is solved when you can say, with units, how fast a quantity is changing at a particular moment, and interpret that result. The derivative can be used to solve related rates problems; that is, finding a rate at which one quantity is changing by relating it to other quantities whose rates of change are known.
Required content · CHA-3.E.1
Learning objective.
CHA-3.E. Interpret related rates in applied contexts.
Essential knowledge.
CHA-3.E.1. The derivative can be used to solve related rates problems; that is, finding a rate at which one quantity is changing by relating it to other quantities whose rates of change are known.
A cone with changing depth
When the equation involves two changing quantities that you do not both need (such as the radius and the depth of water in a cone), use geometry to eliminate one before differentiating.
Figure 4.3. An inverted cone with radius \(5\) ft and height \(10\) ft holds water to depth \(h\), with surface radius \(r\). Similar triangles give \(\dfrac rh=\dfrac5{10}\), so \(r=\dfrac h2\). That relationship lets us write the volume \(V=\tfrac13\pi r^2h\) in terms of \(h\) alone.
Worked example · Water poured into a cone
Water is poured into an inverted conical tank of radius 5 ft and height 10 ft at 8 ft\(^3\)/min. How fast is the water level rising when the water is 6 ft deep?
Let \(V\) be the volume of water, \(h\) its depth, and \(r\) its surface radius. Known: \(\dfrac{dV}{dt}=8\). Find: \(\dfrac{dh}{dt}\) when \(h=6\).
Relationship. \(V=\tfrac13\pi r^2h\). By similar triangles, \(\dfrac rh=\dfrac5{10}\), so \(r=\dfrac h2\). Then \(V=\tfrac13\pi\left(\dfrac h2\right)^2h=\dfrac{\pi}{12}h^3\).
Substitute. \(8=\dfrac\pi4(36)\dfrac{dh}{dt}=9\pi\dfrac{dh}{dt}\), so \(\dfrac{dh}{dt}=\dfrac{8}{9\pi}\approx0.283\).
Answer. The water level is rising at about 0.283 feet per minute when the depth is 6 ft.
A moving shadow (similar triangles)
Worked example · A shadow
A 6-ft-tall person walks away from a 15-ft lamppost at 4 ft/s. How fast is the tip of the person's shadow moving, and how fast is the shadow lengthening?
Let \(x\) be the person's distance from the post and \(s\) the length of the shadow. By similar triangles, \(\dfrac{s}{6}=\dfrac{x+s}{15}\), so \(15s=6x+6s\) and \(s=\dfrac23x\). The tip of the shadow is at distance \(x+s=\dfrac53x\) from the post.
With \(\dfrac{dx}{dt}=4\): the tip moves at \(\dfrac53(4)=\dfrac{20}{3}\approx6.67\) ft/s, and the shadow lengthens at \(\dfrac{ds}{dt}=\dfrac23(4)=\dfrac83\approx2.67\) ft/s. (Both rates are constant, independent of where the person stands.)
Two moving objects
Worked example · Two cars
Car A travels east at 60 mph and car B travels north at 40 mph, both leaving an intersection at the same time. How fast is the distance between them changing after 2 hours?
After 2 hours: \(x=120\) mi, \(y=80\) mi, \(z=\sqrt{120^2+80^2}=\sqrt{20800}=40\sqrt{13}\approx144.2\) mi. From \(z^2=x^2+y^2\): \(z\dfrac{dz}{dt}=x\dfrac{dx}{dt}+y\dfrac{dy}{dt}\), so \[\frac{dz}{dt}=\frac{120(60)+80(40)}{40\sqrt{13}}=\frac{10400}{40\sqrt{13}}=\frac{260}{\sqrt{13}}=20\sqrt{13}\approx72.1\ \text{mph}.\] The cars are separating at about 72.1 miles per hour.
Angles
When the unknown rate is an angle's rate, relate the angle to a length by a trigonometric ratio, and remember that \(\dfrac{d}{d\theta}[\tan\theta]=\sec^2\theta\).
Worked example · A rising balloon
A balloon rises vertically at 5 m/s from a point 100 m from an observer. How fast is the angle of elevation \(\theta\) of the observer's line of sight changing when the balloon is 100 m high?
\(\tan\theta=\dfrac{h}{100}\). Differentiate: \(\sec^2\theta\,\dfrac{d\theta}{dt}=\dfrac1{100}\dfrac{dh}{dt}\). When \(h=100\), \(\tan\theta=1\), \(\theta=\dfrac\pi4\), and \(\sec^2\theta=2\). So \(2\dfrac{d\theta}{dt}=\dfrac{5}{100}\), and \(\dfrac{d\theta}{dt}=0.025\) rad/s.
Interpreting the result
On the free-response section, you are often asked to interpret: "At the instant the depth is 6 feet, the water level is rising at a rate of 0.283 feet per minute." A statement of the sign without units, or a number without the instant, is incomplete.
Exam focus
Draw, name variables, list rates with signs, write an equation true at all times (use similar triangles or geometry to reduce variables).
Differentiate with respect to \(t\), then substitute the instant's values.
Similar triangles and the Pythagorean theorem are the usual relationships; for angles use a trigonometric ratio and \(\sec^2\).
Answer with units, the instant, and the direction of change.
Check reasonableness: does the sign match the story (a draining tank has \(\frac{dh}{dt}\lt0\))?
Practice
A spherical balloon is inflated at 100 cm\(^3\)/s. How fast is the radius increasing when \(r=5\) cm?
Show answer
\(\dfrac{dV}{dt}=4\pi r^2\dfrac{dr}{dt}\), so \(100=4\pi(25)\dfrac{dr}{dt}\) and \(\dfrac{dr}{dt}=\dfrac1\pi\approx0.318\) cm/s.
Water drains from the cone of radius 5 ft and height 10 ft at 6 ft\(^3\)/min. How fast is the depth changing when \(h=4\) ft?
Show answer
\(\dfrac{dV}{dt}=\dfrac\pi4h^2\dfrac{dh}{dt}\), so \(-6=\dfrac\pi4(16)\dfrac{dh}{dt}=4\pi\dfrac{dh}{dt}\) and \(\dfrac{dh}{dt}=-\dfrac{3}{2\pi}\approx-0.477\) ft/min: the depth decreases at about 0.477 ft/min.
A 5-ft-tall person walks at 3 ft/s away from a 20-ft streetlight. How fast is the length of the shadow increasing?
Show answer
\(\dfrac{s}{5}=\dfrac{x+s}{20}\) gives \(20s=5x+5s\), so \(s=\dfrac13x\) and \(\dfrac{ds}{dt}=\dfrac13(3)=1\) ft/s.
A kite is 80 ft high and is blown horizontally away at 5 ft/s. How fast is the string (taut, from the ground at your feet) lengthening when 100 ft of string is out?
Show answer
\(z^2=x^2+80^2\). When \(z=100\), \(x=60\). \(z\dfrac{dz}{dt}=x\dfrac{dx}{dt}\), so \(\dfrac{dz}{dt}=\dfrac{60(5)}{100}=3\) ft/s.
A 10-ft ladder's base is pulled away at 1 ft/s. How fast is the angle between the ladder and the ground changing when the base is 6 ft from the wall? (Use \(\cos\theta=\dfrac x{10}\).)
Show answer
\(-\sin\theta\dfrac{d\theta}{dt}=\dfrac1{10}\dfrac{dx}{dt}\). When \(x=6\), \(\sin\theta=\dfrac{8}{10}\). So \(\dfrac{d\theta}{dt}=-\dfrac{1}{10(0.8)}=-0.125\) rad/s: the angle decreases at 0.125 rad/s.
In the balloon problem, is \(\dfrac{d\theta}{dt}\) larger or smaller when the balloon is 200 m high than when it is 100 m high? Explain with the equation.
Show answer
\(\dfrac{d\theta}{dt}=\dfrac{5}{100}\cos^2\theta\). At 200 m, \(\tan\theta=2\), so \(\cos^2\theta=\dfrac15\) and \(\dfrac{d\theta}{dt}=0.01\), smaller than 0.025 at 100 m. The angle changes more slowly as the balloon gets farther away.
Near a point, a smooth curve looks almost like its tangent line. That simple fact lets you estimate hard values (a square root, a logarithm, an exponential) with arithmetic you can do by hand, and it lets you say whether the estimate is too big or too small.
Required content · CHA-3.F.1, CHA-3.F.2
Learning objective.
CHA-3.F. Approximate a value on a curve using the equation of a tangent line.
Essential knowledge.
CHA-3.F.1. The tangent line is the graph of a locally linear approximation of the function near the point of tangency.
CHA-3.F.2. For a tangent line approximation, the function's behavior near the point of tangency may determine whether a tangent line value is an underestimate or an overestimate of the corresponding function value.
The tangent line as an approximation
The tangent line to \(f\) at \(x=a\) is \(L(x)=f(a)+f'(a)(x-a)\). Because the tangent line is a good model of the curve near the point of tangency, \(L(x)\) is the local linear approximation of \(f\) near \(x=a\): \(f(x)\approx L(x)\) when \(x\) is close to \(a\). The tangent line can be used to approximate function values, and the approximation is generally better the closer \(x\) is to \(a\).
Figure 4.4. Near \(x=4\) the tangent line \(L(x)=2+\tfrac14(x-4)\) stays close to \(f(x)=\sqrt x\). At \(x=5\), \(L(5)=2.25\) overestimates \(\sqrt5\approx2.236\) because the graph is concave down and lies below its tangent line.
Worked example · Approximating a square root
Use the tangent line to \(f(x)=\sqrt x\) at \(x=4\) to approximate \(\sqrt{4.1}\).
\(f(4)=2\) and \(f'(x)=\dfrac1{2\sqrt x}\), so \(f'(4)=\dfrac14\). Then \(L(x)=2+\dfrac14(x-4)\) and \[\sqrt{4.1}\approx L(4.1)=2+\tfrac14(0.1)=2.025.\] (The true value is 2.02485 to five decimal places.)
Worked example · Approximating a cube root
Approximate \(\sqrt[3]{8.3}\) with the tangent line at \(x=8\).
\(f(x)=x^{1/3}\), \(f(8)=2\), \(f'(x)=\tfrac13x^{-2/3}\), \(f'(8)=\tfrac1{12}\). So \(\sqrt[3]{8.3}\approx2+\tfrac1{12}(0.3)=2.025\).
Worked example · Exponential and logarithm
At \(x=0\), \(e^x\) has \(L(x)=1+x\), so \(e^{0.1}\approx1.1\) (true value 1.10517). At \(x=1\), \(\ln x\) has \(L(x)=x-1\), so \(\ln1.05\approx0.05\) (true value 0.04879).
Overestimate or underestimate?
The sign of \(f''\) tells you which side of the tangent line the curve lies on.
If \(f\) is concave up near \(a\) (\(f''\gt0\)), the curve lies above its tangent line, so the tangent-line estimate is an underestimate.
If \(f\) is concave down near \(a\) (\(f''\lt0\)), the curve lies below its tangent line, so the tangent-line estimate is an overestimate.
Check: \(\sqrt x\) is concave down (\(f''=-\tfrac14x^{-3/2}\lt0\)), so 2.025 should be an overestimate of \(\sqrt{4.1}\); the true value, 2.02485, is smaller. And \(e^x\) is concave up, so \(1.1\) underestimates \(e^{0.1}=1.10517\). Both agree.
On the exam, a justification is "\(f''(x)\gt0\) on the interval, so the graph of \(f\) lies above the tangent line, and the tangent-line approximation is an underestimate."
Using given information
Often \(f\) is not given by a formula; you know only \(f(a)\) and \(f'(a)\) (from a table or a differential equation). The local linear approximation is still \(f(a)+f'(a)(x-a)\).
Worked example · From a table
A function \(g\) is twice differentiable with \(g(3)=5\), \(g'(3)=-2\), and \(g''(x)\gt0\) for all \(x\). Approximate \(g(3.2)\) and tell whether the approximation is too large or too small.
\(g(3.2)\approx5+(-2)(0.2)=4.6\). Because \(g''\gt0\), \(g\) is concave up, the graph lies above its tangent line, and 4.6 is an underestimate: the true \(g(3.2)\) is greater than 4.6.
Exam focus
\(L(x)=f(a)+f'(a)(x-a)\); then \(f(x)\approx L(x)\) near \(a\).
Choose \(a\) as the nearby point where \(f(a)\) and \(f'(a)\) are easy to compute.
Concave up \(\Rightarrow\) underestimate. Concave down \(\Rightarrow\) overestimate. Say why.
The farther \(x\) is from \(a\), the worse the approximation can be.
Show the tangent line equation (or the setup) before the number.
Practice
Find the local linearization of \(f(x)=x^3\) at \(x=2\) and use it to approximate \(2.1^3\).
Show answer
\(f(2)=8\), \(f'(2)=12\), so \(L(x)=8+12(x-2)\) and \(2.1^3\approx8+12(0.1)=9.2\). (True value 9.261.)
Use the tangent line to \(f(x)=\ln x\) at \(x=1\) to estimate \(\ln0.9\). Is the estimate too large or too small? Why?
Show answer
\(L(x)=x-1\), so \(\ln0.9\approx-0.1\). Since \(f''(x)=-\dfrac1{x^2}\lt0\), the curve lies below the tangent line, so \(-0.1\) is an overestimate (true value \(\approx-0.10536\)).
A curve has \(f(5)=12\) and \(f'(5)=-3\). Estimate \(f(4.8)\).
Show answer
\(f(4.8)\approx12+(-3)(-0.2)=12.6\).
Use a tangent line to estimate \(\sqrt{99}\).
Show answer
Use \(a=100\): \(f(100)=10\), \(f'(100)=\dfrac1{20}\). \(\sqrt{99}\approx10+\dfrac1{20}(-1)=9.95\). (True value about 9.94987.)
Let \(h\) be a function with \(h(1)=4\), \(h'(1)=2\), and \(h''(x)\lt0\) for all \(x\). Is the tangent-line approximation of \(h(1.5)\) an overestimate or an underestimate? Give the estimate.
Show answer
\(h(1.5)\approx4+2(0.5)=5\). Since \(h''\lt0\), \(h\) is concave down and lies below its tangent line, so 5 is an overestimate.
A student writes \(\sin(0.1)\approx\sin(0)+\cos(0)(0.1)=0.1\). Verify this is the local linearization at \(x=0\) and say whether it over- or underestimates for small positive \(x\).
Show answer
\(f(0)=0\), \(f'(0)=\cos0=1\), so \(L(x)=x\) and the student is right. For \(0\lt x\lt\pi\), \(f''(x)=-\sin x\lt0\), so \(\sin x\) lies below \(L(x)=x\): \(0.1\) overestimates (true value 0.09983).
Sometimes substituting a limit point gives \(\dfrac00\) or \(\dfrac\infty\infty\). These are called indeterminate forms: the form alone does not tell the limit. You have handled them by factoring or conjugates; L'Hospital's rule is a faster tool that uses derivatives.
Required content · LIM-4.A.1, LIM-4.A.2
Learning objective.
LIM-4.A. Determine limits of functions that result in indeterminate forms.
Essential knowledge.
LIM-4.A.1. When the ratio of two functions tends to \(\dfrac{0}{0}\) or \(\dfrac{\infty}{\infty}\) in the limit, such forms are said to be indeterminate.
LIM-4.A.2. Limits of the indeterminate forms \(\dfrac{0}{0}\) or \(\dfrac{\infty}{\infty}\) may be evaluated using L'Hospital's Rule.
The rule
Suppose \(f\) and \(g\) are differentiable near \(a\) (except possibly at \(a\)), and \(g'(x)\ne0\) near \(a\). If \(\displaystyle\lim_{x\to a}\frac{f(x)}{g(x)}\) has the form \(\dfrac00\) or \(\dfrac{\pm\infty}{\pm\infty}\), then \[\lim_{x\to a}\frac{f(x)}{g(x)}=\lim_{x\to a}\frac{f'(x)}{g'(x)},\] provided the limit on the right exists (or is infinite). The same holds for one-sided limits and for \(x\to\pm\infty\). Differentiate the top and the bottom separately; this is not the quotient rule.
Worked example · A \(\frac00\) limit
\(\displaystyle\lim_{x\to0}\frac{\sin x}{x}\): substituting gives \(\tfrac00\). Differentiating top and bottom: \(\displaystyle\lim_{x\to0}\frac{\cos x}{1}=1\).
Worked example · Applying it twice
\(\displaystyle\lim_{x\to0}\frac{e^x-1-x}{x^2}\) is \(\tfrac00\). Then \(\displaystyle\lim_{x\to0}\frac{e^x-1}{2x}\) is still \(\tfrac00\), so apply the rule again: \(\displaystyle\lim_{x\to0}\frac{e^x}{2}=\frac12\).
Worked example · An \(\frac\infty\infty\) limit: exponentials beat polynomials
\(\displaystyle\lim_{x\to\infty}\frac{x^2}{e^x}\) has the form \(\tfrac\infty\infty\). Apply the rule: \(\displaystyle\lim_{x\to\infty}\frac{2x}{e^x}\), still \(\tfrac\infty\infty\); apply again: \(\displaystyle\lim_{x\to\infty}\frac{2}{e^x}=0\). Similarly, \(\displaystyle\lim_{x\to\infty}\frac{\ln x}{x}=\lim_{x\to\infty}\frac{1/x}{1}=0\). Hence exponentials grow faster than any power of \(x\), and any power of \(x\) grows faster than \(\ln x\).
Warnings
Check the form first. The rule applies only to \(\tfrac00\) and \(\tfrac\infty\infty\). For a limit such as \(\displaystyle\lim_{x\to0}\frac{x+1}{x+2}\), whose top is 1 and bottom is 2, applying the rule would give the wrong answer \(1\) instead of \(\tfrac12\).
Do not use the quotient rule; differentiate numerator and denominator separately.
If the new limit does not exist (not even as \(\pm\infty\)), the rule tells you nothing; try another method. For example, \(\displaystyle\lim_{x\to\infty}\frac{x+\sin x}{x}\) is \(\tfrac\infty\infty\), but \(\dfrac{1+\cos x}{1}\) oscillates, so L'Hospital's rule does not decide it; dividing by \(x\) gives \(1\).
Write the form. On the exam, state the form, for example "\(\lim\dfrac{f}{g}=\dfrac00\), so by L'Hospital's rule...".
Worked example · Limits from graph information
Let \(f(3)=g(3)=0\), \(f'(3)=8\) and \(g'(3)=4\), with \(f\) and \(g\) differentiable. Then \(\displaystyle\lim_{x\to3}\frac{f(x)}{g(x)}\) is of the form \(\tfrac00\), and by the rule equals \(\dfrac{f'(3)}{g'(3)}=2\).
Exam focus
Use only on \(\tfrac00\) or \(\tfrac\infty\infty\); state the form.
Differentiate numerator and denominator separately.
Repeat if still indeterminate; stop when the limit is determined.
If the new limit fails to exist, the rule is inconclusive.
Growth ranking: \(\ln x\ll x^n\ll e^x\) as \(x\to\infty\).
Practice
Evaluate \(\displaystyle\lim_{x\to2}\frac{x^2-4}{x-2}\) using L'Hospital's rule.
Show answer
Evaluate \(\displaystyle\lim_{x\to\infty}\frac{3x^2+x}{5x^2-2}\) two ways: with L'Hospital's rule and by dividing by \(x^2\).
Show answer
\(\tfrac\infty\infty\). Rule: \(\dfrac{6x+1}{10x}\), again \(\tfrac\infty\infty\); then \(\dfrac{6}{10}=\dfrac35\). Dividing by \(x^2\): \(\dfrac{3+1/x}{5-2/x^2}\to\dfrac35\).
A student writes \(\displaystyle\lim_{x\to0}\frac{x+3}{x+2}=\lim_{x\to0}\frac{1}{1}=1\). What is wrong, and what is the correct limit?
Show answer
The form is \(\tfrac32\), not indeterminate, so the rule does not apply. Direct substitution gives the limit \(\tfrac32\).
Explain why L'Hospital's rule cannot be used on \(\displaystyle\lim_{x\to\infty}\frac{x+\cos x}{x}\) to find the limit, and find it another way.
Show answer
The form is \(\tfrac\infty\infty\), but \(\dfrac{1-\sin x}{1}\) oscillates and has no limit, so the rule is inconclusive. Divide by \(x\): \(1+\dfrac{\cos x}{x}\to1\), since \(\dfrac{\cos x}x\to0\) by squeeze.
Unit 4 practice · 10 problems
Unit 4 practice: Contextual Applications of Differentiation
Ten problems covering the whole unit. Work each one on paper before
revealing the answer. A calculator is fine for the final arithmetic in
problems 4, 5, and 10.
\(W(t)\) is the number of gallons of water in a pool \(t\) hours after noon, and \(W'(3) = -15\). Interpret this with units.
Show answer
At 3:00 pm, the amount of water in the pool is decreasing at a rate of 15 gallons per hour.
A particle's position is \(x(t) = t^3 - 9t^2 + 24t\) for \(t \ge 0\). When is it at rest?
Show answer
\(v(t) = 3t^2 - 18t + 24 = 3(t - 2)(t - 4)\). At rest at \(t = 2\) and \(t = 4\).
For the same particle, is it speeding up or slowing down at \(t = 1\)?
Show answer
\(v(1) = 9\) and \(a(t) = 6t - 18\), so \(a(1) = -12\). Opposite signs: slowing down.
The edge of a cube grows at 2 cm/s. How fast is the volume increasing when the edge is 5 cm?
Two cars leave the same point at the same time; one drives north at 60 mph, the other east at 80 mph. How fast is the distance between them increasing after one hour?
If you drive 120 miles in exactly 2 hours, your average speed is 60 mph. Somewhere on that trip your speedometer must have read exactly 60, because a speedometer that starts below 60 and ends above it (or the reverse) cannot skip a value. That is the Mean Value Theorem: on a smooth stretch, the instantaneous rate of change equals the average rate of change at some moment. It is one of the central theorems of the course, because it is what lets a derivative's sign tell you about the function.
Required content · FUN-1.B.1
Learning objective.
FUN-1.B. Justify conclusions about functions by applying the Mean Value Theorem over an interval.
Essential knowledge.
FUN-1.B.1. If a function \(f\) is continuous over the interval \([a,b]\) and differentiable over the interval \((a,b)\), then the Mean Value Theorem guarantees a point within that open interval where the instantaneous rate of change equals the average rate of change over the interval.
The theorem
Mean Value Theorem (MVT). If \(f\) is continuous on the closed interval \([a,b]\) and differentiable on the open interval \((a,b)\), then there is at least one number \(c\) in \((a,b)\) such that \[f'(c)=\frac{f(b)-f(a)}{b-a}.\] In words: some tangent line is parallel to the secant line through the endpoints.
Figure 5.1. The Mean Value Theorem. The secant line through the endpoints has slope \(\dfrac{f(b)-f(a)}{b-a}=1.25\). For this \(f\), there are two numbers \(c\) in \((0,5)\) where the tangent line is parallel to the secant (one is enough for the theorem).
Both hypotheses matter, and on the exam you earn a point by verifying them. Continuity on \([a,b]\) closes the gap at the endpoints; differentiability on \((a,b)\) rules out corners and vertical tangents in between.
Rolle's Theorem is the special case \(f(a)=f(b)\): then the secant slope is 0, so \(f'(c)=0\) for some \(c\) in \((a,b)\): somewhere the tangent is horizontal.
Using the theorem on a formula
Worked example · Finding the guaranteed value
Show that \(f(x)=x^3\) satisfies the MVT on \([0,2]\) and find the value of \(c\).
\(f\) is a polynomial, so it is continuous on \([0,2]\) and differentiable on \((0,2)\). The average rate of change is \(\dfrac{f(2)-f(0)}{2-0}=\dfrac{8}{2}=4\). Solve \(f'(c)=3c^2=4\): \(c=\pm\dfrac{2}{\sqrt3}\). Only the positive value is in \((0,2)\), so \(c=\dfrac{2}{\sqrt3}\approx1.155\).
Worked example · The theorem does not apply
Does the MVT apply to \(f(x)=\lvert x\rvert\) on \([-1,3]\)?
\(f\) is continuous everywhere, but it is not differentiable at \(x=0\), which lies in \((-1,3)\). The hypothesis fails, so the theorem does not apply. (In fact the average rate of change is \(\dfrac{3-1}{4}=\dfrac12\), and \(f'(x)\) is only \(1\) or \(-1\), so no \(c\) exists. A failed hypothesis does not always make the conclusion false, but here it does.)
Using the theorem on given information
The most common exam use of the MVT is to conclude that a particular rate occurs.
Worked example · A rate must occur
A car's position is differentiable, and its odometer shows 5 miles at 1:00 p.m. and 65 miles at 2:00 p.m. Explain why the car's speed must equal 60 mph at some time.
Position \(s(t)\) is differentiable, hence continuous. By the MVT on \([1,2]\), there is a time \(c\) in \((1,2)\) with \(s'(c)=\dfrac{65-5}{2-1}=60\). So the velocity is exactly 60 mph at some time between 1:00 and 2:00.
Worked example · From a table
A differentiable function \(f\) has \(f(2)=7\), \(f(6)=19\), and the table below gives \(f(4)\). Show that \(f'(c)=3\) for some \(c\) in \((2,6)\), and decide whether \(f'\) must equal \(\tfrac52\) on \((2,4)\).
\(x\)
2
4
6
\(f(x)\)
7
14.5
19
On \([2,6]\) the average rate is \(\dfrac{19-7}{4}=3\), so \(f'(c)=3\) for some \(c\) in \((2,6)\). On \([2,4]\) the average rate is \(\dfrac{14.5-7}{2}=3.75\), so \(f'\) equals \(3.75\) somewhere in \((2,4)\). The average rate on \([4,6]\) is \(\dfrac{19-14.5}{2}=2.25\), so \(f'\) equals \(2.25\) somewhere in \((4,6)\). The value \(\tfrac52\) is not guaranteed by the MVT on either subinterval.
Consequences you will use again
If \(f'(x)=0\) for all \(x\) in an interval, then \(f\) is constant there (any two points give average rate 0). If \(f'(x)\gt0\) on an interval, then \(f\) is increasing there (any \(x_1\lt x_2\) give a positive average rate, so \(f(x_2)\gt f(x_1)\)). Lessons 5.3 and 5.4 rest on these facts.
Exam focus
State the hypotheses: continuous on \([a,b]\), differentiable on \((a,b)\), and check them.
Conclusion: \(f'(c)=\dfrac{f(b)-f(a)}{b-a}\) for at least one \(c\) in \((a,b)\).
To find \(c\): compute the average rate, solve \(f'(c)=\) that value, and keep only solutions in the open interval.
Rolle's Theorem: if \(f(a)=f(b)\), then \(f'(c)=0\) for some \(c\).
Use the MVT to justify statements such as "the speed was exactly 60 at some time."
Practice
Verify the hypotheses of the MVT for \(f(x)=x^2-4x+1\) on \([1,5]\) and find \(c\).
Show answer
\(f\) is a polynomial, so continuous and differentiable. \(f(5)=6\), \(f(1)=-2\), average rate \(\dfrac{6-(-2)}{4}=2\). \(f'(c)=2c-4=2\), so \(c=3\), which is in \((1,5)\).
Find the number guaranteed by the MVT for \(f(x)=\sqrt x\) on \([1,9]\).
Show answer
The average rate is \(\dfrac{3-1}{8}=\dfrac14\). \(f'(c)=\dfrac1{2\sqrt c}=\dfrac14\), so \(\sqrt c=2\) and \(c=4\).
Does the MVT apply to \(f(x)=\dfrac1{x}\) on \([-1,1]\)? Explain.
Show answer
No. \(f\) is not defined at \(x=0\), so it is not continuous on \([-1,1]\). The hypothesis fails.
A function \(g\) is differentiable for all real numbers, with \(g(1)=3\) and \(g(5)=11\). Must \(g'(c)=2\) for some \(c\) in \((1,5)\)? Justify.
Show answer
Yes. \(g\) is continuous and differentiable, so the MVT applies on \([1,5]\). The average rate is \(\dfrac{11-3}{4}=2\), so \(g'(c)=2\) for some \(c\) in \((1,5)\).
A runner completes a 10 km race in 40 minutes. Use the MVT to explain why the runner's speed was exactly 15 km/h at some moment.
Show answer
Distance \(s(t)\) is continuous and differentiable. The average speed is \(\dfrac{10\text{ km}}{2/3\text{ h}}=15\) km/h. By the MVT there is a time \(c\) in the race with \(s'(c)=15\) km/h.
Show that \(f(x)=x^3-3x+1\) satisfies Rolle's Theorem on \([-\sqrt3,\sqrt3]\) and find the values of \(c\).
Show answer
\(f(\sqrt3)=3\sqrt3-3\sqrt3+1=1\) and \(f(-\sqrt3)=-3\sqrt3+3\sqrt3+1=1\), so \(f(a)=f(b)\). \(f\) is a polynomial. \(f'(c)=3c^2-3=0\) gives \(c=\pm1\), both in \((-\sqrt3,\sqrt3)\).
Optimization, the highest point of a graph, the lowest cost, the fastest moment: all are questions about extreme values. This lesson gives the vocabulary, a theorem that guarantees extremes exist, and the list of places to look for them.
FUN-1.C. Justify conclusions about functions by applying the Extreme Value Theorem.
Essential knowledge.
FUN-1.C.1. If a function \(f\) is continuous over the interval \([a,b]\), then the Extreme Value Theorem guarantees that \(f\) has at least one minimum value and at least one maximum value on \([a,b]\).
FUN-1.C.2. A point on a function where the first derivative equals zero or fails to exist is a critical point of the function.
FUN-1.C.3. All local (relative) extrema occur at critical points of a function, though not all critical points are local extrema.
Local and global extrema
A function \(f\) has a relative (local) maximum at \(x=c\) if \(f(c)\ge f(x)\) for all \(x\) near \(c\); a relative (local) minimum if \(f(c)\le f(x)\) for all \(x\) near \(c\). It has an absolute (global) maximum on an interval if \(f(c)\ge f(x)\) for every \(x\) in the interval, and an absolute (global) minimum if \(f(c)\le f(x)\) for every \(x\) there. A local extremum compares \(f(c)\) only with nearby values; a global extremum compares it with the whole interval. The global maximum and minimum are called the extreme values of \(f\) on that interval.
The Extreme Value Theorem
Extreme Value Theorem (EVT). If \(f\) is continuous on a closed interval \([a,b]\), then \(f\) attains both a global maximum and a global minimum on \([a,b]\).
The hypotheses matter. \(f(x)=\dfrac1x\) on \((0,1]\) has no maximum (it is unbounded) because the interval is not closed; \(f(x)=x\) on \((0,1)\) has neither a maximum nor a minimum; and a discontinuous function can leap over its supposed maximum. On the exam, justification starts with "\(f\) is continuous on the closed interval \([a,b]\), so by the Extreme Value Theorem the extreme values exist."
Critical points
A critical point of \(f\) is a number \(c\) in the domain of \(f\) where \(f'(c)=0\) or \(f'(c)\) does not exist. A critical point is a candidate for an extremum because of the following fact: if \(f\) has a relative extremum at an interior point \(c\), then \(c\) is a critical point of \(f\). (Fermat's Theorem.) The converse is false: a critical point need not be an extremum.
Figure 5.2. \(f(x)=\tfrac14x^4-x^3+1\) on \([-1,4]\). \(f'(x)=x^2(x-3)\), so the critical points are \(x=0\) (a horizontal tangent with no extremum) and \(x=3\) (the global minimum). The global maximum \(f(-1)=2.25\) occurs at an endpoint, so a search for extrema must always check endpoints.
In the figure, \(f'(x)=x^2(x-3)\): the critical point \(x=0\) has a horizontal tangent yet is not an extremum (the graph keeps falling through it), while \(x=3\) is the global minimum. The global maximum on \([-1,4]\) is at the endpoint \(x=-1\), which is not a critical point at all, so endpoints must be checked separately.
Worked example · Finding critical points
Find the critical points of \(f(x)=x^3-6x^2+9x+1\).
\(f'(x)=3x^2-12x+9=3(x-1)(x-3)\), which is defined everywhere and is zero at \(x=1\) and \(x=3\). The critical points are \(x=1\) and \(x=3\).
Worked example · A critical point where \(f'\) is undefined
Find the critical points of \(f(x)=x^{2/3}\).
\(f'(x)=\dfrac23x^{-1/3}=\dfrac{2}{3\sqrt[3]{x}}\) is never zero, but it is undefined at \(x=0\), which is in the domain of \(f\). So \(x=0\) is the only critical point. (The graph has a cusp there, and \(f(0)=0\) is a global minimum.)
Worked example · Not every critical point is an extremum
\(f(x)=x^3\) has \(f'(x)=3x^2=0\) at \(x=0\), so \(x=0\) is a critical point. But \(f\) is increasing on both sides of 0, so \(f(0)\) is neither a maximum nor a minimum.
Worked example · A critical point in a rational function
Find the critical points of \(g(x)=\dfrac{x}{x^2+1}\).
\(g'(x)=\dfrac{(x^2+1)-x(2x)}{(x^2+1)^2}=\dfrac{1-x^2}{(x^2+1)^2}\). The denominator is never zero, and the numerator is zero at \(x=\pm1\). Critical points: \(x=-1\) and \(x=1\).
Looking for extremes: a plan
Confirm that \(f\) is continuous on a closed interval (so the EVT guarantees the extremes).
Find the critical points in the interval.
The global extremes occur at critical points or endpoints. (The next lessons develop how to decide among candidates.)
Exam focus
Critical point: \(f'(c)=0\) or \(f'(c)\) does not exist, with \(c\) in the domain of \(f\).
Relative extrema at interior points occur only at critical points; a critical point may or may not be an extremum.
EVT: continuous on a closed interval \(\Rightarrow\) a global maximum and a global minimum exist.
Global extremes on \([a,b]\) occur at critical points or endpoints.
Do not call a point critical if it is not in the domain of \(f\).
Practice
Find the critical points of \(f(x)=x^4-8x^2\).
Show answer
\(f'(x)=4x^3-16x=4x(x-2)(x+2)\): critical points at \(x=-2,0,2\).
Find the critical points of \(f(x)=x^2e^{-x}\).
Show answer
\(f'(x)=2xe^{-x}-x^2e^{-x}=xe^{-x}(2-x)\). Since \(e^{-x}\gt0\), \(f'(x)=0\) at \(x=0\) and \(x=2\).
Find the critical points of \(f(x)=\lvert x-3\rvert\).
Show answer
\(f'(x)=-1\) for \(x\lt3\) and \(1\) for \(x\gt3\); it does not exist at \(x=3\), which is in the domain. The only critical point is \(x=3\) (a global minimum).
Give an example of a continuous function on \((0,1)\) with no global maximum, and say why the EVT does not contradict it.
Show answer
\(f(x)=\dfrac1x\) on \((0,1)\) is unbounded above. The interval is open, so the EVT's hypothesis of a closed interval fails.
A student says that since \(f'(2)=0\), the function has a relative extremum at \(x=2\). Give a counterexample.
Show answer
\(f(x)=(x-2)^3\) has \(f'(2)=0\), yet it is increasing on both sides of \(x=2\), so there is no extremum there.
The function \(g\) is continuous on \([0,5]\) and has critical points only at \(x=2\). Where can \(g\) attain its global maximum, and why?
Show answer
By the EVT, a global maximum exists. It must occur at an endpoint or at a critical point, so at \(x=0\), \(x=2\) or \(x=5\).
The sign of the first derivative describes the graph's direction: positive slope means rising, negative slope means falling. This lesson turns that into a procedure that produces a clean, justified answer.
Required content · FUN-4.A.1
Learning objective.
FUN-4.A. Justify conclusions about the behavior of a function based on the behavior of its derivatives.
Essential knowledge.
FUN-4.A.1. The first derivative of a function can provide information about the function and its graph, including intervals where the function is increasing or decreasing.
The test
A function \(f\) is increasing on an interval if \(f(x_1)\lt f(x_2)\) whenever \(x_1\lt x_2\) in the interval, and decreasing if \(f(x_1)\gt f(x_2)\) whenever \(x_1\lt x_2\). The MVT (Lesson 5.1) gives the link to derivatives:
If \(f'(x)\gt0\) on an interval, then \(f\) is increasing on that interval.
If \(f'(x)\lt0\) on an interval, then \(f\) is decreasing on that interval.
The first derivative can only change sign at a point where \(f'(x)=0\) or \(f'(x)\) does not exist (or at a point where \(f\) is not defined). So the critical points and points of discontinuity divide the domain into intervals, and on each interval \(f'\) has one sign. Test one point in each interval (or factor \(f'\) and read the signs).
The procedure
Find the domain and \(f'\).
Find every \(x\) where \(f'(x)=0\) or \(f'(x)\) is undefined, and every point where \(f\) is not defined.
Place those numbers on a number line, creating intervals.
Determine the sign of \(f'\) on each interval (a test value or a factored sign chart).
State where \(f\) is increasing and decreasing, giving the reason in terms of \(f'\).
Worked example · A cubic
Find where \(f(x)=x^3-6x^2+9x+1\) is increasing and decreasing.
\(f'(x)=3x^2-12x+9=3(x-1)(x-3)\), zero at \(x=1,3\).
Interval
\((-\infty,1)\)
\((1,3)\)
\((3,\infty)\)
Test value
\(f'(0)=9\)
\(f'(2)=-3\)
\(f'(4)=9\)
Sign of \(f'\)
\(+\)
\(-\)
\(+\)
\(f\)
increasing
decreasing
increasing
\(f\) is increasing on \((-\infty,1)\) and \((3,\infty)\) because \(f'\gt0\) there, and decreasing on \((1,3)\) because \(f'\lt0\) there.
Worked example · An exponential factor
Find where \(g(x)=x^2e^{-x}\) is increasing.
\(g'(x)=2xe^{-x}-x^2e^{-x}=x(2-x)e^{-x}\). Since \(e^{-x}\gt0\), the sign of \(g'\) is the sign of \(x(2-x)\): negative for \(x\lt0\), positive for \(0\lt x\lt2\), negative for \(x\gt2\). So \(g\) is increasing on \((0,2)\), and decreasing on \((-\infty,0)\) and \((2,\infty)\).
Worked example · A gap in the domain
Find where \(h(x)=x+\dfrac1x\) is increasing and decreasing.
The domain is \(x\ne0\). \(h'(x)=1-\dfrac1{x^2}=\dfrac{x^2-1}{x^2}\), zero at \(x=\pm1\) and undefined at \(x=0\) (not in the domain). Intervals: \(x\lt-1\), \(-1\lt x\lt0\), \(0\lt x\lt1\), \(x\gt1\). Signs of \(h'\): \(+,-,-,+\). So \(h\) is increasing on \((-\infty,-1)\) and \((1,\infty)\), and decreasing on \((-1,0)\) and \((0,1)\). Do not write \((-1,1)\) as one interval: \(h\) is not defined at 0.
Reading a derivative's graph or table
Often \(f\) itself is unknown and you are given the graph of \(f'\), or a table of values of \(f'\). The sign of \(f'\) still tells you everything: \(f\) increases where the graph of \(f'\) is above the \(x\)-axis and decreases where it is below.
Worked example · From a table of \(f'\)
For a differentiable function \(f\), values of \(f'\) are given: \(f'(-2)=3\), \(f'(0)=-1\), \(f'(1)=-4\), \(f'(3)=2\). If \(f'\) is continuous and has exactly one zero on each of \((-2,0)\) and \((1,3)\), what can you say about \(f\)?
By continuity, \(f'\) changes sign from \(+\) to \(-\) in \((-2,0)\) and from \(-\) to \(+\) in \((1,3)\). So \(f\) increases, then decreases, then increases again; with only these zeros, the sign pattern is \(+,-,+\).
Exam focus
\(f'\gt0\) on an interval \(\Rightarrow\) \(f\) increasing there; \(f'\lt0\) \(\Rightarrow\) decreasing.
Break the domain at zeros of \(f'\), points where \(f'\) is undefined, and points outside the domain.
Always justify with the sign of \(f'\), not with the picture of \(f\).
Do not merge intervals across a point not in the domain.
Endpoints may be included or excluded in the answer; stay consistent (open intervals are accepted).
Practice
Find where \(f(x)=x^3-12x\) is increasing and decreasing.
Show answer
\(f'(x)=3x^2-12=3(x-2)(x+2)\). Signs: \(+\) on \((-\infty,-2)\), \(-\) on \((-2,2)\), \(+\) on \((2,\infty)\). Increasing on \((-\infty,-2)\) and \((2,\infty)\); decreasing on \((-2,2)\).
Find where \(f(x)=\dfrac{x}{x^2+1}\) is increasing.
Show answer
\(f'(x)=\dfrac{1-x^2}{(x^2+1)^2}\), positive when \(-1\lt x\lt1\). So \(f\) is increasing on \((-1,1)\) and decreasing on \((-\infty,-1)\) and \((1,\infty)\).
Find where \(f(x)=x\ln x\) (for \(x\gt0\)) is decreasing.
Show answer
\(f'(x)=\ln x+1\), which is negative when \(\ln x\lt-1\), that is \(0\lt x\lt e^{-1}\). \(f\) is decreasing on \((0,1/e)\) and increasing on \((1/e,\infty)\).
The derivative of \(f\) is \(f'(x)=(x+3)(x-2)^2\). On which intervals is \(f\) increasing?
Show answer
\((x-2)^2\ge0\) with zero at \(x=2\). So the sign of \(f'\) matches the sign of \(x+3\) except at \(x=2\): negative for \(x\lt-3\), positive for \(x\gt-3\). \(f\) is increasing on \((-3,\infty)\) (including through \(x=2\), where \(f'=0\) but does not change sign), and decreasing on \((-\infty,-3)\).
A student says "\(h'(x)=0\) at \(x=5\), so \(h\) switches from increasing to decreasing there." What is missing?
Show answer
A zero of \(h'\) is only a candidate. The student must show that \(h'\) changes sign at \(x=5\); for example \(h(x)=(x-5)^3\) has \(h'(5)=0\) but keeps increasing.
The graph of \(f'\) lies above the \(x\)-axis on \((-4,-1)\) and \((2,6)\) and below it on \((-1,2)\). On what intervals is \(f\) decreasing? Justify.
Show answer
\(f\) is decreasing on \((-1,2)\) because \(f'\lt0\) there.
Now we turn the sign chart of \(f'\) into a way of finding and classifying extrema. A relative maximum is where the function stops rising and starts falling; a relative minimum is the reverse. The first derivative shows exactly that switch.
Required content · FUN-4.A.2
Learning objective.
FUN-4.A. Justify conclusions about the behavior of a function based on the behavior of its derivatives.
Essential knowledge.
FUN-4.A.2. The first derivative of a function can determine the location of relative (local) extrema of the function.
The first derivative test
Let \(c\) be a critical point of \(f\) (so \(f'(c)=0\) or \(f'(c)\) does not exist), with \(f\) continuous at \(c\).
If \(f'\) changes from positive to negative at \(c\), then \(f\) has a relative maximum at \(c\).
If \(f'\) changes from negative to positive at \(c\), then \(f\) has a relative minimum at \(c\).
If \(f'\) does not change sign at \(c\), then \(f\) has no relative extremum at \(c\).
Why it works: if \(f'\gt0\) just left of \(c\), \(f\) is increasing up to \(c\), and if \(f'\lt0\) just right of \(c\), it decreases after \(c\); so \(f(c)\) is at least as large as every nearby value. Endpoints of a domain interval are handled separately: a relative extremum there is determined by the sign of \(f'\) on one side only.
The exam asks you to justify the classification in a sentence: "\(f\) has a relative maximum at \(x=c\) because \(f'\) changes from positive to negative at \(x=c\)." The statement must include the sign change of \(f'\), not just the word "increasing."
Worked examples
Worked example · A cubic, completely classified
Classify the critical points of \(f(x)=x^3-6x^2+9x+1\).
\(f'(x)=3(x-1)(x-3)\). From Lesson 5.3: \(f'\gt0\) on \((-\infty,1)\), \(f'\lt0\) on \((1,3)\), \(f'\gt0\) on \((3,\infty)\).
At \(x=1\), \(f'\) changes from \(+\) to \(-\): relative maximum, value \(f(1)=1-6+9+1=5\). At \(x=3\), \(f'\) changes from \(-\) to \(+\): relative minimum, value \(f(3)=27-54+27+1=1\).
Worked example · A critical point where \(f'\) is undefined
Classify the critical point of \(f(x)=x^{2/3}\).
\(f'(x)=\dfrac{2}{3\sqrt[3]{x}}\) is negative for \(x\lt0\), positive for \(x\gt0\), and undefined at \(x=0\) (where \(f\) is continuous). \(f'\) changes from negative to positive at \(x=0\), so \(f\) has a relative minimum at \(x=0\), namely \(f(0)=0\). The graph comes to a sharp point (a cusp) there.
Worked example · A critical point that is not an extremum
\(f'(x)=(x+2)(x-1)^2\). Critical points: \(x=-2\) and \(x=1\). The factor \((x-1)^2\) is never negative, so the sign of \(f'\) follows \(x+2\): negative for \(x\lt-2\), positive for \(x\gt-2\) (except zero at \(x=1\)). At \(x=-2\), \(f'\) changes from negative to positive: relative minimum. At \(x=1\), \(f'\) is positive on both sides, so \(f\) has no relative extremum at \(x=1\); its graph flattens and keeps rising.
Worked example · A logarithmic function
Find the relative extremum of \(g(x)=x\ln x\) on \(x\gt0\).
\(g'(x)=\ln x+1=0\) when \(x=e^{-1}\). For \(0\lt x\lt e^{-1}\), \(\ln x\lt-1\) so \(g'\lt0\); for \(x\gt e^{-1}\), \(g'\gt0\). \(g'\) changes from negative to positive, so \(g\) has a relative minimum at \(x=1/e\), with value \(g(1/e)=\dfrac1e\ln\dfrac1e=-\dfrac1e\approx-0.368\). Since this is the only critical point on the whole domain and \(g\) decreases before and increases after, it is also the global minimum.
From the graph of \(f'\)
When you are given the graph of \(f'\) rather than \(f\), the relative extrema of \(f\) appear where the graph of \(f'\) crosses the \(x\)-axis. Crossing from above to below means relative maximum for \(f\); from below to above means relative minimum. Touching the axis without crossing is not an extremum for \(f\).
Exam focus
Classify critical points by the sign change of \(f'\): \(+\to-\) is a maximum, \(-\to+\) is a minimum, no change is neither.
Require \(f\) continuous at the critical point (a cusp is fine, a jump is not).
Write the justification with the sign change of \(f'\), and give the \(y\)-value if asked for the extreme value.
On a graph of \(f'\), extrema of \(f\) are at the \(x\)-intercepts where the graph crosses.
Check domain endpoints separately.
Practice
Use the first derivative test to classify the critical points of \(f(x)=x^3-3x\).
Show answer
\(f'(x)=3(x-1)(x+1)\): \(+\) on \((-\infty,-1)\), \(-\) on \((-1,1)\), \(+\) on \((1,\infty)\). Relative maximum at \(x=-1\) (value 2) because \(f'\) changes from \(+\) to \(-\); relative minimum at \(x=1\) (value \(-2\)) because \(f'\) changes from \(-\) to \(+\).
Find and classify the extrema of \(f(x)=x^4-4x^3\).
Show answer
\(f'(x)=4x^3-12x^2=4x^2(x-3)\). Critical points: 0 and 3. \(f'\lt0\) on both \((-\infty,0)\) and \((0,3)\), so there is no sign change at 0: no extremum. \(f'\) changes from \(-\) to \(+\) at \(x=3\): relative minimum, \(f(3)=81-108=-27\).
Classify the critical points of \(f(x)=\dfrac{x^2}{x-2}\) (domain \(x\ne2\)).
Show answer
\(f'(x)=\dfrac{2x(x-2)-x^2}{(x-2)^2}=\dfrac{x^2-4x}{(x-2)^2}=\dfrac{x(x-4)}{(x-2)^2}\). Critical points \(x=0\) and \(x=4\) (\(x=2\) is not in the domain). Signs: \(+\) for \(x\lt0\), \(-\) on \((0,2)\) and \((2,4)\), \(+\) for \(x\gt4\). Relative maximum at \(x=0\) (\(f(0)=0\)); relative minimum at \(x=4\) (\(f(4)=8\)).
The graph of \(f'\) crosses the \(x\)-axis at \(x=-3\) going upward, touches the axis at \(x=1\) without crossing, and crosses at \(x=4\) going downward. Describe the extrema of \(f\).
Show answer
Relative minimum at \(x=-3\) (\(f'\) changes \(-\) to \(+\)); no extremum at \(x=1\); relative maximum at \(x=4\) (\(f'\) changes \(+\) to \(-\)).
For \(f(x)=xe^{-x}\), find the relative extremum and justify.
Show answer
\(f'(x)=e^{-x}(1-x)\), zero only at \(x=1\); \(f'\gt0\) for \(x\lt1\), \(f'\lt0\) for \(x\gt1\). Relative maximum at \(x=1\) because \(f'\) changes from positive to negative; \(f(1)=1/e\approx0.368\).
A student says "\(f'(3)=0\), and \(f'\gt0\) on both sides, so \(f\) has a relative maximum at \(x=3\)." Correct the mistake.
Show answer
A maximum needs \(f'\) to change from positive to negative. If \(f'\) is positive on both sides, \(f\) is increasing through \(x=3\), so there is no relative extremum there.
Relative extrema answer "highest nearby." Real problems ask "highest of all": the largest profit, the lowest temperature, the maximum height. This lesson gives the dependable method for finding global extrema.
Required content · FUN-4.A.3
Learning objective.
FUN-4.A. Justify conclusions about the behavior of a function based on the behavior of its derivatives.
Essential knowledge.
FUN-4.A.3. Absolute (global) extrema of a function on a closed interval can only occur at critical points or at endpoints.
The candidates test
On a closed interval \([a,b]\) where \(f\) is continuous, the Extreme Value Theorem guarantees a global maximum and minimum, and each occurs either at a critical point in \((a,b)\) or at an endpoint. So:
Find all critical points of \(f\) in \((a,b)\).
Evaluate \(f\) at each critical point and at both endpoints.
The largest value is the global maximum; the smallest is the global minimum.
The list of numbers where you evaluate \(f\) is the list of candidates. The test does not require you to decide which critical points are local maxima or minima.
Worked example · A cubic on a closed interval
Find the global extrema of \(f(x)=x^3-3x\) on \([-2,3]\).
\(f'(x)=3x^2-3=3(x-1)(x+1)\), so the critical points are \(x=\pm1\), both in \([-2,3]\).
\(x\)
\(-2\)
\(-1\)
\(1\)
\(3\)
\(f(x)\)
\(-2\)
\(2\)
\(-2\)
\(18\)
The global maximum is \(18\), at the endpoint \(x=3\). The global minimum is \(-2\), attained at \(x=-2\) and at \(x=1\) (a tie). Notice that the relative maximum at \(x=-1\) (value 2) is not the global maximum.
Worked example · A rational function
Find the global extrema of \(f(x)=\dfrac{x}{x^2+1}\) on \([0,3]\).
\(f'(x)=\dfrac{1-x^2}{(x^2+1)^2}=0\) at \(x=\pm1\); only \(x=1\) is in the interval. Candidates: \(f(0)=0\), \(f(1)=\dfrac12\), \(f(3)=\dfrac{3}{10}\). Global maximum \(\dfrac12\) at \(x=1\); global minimum \(0\) at \(x=0\).
When the interval is not closed
If the interval is open or unbounded, the EVT does not apply and extrema might not exist. If there is a single critical point on the interval, the first derivative test often settles it.
Rule. Suppose \(f\) is continuous on an interval and has exactly one critical point \(c\) in it. If \(f'\) changes from negative to positive at \(c\), \(f(c)\) is the global minimum on the interval; if from positive to negative, \(f(c)\) is the global maximum.
Why: for \(x\lt c\) the function is decreasing toward \(f(c)\), and for \(x\gt c\) it increases away from \(f(c)\); so every other value is larger.
Worked example · An open, unbounded interval
Find the global minimum of \(f(x)=x^2+\dfrac{16}{x}\) on \((0,\infty)\).
\(f'(x)=2x-\dfrac{16}{x^2}=\dfrac{2x^3-16}{x^2}=0\) at \(x^3=8\), \(x=2\), the only critical point. For \(0\lt x\lt2\), \(f'\lt0\); for \(x\gt2\), \(f'\gt0\). So \(f\) has its global minimum on \((0,\infty)\) at \(x=2\), with \(f(2)=4+8=12\). There is no global maximum, since \(f(x)\to\infty\) as \(x\to0^+\) and as \(x\to\infty\).
Worked example · Value and location
A company's profit is \(P(x)=-x^3+9x^2-15x+40\) thousand dollars when it produces \(x\) hundred units, \(0\le x\le7\). Find the production level that gives the greatest profit, and the profit.
\(P'(x)=-3x^2+18x-15=-3(x-1)(x-5)\). Critical points \(x=1,5\). \(P(0)=40\), \(P(1)=-1+9-15+40=33\), \(P(5)=-125+225-75+40=65\), \(P(7)=-343+441-105+40=33\). The greatest profit is 65 thousand dollars, at a production level of 5 hundred units (500 units).
Exam focus
Closed interval, continuous \(f\): evaluate at critical points and endpoints; largest is the global maximum, smallest the global minimum.
Give both the location \(x\) and the value \(f(x)\); read the question for which one is asked.
On an open or unbounded interval with one critical point, use the sign change of \(f'\) to justify a global extremum.
State the justification: "candidates are ..., compare values."
Do not stop at a relative extremum when the question says global.
Practice
Find the global maximum and minimum of \(f(x)=x^3-6x^2+9x\) on \([0,5]\).
Show answer
\(f'(x)=3(x-1)(x-3)\), critical points 1 and 3. Values: \(f(0)=0\), \(f(1)=4\), \(f(3)=0\), \(f(5)=125-150+45=20\). Global maximum 20 at \(x=5\); global minimum 0 at \(x=0\) and \(x=3\).
Find the global extrema of \(f(x)=x^2-4x+3\) on \([0,5]\).
Show answer
\(f'(x)=2x-4=0\) at \(x=2\). \(f(0)=3\), \(f(2)=-1\), \(f(5)=8\). Global maximum 8 at \(x=5\); global minimum \(-1\) at \(x=2\).
Find the global maximum of \(f(x)=xe^{-x}\) on \([0,4]\).
Show answer
\(f'(x)=e^{-x}(1-x)=0\) at \(x=1\). \(f(0)=0\), \(f(1)=e^{-1}\approx0.368\), \(f(4)=4e^{-4}\approx0.073\). Global maximum \(1/e\) at \(x=1\).
Find the global minimum of \(f(x)=x+\dfrac9x\) on \((0,\infty)\) and justify.
Show answer
\(f'(x)=1-\dfrac9{x^2}=0\) at \(x=3\) (only positive critical point). \(f'\lt0\) on \((0,3)\) and \(f'\gt0\) on \((3,\infty)\), so the global minimum is \(f(3)=3+3=6\).
Explain why the candidates test works for \(f(x)=\sqrt x\) on \([0,4]\) even though \(f'(0)\) is undefined.
Show answer
\(f\) is continuous on \([0,4]\), so the EVT applies. There is no critical point inside \((0,4)\) (since \(f'=\dfrac1{2\sqrt x}\gt0\)). The candidates are the endpoints: \(f(0)=0\) (global minimum), \(f(4)=2\) (global maximum).
A student finds that \(f(x)=x^3-3x\) has a relative maximum at \(x=-1\) and says that is the global maximum on \([-2,3]\). What did the student omit, and what is the correct answer?
Show answer
The student did not compare with the endpoint values. \(f(3)=18\gt f(-1)=2\), so the global maximum is 18 at \(x=3\).
Determining concavity of functions over their domains
Two functions can both be increasing and look very different: one climbs faster and faster, the other climbs slower and slower. Concavity describes that difference, and it is controlled by the second derivative.
FUN-4.A. Justify conclusions about the behavior of a function based on the behavior of its derivatives.
Essential knowledge.
FUN-4.A.4. The graph of a function is concave up (down) on an open interval if the function's derivative is increasing (decreasing) on that interval.
FUN-4.A.5. The second derivative of a function provides information about the function and its graph, including intervals of upward or downward concavity.
FUN-4.A.6. The second derivative of a function may be used to locate points of inflection for the graph of the original function.
Concavity
The graph of \(f\) is concave up on an interval if \(f'\) is increasing there (the slopes of the tangent lines get larger as \(x\) increases; the graph lies above its tangent lines). It is concave down if \(f'\) is decreasing (the slopes shrink; the graph lies below its tangent lines).
Figure 5.4. On a concave up interval the slopes of the tangent lines increase from left to right, and the graph lies above its tangent lines. On a concave down interval the slopes decrease, and the graph lies below them.
Since \(f'\) increasing or decreasing is read from the sign of \(f''=(f')'\):
If \(f''(x)\gt0\) on an interval, \(f\) is concave up there.
If \(f''(x)\lt0\) on an interval, \(f\) is concave down there.
The second derivative can change sign only where \(f''(x)=0\) or \(f''(x)\) does not exist (or where \(f\) is not defined). So the procedure parallels Lesson 5.3: find these numbers, split the domain, and read the sign of \(f''\) on each piece.
Points of inflection
A point of inflection is a point on the graph where the concavity changes (and \(f\) is defined, with a tangent line, possibly vertical). At such a point \(f''\) changes sign. \(f''(c)=0\) alone does not guarantee an inflection point: you must check that the sign of \(f''\) changes.
Worked example · A cubic
Find the intervals of concavity and the inflection points of \(f(x)=x^3-3x^2\).
\(f'(x)=3x^2-6x\), \(f''(x)=6x-6=6(x-1)\). \(f''\lt0\) for \(x\lt1\) and \(f''\gt0\) for \(x\gt1\). The graph is concave down on \((-\infty,1)\) and concave up on \((1,\infty)\). Concavity changes at \(x=1\), so there is a point of inflection at \((1,f(1))=(1,-2)\).
Worked example · \(f''=0\) without an inflection
\(f(x)=x^4\): \(f''(x)=12x^2\) is zero at \(x=0\), but \(f''\ge0\) on both sides, so the graph is concave up on both sides. There is no inflection point at \(x=0\).
Worked example · \(f''\) undefined at an inflection point
\(f(x)=\sqrt[3]{x}=x^{1/3}\): \(f'(x)=\tfrac13x^{-2/3}\), \(f''(x)=-\tfrac29x^{-5/3}\), undefined at \(x=0\). For \(x\lt0\), \(f''\gt0\) (concave up); for \(x\gt0\), \(f''\lt0\) (concave down). The graph has a vertical tangent at the origin, and concavity changes there: a point of inflection at \((0,0)\).
Worked example · A product with \(e^x\)
Find the inflection point of \(f(x)=xe^x\).
\(f'(x)=e^x+xe^x=(x+1)e^x\), \(f''(x)=e^x+(x+1)e^x=(x+2)e^x\). Since \(e^x\gt0\), \(f''\) has the sign of \(x+2\): negative for \(x\lt-2\), positive for \(x\gt-2\). The inflection point is at \(x=-2\), \(f(-2)=-2e^{-2}\approx-0.271\).
Concavity from the graph of \(f'\)
If you are given the graph of \(f'\), the concavity of \(f\) is the direction of \(f'\): where the graph of \(f'\) rises, \(f\) is concave up; where it falls, \(f\) is concave down. Inflection points of \(f\) are located at the local extrema of \(f'\) (where its graph turns from rising to falling or the reverse).
Worked example · Table of \(f'\)
A twice differentiable function has the values \(f'(1)=-2\), \(f'(2)=1\), \(f'(3)=3\), \(f'(4)=2\), \(f'(5)=-1\). On which intervals does the table suggest \(f\) is concave up?
\(f'\) increases from 1 to 3 (values \(-2,1,3\)) and decreases from 3 to 5 (values \(3,2,-1\)). The data are consistent with \(f\) concave up on \((1,3)\) and concave down on \((3,5)\), with a possible inflection point near \(x=3\), but a table alone cannot prove it.
Concave up means \(f'\) is increasing and the graph lies above its tangents.
An inflection point needs a sign change of \(f''\); \(f''(c)=0\) is not enough.
Split the domain at zeros and undefined points of \(f''\) and at points outside the domain.
Inflection points of \(f\) correspond to relative extrema of \(f'\).
Practice
Find the intervals of concavity of \(f(x)=x^3-12x\) and any inflection points.
Show answer
\(f''(x)=6x\). Concave down on \((-\infty,0)\), concave up on \((0,\infty)\); inflection point at \((0,0)\).
Find the inflection points of \(f(x)=x^4-6x^2\).
Show answer
\(f''(x)=12x^2-12=12(x-1)(x+1)\). Signs: \(+\) for \(x\lt-1\), \(-\) on \((-1,1)\), \(+\) for \(x\gt1\). Inflection points at \(x=-1\) and \(x=1\); \(f(\pm1)=-5\), so the points are \((-1,-5)\) and \((1,-5)\).
Does \(f(x)=x^5\) have an inflection point at the origin? Justify.
Show answer
\(f''(x)=20x^3\), negative for \(x\lt0\), positive for \(x\gt0\), so \(f''\) changes sign at 0. Yes, \((0,0)\) is an inflection point.
On which intervals is \(g(x)=e^{-x^2}\) concave up?
Show answer
\(g'(x)=-2xe^{-x^2}\), \(g''(x)=(4x^2-2)e^{-x^2}\), which is positive when \(x^2\gt\tfrac12\). Concave up on \((-\infty,-\tfrac1{\sqrt2})\) and \((\tfrac1{\sqrt2},\infty)\); concave down on \((-\tfrac1{\sqrt2},\tfrac1{\sqrt2})\).
The graph of \(f'\) is increasing on \((-3,1)\) and decreasing on \((1,5)\). Describe the concavity of \(f\) and the location of any inflection point.
Show answer
\(f\) is concave up on \((-3,1)\) (since \(f'\) is increasing) and concave down on \((1,5)\). The concavity changes at \(x=1\), so \(f\) has an inflection point at \(x=1\).
A student says "\(f''(2)=0\), so \(f\) has an inflection point at \(x=2\)." Give a counterexample and fix the claim.
Show answer
\(f(x)=(x-2)^4\) has \(f''(2)=0\) but \(f''\ge0\) everywhere, so there is no inflection point. The correct claim requires \(f''\) to change sign at \(x=2\).
Using the second derivative test to determine extrema
At a critical point where the tangent is horizontal, the graph is either at the top of a hill (concave down) or the bottom of a valley (concave up). Concavity at the critical point therefore classifies it, and you only need to evaluate \(f''\) at that one point instead of building a full sign chart.
Required content · FUN-4.A.7, FUN-4.A.8
Learning objective.
FUN-4.A. Justify conclusions about the behavior of a function based on the behavior of its derivatives.
Essential knowledge.
FUN-4.A.7. The second derivative of a function may determine whether a critical point is the location of a relative (local) maximum or minimum.
FUN-4.A.8. When a continuous function has only one critical point on an interval on its domain and the critical point corresponds to a relative (local) extremum of the function on the interval, then that critical point also corresponds to the absolute (global) extremum of the function on the interval.
The second derivative test
Suppose \(f'(c)=0\) and \(f''\) exists near \(c\).
If \(f''(c)\lt0\), then \(f\) has a relative maximum at \(x=c\) (concave down at a flat point).
If \(f''(c)\gt0\), then \(f\) has a relative minimum at \(x=c\) (concave up at a flat point).
If \(f''(c)=0\), the test is inconclusive: \(f\) may have a maximum, a minimum, or neither at \(c\). Use the first derivative test.
Limits of the test: it applies only where \(f'(c)=0\), so it cannot classify a critical point where \(f'\) is undefined.
Worked example · A cubic
Use the second derivative test on \(f(x)=x^3-3x^2\).
\(f'(x)=3x^2-6x=3x(x-2)\), so the critical points are \(x=0\) and \(x=2\). \(f''(x)=6x-6\). \(f''(0)=-6\lt0\): relative maximum at \(x=0\), value \(f(0)=0\). \(f''(2)=6\gt0\): relative minimum at \(x=2\), value \(f(2)=-4\). (These match Figure 5.3 of the next lesson.)
Worked example · Another cubic
Classify the critical points of \(f(x)=2x^3-9x^2+12x\).
\(f(x)=x^4\) has \(f'(0)=0\) and \(f''(0)=0\), so the second derivative test says nothing. The first derivative test shows \(f'=4x^3\) changes from negative to positive at 0: a relative minimum. For \(g(x)=x^3\), \(g'(0)=0\), \(g''(0)=0\) as well, yet \(g'\) does not change sign and there is no extremum. The same inconclusive value of \(f''\) gives different answers.
Worked example · A rational function on \((0,\infty)\)
Classify the critical point of \(f(x)=x+\dfrac4x\) for \(x\gt0\).
\(f'(x)=1-\dfrac4{x^2}=0\) at \(x=2\). \(f''(x)=\dfrac{8}{x^3}\), so \(f''(2)=1\gt0\). Relative minimum at \(x=2\), value \(f(2)=4\). Since \(f''\gt0\) on the entire domain, the graph is concave up everywhere, and this is also the global minimum on \((0,\infty)\).
Choosing between the tests
Use the second derivative test when \(f''\) is easy to compute and nonzero at the critical points. Use the first derivative test when \(f''(c)=0\), when \(f'(c)\) is undefined, or when you are given only information about the sign of \(f'\). In either case, the justification you write must name the test and the facts it uses: "\(f'(2)=0\) and \(f''(2)=6\gt0\), so \(f\) has a relative minimum at \(x=2\)."
Exam focus
At a critical point with \(f'(c)=0\): \(f''(c)\lt0\) gives a maximum, \(f''(c)\gt0\) gives a minimum.
\(f''(c)=0\) means the test fails; fall back on the sign of \(f'\).
The test cannot be used where \(f'(c)\) does not exist.
A function that is concave up on its whole domain has a global minimum at any flat point.
Justify with both \(f'(c)=0\) and the sign of \(f''(c)\).
Practice
Use the second derivative test to classify the critical points of \(f(x)=x^3-12x\).
Show answer
\(f'(x)=3x^2-12=0\) at \(x=\pm2\); \(f''(x)=6x\). \(f''(-2)=-12\lt0\): relative maximum at \(x=-2\) (\(f=16\)). \(f''(2)=12\gt0\): relative minimum at \(x=2\) (\(f=-16\)).
Classify the critical points of \(f(x)=x^4-8x^2\) with the second derivative test where possible.
Show answer
For \(f(x)=x^5\) the second derivative test is inconclusive at \(x=0\). Verify that, and decide what \(f\) does at 0.
Show answer
\(f'(0)=0\) and \(f''(x)=20x^3\) gives \(f''(0)=0\): inconclusive. \(f'(x)=5x^4\ge0\) with no sign change, so there is no extremum at 0.
For \(g(x)=xe^{-x}\), use the second derivative test at the critical point.
Show answer
\(g'(x)=e^{-x}(1-x)=0\) at \(x=1\). \(g''(x)=-e^{-x}(1-x)-e^{-x}=e^{-x}(x-2)\). \(g''(1)=-e^{-1}\lt0\): relative maximum at \(x=1\), value \(1/e\).
Why can't the second derivative test classify the critical point of \(f(x)=\lvert x\rvert\) at \(x=0\)?
Show answer
\(f'(0)\) does not exist, so \(f'(0)\ne0\); the test requires a horizontal tangent. Use the first derivative test: \(f'\) changes from \(-1\) to \(1\), so \(x=0\) is a relative minimum.
A function has \(f'(4)=0\) and \(f''(4)=0\). A student concludes there is no extremum at \(x=4\). Is that right?
Show answer
No. The second derivative test is inconclusive when \(f''(c)=0\). Examples: \(f(x)=(x-4)^4\) has a minimum at 4, \(f(x)=-(x-4)^4\) a maximum, and \(f(x)=(x-4)^3\) neither. The first derivative test is needed.
Sketching graphs of functions and their derivatives
The information in \(f\), \(f'\) and \(f''\) is tightly connected. If you know one graph, you can sketch the others. This lesson is a translation guide between a function and its derivatives, in both directions.
Required content · FUN-4.A.9, FUN-4.A.10
Learning objective.
FUN-4.A. Justify conclusions about the behavior of a function based on the behavior of its derivatives.
Essential knowledge.
FUN-4.A.9. Key features of functions and their derivatives can be identified and related to their graphical, numerical, and analytical representations.
FUN-4.A.10. Graphical, numerical, and analytical information from \(f'\) and \(f''\) can be used to predict and explain the behavior of \(f\).
From \(f\) to \(f'\) and \(f''\)
The graph of \(f'\) records the slope of \(f\). Look at \(f\) and ask four questions at each \(x\):
Feature of \(f\)
Graph of \(f'\)
\(f\) increasing
\(f'\) is above the \(x\)-axis (positive)
\(f\) decreasing
\(f'\) is below the \(x\)-axis (negative)
relative maximum or minimum of \(f\) (smooth)
\(f'\) crosses the \(x\)-axis (zero)
\(f\) steepest or flattest (inflection point)
\(f'\) has a relative maximum or minimum
\(f\) concave up
\(f'\) is increasing
\(f\) concave down
\(f'\) is decreasing
corner or cusp of \(f\)
\(f'\) has a jump or is undefined
vertical tangent
\(f'\) is unbounded there
Figure 5.3. \(f(x)=x^3-3x^2\) with \(f'(x)=3x(x-2)\) and \(f''(x)=6(x-1)\). \(f\) has a relative maximum at \(x=0\) (where \(f'\) goes from positive to negative) and a relative minimum at \(x=2\) (where \(f'\) goes from negative to positive). \(f\) changes concavity at \(x=1\), where \(f''\) changes sign and \(f'\) has its minimum.
The figure shows \(f(x)=x^3-3x^2\) and its derivatives. The graph of \(f\) rises to the relative maximum at \(x=0\) and falls to the relative minimum at \(x=2\); \(f'\) is positive, zero, negative, zero, positive in that order. \(f\) switches from concave down to concave up at \(x=1\), exactly where \(f'\) is at its lowest and \(f''\) crosses zero.
From \(f'\) to \(f\)
Reverse the dictionary. The graph of \(f'\) tells you the direction of \(f\) but not its vertical position: \(f\) is determined only up to adding a constant, so you need a known point of \(f\) (an initial value) to place the graph.
Worked example · Sketching from the derivative's behavior
The function \(f\) is continuous with \(f(0)=2\). Its derivative satisfies: \(f'(x)\gt0\) for \(x\lt-1\); \(f'(-1)=0\); \(f'(x)\lt0\) for \(-1\lt x\lt3\); \(f'(3)=0\); \(f'(x)\gt0\) for \(x\gt3\); \(f'\) has a relative minimum at \(x=1\). Describe the graph of \(f\).
\(f'\) changes from positive to negative at \(x=-1\): relative maximum of \(f\). \(f'\) changes from negative to positive at \(x=3\): relative minimum. \(f'\) has a relative minimum at \(x=1\), so \(f'\) is decreasing before 1 and increasing after: \(f\) is concave down on \((-\infty,1)\) and concave up on \((1,\infty)\) (assuming it only reaches its minimum there), and \(x=1\) is an inflection point. The graph passes through \((0,2)\), rising to a peak at \(x=-1\), falling steeply through the inflection at \(x=1\), bottoming out at \(x=3\), and then rising.
Worked example · A corner
Let \(f(x)=\lvert x^2-1\rvert\). Describe \(f'\).
\(f(x)=x^2-1\) for \(\lvert x\rvert\ge1\) and \(f(x)=1-x^2\) for \(\lvert x\rvert\lt1\). So \(f'(x)=2x\) for \(\lvert x\rvert\gt1\) and \(f'(x)=-2x\) for \(\lvert x\rvert\lt1\). At \(x=\pm1\), \(f\) has corners, and the one-sided derivatives are different (\(-2\) and \(2\) at \(x=-1\); \(-2\) and \(2\) at \(x=1\)), so \(f'\) does not exist at \(\pm1\): its graph has jumps. \(f'(0)=0\), where \(f\) has a relative maximum.
Common sketch-checking questions
On the exam you are often given the graph of \(f'\) and asked about \(f\): intervals of increase, relative extrema, concavity, inflection points. Work from the dictionary and always answer with the reason ("because \(f'\) changes from positive to negative at ..." or "because \(f'\) is increasing on ...").
Exam focus
\(f'\)'s sign gives the direction of \(f\); \(f'\)'s zeros (with sign change) give extrema of \(f\).
\(f'\)'s own increasing or decreasing gives the concavity of \(f\).
Extrema of \(f'\) give inflection points of \(f\).
Going from \(f'\) back to \(f\), you need an initial value to fix the vertical position.
Corners, cusps and vertical tangents in \(f\) show up as jumps or breaks in \(f'\).
Practice
For \(f(x)=x^2\), describe the graphs of \(f'\) and \(f''\).
Show answer
\(f'(x)=2x\) is a line through the origin, negative for \(x\lt0\) and positive for \(x\gt0\), matching the fall and rise of the parabola. \(f''(x)=2\) is a positive constant, so \(f\) is concave up everywhere.
The graph of \(f\) has a relative maximum at \(x=2\) and a relative minimum at \(x=5\), both smooth. What is true of \(f'\) at those \(x\)-values?
Show answer
\(f'(2)=0\) and \(f'(5)=0\); \(f'\) crosses the \(x\)-axis from positive to negative at \(x=2\) and from negative to positive at \(x=5\).
The graph of \(f\) is increasing and concave down on \((0,4)\). What do you know about \(f'\) on that interval?
Show answer
\(f'\gt0\) (because \(f\) is increasing) and \(f'\) is decreasing (because \(f\) is concave down). The graph of \(f'\) is positive and sloping downward.
The graph of \(f'\) has a relative maximum at \(x=-2\) and a relative minimum at \(x=3\). What does this say about \(f\)?
Show answer
\(f'\) changes from increasing to decreasing at \(x=-2\) and from decreasing to increasing at \(x=3\), so \(f''\) changes sign at both: \(f\) has inflection points at \(x=-2\) and \(x=3\).
Is it possible for \(f\) to have two different graphs with the same derivative? Explain.
Show answer
Yes. If \(f'=g'\) then \(f-g\) is constant. For example \(f(x)=x^2\) and \(g(x)=x^2+5\) have the same derivative \(2x\), and the graphs are vertical translations. An initial value fixes which one is meant.
\(f(x)=\lvert x\rvert\). Sketch or describe \(f'\) and explain what happens at \(x=0\).
Show answer
\(f'(x)=-1\) for \(x\lt0\) and \(1\) for \(x\gt0\): two horizontal rays at different heights. At \(x=0\) the graph of \(f\) has a corner, so \(f'(0)\) does not exist.
This lesson puts everything in Unit 5 together. A typical exam question gives you one of \(f\), \(f'\) or \(f''\) (as a formula, a graph, or a table) and asks for the extrema, intervals of increase, concavity, and inflection points of the function, with justification. The graphs of \(f\), \(f'\) and \(f''\) are connected: the sign and behavior of one determines properties of the previous.
Required content · FUN-4.A.11
Learning objective.
FUN-4.A. Justify conclusions about the behavior of a function based on the behavior of its derivatives.
Essential knowledge.
FUN-4.A.11. Key features of the graphs of \(f\), \(f'\), and \(f''\) are related to one another.
The summary
If
then \(f\) is
\(f'\gt0\)
increasing
\(f'\lt0\)
decreasing
\(f'\) changes \(+\to-\) at \(c\)
relative maximum at \(c\)
\(f'\) changes \(-\to+\) at \(c\)
relative minimum at \(c\)
\(f''\gt0\) (\(f'\) increasing)
concave up
\(f''\lt0\) (\(f'\) decreasing)
concave down
\(f''\) changes sign at \(c\) (\(f'\) has an extremum at \(c\))
point of inflection at \(c\)
A key point: the extrema of \(f'\) are the inflection points of \(f\), and the extrema of \(f\) are the zeros (sign changes) of \(f'\). A relative extremum of \(f'\) at \(x=c\) means \(f''\) changes sign there.
Worked example · Everything from \(f'\)
The function \(f\) is continuous with \(f'(x)=x^2(x-3)\). Find the intervals where \(f\) is increasing, the relative extrema, the intervals of concavity and the inflection points.
Increasing. \(x^2\ge0\), so the sign of \(f'\) is the sign of \(x-3\) (except zero at \(x=0\)). \(f'\lt0\) for \(x\lt3\) (\(x\ne0\)) and \(f'\gt0\) for \(x\gt3\). \(f\) is increasing on \((3,\infty)\) and decreasing on \((-\infty,3)\) (including through \(x=0\), where \(f'=0\) but the sign does not change).
Extrema. \(f'\) changes from negative to positive at \(x=3\): relative minimum. At \(x=0\), no sign change, so no extremum.
Concavity. \(f''(x)=3x^2-6x=3x(x-2)\). \(f''\gt0\) for \(x\lt0\) and \(x\gt2\); \(f''\lt0\) on \((0,2)\). Concave up on \((-\infty,0)\) and \((2,\infty)\), concave down on \((0,2)\). Inflection points at \(x=0\) and \(x=2\), since \(f''\) changes sign at each.
(This is the derivative of \(f(x)=\tfrac14x^4-x^3+1\) from Lesson 5.2. Check: \(f(0)=1\), \(f(2)=-3\), \(f(3)=-5.75\).)
Worked example · Starting from a graph or table
The table gives values of a twice differentiable function's derivative.
\(x\)
0
1
2
3
4
5
\(f'(x)\)
\(-3\)
\(-1\)
\(0\)
\(1\)
\(2\)
\(1\)
Suppose \(f'\) is continuous, \(f'(x)\lt0\) for \(x\lt2\) and \(f'(x)\gt0\) for \(x\gt2\) on \((0,5)\). Then \(f\) has a relative minimum at \(x=2\) because \(f'\) changes from negative to positive there. \(f'\) is increasing from \(x=0\) to \(x=4\) and decreasing from 4 to 5, so \(f\) is concave up on \((0,4)\) and concave down on \((4,5)\), suggesting an inflection point near \(x=4\).
Worked example · A function with a parameter
For what value of \(k\) does \(f(x)=x^3+kx^2\) have an inflection point at \(x=2\)?
\(f''(x)=6x+2k=0\) at \(x=2\) gives \(12+2k=0\), \(k=-6\). With \(k=-6\), \(f''(x)=6x-12=6(x-2)\) changes sign at \(x=2\), so it is an inflection point.
Justification language
The College Board expects the reason in terms of the right derivative. Useful templates:
"\(f\) is increasing on \((a,b)\) because \(f'(x)\gt0\) on \((a,b)\)."
"\(f\) has a relative minimum at \(x=c\) because \(f'\) changes from negative to positive at \(x=c\)."
"The graph of \(f\) is concave down on \((a,b)\) because \(f''\lt0\) there." (or "because \(f'\) is decreasing there")
"The graph of \(f\) has a point of inflection at \(x=c\) because \(f''\) changes sign at \(x=c\)."
Exam focus
Move between \(f\), \(f'\) and \(f''\) with the summary table.
Extrema of \(f\) come from sign changes of \(f'\); inflection points of \(f\) come from sign changes of \(f''\) (extrema of \(f'\)).
A zero of \(f'\) or \(f''\) is only a candidate; require the sign change.
Write justifications in terms of the correct derivative.
Tables support conclusions only if sign information is given or can be inferred (with continuity).
Practice
Given \(f'(x)=(x+1)(x-2)\), find the relative extrema and concavity of \(f\).
Show answer
\(f'\) changes \(+\) to \(-\) at \(x=-1\) (relative max) and \(-\) to \(+\) at \(x=2\) (relative min). \(f''(x)=2x-1\): concave down for \(x\lt\tfrac12\), concave up for \(x\gt\tfrac12\), inflection at \(x=\tfrac12\).
The graph of \(f'\) is a parabola opening up with \(x\)-intercepts at 1 and 5. Describe \(f\).
Show answer
\(f'\gt0\) for \(x\lt1\) and \(x\gt5\), \(f'\lt0\) on \((1,5)\). So \(f\) has a relative maximum at \(x=1\) and a relative minimum at \(x=5\). The vertex of the parabola is at \(x=3\) where \(f'\) is at a minimum, so \(f\) has an inflection point at \(x=3\).
Find \(k\) so that \(f(x)=x^3-kx\) has a relative minimum at \(x=2\).
Show answer
\(f'(x)=3x^2-k=0\) at \(x=2\): \(12-k=0\), \(k=12\). Check: \(f''(2)=12\gt0\), a relative minimum.
Given \(f''(x)=6-2x\) and \(f'(3)=0\), classify \(x=3\).
Show answer
\(f''(3)=0\), so the second derivative test fails. But \(f''\gt0\) for \(x\lt3\) means \(f'\) is increasing up to 0 at \(x=3\), so \(f'\lt0\) just before 3; \(f''\lt0\) for \(x\gt3\) means \(f'\) decreases after 3, so \(f'\lt0\) just after as well. No sign change: no extremum at 3, but \(f''\) changes sign, so it is an inflection point.
A twice differentiable \(f\) satisfies \(f'(1)=0\) and \(f''(x)=4-x^2\) on \((-2,2)\). What is \(f\) doing at \(x=1\)?
Show answer
\(f''(1)=3\gt0\), so \(f\) has a relative minimum at \(x=1\) by the second derivative test.
A student looks at a graph of \(f\) and says it has an inflection point because \(f\) changes from decreasing to increasing there. What is the error?
Show answer
That is a relative minimum, which is determined by \(f'\) changing sign. An inflection point is where concavity changes (\(f''\) changes sign).
How do you build the cheapest container, the strongest beam, or the fence that encloses the most area? Optimization problems ask for the greatest or least value of a quantity subject to a constraint. The hard part is not the calculus, which you already have, but turning a story into a function of one variable. This lesson is about that translation.
Required content · FUN-4.B.1
Learning objective.
FUN-4.B. Calculate minimum and maximum values in applied contexts or analysis of functions.
Essential knowledge.
FUN-4.B.1. The derivative can be used to solve optimization problems; that is, finding a minimum or maximum value of a function on a given interval.
The setup
Optimization problems are solved in stages, and the first stage is modeling. The derivative can be used to solve optimization problems; that is, finding a maximum or minimum value in a situation described by a function. Before differentiating:
Draw and label a picture with variables for every changing quantity.
Identify the objective: the quantity to maximize or minimize (area, volume, cost, distance, time).
Identify the constraint: the given fixed quantity (a perimeter, a volume, a budget).
Use the constraint to write the objective as a function of a single variable.
State the domain of the variable: what values make sense in the real situation (lengths are positive; a side cannot exceed what the constraint allows).
Then the extremum is found with the tools of Lessons 5.2 to 5.7, and the final answer must be checked against the domain and the question.
Worked example · The pen with the most area
A rancher has 100 feet of fence to make a rectangular pen. What dimensions give the greatest area?
Let the sides be \(x\) and \(y\). Objective: maximize \(A=xy\). Constraint: \(2x+2y=100\), so \(y=50-x\). Then \(A(x)=x(50-x)=50x-x^2\) with domain \(0\lt x\lt50\) (lengths must be positive).
\(A'(x)=50-2x=0\) at \(x=25\). \(A''(x)=-2\lt0\), so it is a maximum, and since there is only one critical point it is the global maximum on the domain. Then \(y=25\). The best pen is a 25 ft by 25 ft square with area \(625\) ft\(^2\).
Worked example · A river on one side
A farmer has 200 feet of fence to enclose a rectangular field along a straight river. The river side needs no fence. What dimensions give the greatest area?
Let \(x\) be the length of each side perpendicular to the river and \(y\) the side parallel to it. Constraint: \(2x+y=200\), so \(y=200-2x\). Objective: \(A(x)=x(200-2x)=200x-2x^2\), with domain \(0\lt x\lt100\).
\(A'(x)=200-4x=0\) at \(x=50\); \(A''=-4\lt0\), a maximum. \(y=200-100=100\). The field is 50 ft by 100 ft (with the 100 ft side along the river), area \(5000\) ft\(^2\).
Worked example · Two numbers
Find two nonnegative numbers with sum 20 whose product is largest.
Let the numbers be \(x\) and \(20-x\), with \(0\le x\le20\). \(P(x)=x(20-x)=20x-x^2\). On the closed interval, the candidates are \(x=10\) (where \(P'=20-2x=0\)) and the endpoints. \(P(0)=0=P(20)\), \(P(10)=100\). The largest product is 100, with both numbers equal to 10.
Choosing the domain and the endpoints
Because you may use the candidates test (closed interval) or the first derivative test (open interval), the domain is part of the answer. If the real-world domain is closed (for example \(0\le x\le20\) above), check endpoints. If it is open, argue from the sign of \(f'\) or \(f''\) as in Lesson 5.5.
Check the answer
Always ask: Does the answer make sense? Is it inside the domain? Did I answer what was asked (the dimensions, the maximum area, or both)? Are the units included?
Exam focus
Draw, name variables, find the objective and the constraint.
Eliminate variables with the constraint so the objective depends on one variable.
Write the domain from the situation.
Optimize with critical points, endpoints, and the first or second derivative test.
Answer the question asked, with units.
Practice
Write the objective function (in one variable) for the rectangle with perimeter 60 ft that has the greatest area. Give the domain.
Show answer
Sides \(x\) and \(30-x\): \(A(x)=x(30-x)\), \(0\lt x\lt30\).
A rectangle has area 36 square units. Write its perimeter as a function of its width \(w\), to be minimized.
Show answer
Two numbers have sum 10. Find them if the sum of their squares is to be minimized.
Show answer
\(S(x)=x^2+(10-x)^2=2x^2-20x+100\). \(S'(x)=4x-20=0\) at \(x=5\); \(S''=4\gt0\): minimum. The numbers are 5 and 5 (sum of squares 50).
A farmer fences a rectangular field along a river (no fence on the river side) with 400 feet of fence. Find the dimensions that maximize the area.
Show answer
\(y=400-2x\), \(A=400x-2x^2\), \(A'=400-4x=0\) at \(x=100\), \(y=200\). Dimensions 100 ft by 200 ft, area 20,000 ft\(^2\).
A rectangle is inscribed in the region under \(y=12-x^2\) above the \(x\)-axis, with its base on the \(x\)-axis and symmetric about the \(y\)-axis. Write its area as a function of \(x\), the right-hand \(x\)-coordinate of its top corner, with its domain.
Show answer
Width \(2x\), height \(12-x^2\): \(A(x)=2x(12-x^2)=24x-2x^3\), with \(0\lt x\lt\sqrt{12}\).
A student maximizes \(A=xy\) without using the constraint and finds that \(A\) has no maximum. What went wrong?
Show answer
The constraint links \(x\) and \(y\); without it, \(A=xy\) is unbounded. Substituting the constraint \(y=50-x\) (or similar) gives a function of one variable that has a maximum.
With the model in hand, the calculus finishes the job. The derivative can be used to solve optimization problems; that is, finding a maximum or minimum value in a situation described by a function: boxes, cans, distances, and costs. In each case you find the critical points of the objective, justify that one of them gives the extremum you want, and report the answer in context.
Required content · FUN-4.C.1
Learning objective.
FUN-4.C. Interpret minimum and maximum values calculated in applied contexts.
Essential knowledge.
FUN-4.C.1. Minimum and maximum values of a function take on specific meanings in applied contexts.
The box from a sheet
Figure 5.5. Left: squares of side \(x\) are cut from the corners of a 12 inch by 12 inch sheet and the flaps are folded up. Right: the volume \(V(x)=x(12-2x)^2\) on \(0\lt x\lt6\). The maximum, 128 cubic inches, occurs at \(x=2\), where \(V'(x)=0\).
Worked example · The open box
A 12 inch by 12 inch sheet of cardboard has a square of side \(x\) cut from each corner, and the sides are folded up to make an open box. Find the \(x\) that maximizes the volume, and the maximum volume.
The base is \((12-2x)\) by \((12-2x)\) and the height is \(x\), so \(V(x)=x(12-2x)^2\), with domain \(0\le x\le6\) (the cut cannot be more than half the side).
By the product and chain rules, \(V'(x)=(12-2x)^2+x\cdot2(12-2x)(-2)=(12-2x)\big[(12-2x)-4x\big]=(12-2x)(12-6x)\). The zeros are \(x=6\) and \(x=2\).
Candidates on \([0,6]\): \(V(0)=0\), \(V(2)=2(8)^2=128\), \(V(6)=0\). The maximum volume is 128 cubic inches, at \(x=2\) inches. The box is 8 by 8 by 2.
The can with least material
Worked example · The cylindrical can
A can must hold \(1000\) cm\(^3\). Find the radius and height that minimize the surface area.
Volume: \(\pi r^2h=1000\), so \(h=\dfrac{1000}{\pi r^2}\). Surface area (top, bottom and side): \(S=2\pi r^2+2\pi rh=2\pi r^2+\dfrac{2000}{r}\), domain \(r\gt0\).
\(S'(r)=4\pi r-\dfrac{2000}{r^2}=0\) gives \(r^3=\dfrac{500}{\pi}\), \(r=\sqrt[3]{500/\pi}\approx5.419\) cm. Check: \(S'\lt0\) for smaller \(r\) and \(S'\gt0\) for larger \(r\) (or \(S''=4\pi+\dfrac{4000}{r^3}\gt0\)), so this is the global minimum on \((0,\infty)\). Then \(h=\dfrac{1000}{\pi r^2}\approx10.839\) cm, which is exactly \(2r\): the best can has height equal to its diameter.
Distance problems
Worked example · Closest point on a curve
Find the point on \(y=\sqrt x\) closest to \((4,0)\).
The distance squared from \((x,\sqrt x)\) to \((4,0)\) is \(D(x)=(x-4)^2+x\), for \(x\ge0\). (Minimizing the square of the distance has the same minimizer as minimizing the distance, and it avoids a square root.) \(D'(x)=2(x-4)+1=2x-7=0\) at \(x=\tfrac72\). \(D''=2\gt0\): minimum. The point is \(\left(\tfrac72,\sqrt{3.5}\right)\approx(3.5,1.871)\), at distance \(\sqrt{D(3.5)}=\sqrt{0.25+3.5}=\sqrt{3.75}\approx1.936\).
Worked example · Cost
A company's average cost per unit when it makes \(x\) units is \(A(x)=\dfrac{4000}{x}+0.5x\) dollars, \(x\gt0\). Find the production level that minimizes the average cost.
\(A'(x)=-\dfrac{4000}{x^2}+0.5=0\) gives \(x^2=8000\), \(x\approx89.44\). \(A''(x)=\dfrac{8000}{x^3}\gt0\), so a minimum on \(x\gt0\). Since units are whole, compare \(A(89)=44.944+44.5=89.44\) and \(A(90)=44.444+45=89.44\): about 89 or 90 units, with an average cost near \$89.44 per unit.
Justifying the extremum
A correct solution always includes a reason that the critical point gives the maximum or minimum asked for: the candidates test on a closed interval, a sign change of \(f'\), or the second derivative test with \(f''\) of one sign on the domain. A calculator can find the critical point, but the justification is yours to write.
Exam focus
Find the critical points of the single-variable objective on its domain.
Justify the extremum: candidates test (closed), sign of \(f'\) or \(f''\) (open).
Minimizing the square of a distance gives the same point as minimizing the distance.
Return to the question: dimensions, value, or both; include units.
Check endpoints of the domain when the interval is closed.
Practice
A rectangle has perimeter 48 inches. Find its dimensions of largest area.
Show answer
\(A=x(24-x)\), \(A'=24-2x=0\), \(x=12\). A 12 by 12 square, area 144 square inches.
A rectangle has area 100 square feet. What dimensions give the smallest perimeter?
Show answer
\(P=2w+\dfrac{200}{w}\), \(P'=2-\dfrac{200}{w^2}=0\), \(w=10\). A 10 by 10 square, perimeter 40 ft. \(P''=\dfrac{400}{w^3}\gt0\) confirms the minimum.
Find the point on the line \(y=2x+1\) closest to the origin.
Show answer
\(D=x^2+(2x+1)^2=5x^2+4x+1\), \(D'=10x+4=0\), \(x=-0.4\), \(y=0.2\). The closest point is \((-0.4,0.2)\), distance \(\sqrt{0.16+0.04}=\sqrt{0.2}\approx0.447\).
An open box has a square base and volume 32 cubic feet. Find the dimensions that use the least material (base and four sides).
Show answer
Base side \(x\), height \(h=\dfrac{32}{x^2}\). Area \(S=x^2+4xh=x^2+\dfrac{128}{x}\). \(S'=2x-\dfrac{128}{x^2}=0\), \(x^3=64\), \(x=4\), \(h=2\). The box is 4 ft by 4 ft by 2 ft, with surface area \(16+32=48\) ft\(^2\). \(S''=2+\dfrac{256}{x^3}\gt0\).
A ball is thrown so that its height is \(h(t)=-16t^2+64t+5\) feet. Find its maximum height.
Show answer
\(h'(t)=-32t+64=0\) at \(t=2\). \(h''=-32\lt0\): maximum. \(h(2)=-64+128+5=69\) ft.
Two positive numbers have product 64. Find them if their sum is a minimum, and justify.
Show answer
\(S(x)=x+\dfrac{64}{x}\), \(x\gt0\). \(S'=1-\dfrac{64}{x^2}=0\), \(x=8\); \(S''=\dfrac{128}{x^3}\gt0\), so a minimum. The numbers are 8 and 8, with sum 16.
Implicit differentiation (Lesson 3.2) gave us slopes of curves that are not graphs of functions. Now we use the same slopes to answer the questions of this unit for such curves: where are the tangents horizontal or vertical, and is a point on the curve a local high point or low point? The derivative of an implicitly defined relation carries the same information as the derivative of a function: its zeros give horizontal tangents, and where it is undefined (with the curve still passing through) the tangent is vertical.
FUN-4.D. Determine critical points of implicit relations.
FUN-4.E. Justify conclusions about the behavior of an implicitly defined function based on evidence from its derivatives.
Essential knowledge.
FUN-4.D.1. A point on an implicit relation where the first derivative equals zero or does not exist is a critical point of the function.
FUN-4.E.1. Applications of derivatives can be extended to implicitly defined functions.
FUN-4.E.2. Second derivatives involving implicit differentiation may be relations of \(x\), \(y\), and \(\dfrac{dy}{dx}\).
Horizontal and vertical tangents
For a curve defined by an equation in \(x\) and \(y\), implicit differentiation produces \[\frac{dy}{dx}=\frac{N(x,y)}{D(x,y)}.\] - Horizontal tangent where \(N=0\) and \(D\ne0\), at a point on the curve. - Vertical tangent where \(D=0\) and \(N\ne0\), at a point on the curve.
Both conditions must be combined with the original equation, because the point must lie on the curve.
Worked example · The circle
For \(x^2+y^2=25\), \(\dfrac{dy}{dx}=-\dfrac xy\). Horizontal tangents: \(x=0\), so \(y=\pm5\): the points \((0,\pm5)\). Vertical tangents: \(y=0\), so \(x=\pm5\): the points \((\pm5,0)\).
Second derivative: \(\dfrac{d^2y}{dx^2}=-\dfrac{y-x\frac{dy}{dx}}{y^2}=-\dfrac{y+\frac{x^2}{y}}{y^2}=-\dfrac{y^2+x^2}{y^3}=-\dfrac{25}{y^3}\). At \((0,5)\) this is \(-\dfrac{25}{125}=-0.2\lt0\): concave down, so the top of the circle is a local maximum of \(y\). At \((0,-5)\) it is \(+0.2\gt0\): concave up, a local minimum.
Worked example · Horizontal and vertical tangents of a tilted curve
Find all points on \(x^2-xy+y^2=3\) where the tangent line is horizontal or vertical.
Differentiate: \(2x-y-x\dfrac{dy}{dx}+2y\dfrac{dy}{dx}=0\), so \(\dfrac{dy}{dx}=\dfrac{y-2x}{2y-x}\).
Horizontal: \(y=2x\). Substitute: \(x^2-2x^2+4x^2=3x^2=3\), \(x=\pm1\). Points \((1,2)\) and \((-1,-2)\). (The denominator there is \(2y-x=3x\ne0\).)
Vertical: \(x=2y\). Substitute: \(4y^2-2y^2+y^2=3y^2=3\), \(y=\pm1\). Points \((2,1)\) and \((-2,-1)\). (The numerator there is \(y-2x=-3y\ne0\).)
Classifying the horizontal-tangent points. With \(N=y-2x\) and \(D=2y-x\), \(\dfrac{d^2y}{dx^2}=\dfrac{N'D-ND'}{D^2}\). At a horizontal tangent \(N=0\) and \(N'=\dfrac{dy}{dx}-2=-2\). At \((1,2)\), \(D=3\): \(\dfrac{d^2y}{dx^2}=\dfrac{(-2)(3)}{9}=-\dfrac23\lt0\), so it is a local maximum of \(y\). At \((-1,-2)\), \(D=-3\): \(\dfrac{d^2y}{dx^2}=\dfrac{(-2)(-3)}{9}=\dfrac23\gt0\), a local minimum.
Interpreting the second derivative
A relation's second derivative is found by differentiating \(\dfrac{dy}{dx}\) implicitly, and then substituting \(\dfrac{dy}{dx}\) (and, where needed, the equation of the curve) to get an expression in \(x\) and \(y\). At a point where \(\dfrac{dy}{dx}=0\), the sign of \(\dfrac{d^2y}{dx^2}\) tells you if the point is a local maximum (negative) or local minimum (positive) of \(y\), exactly like the second derivative test. The sign also tells you the concavity of the curve at that point.
Worked example · Concavity at a point
A curve satisfies \(x^3+y^3=9\). Find \(\dfrac{dy}{dx}\) and \(\dfrac{d^2y}{dx^2}\) at \((1,2)\).
\(3x^2+3y^2\dfrac{dy}{dx}=0\), so \(\dfrac{dy}{dx}=-\dfrac{x^2}{y^2}\); at \((1,2)\), \(-\dfrac14\). \(\dfrac{d^2y}{dx^2}=-\dfrac{2xy^2-x^2\cdot2y\frac{dy}{dx}}{y^4}\). At \((1,2)\): numerator \(2(1)(4)-1\cdot2(2)(-\tfrac14)=8+1=9\); so \(\dfrac{d^2y}{dx^2}=-\dfrac{9}{16}\). The curve is concave down there.
Exam focus
Horizontal tangent: numerator of \(\dfrac{dy}{dx}\) is zero (denominator not); vertical: denominator is zero (numerator not).
Combine each condition with the original equation to get the points.
Find \(\dfrac{d^2y}{dx^2}\) by differentiating \(\dfrac{dy}{dx}\) implicitly and substituting.
At a flat point, \(\dfrac{d^2y}{dx^2}\lt0\) means a local maximum, \(\gt0\) a local minimum.
Check that candidate points are actually on the curve.
Practice
Find the points of horizontal tangency of \(x^2+y^2-4y=0\).
Show answer
\(2x+2y\dfrac{dy}{dx}-4\dfrac{dy}{dx}=0\), \(\dfrac{dy}{dx}=\dfrac{x}{2-y}\). Horizontal when \(x=0\): \(y^2-4y=0\), \(y=0\) or \(4\). Points \((0,0)\) and \((0,4)\).
Find where \(x^2+y^2=25\) has slope \(\tfrac34\).
Show answer
For \(x^2+xy+y^2=3\), find \(\dfrac{dy}{dx}\) and locate the points with horizontal tangents.
Show answer
\(2x+y+x\dfrac{dy}{dx}+2y\dfrac{dy}{dx}=0\), \(\dfrac{dy}{dx}=-\dfrac{2x+y}{x+2y}\). Horizontal: \(y=-2x\). \(x^2-2x^2+4x^2=3x^2=3\), \(x=\pm1\). Points \((1,-2)\) and \((-1,2)\).
For the curve \(x^3+y^3=9\), explain why the tangent at \((0,\sqrt[3]{9})\) is horizontal.
Show answer
\(\dfrac{dy}{dx}=-\dfrac{x^2}{y^2}=0\) when \(x=0\), with \(y=\sqrt[3]9\ne0\).
For \(x^2+y^2=25\), use \(\dfrac{d^2y}{dx^2}=-\dfrac{25}{y^3}\) to say whether the point \((3,4)\) is on a concave up or concave down portion.
Show answer
\(-\dfrac{25}{64}\lt0\): concave down.
A student says there is a vertical tangent to \(y^2=x\) at \((0,0)\) because \(\dfrac{dy}{dx}=\dfrac{1}{2y}\) is undefined there. Is the reasoning right?
Show answer
Yes. \(2y\dfrac{dy}{dx}=1\) gives \(\dfrac{dy}{dx}=\dfrac1{2y}\), which is undefined at \(y=0\), and the numerator 1 is not zero, and \((0,0)\) is on the curve. The parabola \(x=y^2\) has a vertical tangent at its vertex.
Unit 5 practice · 10 problems
Unit 5 practice: Analytical Applications of Differentiation
Ten problems covering the whole unit: one of the two heaviest-weighted
units on the exam. Work each one on paper before revealing the answer. No
calculator needed.
Find the absolute maximum and minimum of \(f(x) = x^3 - 12x\) on \([-3, 3]\).
Show answer
\(f'(x) = 3x^2 - 12 = 0\) at \(x = \pm 2\). Candidates: \(f(-3) = 9\), \(f(-2) = 16\), \(f(2) = -16\), \(f(3) = -9\). Absolute max 16 at \(x = -2\); absolute min \(-16\) at \(x = 2\).
Find the value of \(c\) guaranteed by the Mean Value Theorem for \(f(x) = \sqrt{x}\) on \([1, 9]\).
Show answer
Average rate \(= \dfrac{3 - 1}{8} = \dfrac{1}{4}\). \(f'(c) = \dfrac{1}{2\sqrt{c}} = \dfrac{1}{4} \Rightarrow c = 4\).
Find the intervals of concavity and the points of inflection of \(g(x) = x^4 - 8x^2 + 3\).
Show answer
\(g''(x) = 12x^2 - 16 = 0 \Rightarrow x = \pm\dfrac{2}{\sqrt{3}}\). Concave up for \(|x| > \dfrac{2}{\sqrt{3}}\), concave down on \(\left(-\dfrac{2}{\sqrt{3}}, \dfrac{2}{\sqrt{3}}\right)\); inflection points at \(x = \pm\dfrac{2}{\sqrt{3}} \approx \pm 1.155\).
\(f'(x) = x^2(x - 4)(x + 2)\). At which \(x\) does \(f\) have a local maximum? A local minimum?
Show answer
\(x^2 \ge 0\) never changes sign. \((x - 4)(x + 2)\) is positive for \(x < -2\), negative on \((-2, 4)\), positive for \(x > 4\). Local max at \(x = -2\) (\(+\) to \(-\)); local min at \(x = 4\) (\(-\) to \(+\)); nothing at \(x = 0\).
Use the second derivative test to classify the critical points of \(h(x) = 2x^3 - 3x^2 - 36x + 5\).
Show answer
\(h'(x) = 6(x - 3)(x + 2)\); critical points \(3\) and \(-2\). \(h''(x) = 12x - 6\): \(h''(3) = 30 > 0\) ⇒ local min at 3; \(h''(-2) = -30 < 0\) ⇒ local max at \(-2\).
A rectangle has its base on the \(x\)-axis and its upper corners on \(y = 12 - x^2\). Find the maximum possible area.
Show answer
\(A(x) = 2x(12 - x^2) = 24x - 2x^3\); \(A'(x) = 24 - 6x^2 = 0 \Rightarrow x = 2\); \(A''(2) < 0\). Maximum area \(= 4 \cdot 8 = 32\) (a 4 by 8 rectangle).
An open-top box with a square base must hold 32 in³. Find the dimensions that minimize the surface area.
Show answer
\(x^2 h = 32 \Rightarrow h = \dfrac{32}{x^2}\). \(S = x^2 + 4xh = x^2 + \dfrac{128}{x}\); \(S' = 2x - \dfrac{128}{x^2} = 0 \Rightarrow x = 4\), \(h = 2\). Minimum surface area \(= 48\) in².
\(f\) is twice differentiable on \([0, 6]\), with \(f' > 0\) on \((0, 2)\), \(f' < 0\) on \((2, 6)\), \(f'' < 0\) on \((0, 4)\), and \(f'' > 0\) on \((4, 6)\). Describe the graph of \(f\).
Show answer
\(f\) increases to a local maximum at \(x = 2\), then decreases. It is concave down on \((0, 4)\) and concave up on \((4, 6)\), with a point of inflection at \(x = 4\).
Find all points on \(x^2 + 4y^2 = 16\) where the tangent line is horizontal, and all points where it is vertical.
Find and classify the local extrema of \(f(x) = \dfrac{x}{x^2 + 1}\).
Show answer
\(f'(x) = \dfrac{1 - x^2}{(x^2 + 1)^2}\), zero at \(x = \pm 1\); positive on \((-1, 1)\), negative outside. Local minimum \(f(-1) = -\tfrac{1}{2}\); local maximum \(f(1) = \tfrac{1}{2}\).
Up to now we went from a quantity to its rate: position to velocity, volume to flow rate. The integral goes the other way. If you know how fast something is changing, you can add up the changes to find how much it changed in total. Water flowing into a tank, cars crossing a bridge, a population growing: in each case the amount of change is the accumulation of the rate.
CHA-4.A. Interpret the meaning of areas associated with the graph of a rate of change in context.
Essential knowledge.
CHA-4.A.1. The area of the region between the graph of a rate of change function and the \(x\)-axis gives the accumulation of change.
CHA-4.A.2. In some cases, accumulation of change can be evaluated by using geometry.
CHA-4.A.3. If a rate of change is positive (negative) over an interval, then the accumulated change is positive (negative).
CHA-4.A.4. The unit for the area of a region defined by rate of change is the unit for the rate of change multiplied by the unit for the independent variable.
Rate times time
If a rate is constant, accumulation is multiplication: water pouring at 5 gallons per minute for 3 minutes adds \(5\cdot3=15\) gallons. On a graph of the rate against time, that is the area of a rectangle with height 5 and width 3. The area under a rate graph, between the graph and the horizontal axis, is the accumulated change in the quantity.
When the rate varies, the same idea holds with area under a curve. The area of the region between a rate-of-change graph and the horizontal axis is the accumulated (net) change in the quantity over the interval. Areas of rectangles and triangles can be found with geometry; for a curved rate graph, we approximate the area (Lesson 6.2) or find it exactly (Lesson 6.4 onward).
Figure 6.1. A tank fills at rate \(r(t)=3+3\sin\left(\dfrac{\pi t}{6}\right)\) gallons per hour. The area of the shaded region between the graph of \(r\) and the \(t\)-axis from \(t=0\) to \(t=6\) is the total change in the amount of water: \(\displaystyle\int_0^6r(t)\,dt=18+\dfrac{36}{\pi}\approx29.46\) gallons.
Units and meaning
The units of the accumulated change are the units of the rate multiplied by the units of the independent variable. Gallons per hour times hours gives gallons; meters per second times seconds gives meters; people per year times years gives people. This unit check is a quick way to confirm that an answer to an application problem makes sense.
If the rate is negative, the quantity is decreasing, and that region of the graph below the axis counts as negative change. If \(r(t)\) is the rate at which water leaves a tank, it is simplest to define \(r\) as the rate at which water enters and let negative values mean outflow. The net change is then the signed area: the area above the axis minus the area below it.
The definite integral
The definite integral of a rate \(r\) from \(t=a\) to \(t=b\), written \(\displaystyle\int_a^br(t)\,dt\), is defined to be the net accumulation, the signed area between the graph of \(r\) and the \(t\)-axis. The accumulated change in the amount of a quantity \(Q\) from \(t=a\) to \(t=b\) is \[Q(b)-Q(a)=\int_a^bQ'(t)\,dt,\qquad\text{so}\qquad Q(b)=Q(a)+\int_a^bQ'(t)\,dt.\]
Worked example · Water entering a tank
Water enters a tank at the rate \(r(t)=3+3\sin\left(\dfrac{\pi t}{6}\right)\) gallons per hour for \(0\le t\le6\) hours, as in the figure. The tank holds 20 gallons at \(t=0\). How much water is in the tank at \(t=6\)?
The amount added is the area under the graph: \(\displaystyle\int_0^6\left(3+3\sin\frac{\pi t}6\right)dt=\left[3t-\frac{18}{\pi}\cos\frac{\pi t}6\right]_0^6=\left(18+\frac{18}\pi\right)-\left(-\frac{18}{\pi}\right)=18+\frac{36}\pi\approx29.46\) gallons. (The antiderivative method is developed in Lesson 6.7; for now, a calculator's numerical integral gives the same value.) So \(Q(6)=20+29.46=49.46\) gallons.
Worked example · Using geometry
A cyclist's speed in miles per hour increases steadily from 0 to 18 mph during the first half hour, stays at 18 mph for the next hour, and then decreases steadily to 0 over the last half hour. How far did the cyclist travel?
The distance is the area under the speed graph: two triangles and a rectangle. The triangles each have base 0.5 h and height 18 mph: area \(\tfrac12(0.5)(18)=4.5\) miles each. The rectangle: \(1\cdot18=18\) miles. Total \(4.5+18+4.5=27\) miles.
Worked example · Rate positive and negative
The rate of change of the number of people at a festival is \(P'(t)\) people per hour, with \(P'\) positive on \((0,3)\) (area 1200 above the axis) and negative on \((3,5)\) (area 700 below the axis). There are 500 people at \(t=0\). How many are there at \(t=5\)?
Net change \(=1200-700=500\). At \(t=5\): \(500+500=1000\) people. (The maximum crowd occurred at \(t=3\) with \(500+1200=1700\) people.)
Calculator
On the calculator. Calculators can compute a numerical integral, \(\displaystyle\int_a^b f(x)\,dx\), of a function you type in. When a question allows a calculator, write the integral setup (limits and integrand) in your work, not just the number.
Exam focus
The area under a rate graph is the accumulated change of the quantity.
Units of the accumulation = units of the rate times units of the input.
Signed area: regions below the axis count negative.
\(Q(b)=Q(a)+\int_a^bQ'(t)\,dt\).
Geometry (rectangles, triangles, semicircles) gives exact areas for simple graphs.
Practice
Water flows into a pool at a constant 12 gallons per minute for 25 minutes. Express the accumulated volume as an area and compute it.
Show answer
The graph of the rate is a horizontal line at height 12 from \(t=0\) to 25; the area is a rectangle \(12\cdot25=300\) gallons.
A car's velocity (miles per hour) is 30 for the first 2 hours and then decreases linearly to 0 over the next 1 hour. Find the distance traveled.
Show answer
Rectangle: \(30\cdot2=60\) miles. Triangle: \(\tfrac12(1)(30)=15\) miles. Total 75 miles.
The rate at which a population changes is \(P'(t)=100-20t\) people per year. Find the net change in population from \(t=0\) to \(t=5\) using the area under the line.
Show answer
The line goes from 100 at \(t=0\) to 0 at \(t=5\): a triangle with base 5 and height 100, so the area is \(\tfrac12(5)(100)=250\) people.
Water leaves a tank at \(r(t)=-4\) gallons per minute (negative for outflow) for 10 minutes; the tank had 90 gallons. How much is left?
Show answer
Net change \(=-4\cdot10=-40\); \(90-40=50\) gallons.
Explain why, if \(v(t)\lt0\) on an interval, the integral of \(v\) over that interval is negative, and what that means for position.
Show answer
The region is below the axis, so its signed area is negative. Velocity is negative when moving in the negative direction, so position decreases: the net displacement is negative.
A rate graph has area 8 above the axis from 0 to 4 and area 3 below the axis from 4 to 6. The quantity is 10 at \(t=0\). Find its value at \(t=6\) and at \(t=4\).
Show answer
Most rate graphs are not made of rectangles and triangles, and the data we are given are often a table, not a formula. To estimate the area under a curve, cut the interval into pieces, replace each piece with a rectangle (or trapezoid) whose area is easy, and add. These are Riemann sums, and they are the foundation of the integral.
LIM-5.A. Approximate a definite integral using geometric and numerical methods.
Essential knowledge.
LIM-5.A.1. Definite integrals can be approximated for functions that are represented graphically, numerically, analytically, and verbally.
LIM-5.A.2. Definite integrals can be approximated using a left Riemann sum, a right Riemann sum, a midpoint Riemann sum, or a trapezoidal sum; approximations can be computed using either uniform or nonuniform partitions.
LIM-5.A.3. Definite integrals can be approximated using numerical methods, with or without technology.
LIM-5.A.4. Depending on the behavior of a function, it may be possible to determine whether an approximation for a definite integral is an underestimate or overestimate for the value of the definite integral.
Rectangles
Divide \([a,b]\) into \(n\) subintervals. On a subinterval of width \(\Delta x\), a rectangle with height \(f(x^*)\) has area \(f(x^*)\Delta x\), where \(x^*\) is a sample point chosen in the subinterval:
Left Riemann sum (LRAM): \(x^*\) is the left endpoint.
Right Riemann sum (RRAM): \(x^*\) is the right endpoint.
Midpoint Riemann sum (MRAM): \(x^*\) is the midpoint.
The sum is \(\displaystyle\sum f(x_i^*)\Delta x_i\). The subintervals need not have equal width, especially for table data.
Figure 6.2. Approximations of \(\displaystyle\int_0^4\left(\tfrac14x^2+1\right)dx\) with four subintervals of width 1. The function is increasing, so the left sum is an underestimate and the right sum an overestimate. It is concave up, so the midpoint sum underestimates and the trapezoid sum overestimates. The exact value is \(\dfrac{28}{3}\approx9.33\).
Trapezoids
A trapezoidal sum uses, on each subinterval, a trapezoid joining the points \((x_{i-1},f(x_{i-1}))\) and \((x_i,f(x_i))\): its area is \(\tfrac12\big[f(x_{i-1})+f(x_i)\big]\Delta x_i\). When all subintervals have the same width, the trapezoidal sum is the average of the left and right sums.
Over- and underestimates
Use the shape of the graph to say whether a sum is too big or too small:
If \(f\) is increasing on the interval, a left sum is an underestimate and a right sum is an overestimate. If \(f\) is decreasing, the reverse.
If \(f\) is concave up, a trapezoidal sum is an overestimate and a midpoint sum an underestimate. If \(f\) is concave down, a trapezoidal sum underestimates and a midpoint sum overestimates.
Worked example · Four methods
For \(f(x)=\tfrac14x^2+1\) on \([0,4]\) with \(n=4\) (\(\Delta x=1\)), the function values are \(f(0)=1\), \(f(0.5)=1.0625\), \(f(1)=1.25\), \(f(1.5)=1.5625\), \(f(2)=2\), \(f(2.5)=2.5625\), \(f(3)=3.25\), \(f(3.5)=4.0625\), \(f(4)=5\).
Left: \(1(1+1.25+2+3.25)=7.5\)
Right: \(1(1.25+2+3.25+5)=11.5\)
Midpoint: \(1(1.0625+1.5625+2.5625+4.0625)=9.25\)
Trapezoid: \(\tfrac12(7.5+11.5)=9.5\)
The exact area is \(\displaystyle\int_0^4\left(\tfrac14x^2+1\right)dx=\left[\tfrac{x^3}{12}+x\right]_0^4=\tfrac{28}{3}\approx9.33\). As predicted: increasing, so left (7.5) is under and right (11.5) is over; concave up, so midpoint (9.25) is under and trapezoid (9.5) is over.
Sums from a table
Worked example · Unequal subintervals
The rate \(R(t)\) (gallons per minute) at which water enters a tank is measured at selected times.
\(t\) (minutes)
0
2
5
9
\(R(t)\)
3
5
6
4
Estimate the total water entering during \(0\le t\le9\), using a left sum, a right sum and a trapezoidal sum.
Subinterval widths: 2, 3, 4. Left: \(3(2)+5(3)+6(4)=6+15+24=45\) gallons. Right: \(5(2)+6(3)+4(4)=10+18+16=44\) gallons. Trapezoid: \(\tfrac{3+5}{2}(2)+\tfrac{5+6}{2}(3)+\tfrac{6+4}{2}(4)=8+16.5+20=44.5\) gallons. (Data do not tell us whether \(R\) is monotonic, so we cannot say which sum is an over- or underestimate.)
Worked example · Interpreting the sum
In the previous example, interpret the right sum: "The amount of water entering the tank over the 9 minutes is approximately 44 gallons, found by using 5, 6 and 4 gallons per minute as the rate on the three subintervals."
Exam focus
A Riemann sum is \(\sum f(x_i^*)\Delta x_i\) with left, right or midpoint heights; trapezoids use the average of the end heights.
Widths may differ; use each subinterval's own width.
Increasing \(f\): left under, right over. Concave up: trapezoid over, midpoint under.
Show the sum with each term (not just the total) so a point can be awarded for setup.
Interpret an answer with units.
Practice
Use a left Riemann sum with four subintervals to approximate \(\int_0^2 x^2\,dx\).
Show answer
\(\Delta x=0.5\). \(0.5\big[f(0)+f(0.5)+f(1)+f(1.5)\big]=0.5(0+0.25+1+2.25)=1.75\). (Exact value \(\tfrac83\approx2.667\); an underestimate since \(x^2\) is increasing there.)
Use a right sum with the same data and say whether it is an overestimate.
Show answer
\(0.5\big[f(0.5)+f(1)+f(1.5)+f(2)\big]=0.5(0.25+1+2.25+4)=3.75\). An overestimate, since \(f\) is increasing.
A midpoint sum with \(n=2\) subintervals on \([0,4]\) for \(f(x)=x^2\): compute.
Show answer
Midpoints 1 and 3: \(2\big[f(1)+f(3)\big]=2(1+9)=20\). Exact is \(\tfrac{64}{3}\approx21.33\), so this is an underestimate (concave up).
Velocity (m/s) at \(t=0,4,10,12\) is \(2,6,9,5\). Use a trapezoidal sum to estimate the distance traveled in 12 seconds.
Show answer
A table gives \(f\) decreasing: \(f(0)=10\), \(f(1)=7\), \(f(2)=3\), \(f(3)=1\). Find the left and right sums for \(\int_0^3f\) with \(n=3\), and state which is the overestimate.
Show answer
Left: \(10+7+3=20\). Right: \(7+3+1=11\). Since \(f\) is decreasing, the left sum (20) is the overestimate and the right sum (11) is the underestimate.
For a function that is concave down on \([a,b]\), is the trapezoidal sum an over- or underestimate of the integral? Justify.
Show answer
Underestimate. The chord joining two points of a concave down graph lies below the graph, so each trapezoid lies below the curve.
Riemann sums get better as the rectangles get thinner. The definite integral is defined as what the sums approach when the number of subintervals grows without bound. This lesson writes sums in compact notation and links them to the integral symbol.
LIM-5.B. Interpret the limiting case of the Riemann sum as a definite integral.
LIM-5.C. Represent the limiting case of the Riemann sum as a definite integral.
Essential knowledge.
LIM-5.B.1. The limit of an approximating Riemann sum can be interpreted as a definite integral.
LIM-5.B.2. A Riemann sum, which requires a partition of an interval \(I\), is the sum of products, each of which is the value of the function at a point in a subinterval multiplied by the length of that subinterval of the partition.
LIM-5.C.1. The definite integral of a continuous function \(f\) over the interval \([a,b]\), denoted by \(\displaystyle\int_a^b f(x)\,dx\), is the limit of Riemann sums as the widths of the subintervals approach 0. That is, \(\displaystyle\int_a^b f(x)\,dx=\lim_{\max\Delta x_i\to 0}\sum_{i=1}^{n} f(x_i^*)\,\Delta x_i\), where \(n\) is the number of subintervals, \(\Delta x_i\) is the width of the \(i\)th subinterval, and \(x_i^*\) is a value in the \(i\)th subinterval.
LIM-5.C.2. A definite integral can be translated into the limit of a related Riemann sum, and the limit of a Riemann sum can be written as a definite integral.
Summation notation
The symbol \(\displaystyle\sum_{i=1}^{n}a_i\) means \(a_1+a_2+\cdots+a_n\). A Riemann sum with equal subintervals of width \(\Delta x=\dfrac{b-a}{n}\) and right endpoints \(x_i=a+i\,\Delta x\) is \[\sum_{i=1}^{n}f(x_i)\,\Delta x.\] For a left sum use \(x_{i-1}\) instead (sum from \(i=1\) to \(n\) of \(f(a+(i-1)\Delta x)\Delta x\)).
The definite integral as a limit
The definite integral of \(f\) from \(a\) to \(b\) is the limit of Riemann sums as the number of subintervals increases without bound (and the widths go to 0), provided it exists: \[\int_a^bf(x)\,dx=\lim_{n\to\infty}\sum_{i=1}^nf(x_i^*)\,\Delta x.\] Every continuous function on \([a,b]\) is integrable, and the limit does not depend on whether we use left, right or midpoint sample points. In the notation \(\displaystyle\int_a^bf(x)\,dx\): the numbers \(a\) and \(b\) are the limits of integration, \(f(x)\) is the integrand, and \(dx\) names the variable of integration (it comes from \(\Delta x\)). The integral is a number (for fixed \(a\), \(b\)); the variable name does not matter: \(\int_a^bf(x)\,dx=\int_a^bf(t)\,dt\).
Writing a limit of a sum as an integral
To translate a limit of a Riemann sum into an integral, identify \(\Delta x\), the starting point \(a\), the function.
Worked example · From limit to integral
Write \(\displaystyle\lim_{n\to\infty}\sum_{k=1}^{n}\left(2+\frac{3k}{n}\right)^2\cdot\frac3n\) as a definite integral and evaluate.
Compare with \(\sum f(x_k)\Delta x\): \(\Delta x=\dfrac3n=\dfrac{b-a}{n}\), so \(b-a=3\). The sample points are \(x_k=2+\dfrac{3k}{n}=a+k\,\Delta x\), so \(a=2\) and \(b=5\), and \(f(x)=x^2\). The limit equals \(\displaystyle\int_2^5x^2\,dx=\frac{125-8}{3}=39\).
Worked example · From integral to a sum
Write a right Riemann sum with \(n\) subintervals for \(\displaystyle\int_1^3\frac1x\,dx\).
When the graph is a line, a circle or a combination, the integral can be found as signed area with area formulas.
Worked example · A quarter circle
\(\displaystyle\int_0^4\sqrt{16-x^2}\,dx\) is the area of a quarter of the circle \(x^2+y^2=16\): \(\tfrac14\pi(4)^2=4\pi\approx12.57\).
Worked example · Absolute value
\(\displaystyle\int_{-3}^{3}\lvert x\rvert\,dx\) is the area of two triangles, each with base 3 and height 3: \(2\cdot\tfrac12(3)(3)=9\).
Worked example · Interpreting an integral
\(\displaystyle\int_0^{12}R(t)\,dt\), where \(R(t)\) is the rate (in thousands of gallons per day) at which a reservoir is refilled, equals the total number of thousand gallons added during the first 12 days.
How many subintervals?
A Riemann sum with a small \(n\) gives a rough estimate. Increasing \(n\) shrinks the error. On the calculator, numerical integration does this automatically with a very large effective \(n\). For a function whose graph is not made of simple shapes, the exact value must wait for the Fundamental Theorem (Lesson 6.4).
Exam focus
\(\sum_{i=1}^n a_i=a_1+\cdots+a_n\).
\(\int_a^bf(x)\,dx=\lim_{n\to\infty}\sum f(x_i^*)\Delta x\), with \(\Delta x=\dfrac{b-a}{n}\).
To convert a limit of a sum to an integral: find \(\Delta x\) (gives \(b-a\)), then the first sample point (gives \(a\)), then \(f\).
The variable of integration is a dummy variable.
Use geometry for integrals of lines, circles and absolute values.
Practice
Write \(\displaystyle\sum_{i=1}^{4}(2i+1)\) as an explicit sum and evaluate.
Show answer
\(3+5+7+9=24\).
Write \(\displaystyle\lim_{n\to\infty}\sum_{i=1}^n\left(1+\frac{2i}{n}\right)^3\frac2n\) as an integral.
Show answer
Use geometry to evaluate \(\displaystyle\int_0^6(6-x)\,dx\).
Show answer
Triangle with base 6 and height 6: \(\tfrac12(6)(6)=18\).
Use geometry to evaluate \(\displaystyle\int_{-2}^{2}\sqrt{4-x^2}\,dx\).
Show answer
Upper half of a circle of radius 2: \(\tfrac12\pi(2)^2=2\pi\).
A student says \(\displaystyle\int_0^3x\,dx\) and \(\displaystyle\int_0^3t\,dt\) are different. Is that right?
Show answer
No. The integration variable is a dummy variable; both equal \(\tfrac12(3)(3)=4.5\).
Express as a definite integral the limit \(\displaystyle\lim_{n\to\infty}\sum_{k=1}^{n}\sqrt{k/n}\cdot\frac1n\), and interpret \(\Delta x\).
Show answer
The Fundamental Theorem and accumulation functions
Differentiation and integration seem unrelated: one finds slopes, the other finds areas. The Fundamental Theorem of Calculus says they are inverse processes. This lesson uses the first half: the derivative of an accumulation function is the function being accumulated.
Required content · FUN-5.A.1, FUN-5.A.2
Learning objective.
FUN-5.A. Represent accumulation functions using definite integrals.
Essential knowledge.
FUN-5.A.1. The definite integral can be used to define new functions.
FUN-5.A.2. If \(f\) is a continuous function on an interval containing \(a\), then \(\dfrac{d}{dx}\!\left(\displaystyle\int_a^x f(t)\,dt\right)=f(x)\), where \(x\) is in the interval.
Accumulation functions
Let \(f\) be continuous and let \(a\) be a constant. The accumulation function \[g(x)=\int_a^xf(t)\,dt\] gives the net area between the graph of \(f\) and the \(t\)-axis from \(t=a\) to \(t=x\). It is a function of \(x\): as the upper limit \(x\) moves, the accumulated area changes. (The variable of integration is \(t\) so it does not collide with the limit \(x\).) Note that \(g(a)=\int_a^af(t)\,dt=0\).
The Fundamental Theorem, part one
FTC (part 1). If \(f\) is continuous on an interval containing \(a\), then \(g(x)=\displaystyle\int_a^xf(t)\,dt\) is differentiable and \[\frac{d}{dx}\int_a^xf(t)\,dt=f(x).\]
Why: the change \(g(x+h)-g(x)\) is the area of a thin strip of height about \(f(x)\) and width \(h\), so \(\dfrac{g(x+h)-g(x)}{h}\approx f(x)\), and the limit as \(h\to0\) is \(f(x)\). In words, the rate at which area accumulates equals the height of the function at the edge.
Figure 6.4. Left: \(f(t)=(t-1)(t-3)\). Right: \(g(x)=\int_0^xf(t)\,dt=\tfrac13x^3-2x^2+3x\), whose derivative is \(f\). \(g\) increases while \(f\gt0\), reaches a relative maximum at \(x=1\) where \(f\) changes from positive to negative, decreases while \(f\lt0\), and has a relative minimum at \(x=3\). \(g\) has an inflection point at \(x=2\), where \(f\) has its minimum (so \(g'=f\) changes from decreasing to increasing).
The figure shows \(f(t)=(t-1)(t-3)\) and \(g(x)=\displaystyle\int_0^xf(t)\,dt=\tfrac13x^3-2x^2+3x\); indeed \(g'(x)=x^2-4x+3=f(x)\). Where \(f\) is positive, \(g\) rises; where \(f\) is negative, \(g\) falls.
Using the theorem
Worked example · Direct use
\(\displaystyle\frac{d}{dx}\int_1^x\sin(t^2)\,dt=\sin(x^2)\), and \(\displaystyle\frac{d}{dx}\int_0^x\sqrt{1+t^3}\,dt=\sqrt{1+x^3}\). You do not need an antiderivative: the theorem gives the derivative directly.
Worked example · The chain rule version
If the upper limit is a function \(u(x)\), use the chain rule: \(\displaystyle\frac{d}{dx}\int_a^{u(x)}f(t)\,dt=f\big(u(x)\big)\cdot u'(x)\).
Find \(\displaystyle\frac{d}{dx}\int_2^{x^2}\ln t\,dt\). With \(u=x^2\), \(u'=2x\): the answer is \(\ln(x^2)\cdot2x=4x\ln x\) for \(x\gt0\).
Worked example · Lower limit variable
If the variable is in the lower limit, reverse the limits (changing the sign): \(\displaystyle\frac{d}{dx}\int_x^5e^{t^2}\,dt=-\frac{d}{dx}\int_5^xe^{t^2}\,dt=-e^{x^2}\).
Worked example · Values of an accumulation function
\(g(x)=\displaystyle\int_0^xf(t)\,dt\), where the graph of \(f\) is made of line segments: from \((0,0)\) up to \((2,4)\), then down to \((4,0)\), then down to \((6,-2)\). Find \(g(2)\), \(g(4)\), \(g(6)\), \(g'(4)\) and \(g'(6)\).
\(g(2)=\) triangle area \(\tfrac12(2)(4)=4\). \(g(4)=4+\tfrac12(2)(4)=8\). From 4 to 6 the graph is a triangle below the axis with base 2 and height 2: area \(-2\), so \(g(6)=8-2=6\). By the theorem, \(g'(4)=f(4)=0\) and \(g'(6)=f(6)=-2\).
Accumulation with a different starting point
If \(h(x)=\displaystyle\int_3^xf(t)\,dt\) and \(g(x)=\displaystyle\int_0^xf(t)\,dt\), then \(h(x)=g(x)-g(3)\): they differ by a constant, and have the same derivative \(f(x)\).
Exam focus
\(g(x)=\int_a^xf(t)\,dt\) has \(g(a)=0\) and \(g'(x)=f(x)\).
With an upper limit \(u(x)\): \(\dfrac{d}{dx}\int_a^{u(x)}f(t)\,dt=f(u(x))\,u'(x)\).
A variable lower limit flips the sign.
No antiderivative is needed to differentiate an accumulation function.
Evaluate \(g\) at a point as signed area from the starting point.
Practice
Find \(\dfrac{d}{dx}\displaystyle\int_0^xe^{-t^2}\,dt\).
Show answer
\(e^{-x^2}\).
Find \(\dfrac{d}{dx}\displaystyle\int_1^{x^3}\cos t\,dt\).
Show answer
\(\cos(x^3)\cdot3x^2=3x^2\cos(x^3)\).
Find \(\dfrac{d}{dx}\displaystyle\int_x^{4}\sqrt{t^2+1}\,dt\).
Show answer
\(-\sqrt{x^2+1}\).
If \(g(x)=\displaystyle\int_2^x\frac{1}{1+t^2}\,dt\), find \(g(2)\) and \(g'(2)\).
Show answer
\(g(2)=0\); \(g'(x)=\dfrac1{1+x^2}\), so \(g'(2)=\dfrac15\).
The graph of \(f\) is a horizontal segment from \((0,3)\) to \((4,3)\), then a line down to \((6,0)\). Let \(g(x)=\displaystyle\int_0^xf(t)\,dt\). Find \(g(4)\) and \(g(6)\).
Show answer
\(g(4)=3\cdot4=12\). From 4 to 6 the area is a triangle with base 2, height 3: 3. \(g(6)=15\).
A student claims \(\dfrac{d}{dx}\displaystyle\int_0^{x^2}t\,dt=x^2\). Find the error and the correct derivative.
Show answer
The student forgot the chain rule. \(\dfrac{d}{dx}\displaystyle\int_0^{x^2}t\,dt=(x^2)\cdot2x=2x^3\). (Check: the integral equals \(\tfrac{x^4}{2}\), whose derivative is \(2x^3\).)
An accumulation function \(g(x)=\int_a^xf(t)\,dt\) is a new function, and its graph can be described using only the graph of \(f\), without ever computing an antiderivative. Because \(g'=f\) and \(g''=f'\), everything you learned in Unit 5 about extrema, increasing and decreasing, concavity and inflection points carries over, with \(f\) playing the role of the derivative.
Required content · FUN-5.A.3
Learning objective.
FUN-5.A. Represent accumulation functions using definite integrals.
Essential knowledge.
FUN-5.A.3. Graphical, numerical, analytical, and verbal representations of a function \(f\) provide information about the function \(g\) defined as \(g(x)=\displaystyle\int_a^x f(t)\,dt\).
The dictionary
For \(g(x)=\displaystyle\int_a^xf(t)\,dt\):
Behavior of \(f\)
Behavior of \(g\)
\(f(x)\gt0\)
\(g\) is increasing
\(f(x)\lt0\)
\(g\) is decreasing
\(f\) changes from \(+\) to \(-\) at \(c\)
\(g\) has a relative maximum at \(c\)
\(f\) changes from \(-\) to \(+\) at \(c\)
\(g\) has a relative minimum at \(c\)
\(f\) is increasing
\(g\) is concave up
\(f\) is decreasing
\(g\) is concave down
\(f\) has a relative extremum at \(c\)
\(g\) has a point of inflection at \(c\)
The values of \(g\) are accumulated areas, so you can compute them by geometry when the graph of \(f\) is made of simple shapes.
Worked example · A piecewise linear \(f\)
The graph of \(f\) consists of line segments connecting \((0,-2)\), \((2,2)\), \((5,2)\), \((7,-2)\). Let \(g(x)=\displaystyle\int_0^xf(t)\,dt\). Analyze \(g\) on \([0,7]\).
Zeros of \(f\). The first segment crosses zero at \(x=1\); the last segment crosses zero at \(x=6\).
Values of \(g\). \(g(0)=0\). From 0 to 1, a triangle below the axis with base 1 and height 2: area \(-1\), \(g(1)=-1\). From 1 to 2, a triangle above: \(+1\), \(g(2)=0\). From 2 to 5, a rectangle \(3\times2=6\): \(g(5)=6\). From 5 to 6, a triangle above, \(+1\): \(g(6)=7\). From 6 to 7 a triangle below, \(-1\): \(g(7)=6\).
Extrema. \(f\) changes from negative to positive at \(x=1\): \(g\) has a relative minimum there, \(g(1)=-1\). \(f\) changes from positive to negative at \(x=6\): relative maximum, \(g(6)=7\). Comparing with the endpoints \(g(0)=0\) and \(g(7)=6\), the absolute maximum of \(g\) on \([0,7]\) is 7 at \(x=6\), and the absolute minimum is \(-1\) at \(x=1\).
Concavity. \(f\) is increasing on \((0,2)\): \(g\) is concave up there. \(f\) is constant on \((2,5)\): \(g\) is a straight line with slope 2. \(f\) is decreasing on \((5,7)\): \(g\) is concave down. Concavity switches from up to down across the straight stretch, but because \(g''=f'\) is zero on all of \((2,5)\), no single point is the place where it changes, so we do not name an inflection point.
Worked example · Justifying a statement
Is \(g\) increasing at \(x=3\)? Because \(g'(3)=f(3)=2\gt0\), yes. Is \(g\) concave up there? \(g''(3)=f'(3)=0\) (the graph of \(f\) is horizontal), so \(g\) is neither concave up nor concave down at \(x=3\); it is a straight line segment there.
Worked example · Reading a description of \(f\)
Suppose \(f\) is a continuous function with \(f(0)=0\), \(f\) increasing on \((0,3)\), decreasing on \((3,8)\), and \(f(5)=0\). Let \(g(x)=\displaystyle\int_0^xf(t)\,dt\). Where does \(g\) have a relative maximum?
\(f\gt0\) on \((0,5)\) (it rises from 0 and falls to 0 at \(x=5\)) and \(f\lt0\) on \((5,8)\). So \(f\) changes from positive to negative at \(x=5\): \(g\) has a relative maximum at \(x=5\). \(g\) has an inflection point at \(x=3\), where \(f\) changes from increasing to decreasing.
Writing the justification
Use \(f\) (not \(g\)) in the reason: "\(g\) has a relative maximum at \(x=6\) because \(g'(x)=f(x)\) changes from positive to negative at \(x=6\)." "The graph of \(g\) is concave down on \((5,7)\) because \(g'=f\) is decreasing there."
Exam focus
\(g'=f\) and \(g''=f'\): read extrema from the sign changes of \(f\), concavity from the direction of \(f\).
Inflection points of \(g\) are at relative extrema of \(f\).
Find values of \(g\) by adding signed areas from the starting point.
Check endpoints for absolute extrema of \(g\).
Justify in terms of \(f\).
Practice
For \(g(x)=\displaystyle\int_0^x(t^2-4)\,dt\), find where \(g\) has relative extrema on \((-\infty,\infty)\).
Show answer
\(g'(x)=x^2-4\), zero at \(x=\pm2\). \(g'\) changes \(+\) to \(-\) at \(x=-2\) (relative maximum) and \(-\) to \(+\) at \(x=2\) (relative minimum).
For the same \(g\), find the intervals where \(g\) is concave up.
Show answer
\(g''(x)=2x\gt0\) for \(x\gt0\). \(g\) is concave up on \((0,\infty)\).
The graph of \(f\) is a semicircle of radius 2 above the \(t\)-axis from \(t=-2\) to \(t=2\). Find \(\displaystyle\int_{-2}^{2}f(t)\,dt\).
Show answer
Area of a semicircle: \(\tfrac12\pi(2)^2=2\pi\).
Let \(g(x)=\displaystyle\int_1^xf(t)\,dt\), where \(f\) is positive on \((1,4)\) and negative on \((4,9)\). Where is \(g\) increasing, and where on \([1,9]\) is its absolute maximum?
Show answer
\(g\) is increasing on \((1,4)\) and decreasing on \((4,9)\), since \(g'=f\). The candidates for the absolute maximum are \(x=1\), \(4\) and \(9\). \(g(1)=0\) and \(g\) increases up to \(x=4\), so \(g(4)\gt g(1)\); \(g\) then decreases, so \(g(4)\gt g(9)\). The absolute maximum is at \(x=4\).
A rate of entry \(E(t)\) (people per minute) into a stadium is positive for \(0\le t\le 30\) and the exit rate is \(X(t)\). Write a function for the number of people in the stadium at time \(x\) given that 500 people were there at \(t=0\), and explain when the number is largest.
Show answer
\(N(x)=500+\displaystyle\int_0^x\big(E(t)-X(t)\big)\,dt\). \(N'(x)=E(x)-X(x)\); the maximum occurs where \(E-X\) changes from positive to negative (or at an endpoint).
Given \(g'(x)=f(x)\) with \(f\) increasing on \((0,2)\) and decreasing on \((2,6)\), what does \(g\) do at \(x=2\)?
Show answer
\(g''=f'\) changes from positive to negative at \(x=2\), so \(g\) has a point of inflection at \(x=2\) (its graph changes from concave up to concave down).
Definite integrals obey a short list of rules that follow from their meaning as signed area. They let you combine, split and compare integrals when you only know a few values, which is how many free-response questions are built.
FUN-6.A. Calculate a definite integral using areas and properties of definite integrals.
Essential knowledge.
FUN-6.A.1. In some cases, a definite integral can be evaluated by using geometry and the connection between the definite integral and area.
FUN-6.A.2. Properties of definite integrals include the integral of a constant times a function, the integral of the sum of two functions, reversal of limits of integration, and the integral of a function over adjacent intervals.
FUN-6.A.3. The definition of the definite integral may be extended to functions with removable or jump discontinuities.
The properties
For \(f\) and \(g\) continuous on an interval containing \(a\), \(b\), \(c\), and a constant \(k\):
Same limits: \(\displaystyle\int_a^af(x)\,dx=0\). (A region of width zero has no area.)
Reversing limits: \(\displaystyle\int_a^bf(x)\,dx=-\int_b^af(x)\,dx\). (Going backward flips the sign.)
Sum and difference: \(\displaystyle\int_a^b\big[f(x)\pm g(x)\big]dx=\int_a^bf(x)\,dx\pm\int_a^bg(x)\,dx\).
Additivity over intervals: \(\displaystyle\int_a^bf(x)\,dx=\int_a^cf(x)\,dx+\int_c^bf(x)\,dx\) for any \(c\) (inside or outside \([a,b]\)).
Comparison: if \(f(x)\ge g(x)\) on \([a,b]\) with \(a\lt b\), then \(\displaystyle\int_a^bf\ge\int_a^bg\). In particular, if \(f\ge0\), the integral is nonnegative, and if \(m\le f(x)\le M\), then \(m(b-a)\le\displaystyle\int_a^bf\le M(b-a)\).
There is no rule for the integral of a product or a quotient: \(\int fg\ne\int f\cdot\int g\).
Properties 3 and 4 mean the integral is linear. Property 5 is the one used most with graphs and tables: you can split an interval at any point, for example at a place where the formula or the sign of \(f\) changes.
Figure 6.3. For \(f(x)=(x-1)(x-3)\), the integral counts area above the \(x\)-axis as positive and area below as negative. \(\displaystyle\int_0^1f=\tfrac43\), \(\displaystyle\int_1^3f=-\tfrac43\), \(\displaystyle\int_3^4f=\tfrac43\), so \(\displaystyle\int_0^3f=0\) and \(\displaystyle\int_0^4f=\tfrac43\), while the total (unsigned) area between the graph and the axis from 0 to 4 is 4.
Examples
Worked example · Combining given values
Given \(\displaystyle\int_1^5f(x)\,dx=7\), \(\displaystyle\int_1^3f(x)\,dx=2\) and \(\displaystyle\int_1^5g(x)\,dx=-4\), find: (a) \(\displaystyle\int_3^5f\), (b) \(\displaystyle\int_5^1f\), (c) \(\displaystyle\int_1^5\big(3f-2g\big)\), (d) \(\displaystyle\int_1^53\,dx\).
(a) By additivity, \(\int_3^5f=\int_1^5f-\int_1^3f=7-2=5\). (b) Reversing the limits, \(\int_5^1f=-7\). (c) \(3(7)-2(-4)=21+8=29\). (d) \(\int_1^53\,dx=3(5-1)=12\).
Worked example · Signed area from the graph
The figure shows \(f(x)=(x-1)(x-3)\). The areas of the three shaded regions are each \(\tfrac43\). Then \(\int_0^1f=\tfrac43\), \(\int_1^3f=-\tfrac43\) (below the axis), \(\int_3^4f=\tfrac43\). By additivity, \(\int_0^4f=\tfrac43-\tfrac43+\tfrac43=\tfrac43\), and \(\int_0^3f=0\). But the total area (always positive) between the curve and the axis on \([0,4]\) is \(\int_0^4\lvert f\rvert\,dx=4\).
Split at \(x=1\): \(\displaystyle\int_0^12x\,dx=1\) (triangle area, \(\tfrac12(1)(2)\)). \(\displaystyle\int_1^3(3-x)\,dx\): the graph goes from height 2 at \(x=1\) to height 0 at \(x=3\), a triangle with base 2 and height 2: area 2. Total \(=3\).
Worked example · Bounding an integral
Show that \(\displaystyle\int_0^2\frac{1}{1+x^2}\,dx\) lies between \(0.4\) and \(2\).
On \([0,2]\), \(\dfrac1{1+x^2}\) decreases from 1 to \(\dfrac15=0.2\). So \(0.2\le f(x)\le1\), and \(0.2(2)\le\displaystyle\int_0^2f\le1(2)\), giving \(0.4\le\displaystyle\int_0^2f\le2\).
Worked example · Even and odd symmetry
\(\displaystyle\int_{-2}^{2}x^3\,dx=0\), because \(x^3\) is odd: the area to the left of the axis cancels the area to the right. For an even function such as \(x^2\), \(\displaystyle\int_{-2}^2x^2\,dx=2\int_0^2x^2\,dx\).
The Fundamental Theorem of Calculus and definite integrals
The second half of the Fundamental Theorem turns the definite integral from a limit of sums into a subtraction: find an antiderivative, evaluate it at the two limits, and subtract. This is how almost every exact integral in the course is computed.
FUN-6.B. Evaluate definite integrals analytically using the Fundamental Theorem of Calculus.
Essential knowledge.
FUN-6.B.1. An antiderivative of a function \(f\) is a function \(g\) whose derivative is \(f\).
FUN-6.B.2. If a function \(f\) is continuous on an interval containing \(a\), the function defined by \(F(x)=\displaystyle\int_a^x f(t)\,dt\) is an antiderivative of \(f\) for \(x\) in the interval.
FUN-6.B.3. If \(f\) is continuous on the interval \([a,b]\) and \(F\) is an antiderivative of \(f\), then \(\displaystyle\int_a^b f(x)\,dx=F(b)-F(a)\).
The theorem
FTC (part 2). If \(f\) is continuous on \([a,b]\) and \(F\) is any antiderivative of \(f\) (that is, \(F'=f\)), then \[\int_a^bf(x)\,dx=F(b)-F(a).\] We write \(F(b)-F(a)\) as \(\Big[F(x)\Big]_a^b\). Why it holds: \(g(x)=\int_a^xf(t)\,dt\) is an antiderivative of \(f\) by part 1, and any two antiderivatives differ by a constant, so \(F(x)=g(x)+C\); then \(F(b)-F(a)=g(b)-g(a)=g(b)-0\). The constant \(C\) cancels, so you never need to add \(+C\) in a definite integral.
Net change
Since \(f=F'\), the theorem says \(\displaystyle\int_a^bF'(x)\,dx=F(b)-F(a)\): the integral of a rate of change is the net change of the quantity. In motion: if \(v(t)=s'(t)\) then \(\displaystyle\int_a^bv(t)\,dt=s(b)-s(a)\) is the displacement. The total distance traveled is \(\displaystyle\int_a^b\lvert v(t)\rvert\,dt\): split where \(v\) changes sign and add the absolute values.
A particle moves on a line with velocity \(v(t)=t^2-6t+8=(t-2)(t-4)\) m/s for \(0\le t\le6\). Find the displacement and the total distance.
An antiderivative is \(F(t)=\tfrac{t^3}{3}-3t^2+8t\). \(F(2)=\tfrac{20}{3}\), \(F(4)=\tfrac{16}{3}\), \(F(6)=12\). Displacement: \(F(6)-F(0)=12\) m. Signs of \(v\): positive on \((0,2)\), negative on \((2,4)\), positive on \((4,6)\). Distance: \(\big(F(2)-F(0)\big)+\big|F(4)-F(2)\big|+\big(F(6)-F(4)\big)=\tfrac{20}{3}+\tfrac43+\tfrac{20}{3}=\tfrac{44}{3}\approx14.67\) m.
Worked example · A rate in context
The rate at which water leaks from a tank is \(L(t)=40e^{-0.1t}\) liters per hour. How much leaks out during the first 5 hours?
Forgetting to subtract \(F(a)\), or reversing the order.
Using \(\int v\,dt\) as distance when \(v\) changes sign: that gives displacement.
Skipping the check for discontinuity: the theorem needs \(f\) continuous on \([a,b]\). For \(\displaystyle\int_{-1}^1\frac{1}{x^2}\,dx\), blindly computing \([-1/x]_{-1}^{1}=-2\) is wrong, because \(\tfrac1{x^2}\) is undefined at 0 and unbounded near it (the area is infinite, and the "negative area" is impossible for a positive function).
Exam focus
\(\int_a^bf=F(b)-F(a)\) for any antiderivative \(F\) of continuous \(f\).
The integral of a rate is net change; \(Q(b)=Q(a)+\int_a^bQ'\).
Displacement: \(\int v\,dt\). Total distance: \(\int\lvert v\rvert\,dt\).
Show the antiderivative and the evaluation; a bare number with no setup earns little.
Check that the integrand is continuous on the whole interval.
Practice
Evaluate \(\displaystyle\int_0^2(x^3+1)\,dx\).
Show answer
\(\Big[\tfrac{x^4}{4}+x\Big]_0^2=(4+2)-0=6\).
Evaluate \(\displaystyle\int_1^9\frac{1}{\sqrt x}\,dx\).
Show answer
\(\big[2\sqrt x\big]_1^9=6-2=4\).
Evaluate \(\displaystyle\int_0^{\pi}\sin x\,dx\).
Show answer
\(\big[-\cos x\big]_0^\pi=1-(-1)=2\).
A particle's velocity is \(v(t)=4-2t\) for \(0\le t\le5\). Find the displacement and total distance.
Show answer
\(F(t)=4t-t^2\). Displacement \(F(5)-F(0)=-5\). \(v\gt0\) on \((0,2)\), \(\lt0\) on \((2,5)\): \(F(2)=4\). Distance \(=4+|F(5)-F(2)|=4+|-5-4|=4+9=13\).
The rate of change of a population is \(P'(t)=500e^{0.2t}\) people per year. Find the change from \(t=0\) to \(t=3\).
Show answer
A student computes \(\displaystyle\int_{-2}^{3}\frac{1}{x}\,dx=\big[\ln x\big]_{-2}^3=\ln3-\ln(-2)\). What is wrong?
Show answer
\(\dfrac1x\) is not defined (or continuous) at \(x=0\) in the interval, so the theorem does not apply; and \(\ln(-2)\) is undefined. The integral is improper and diverges.
To use the Fundamental Theorem you need antiderivatives, the reverse of differentiation. This lesson builds the table of basic antiderivatives and the notation for the family of all antiderivatives, the indefinite integral.
FUN-6.C. Determine antiderivatives of functions and indefinite integrals, using knowledge of derivatives.
Essential knowledge.
FUN-6.C.1. \(\displaystyle\int f(x)\,dx\) is an indefinite integral of the function \(f\) and can be expressed as \(\displaystyle\int f(x)\,dx=F(x)+C\), where \(F'(x)=f(x)\) and \(C\) is any constant.
FUN-6.C.2. Differentiation rules provide the foundation for finding antiderivatives.
FUN-6.C.3. Many functions do not have closed-form antiderivatives.
Antiderivatives and the constant
A function \(F\) is an antiderivative of \(f\) on an interval if \(F'(x)=f(x)\) there. If \(F\) is one antiderivative, then so is \(F+C\) for any constant \(C\), because a constant has derivative zero. By the Mean Value Theorem (Lesson 5.1), any two antiderivatives of \(f\) on an interval differ by a constant. So the whole family is written with the indefinite integral: \[\int f(x)\,dx=F(x)+C.\] Here \(C\) is the constant of integration. An indefinite integral is a family of functions; a definite integral is a number. Differentiation and antidifferentiation undo each other: \(\dfrac{d}{dx}\displaystyle\int f(x)\,dx=f(x)\).
The basic rules
Each of these is read off from a derivative you already know (check by differentiating the right side).
\(f(x)\)
\(\displaystyle\int f(x)\,dx\)
\(k\) (constant)
\(kx+C\)
\(x^n\), \(n\ne-1\)
\(\dfrac{x^{n+1}}{n+1}+C\)
\(\dfrac1x\)
\(\ln\lvert x\rvert+C\)
\(e^x\)
\(e^x+C\)
\(a^x\) (\(a\gt0\), \(a\ne1\))
\(\dfrac{a^x}{\ln a}+C\)
\(\sin x\)
\(-\cos x+C\)
\(\cos x\)
\(\sin x+C\)
\(\sec^2x\)
\(\tan x+C\)
\(\csc^2x\)
\(-\cot x+C\)
\(\sec x\tan x\)
\(\sec x+C\)
\(\csc x\cot x\)
\(-\csc x+C\)
\(\dfrac1{\sqrt{1-x^2}}\)
\(\arcsin x+C\)
\(\dfrac{1}{1+x^2}\)
\(\arctan x+C\)
Constants factor out and sums split: \(\int\big[af(x)+bg(x)\big]dx=a\int f+b\int g\). To use the power rule on a root or a reciprocal power, rewrite first: \(\sqrt x=x^{1/2}\), \(\dfrac1{x^3}=x^{-3}\).
In the symbol \(\displaystyle\int f(x)\,dx\) the \(dx\) is not decoration: it tells the variable of integration. Do not forget \(+C\) in an indefinite integral; leaving it out loses credit on the exam. The antiderivative of \(\dfrac1x\) is \(\ln\lvert x\rvert\) (with absolute value) so that it is valid for negative \(x\).
Exam focus
\(\int f\,dx\) is a family of antiderivatives \(F(x)+C\); check by differentiating.
Power rule for integrals: raise the exponent by 1, divide by the new exponent (not for \(n=-1\)).
Know the table: \(\sin\), \(\cos\), \(\sec^2\), \(e^x\), \(1/x\), arcsine and arctangent forms.
Rewrite roots, quotients and powers before integrating.
An initial condition determines \(C\); always add \(+C\) on an indefinite integral.
Practice
Find \(\displaystyle\int(4x^3-6x+7)\,dx\).
Show answer
\(x^4-3x^2+7x+C\).
Find \(\displaystyle\int\left(\frac{5}{x}+3e^x\right)dx\).
Show answer
\(5\ln\lvert x\rvert+3e^x+C\).
Find \(\displaystyle\int\frac{2}{\sqrt{x}}\,dx\).
Show answer
The chain rule makes derivatives of composite functions come out with an "inside derivative" factor. Reading that backward gives a technique for integrals: when an integrand contains a function and (a constant multiple of) its inner derivative, substitution (also called u-substitution) turns a hard integral into a basic one.
Required content · FUN-6.D.1, FUN-6.D.2
Learning objective.
FUN-6.D. For integrands requiring substitution or rearrangements into equivalent forms: (a) determine indefinite integrals; (b) evaluate definite integrals.
Essential knowledge.
FUN-6.D.1. Substitution of variables is a technique for finding antiderivatives.
FUN-6.D.2. For a definite integral, substitution of variables requires corresponding changes to the limits of integration.
The idea
Recall \(\dfrac{d}{dx}F\big(u(x)\big)=F'\big(u(x)\big)\,u'(x)\). Therefore \[\int f\big(u(x)\big)\,u'(x)\,dx=\int f(u)\,du=F(u)+C=F\big(u(x)\big)+C.\] In practice: choose \(u\) (usually the "inside" function, or something whose derivative appears), compute \(du=u'(x)\,dx\), replace everything in terms of \(u\), integrate, and substitute back.
The steps
Choose \(u=g(x)\), typically the inner function of a composite, the expression under a root, in a denominator, or in an exponent.
Compute \(du=g'(x)\,dx\) and solve for \(dx\) or for the leftover piece (a constant multiple is fine to adjust).
Rewrite the entire integral in terms of \(u\). No \(x\) may remain.
Integrate with respect to \(u\).
Replace \(u\) by \(g(x)\) (for an indefinite integral).
Worked example · The basic pattern
\(\displaystyle\int2x\cos(x^2)\,dx\). Let \(u=x^2\), \(du=2x\,dx\). Then \(\displaystyle\int\cos u\,du=\sin u+C=\sin(x^2)+C\).
Worked example · An adjusting constant
\(\displaystyle\int x\,e^{x^2+1}\,dx\). Let \(u=x^2+1\), \(du=2x\,dx\), so \(x\,dx=\tfrac12du\). Then \(\tfrac12\displaystyle\int e^u\,du=\tfrac12e^u+C=\tfrac12e^{x^2+1}+C\).
Worked example · A quotient with a logarithm
\(\displaystyle\int\frac{\ln x}{x}\,dx\). Let \(u=\ln x\), \(du=\dfrac1x\,dx\). Then \(\displaystyle\int u\,du=\tfrac{u^2}{2}+C=\tfrac12(\ln x)^2+C\).
Worked example · Tangent
\(\displaystyle\int\tan x\,dx=\int\frac{\sin x}{\cos x}\,dx\). Let \(u=\cos x\), \(du=-\sin x\,dx\). Then \(-\displaystyle\int\frac{du}{u}=-\ln\lvert u\rvert+C=-\ln\lvert\cos x\rvert+C\).
Definite integrals: change the limits
For \(\displaystyle\int_a^bf(u(x))\,u'(x)\,dx\), you can either substitute back to \(x\) before evaluating, or (usually easier) change the limits to \(u(a)\) and \(u(b)\) and never return to \(x\).
Worked example · New limits
\(\displaystyle\int_0^2x\sqrt{x^2+1}\,dx\). Let \(u=x^2+1\), \(du=2x\,dx\). When \(x=0\), \(u=1\); when \(x=2\), \(u=5\). Then \[\tfrac12\int_1^5\sqrt u\,du=\tfrac12\Big[\tfrac23u^{3/2}\Big]_1^5=\tfrac13\big(5\sqrt5-1\big)\approx3.393.\]
Worked example · Trigonometric limits
\(\displaystyle\int_0^{\pi/2}\sin x\cos^3x\,dx\). Let \(u=\cos x\), \(du=-\sin x\,dx\). At \(x=0\), \(u=1\); at \(x=\pi/2\), \(u=0\). The integral is \(-\displaystyle\int_1^0u^3\,du=\displaystyle\int_0^1u^3\,du=\tfrac14\).
Worked example · A rate application
The rate at which a company's revenue grows is \(R'(t)=\dfrac{12t}{t^2+4}\) thousand dollars per month. Find the revenue gained from \(t=0\) to \(t=2\).
\(u=t^2+4\), \(du=2t\,dt\): \(\displaystyle\int\frac{12t}{t^2+4}\,dt=6\int\frac{du}{u}=6\ln u\). From \(u=4\) to \(u=8\): \(6\ln8-6\ln4=6\ln2\approx4.159\) thousand dollars.
When substitution works, and checking
A good \(u\) makes the leftover integrand a simple function of \(u\). If \(x\) remains after substitution, try another \(u\) (or another technique). Always check an indefinite result by differentiating.
Exam focus
Look for a function and its derivative (up to a constant) in the integrand.
\(du=u'(x)\,dx\); adjust constants as needed.
Convert everything to \(u\); for definite integrals, convert the limits too.
Substitute back for indefinite integrals; check by differentiating.
\(\int\tan x\,dx=-\ln\lvert\cos x\rvert+C\).
Practice
Find \(\displaystyle\int(3x+1)^5\,dx\).
Show answer
Find \(\displaystyle\int\frac{e^x}{1+e^x}\,dx\).
Show answer
\(u=1+e^x\), \(du=e^xdx\): \(\ln(1+e^x)+C\).
Evaluate \(\displaystyle\int_0^1\frac{x}{x^2+1}\,dx\).
Show answer
\(u=x^2+1\), \(\tfrac12du=x\,dx\); limits 1 to 2: \(\tfrac12\ln2\approx0.3466\).
Evaluate \(\displaystyle\int_1^e\frac{(\ln x)^2}{x}\,dx\).
Show answer
\(u=\ln x\), limits 0 to 1: \(\displaystyle\int_0^1u^2\,du=\tfrac13\).
A student integrates \(\displaystyle\int_0^1 2x(x^2+1)^3\,dx\) with \(u=x^2+1\) but keeps the limits 0 and 1 for \(u\), getting \(\big[\tfrac{u^4}{4}\big]_0^1\). What is the mistake and the correct answer?
Show answer
The limits must be changed to the \(u\)-values: \(u(0)=1\), \(u(1)=2\). The correct result is \(\Big[\tfrac{u^4}{4}\Big]_1^2=\tfrac{16}{4}-\tfrac14=\tfrac{15}{4}\).
Some integrals do not match the basic table or a simple substitution until the integrand is rewritten. Two algebraic moves cover many of them: long division for improper rational functions (degree of the top at least the degree of the bottom) and completing the square for quadratic denominators or radicands.
Required content · FUN-6.D.3
Learning objective.
FUN-6.D. For integrands requiring substitution or rearrangements into equivalent forms: (a) determine indefinite integrals; (b) evaluate definite integrals.
Essential knowledge.
FUN-6.D.3. Techniques for finding antiderivatives include rearrangements into equivalent forms, such as long division and completing the square.
Long division
If a rational function has numerator degree at least the denominator degree, divide first to get a polynomial plus a proper fraction; the pieces are integrable.
Worked example · A linear denominator
\(\displaystyle\int\frac{x^2+1}{x-1}\,dx\). Divide: \(x^2+1=(x-1)(x+1)+2\), so \(\dfrac{x^2+1}{x-1}=x+1+\dfrac{2}{x-1}\). Then \[\int\left(x+1+\frac{2}{x-1}\right)dx=\frac{x^2}{2}+x+2\ln\lvert x-1\rvert+C.\]
Worked example · A definite integral
\(\displaystyle\int_2^3\frac{x}{x-1}\,dx\). \(\dfrac{x}{x-1}=1+\dfrac1{x-1}\). The integral is \(\Big[x+\ln\lvert x-1\rvert\Big]_2^3=(3+\ln2)-(2+\ln1)=1+\ln2\approx1.693\).
Completing the square
A quadratic \(x^2+bx+c\) can be written \((x+\tfrac b2)^2+\big(c-\tfrac{b^2}{4}\big)\). When the result is \((x-h)^2+k^2\) in a denominator, the arctangent form applies; when \(k^2-(x-h)^2\) is under a square root in a denominator, the arcsine form applies. Use \(u=x-h\) (a shift).
The forms to recognize: \[\int\frac{du}{u^2+a^2}=\frac1a\arctan\frac ua+C,\qquad\int\frac{du}{\sqrt{a^2-u^2}}=\arcsin\frac ua+C.\]
Worked example · Arctangent
\(\displaystyle\int\frac{dx}{x^2+4x+13}\). Complete the square: \(x^2+4x+13=(x+2)^2+9\). With \(u=x+2\): \[\int\frac{du}{u^2+9}=\frac13\arctan\frac u3+C=\frac13\arctan\frac{x+2}{3}+C.\]
Worked example · Arcsine
\(\displaystyle\int\frac{dx}{\sqrt{8x-x^2-7}}\). \(8x-x^2-7=9-(x-4)^2\). With \(u=x-4\): \(\displaystyle\int\frac{du}{\sqrt{9-u^2}}=\arcsin\frac u3+C=\arcsin\frac{x-4}{3}+C\).
Worked example · Definite, with numbers
\(\displaystyle\int_0^1\frac{dx}{x^2+2x+2}\). \(x^2+2x+2=(x+1)^2+1\). The integral is \(\Big[\arctan(x+1)\Big]_0^1=\arctan2-\arctan1=\arctan2-\dfrac\pi4\approx1.1071-0.7854=0.3218\).
Worked example · Combine with substitution
\(\displaystyle\int\frac{2x+6}{x^2+6x+13}\,dx\). The numerator is the derivative of the denominator, so with \(u=x^2+6x+13\): \(\ln(x^2+6x+13)+C\) (the quadratic is always positive).
Exam focus
Improper rational function: divide first.
Quadratic denominator with no real factors: complete the square and use arctangent; \(\sqrt{a^2-u^2}\): arcsine.
Shift with \(u=x-h\); watch the \(\tfrac1a\) factor in the arctangent form.
If the numerator is the derivative of the denominator, the answer is a logarithm.
Check by differentiating.
Practice
Find \(\displaystyle\int\frac{x^2}{x+1}\,dx\).
Show answer
\(x^2=(x+1)(x-1)+1\), so \(\dfrac{x^2}{x+1}=x-1+\dfrac1{x+1}\). Integral: \(\tfrac{x^2}{2}-x+\ln\lvert x+1\rvert+C\).
Find \(\displaystyle\int\frac{dx}{x^2+6x+10}\).
Show answer
\((x+3)^2+1\): \(\arctan(x+3)+C\).
Find \(\displaystyle\int\frac{dx}{\sqrt{4-(x-1)^2}}\).
Show answer
\(\arcsin\dfrac{x-1}{2}+C\).
Evaluate \(\displaystyle\int_0^2\frac{dx}{x^2+4}\).
Show answer
You now have several tools for finding antiderivatives. The final skill is choosing, quickly, the right one. This lesson is a decision guide plus practice mixing the methods.
A decision guide
Look at the integrand and ask, in order:
Is it a basic form? Check the table in Lesson 6.8: powers, \(e^x\), \(\sin x\), \(\cos x\), \(\sec^2x\), \(\dfrac1x\), \(\dfrac1{1+x^2}\), \(\dfrac1{\sqrt{1-x^2}}\). Rewrite roots and reciprocals as powers first.
Is there a function and its derivative (up to a constant)? Try substitution (\(\dfrac{\ln x}{x}\), \(xe^{x^2}\), \(\dfrac{2x}{x^2+1}\), \(\sin x\cos^2x\)).
Is it a rational function? If the degree of the top is at least the degree of the bottom, divide. If the denominator is an irreducible quadratic, complete the square.
Is the result a number you can check? Differentiate your antiderivative to verify, or use a calculator's numerical integral to confirm a definite integral.
Some integrals combine moves (substitute first, then parts), and some have no elementary antiderivative at all, such as \(\displaystyle\int e^{-x^2}\,dx\) or \(\displaystyle\int\frac{\sin x}{x}\,dx\). In such cases the Fundamental Theorem (part 1) still lets you define and differentiate the function \(\int_0^xe^{-t^2}\,dt\), and a calculator or Riemann sum evaluates definite integrals.
The first has a function and (half) its derivative: \(u=x^2+1\) gives \(\tfrac12\ln2\). The second is the arctangent form: \(\arctan1-\arctan0=\tfrac\pi4\). The sum is \(\tfrac12\ln2+\tfrac\pi4\approx0.3466+0.7854=1.1320\).
Worked example · A division that makes substitution possible
\(\displaystyle\int\frac{x^2+3x}{x+1}\,dx\): divide: \(x^2+3x=(x+1)(x+2)-2\), so the integrand is \(x+2-\dfrac2{x+1}\). The integral is \(\tfrac{x^2}2+2x-2\ln\lvert x+1\rvert+C\).
Exam focus
Try, in order: basic form, rewrite, substitution, division or completing the square.
Substitution is the first thing to look for when a function and its derivative appear.
Some functions have no elementary antiderivative; use the FTC to define them and a calculator to evaluate definite integrals.
Always verify by differentiating (indefinite) or with a numerical integral (definite).
Practice
Which method fits \(\displaystyle\int\frac{e^{\sqrt x}}{\sqrt x}\,dx\)? Evaluate it.
Show answer
Substitution with \(u=\sqrt x\), \(du=\dfrac{1}{2\sqrt x}dx\): \(2\displaystyle\int e^u\,du=2e^{\sqrt x}+C\).
Evaluate \(\displaystyle\int\frac{x^2+4}{x}\,dx\).
Show answer
Evaluate \(\displaystyle\int\sin^2x\cos x\,dx\).
Show answer
\(u=\sin x\): \(\tfrac{\sin^3x}{3}+C\).
Evaluate \(\displaystyle\int\frac{x^2}{x+1}\,dx\) by choosing a method.
Show answer
Long division: \(x-1+\dfrac1{x+1}\); \(\tfrac{x^2}{2}-x+\ln\lvert x+1\rvert+C\).
Explain why \(\displaystyle\int e^{-x^2}\,dx\) cannot be found by these methods, and how you can still work with \(g(x)=\displaystyle\int_0^xe^{-t^2}\,dt\).
Show answer
The function has no elementary antiderivative. By FTC part 1, \(g'(x)=e^{-x^2}\), so you can analyze \(g\) (increasing everywhere, concave down for \(x\gt0\)), and a calculator evaluates \(g(b)\) numerically.
Unit 6 practice · 10 problems
Unit 6 practice: Integration and Accumulation of Change
Ten problems covering the whole unit. Work each one on paper before
revealing the answer. No calculator needed.
Compute the left Riemann sum for \(f(x) = x^2 + 1\) on \([0, 4]\) with four equal subintervals. Is it an over- or underestimate?
Show answer
\(L_4 = f(0) + f(1) + f(2) + f(3) = 1 + 2 + 5 + 10 = 18\). \(f\) is increasing, so this underestimates.
Use a trapezoidal sum with the table to estimate \(\displaystyle\int_0^6 v(t)\,dt\): \(t = 0, 1, 3, 6\); \(v(t) = 10, 14, 20, 26\).
Divide: \(x^2 + 2x + 3 = (x + 1)^2 + 2\), so the integrand is \(x + 1 + \dfrac{2}{x + 1}\). Result: \(\dfrac{x^2}{2} + x + 2\ln|x + 1| + C\).
Given \(\displaystyle\int_{-2}^{4} f(x)\,dx = 10\) and \(\displaystyle\int_{1}^{4} f(x)\,dx = 3\), find \(\displaystyle\int_{-2}^{1}\big(2f(x) + 3\big)\,dx\).
Show answer
\(\displaystyle\int_{-2}^{1} f = 10 - 3 = 7\). Then \(2(7) + 3(3) = 23\).
\(f(t) = 2\) on \([0, 3]\), then drops linearly from \((3, 2)\) to \((5, -2)\), then stays at \(-2\) on \([5, 7]\). Let \(g(x) = \displaystyle\int_0^x f(t)\,dt\). Find \(g(5)\), and the \(x\) in \([0, 7]\) where \(g\) is greatest.
Show answer
\(g(3) = 6\). From 3 to 4, \(f\) goes from 2 to 0 (area \(+1\)); from 4 to 5, from 0 to \(-2\) (area \(-1\)). So \(g(5) = 6\). \(g\) is greatest where \(f\) changes from positive to negative: \(x = 4\), with \(g(4) = 7\).
Many real systems are easier to describe by how they change than by a formula for what they are. A hot drink cools faster when it is much hotter than the room. A population grows faster when there are more individuals. Each of these statements is an equation involving a derivative: a differential equation.
Required content · FUN-7.A.1
Learning objective.
FUN-7.A. Interpret verbal statements of problems as differential equations involving a derivative expression.
Essential knowledge.
FUN-7.A.1. Differential equations relate a function to one or more of its derivatives.
What a differential equation is
A differential equation is an equation involving an unknown function and one or more of its derivatives, such as \(\dfrac{dy}{dx}=3y\) or \(\dfrac{dP}{dt}=0.2P\). A solution is a function that makes the equation true for all \(x\) in an interval. Differential equations describe how a quantity changes, and solving one means finding the quantity itself. A differential equation is first order if it involves only the first derivative.
Writing an equation from a description
The key is to translate language about rates into symbols:
"The rate of change of \(y\)" \(\to\dfrac{dy}{dt}\).
"is proportional to \(y\)" \(\to\dfrac{dy}{dt}=ky\) for some constant \(k\).
"is proportional to the difference between \(y\) and 20" \(\to\dfrac{dy}{dt}=k(y-20)\).
"is inversely proportional to \(y\)" \(\to\dfrac{dy}{dt}=\dfrac ky\).
"equals inflow minus outflow" \(\to\dfrac{dQ}{dt}=\text{(rate in)}-\text{(rate out)}\).
Pay attention to signs: a decreasing quantity has \(\dfrac{dy}{dt}\lt0\). If a positive quantity decays at a rate proportional to its size, \(\dfrac{dy}{dt}=-ky\) with \(k\gt0\).
Worked example · Proportional growth
A bacterial culture grows at a rate proportional to its size \(B\). Write a differential equation.
\(\dfrac{dB}{dt}=kB\), where \(k\gt0\) is the constant of proportionality.
Worked example · Cooling
The rate at which the temperature \(T\) of a cup of coffee changes is proportional to the difference between \(T\) and the room temperature of \(70^\circ\)F. Write the equation.
\(\dfrac{dT}{dt}=k(T-70)\). If the coffee is hotter than the room, \(T-70\gt0\) and the temperature must decrease, so \(k\lt0\).
Worked example · A tank with inflow and outflow
A tank holds 200 gallons of water containing salt. Brine with 0.5 pounds of salt per gallon flows in at 4 gallons per minute, and the well-mixed solution flows out at 4 gallons per minute. Let \(S(t)\) be the pounds of salt at time \(t\). Write a differential equation.
Salt in: \(0.5\cdot4=2\) pounds per minute. Salt out: concentration \(\dfrac{S}{200}\) pounds per gallon times 4 gallons per minute, which is \(\dfrac{4S}{200}=\dfrac{S}{50}\). So \(\dfrac{dS}{dt}=2-\dfrac{S}{50}\).
Worked example · Rate depends on position
A particle moves along a line so that its velocity is inversely proportional to its position \(y\). Then \(\dfrac{dy}{dt}=\dfrac ky\).
Interpreting an equation
Once an equation is written, you can often learn something without solving it. In \(\dfrac{dT}{dt}=-0.1(T-70)\), when \(T=90\) the rate is \(-2\) degrees per minute (cooling), when \(T=70\) the rate is 0 (equilibrium), and when \(T=50\) the rate is \(+2\) (warming toward the room).
Exam focus
A differential equation relates a function to its derivatives; a solution is a function that satisfies it.
Translate "rate" to \(\dfrac{dy}{dt}\), "proportional to" to a constant multiple, and attend to the sign.
Inflow minus outflow models mixing and accumulation.
Evaluate the right-hand side at given values to interpret the rate.
Practice
The rate of change of a population \(P\) is proportional to \(P\). Write the equation. If \(P\) is decreasing, what is the sign of the constant?
Show answer
\(\dfrac{dP}{dt}=kP\). If \(P\) is decreasing (and positive), \(k\lt0\).
Write a differential equation for the amount \(A\) of a drug in the body if the body eliminates the drug at a rate proportional to the amount present.
Show answer
\(\dfrac{dA}{dt}=-kA\), with \(k\gt0\).
A water tank has \(V\) gallons. Water enters at 20 gallons per minute and leaves at a rate proportional to the square root of \(V\). Write the differential equation.
Show answer
\(\dfrac{dV}{dt}=20-k\sqrt V\).
For \(\dfrac{dy}{dt}=0.5(100-y)\), find the rate when \(y=60\) and when \(y=100\). What is happening to \(y\) in each case?
Show answer
At \(y=60\): \(0.5(40)=20\), increasing. At \(y=100\): \(0\), no change (equilibrium).
The salt example above has \(\dfrac{dS}{dt}=2-\dfrac S{50}\). Find the value of \(S\) at which the amount of salt stops changing.
Show answer
\(2-\dfrac S{50}=0\) gives \(S=100\) pounds.
A student writes "the rate of change of \(y\) is proportional to \(x\)" as \(\dfrac{dy}{dx}=kxy\). What is the error?
Show answer
Proportional to \(x\) means \(\dfrac{dy}{dx}=kx\). The extra factor \(y\) would say the rate is proportional to the product of \(x\) and \(y\).
You can often check that a function solves a differential equation without solving the equation at all: substitute the function and its derivatives and see if both sides agree. This is the verification skill, and it is tested directly.
Required content · FUN-7.B.1, FUN-7.B.2
Learning objective.
FUN-7.B. Verify solutions to differential equations.
Essential knowledge.
FUN-7.B.1. Derivatives can be used to verify that a function is a solution to a given differential equation.
FUN-7.B.2. There may be infinitely many solutions to a differential equation.
The method
To verify that \(y=f(x)\) is a solution of a differential equation:
Compute the derivatives of \(f\) that appear in the equation.
Substitute \(y\) and its derivatives into both sides of the equation.
Simplify and confirm that the two sides are equal for all \(x\) in the interval.
If the problem also gives an initial condition such as \(y(0)=1\), check that \(f\) satisfies it as well. Such a problem is an initial value problem, and the solution that satisfies the condition is a particular solution. A solution with an arbitrary constant, such as \(y=Ce^{2x}\), is a general solution.
Worked example · Verifying a family
Show that \(y=Ce^{2x}\) is a solution of \(\dfrac{dy}{dx}=2y\) for every constant \(C\).
\(\dfrac{dy}{dx}=2Ce^{2x}\), and \(2y=2Ce^{2x}\). The two sides are equal, so the function is a solution.
Worked example · Verifying with an initial condition
Is \(y=x+1\) a solution of the initial value problem \(\dfrac{dy}{dx}=y-x\), \(y(0)=1\)?
\(\dfrac{dy}{dx}=1\) and \(y-x=(x+1)-x=1\): equal. And \(y(0)=1\). Yes.
Worked example · A product on the right
Show that \(y=x^3\) satisfies \(x\dfrac{dy}{dx}=3y\) for \(x\gt0\).
\(\dfrac{dy}{dx}=3x^2\), so \(x\dfrac{dy}{dx}=3x^3\), and \(3y=3x^3\). Equal.
Worked example · A trigonometric solution
Show that \(y=\sin x\) is a solution of \(\dfrac{dy}{dx}=\sqrt{1-y^2}\) for \(0\lt x\lt\dfrac\pi2\).
\(\dfrac{dy}{dx}=\cos x\) and \(\sqrt{1-\sin^2x}=\sqrt{\cos^2x}=\cos x\) (since \(\cos x\gt0\) on that interval). The sides agree.
Worked example · Finding the constant so that the solution works
For what value of \(k\) is \(y=e^{kx}\) a solution of \(\dfrac{dy}{dx}=-4y\)?
\(\dfrac{dy}{dx}=ke^{kx}\) and \(-4y=-4e^{kx}\). So \(k=-4\).
Worked example · A function that is not a solution
Is \(y=x^2\) a solution of \(\dfrac{dy}{dx}=y\)?
\(\dfrac{dy}{dx}=2x\) but \(y=x^2\): these are equal only at \(x=0\) and \(x=2\), not for all \(x\). Not a solution.
Using the equation without solving it
A differential equation gives the derivative in terms of \(x\) and \(y\), so at any point you can find the slope. If \(\dfrac{dy}{dx}=x^2-y\) and a solution passes through \((1,3)\), the slope there is \(1-3=-2\). Differentiating the equation implicitly gives the second derivative: \(\dfrac{d^2y}{dx^2}=2x-\dfrac{dy}{dx}=2x-(x^2-y)\), which at \((1,3)\) is \(2-(-2)=4\gt0\), so the solution curve is concave up there.
Exam focus
Substitute the function and its derivatives into both sides; they must agree for all \(x\).
Check initial conditions too.
A general solution has a constant; a particular solution is fixed by an initial condition.
Slope at a point: put the point into the equation. Concavity: differentiate the equation to get \(\dfrac{d^2y}{dx^2}\).
A function that only matches at certain points is not a solution.
Practice
Verify that \(y=3e^{-x}\) is a solution of \(\dfrac{dy}{dx}=-y\) with \(y(0)=3\).
Show answer
\(y'=-3e^{-x}=-y\) and \(y(0)=3\).
Is \(y=2x^2\) a solution of \(x\dfrac{dy}{dx}=2y\)?
Show answer
\(x\cdot4x=4x^2\), and \(2y=4x^2\). Yes.
Show that \(y=\ln x\) is a solution of \(x\dfrac{dy}{dx}=1\) for \(x\gt0\).
Show answer
\(\dfrac{dy}{dx}=\dfrac1x\), so \(x\cdot\dfrac1x=1\).
Verify that \(y=e^{x}+x\) satisfies \(\dfrac{dy}{dx}=y-x+1\).
Show answer
A student says \(y=e^{-x}\) is a solution of \(\dfrac{dy}{dx}=y\) "because the derivative of \(e^{-x}\) is also an exponential." What is wrong?
Show answer
The derivative is \(-e^{-x}=-y\), not \(y\). The two sides must be equal, not merely similar. \(e^{-x}\) solves \(\dfrac{dy}{dx}=-y\).
Most differential equations cannot be solved with a formula, but you can still see their solutions. A slope field (or direction field) draws the slope that the equation assigns at many points; solution curves must follow these little segments, like a boat on a current. Slope fields are one of the most frequently tested topics on the AP exam.
FUN-7.C. Estimate solutions to differential equations.
Essential knowledge.
FUN-7.C.1. A slope field is a graphical representation of a differential equation on a finite set of points in the plane.
FUN-7.C.2. Slope fields provide information about the behavior of solutions to first-order differential equations.
FUN-7.C.3. Solutions to differential equations are functions or families of functions.
What a slope field is
For \(\dfrac{dy}{dx}=f(x,y)\), a slope field is a grid of short line segments. At each point \((x,y)\), the segment has slope \(f(x,y)\). A solution curve passes through the field tangent to every segment it meets. If the equation is given, you build the field by computing slopes at grid points; if a slope field is given, you can read approximate slopes and sketch solutions.
Figure 7.1. The slope field for \(\dfrac{dy}{dx}=x+y\). Each short segment has the slope the equation assigns to its point. The solution through \((0,1)\), \(y=2e^x-x-1\), follows the segments. The slope is zero along the line \(y=-x\) (where the segments are horizontal) and the same along any line \(x+y=\) constant.
Sketching a slope field
Choose grid points (often integer coordinates).
At each point, substitute \(x\) and \(y\) into \(f(x,y)\).
Draw a short segment with that slope: horizontal for 0, steeply up for large positive, steeply down for large negative; undefined slope gives a vertical segment.
Worked example · Computing slopes
For \(\dfrac{dy}{dx}=x+y\), the slopes at several points are:
Point
\((0,0)\)
\((1,0)\)
\((0,1)\)
\((-1,1)\)
\((1,-2)\)
Slope \(x+y\)
0
1
1
0
\(-1\)
The slope is the same along each line \(x+y=c\). In the figure, the segments are horizontal along the line \(y=-x\), and steeper farther from it.
Patterns to recognize
If \(\dfrac{dy}{dx}\) depends only on \(x\), the segments in each vertical column are parallel; solutions are vertical shifts of each other.
If \(\dfrac{dy}{dx}\) depends only on \(y\), the segments in each horizontal row are parallel (an autonomous equation).
Equilibrium solutions are horizontal lines \(y=c\) where \(\dfrac{dy}{dx}=0\) for all \(x\) (the slope is 0 along the whole line).
Solutions never cross (for nice equations), because each point has one slope.
Worked example · Matching an equation to a field
Which equation has a slope field with horizontal segments along the \(x\)-axis and along \(y=2\) and with positive slope between them? \(\dfrac{dy}{dx}=y(2-y)\): the slope is 0 at \(y=0\) and \(y=2\), positive for \(0\lt y\lt2\), negative for \(y\lt0\) and \(y\gt2\). The solutions at \(y=0\) and \(y=2\) are equilibria.
Reasoning with a slope field
Given an initial condition, sketch the solution by starting at the point and following the segments, drawing a curve tangent to each. From the field you can read:
Increasing or decreasing: the sign of the segments the curve passes through.
Relative extrema: where the solution passes through a segment of slope 0 and the slope changes sign.
Concavity: the solution is concave up where the slopes along it increase as \(x\) increases.
Limits: the behavior of the curve as \(x\to\infty\) (approach an equilibrium, or grow without bound).
Worked example · Slope and concavity from the equation
For \(\dfrac{dy}{dx}=x+y\) with \(y(0)=1\): the slope at \((0,1)\) is 1, so \(y\) is increasing there. Differentiate the equation: \(\dfrac{d^2y}{dx^2}=1+\dfrac{dy}{dx}\). At \((0,1)\): \(1+1=2\gt0\). The solution is concave up at the point, so near \((0,1)\) the curve lies above its tangent line, and the tangent line approximation \(y\approx1+x\) underestimates the solution. (The exact solution is \(y=2e^x-x-1\), which gives \(y(0.2)\approx1.2428\gt1.2\).)
Worked example · Extrema of a solution
For \(\dfrac{dy}{dx}=x(y-1)\) with a solution \(y=f(x)\) through \((0,3)\): the slope at \((0,3)\) is \(0\). \(\dfrac{d^2y}{dx^2}=(y-1)+x\dfrac{dy}{dx}\); at \((0,3)\) it is \(2\gt0\). So \(f\) has a relative minimum at \(x=0\) (slope zero and concave up).
Worked example · Equilibrium and long-term behavior
For \(\dfrac{dy}{dx}=y(4-y)\) and \(y(0)=1\): the slope is positive for \(0\lt y\lt4\), so the solution increases; it cannot cross the equilibrium line \(y=4\), so it approaches it: \(\displaystyle\lim_{x\to\infty}y=4\).
Exam focus
A slope field shows the slope \(f(x,y)\) at each grid point; solutions follow the segments.
Equations in \(x\) only give columns of parallel segments; in \(y\) only, rows.
Equilibrium solutions: \(y=c\) with \(f(x,c)=0\) for all \(x\).
Sketch a solution by tracing the field through the initial point.
Use \(\dfrac{d^2y}{dx^2}\) (differentiate the equation implicitly) for concavity at a point.
Practice
For \(\dfrac{dy}{dx}=x-y\), find the slopes at \((0,0)\), \((1,0)\), \((0,1)\), \((2,2)\).
Show answer
\(0\), \(1\), \(-1\), \(0\).
For \(\dfrac{dy}{dx}=\dfrac{x}{y}\), describe the slope along the \(x\)-axis and along the \(y\)-axis.
Show answer
Along the \(x\)-axis (\(y=0\), \(x\ne0\)) the slope is undefined (vertical segments). Along the \(y\)-axis (\(x=0\), \(y\ne0\)) the slope is 0 (horizontal segments).
For \(\dfrac{dy}{dx}=y-2\), find the equilibrium solution and say what solutions do above and below it.
Show answer
\(y=2\). If \(y\gt2\), \(\dfrac{dy}{dx}\gt0\), so solutions increase (move away from 2); if \(y\lt2\), they decrease (move away). Unstable equilibrium.
For \(\dfrac{dy}{dx}=x^2-y\) with \(y(0)=1\), find the slope and concavity at \((0,1)\).
Show answer
Slope: \(0-1=-1\). \(\dfrac{d^2y}{dx^2}=2x-\dfrac{dy}{dx}\); at \((0,1)\): \(0-(-1)=1\gt0\). Decreasing and concave up.
A solution of \(\dfrac{dy}{dx}=\dfrac{y}{x}\) passes through \((1,2)\). Use the slope field idea to describe the solution.
Show answer
The slope at \((1,2)\) is \(2\); the slope at a point equals the slope of the line from the origin to that point, so the segments point away from the origin. The solution through \((1,2)\) is the line \(y=2x\).
A student says that solution curves of a slope field can cross if they start at different points. Is that right?
Show answer
No. At any point the equation assigns one slope (when \(f\) is continuous and nicely behaved), so there is one tangent direction there. Two solutions through the same point would have to follow the same segment and therefore coincide.
Finding general solutions using separation of variables
Slope fields show the shape of solutions; to get a formula we need to solve the equation. For a large and important class of equations, separation of variables works: gather everything with \(y\) on one side and everything with \(x\) on the other, then integrate both sides.
Required content · FUN-7.D.1, FUN-7.D.2
Learning objective.
FUN-7.D. Determine general solutions to differential equations.
Essential knowledge.
FUN-7.D.1. Some differential equations can be solved by separation of variables.
FUN-7.D.2. Antidifferentiation can be used to find general solutions to differential equations.
Separable equations
A first-order equation is separable if it can be written \(\dfrac{dy}{dx}=g(x)\,h(y)\), a product of a function of \(x\) alone and a function of \(y\) alone. Then \[\frac{1}{h(y)}\,dy=g(x)\,dx\quad\Longrightarrow\quad\int\frac{1}{h(y)}\,dy=\int g(x)\,dx.\] Integrating gives a family of solutions with a constant \(C\) (a general solution), typically written implicitly, and solved for \(y\) when possible. Put the constant on one side only (a single \(+C\)), because the two constants merge.
The steps
Check the equation is separable, and note any values with \(h(y)=0\) (these give constant solutions that dividing might lose).
Separate: all \(y\)'s with \(dy\), all \(x\)'s with \(dx\).
Integrate both sides, adding \(+C\) once.
Solve for \(y\) if possible.
Worked example · Exponential type
\(\dfrac{dy}{dx}=2xy\). Separate: \(\dfrac{dy}{y}=2x\,dx\). Integrate: \(\ln\lvert y\rvert=x^2+C\). Exponentiate: \(\lvert y\rvert=e^Ce^{x^2}\), so \(y=Ae^{x^2}\) where \(A=\pm e^C\) is any nonzero constant. (\(A=0\) gives the constant solution \(y=0\), which also satisfies the equation.)
Worked example · Circles
\(\dfrac{dy}{dx}=-\dfrac{x}{y}\). Separate: \(y\,dy=-x\,dx\). Integrate: \(\tfrac{y^2}{2}=-\tfrac{x^2}{2}+C\), so \(x^2+y^2=2C\): circles centered at the origin (an implicit general solution).
Worked example · Trigonometric right side
\(\dfrac{dy}{dx}=y^2\cos x\). Separate: \(\dfrac{dy}{y^2}=\cos x\,dx\). Integrate: \(-\dfrac1y=\sin x+C\), so \(y=-\dfrac{1}{\sin x+C}\). (The constant solution \(y=0\) is also a solution that was lost in the division.)
Worked example · Arctangent
\(\dfrac{dy}{dx}=1+y^2\). Separate: \(\dfrac{dy}{1+y^2}=dx\). Integrate: \(\arctan y=x+C\), so \(y=\tan(x+C)\).
Worked example · Rate proportional to the difference
\(\dfrac{dy}{dt}=k(y-20)\). Separate: \(\dfrac{dy}{y-20}=k\,dt\); \(\ln\lvert y-20\rvert=kt+C\); \(y-20=Ae^{kt}\); \(y=20+Ae^{kt}\). (The cooling model of Lesson 7.8.)
Which equations are separable?
\(\dfrac{dy}{dx}=x+y\) is not separable: the right side is a sum, not a product of a function of \(x\) and a function of \(y\). Its solution needs another method (the slope field and Euler's method still work). On the AP exam, an equation like \(\dfrac{dy}{dx}=(x+1)y\) or \(\dfrac{dy}{dx}=\dfrac{x}{y^2}\) is separable, while a sum like \(x+y\) is not.
Exam focus
Separable: \(\dfrac{dy}{dx}=g(x)h(y)\); write \(\dfrac{dy}{h(y)}=g(x)\,dx\) and integrate both sides.
One constant \(C\), on one side; solve for \(y\) when possible.
Dividing by \(h(y)\) can lose constant solutions where \(h(y)=0\).
\(\dfrac{dy}{dx}=x+y\) is not separable.
Check the solution by differentiating.
Practice
Find the general solution of \(\dfrac{dy}{dx}=\dfrac{x}{y}\).
Show answer
\(y\,dy=x\,dx\), \(\tfrac{y^2}2=\tfrac{x^2}2+C\), so \(y^2=x^2+C_1\).
Find the general solution of \(\dfrac{dy}{dx}=3y\).
Show answer
\(\ln\lvert y\rvert=3x+C\), \(y=Ae^{3x}\).
Find the general solution of \(\dfrac{dy}{dx}=\dfrac{x^2}{y}\), written as \(y^2=\ldots\)
Show answer
A general solution contains an arbitrary constant. An initial condition such as \(y(2)=5\) picks out one curve from the family, a particular solution. This lesson finishes the process: separate, integrate, apply the condition, solve for \(y\), and state the domain on which the solution is valid.
FUN-7.E. Determine particular solutions to differential equations.
Essential knowledge.
FUN-7.E.1. A general solution may describe infinitely many solutions to a differential equation. There is only one particular solution passing through a given point.
FUN-7.E.2. The function \(F\) defined by \(F(x)=y_0+\displaystyle\int_a^x f(t)\,dt\) is a particular solution to the differential equation \(\dfrac{dy}{dx}=f(x)\), satisfying \(F(a)=y_0\).
FUN-7.E.3. Solutions to differential equations may be subject to domain restrictions.
The process
Use the initial condition right after integrating to find \(C\), before you simplify too much. Then solve for \(y\), taking care to choose the sign or branch that agrees with the initial condition. Finally decide the largest interval containing the initial \(x\)-value on which the formula is defined and satisfies the equation.
Worked example · An exponential solution
Solve \(\dfrac{dy}{dx}=2xy\), \(y(0)=3\).
\(\dfrac{dy}{y}=2x\,dx\); \(\ln\lvert y\rvert=x^2+C\). With \(y(0)=3\): \(\ln3=C\). So \(\ln\lvert y\rvert=x^2+\ln3\); since \(y(0)\gt0\), \(y=3e^{x^2}\), valid for all real \(x\).
Worked example · A root and a sign
Solve \(\dfrac{dy}{dx}=-\dfrac xy\), \(y(0)=-4\).
\(y\,dy=-x\,dx\); \(\tfrac{y^2}2=-\tfrac{x^2}2+C\); \(x^2+y^2=2C\). With \((0,-4)\): \(2C=16\). So \(y^2=16-x^2\). The initial \(y\) is negative, so \(y=-\sqrt{16-x^2}\), valid for \(-4\lt x\lt4\) (the lower semicircle; at \(x=\pm4\), \(y=0\) and the equation is undefined).
Worked example · A domain restriction
Solve \(\dfrac{dy}{dx}=xe^{y}\), \(y(0)=0\).
\(e^{-y}\,dy=x\,dx\); \(-e^{-y}=\tfrac{x^2}2+C\). With \(y(0)=0\): \(-1=C\). So \(e^{-y}=1-\tfrac{x^2}2\), and \(y=-\ln\left(1-\tfrac{x^2}{2}\right)\). The logarithm requires \(1-\tfrac{x^2}2\gt0\), so the domain is \(-\sqrt2\lt x\lt\sqrt2\). As \(x\to\sqrt2\), \(y\to\infty\): the solution blows up.
\(y\,dy=(x+1)\,dx\); \(\tfrac{y^2}{2}=\tfrac{x^2}{2}+x+C\). At \((1,3)\): \(\tfrac92=\tfrac12+1+C\), so \(C=3\). \(y^2=x^2+2x+6\). Since \(y(1)\gt0\), \(y=\sqrt{x^2+2x+6}=\sqrt{(x+1)^2+5}\), defined for all \(x\) (and \(y\gt0\) so the equation is defined).
Worked example · Using the solution
Given \(\dfrac{dy}{dx}=\dfrac{y}{x}\) with \(y(1)=2\) and \(x\gt0\): \(\dfrac{dy}{y}=\dfrac{dx}x\), \(\ln y=\ln x+C\); at \((1,2)\), \(C=\ln2\); \(y=2x\). Then \(y(5)=10\).
Common mistakes
Forgetting \(+C\), or putting it on both sides.
Applying the initial condition before integrating (it belongs after).
Using the wrong sign when taking a square root; the initial condition chooses the sign.
Ignoring the domain; a particular solution is valid only on an interval containing the initial \(x\)-value on which it is differentiable and the equation makes sense.
Exam focus
Separate, integrate, apply the initial condition to find \(C\), then solve for \(y\).
The initial condition determines the sign of a root and the sign of \(\lvert y\rvert\).
State the domain: the interval containing the initial \(x\) on which the solution exists.
Check by substituting back into the equation and the initial condition.
Practice
Solve \(\dfrac{dy}{dx}=\dfrac{y}{3}\), \(y(0)=6\).
Show answer
Solve \(\dfrac{dy}{dx}=y^2\), \(y(0)=1\), and state the domain.
Show answer
\(\dfrac{dy}{y^2}=dx\); \(-\dfrac1y=x+C\); \(-1=C\). \(y=\dfrac1{1-x}\), valid for \(x\lt1\) (the solution blows up at \(x=1\)).
Solve \(\dfrac{dy}{dx}=\dfrac{\cos x}{y}\), \(y(0)=2\).
Show answer
\(y\,dy=\cos x\,dx\); \(\tfrac{y^2}{2}=\sin x+C\); \(2=C\); \(y^2=2\sin x+4\); \(y=\sqrt{2\sin x+4}\) (positive, since \(y(0)=2\gt0\)), defined for all \(x\).
A student solves \(\dfrac{dy}{dx}=\dfrac{x}{y}\), \(y(0)=-2\), and writes \(y=\sqrt{x^2+4}\). What is wrong?
Show answer
\(y^2=x^2+4\) is right, but \(y(0)=-2\) is negative, so the solution is \(y=-\sqrt{x^2+4}\).
Growth and decay that are proportional to the amount present are among the most important models in science: unconstrained populations, radioactive decay, compound interest, and cooling objects. Separation of variables solves them all with one pattern.
FUN-7.F. Interpret the meaning of a differential equation and its variables in context.
FUN-7.G. Determine general and particular solutions for problems involving differential equations in context.
Essential knowledge.
FUN-7.F.1. Specific applications of finding general and particular solutions to differential equations include motion along a line and exponential growth and decay.
FUN-7.F.2. The model for exponential growth and decay that arises from the statement \"The rate of change of a quantity is proportional to the size of the quantity\" is \(\dfrac{dy}{dt}=ky\).
FUN-7.G.1. The exponential growth and decay model, \(\dfrac{dy}{dt}=ky\), with initial condition \(y=y_0\) when \(t=0\), has solutions of the form \(y=y_0e^{kt}\).
The model
If a quantity \(y\) changes at a rate proportional to itself, \(\dfrac{dy}{dt}=ky\), the solutions are the exponential functions \[y=y_0e^{kt},\qquad y_0=y(0).\] (Separate: \(\dfrac{dy}{y}=k\,dt\); \(\ln\lvert y\rvert=kt+C\); \(y=Ae^{kt}\), and \(A=y(0)\).) If \(k\gt0\) the quantity grows; if \(k\lt0\), it decays. The doubling time of a growing quantity is \(\dfrac{\ln2}{k}\); the half-life of a decaying quantity is \(\dfrac{\ln2}{\lvert k\rvert}\), and neither depends on the starting amount.
Worked example · Population growth
A population grows at a rate \(\dfrac{dP}{dt}=0.03P\) (\(t\) in years), with \(P(0)=500\). Find \(P(t)\), the population after 50 years, and the doubling time.
\(P(t)=500e^{0.03t}\). \(P(50)=500e^{1.5}\approx2240.8\). Doubling time \(\dfrac{\ln2}{0.03}\approx23.1\) years.
Worked example · Finding \(k\) from data
A bacterial culture grows at a rate proportional to its size. It has 200 cells at \(t=0\) and 500 cells at \(t=3\) hours. How many cells at \(t=5\)?
\(y=200e^{kt}\). \(500=200e^{3k}\) gives \(e^{3k}=2.5\), \(k=\dfrac{\ln2.5}{3}\approx0.3054\). At \(t=5\): \(y=200e^{5k}=200(2.5)^{5/3}\approx921\) cells.
Worked example · Decay and half-life
Carbon-14 has a half-life of 5730 years. What fraction remains after 2000 years?
\(\dfrac{dy}{dt}=-ky\), with \(y=y_0e^{-kt}\); half-life: \(e^{-5730k}=\tfrac12\), so \(k=\dfrac{\ln2}{5730}\approx1.21\times10^{-4}\). After 2000 years: \(e^{-2000k}=2^{-2000/5730}\approx0.785\), about 78.5 percent remains.
Newton's law of cooling
Newton's law of cooling says an object's temperature changes at a rate proportional to the difference between its temperature and the temperature \(A\) of its surroundings: \(\dfrac{dT}{dt}=k(T-A)\). The solution is \(T=A+(T_0-A)e^{kt}\), with \(k\lt0\): the temperature approaches \(A\).
Worked example · Cooling coffee
A cup of coffee at \(90^\circ\)C is placed in a room at \(20^\circ\)C. Suppose \(\dfrac{dT}{dt}=-0.2(T-20)\) with \(t\) in minutes. Find \(T(t)\), the temperature after 5 minutes, and when the coffee reaches \(40^\circ\)C.
\(T-20=70e^{-0.2t}\), so \(T(t)=20+70e^{-0.2t}\). \(T(5)=20+70e^{-1}\approx45.75^\circ\)C. Set \(20+70e^{-0.2t}=40\): \(e^{-0.2t}=\tfrac27\), \(t=5\ln3.5\approx6.26\) minutes.
Worked example · Interpreting the equation without solving
For \(\dfrac{dy}{dt}=-0.5y\), is the rate of decrease larger when \(y=10\) or when \(y=2\)? At \(y=10\), \(\dfrac{dy}{dt}=-5\); at \(y=2\), \(-1\). The decay is faster when the quantity is larger: the rate is proportional to the amount present.
Reading problems
Look for "proportional to the amount present" (pure exponential) and "proportional to the difference from a fixed value" (shifted exponential). In the second case, the variable \(y-A\) is the one that grows or decays exponentially, and \(y\) approaches the equilibrium \(A\) if \(k\lt0\).
Exam focus
\(\dfrac{dy}{dt}=ky\) has solutions \(y=y_0e^{kt}\); \(k\gt0\) growth, \(k\lt0\) decay.
Doubling time \(\dfrac{\ln2}{k}\); half-life \(\dfrac{\ln2}{\lvert k\rvert}\).
Use two data points to find \(k\): divide, take logarithms.
A student claims that if \(\dfrac{dy}{dt}=ky\), the quantity doubles every year if it doubles in the first year, regardless of the starting amount. Is that right?
Show answer
Yes. The doubling time \(\dfrac{\ln2}{k}\) does not depend on \(y_0\), so every doubling takes the same time.
A loan balance \(B\) has \(\dfrac{dB}{dt}=0.05B-1200\) (5 percent interest, payments of \$1200 per year). Find the equilibrium balance and say what happens above and below it.
Show answer
\(0.05B-1200=0\): \(B=24{,}000\). If \(B\gt24{,}000\), \(\dfrac{dB}{dt}\gt0\) and the debt grows without bound; if \(B\lt24{,}000\), it decreases (and the loan is paid off).
Unit 7 practice · 10 problems
Unit 7 practice: Differential Equations
Ten problems covering the whole unit. Work each one on paper before
revealing the answer. A calculator is fine for the final arithmetic in
problems 8 and 9.
Write a differential equation: the temperature \(T\) of a pie changes at a rate proportional to the difference between \(T\) and the 75° room.
Show answer
Translate phrase by phrase: the rate of change of \(T\) is \(\dfrac{dT}{dt}\); proportional to introduces a constant \(k\); the difference between \(T\) and the room is \(T - 75\). So \(\dfrac{dT}{dt} = k(T - 75)\), with \(k\) negative as the pie cools.
Verify that \(y = 2e^{3x} - 1\) is a solution of \(y' = 3y + 3\).
A population satisfies \(\dfrac{dP}{dt} = 0.02P\) with \(P(0) = 1500\). When does it reach 3000?
Show answer
\(P = 1500e^{0.02t} = 3000 \Rightarrow t = \dfrac{\ln 2}{0.02} \approx 34.7\) years.
In a slope field, every segment in a given vertical column is parallel, and the segments along the \(y\)-axis are horizontal. Which equation matches?
If \(y' = y\), the slope depends on \(y\) alone, so the segments would be parallel along horizontal rows, not vertical columns, and along the \(y\)-axis the slopes would run from negative to positive instead of all being horizontal.
Parallel segments within each vertical column mean the slope depends on \(x\) only, and horizontal segments where \(x = 0\) mean the slope is zero there. \(y' = x\) satisfies both: it is constant on any vertical line and equals zero exactly on the \(y\)-axis.
\(y' = xy\) does give zero slope on the \(y\)-axis (where \(x = 0\)), which is why it tempts, but within a vertical column the slope still changes with \(y\), so the segments in a column would not be parallel.
\(y' = x - y\) depends on both variables, so the segments in a column tilt differently as \(y\) changes, and on the \(y\)-axis the slope is \(-y\), which is horizontal only at the origin.
Finding the average value of a function on an interval
The average of a list of numbers is their sum divided by how many there are. A function takes infinitely many values on an interval, but the integral gives a way to "add them all up" and divide by the length of the interval. The result is the average value of the function.
Required content · CHA-4.B.1
Learning objective.
CHA-4.B. Determine the average value of a function using definite integrals.
Essential knowledge.
CHA-4.B.1. The average value of a continuous function \(f\) over an interval \([a,b]\) is \(\dfrac{1}{b-a}\displaystyle\int_a^b f(x)\,dx\).
The definition
The average value of \(f\) on \([a,b]\) is \[f_{\text{avg}}=\frac{1}{b-a}\int_a^bf(x)\,dx.\] Why this makes sense: a Riemann sum divided by the length gives \(\dfrac{1}{b-a}\sum f(x_i)\Delta x=\dfrac1n\sum f(x_i)\) (when \(\Delta x=\dfrac{b-a}{n}\)), the average of \(n\) equally spaced values. Geometrically, the average value is the height of the rectangle on \([a,b]\) whose area equals the area under the curve.
Figure 8.1. The average value of \(f(x)=x(4-x)\) on \([0,4]\) is \(\dfrac14\displaystyle\int_0^4x(4-x)\,dx=\dfrac{8}{3}\). The rectangle of height \(\tfrac83\) and width 4 has the same area as the region under the curve. The graph of \(f\) crosses the average height at \(x=2\pm\tfrac{2}{\sqrt3}\), where \(f(c)=\tfrac83\) as the Mean Value Theorem for integrals promises.
The Mean Value Theorem for integrals
If \(f\) is continuous on \([a,b]\), there is a number \(c\) in \([a,b]\) with \(f(c)=f_{\text{avg}}\). The graph must cross its average height at least once, because a continuous function cannot stay entirely above (or entirely below) its own average. In the figure \(f(x)=x(4-x)\) equals \(\tfrac83\) at two points.
Worked example · A polynomial
Find the average value of \(f(x)=x(4-x)\) on \([0,4]\), and the \(x\)-values where \(f\) equals it.
\(\displaystyle\int_0^4(4x-x^2)\,dx=\Big[2x^2-\tfrac{x^3}3\Big]_0^4=32-\tfrac{64}3=\tfrac{32}3\). Average \(=\dfrac14\cdot\dfrac{32}{3}=\dfrac83\). Solve \(x(4-x)=\tfrac83\): \(x^2-4x+\tfrac83=0\), \(x=2\pm\sqrt{4-\tfrac83}=2\pm\tfrac{2}{\sqrt3}\approx0.845,\,3.155\).
Worked example · A trigonometric function
The average value of \(\sin x\) on \([0,\pi]\): \(\dfrac1\pi\displaystyle\int_0^\pi\sin x\,dx=\dfrac{2}{\pi}\approx0.637\).
Worked example · In context
The temperature \(T(t)=60+20\sin\left(\dfrac{\pi t}{12}\right)\) degrees Fahrenheit models the temperature \(t\) hours after 6 a.m. Find the average temperature from \(t=0\) to \(t=12\).
Worked example · Average rate of change versus average value
The average value of \(f\) on \([a,b]\) is \(\dfrac1{b-a}\displaystyle\int_a^bf\); the average rate of change of \(f\) is \(\dfrac{f(b)-f(a)}{b-a}\). These are different. If \(f=F'\), the second equals the average value of \(F'=f\), by the Fundamental Theorem: the average value of a rate is the average rate of change of the quantity.
Example: if the velocity is \(v(t)\), the average value of \(v\) on \([a,b]\) is \(\dfrac{s(b)-s(a)}{b-a}\), the average velocity.
From a table
If \(f\) is given by a table, approximate the integral with a Riemann or trapezoidal sum and divide by the length.
Worked example · A table
Values of a flow rate \(r(t)\) in liters per minute are \(r(0)=4\), \(r(2)=6\), \(r(5)=7\), \(r(9)=3\). Estimate the average flow rate on \([0,9]\) using a trapezoidal sum.
\(\displaystyle\int_0^9r\approx\tfrac{4+6}{2}(2)+\tfrac{6+7}{2}(3)+\tfrac{7+3}{2}(4)=10+19.5+20=49.5\). Average \(\approx\dfrac{49.5}{9}=5.5\) liters per minute.
Exam focus
\(f_{\text{avg}}=\dfrac1{b-a}\int_a^bf(x)\,dx\); units are those of \(f\).
Geometric meaning: height of the rectangle with the same area.
The MVT for integrals: a continuous \(f\) equals its average at some \(c\).
Average value of a rate = average rate of change of the quantity.
For table data, estimate the integral then divide by \(b-a\).
Practice
Find the average value of \(f(x)=x^2\) on \([0,3]\).
Show answer
\(\dfrac13\cdot\dfrac{27}{3}=3\).
Find the average value of \(f(x)=e^x\) on \([0,2]\).
Show answer
\(\dfrac12(e^2-1)\approx3.195\).
Find the average value of \(v(t)=3t^2-2t\) on \([1,4]\) and interpret if \(v\) is velocity in m/s.
Show answer
\(\dfrac13\Big[t^3-t^2\Big]_1^4=\dfrac13(48-0)=16\) m/s: the average velocity over the three seconds (the displacement divided by 3 s).
The function \(f\) is continuous with \(\displaystyle\int_2^8f(x)\,dx=30\). Find the average value of \(f\) on \([2,8]\), and say what the MVT for integrals guarantees.
Show answer
\(\dfrac{30}{6}=5\). There is a \(c\) in \([2,8]\) with \(f(c)=5\).
Estimate the average value of \(g\) on \([0,10]\) using a left sum with the data \(g(0)=2,g(2)=5,g(5)=6,g(10)=4\).
Show answer
Left sum: \(2(2)+5(3)+6(5)=49\). Average \(\approx\dfrac{49}{10}=4.9\).
A student says "the average value of \(f(x)=x\) on \([0,4]\) is \(\dfrac{f(0)+f(4)}2\) always, for any \(f\)." Is that always true?
Show answer
No. It is true for \(f(x)=x\) (average 2) because the graph is linear, but for \(f(x)=x^2\) on \([0,4]\) the average is \(\dfrac{16}{3}\approx5.33\), not \(\dfrac{0+16}2=8\).
Position, velocity and acceleration with integrals
In Lesson 4.2 you went from position to velocity to acceleration by differentiating. Now you can go the other way with integrals, and answer questions about total distance, position at a given time, and speed from acceleration.
Required content · CHA-4.C.1
Learning objective.
CHA-4.C. Determine values for positions and rates of change using definite integrals in problems involving rectilinear motion.
Essential knowledge.
CHA-4.C.1. For a particle in rectilinear motion over an interval of time, the definite integral of velocity represents the particle's displacement over the interval of time, and the definite integral of speed represents the particle's total distance traveled over the interval of time.
The relationships
For motion along a line with position \(s(t)\), velocity \(v(t)=s'(t)\), and acceleration \(a(t)=v'(t)\): \[v(t)=v(t_0)+\int_{t_0}^{t}a(u)\,du,\qquad s(t)=s(t_0)+\int_{t_0}^{t}v(u)\,du.\] - Displacement from \(t=a\) to \(t=b\): \(\displaystyle\int_a^bv(t)\,dt=s(b)-s(a)\). It is signed. - Total distance traveled: \(\displaystyle\int_a^b\lvert v(t)\rvert\,dt\). Split the interval at the times where \(v=0\) and change sign. - Change in velocity from \(a\) to \(b\): \(\displaystyle\int_a^ba(t)\,dt=v(b)-v(a)\). - Average velocity \(=\dfrac{s(b)-s(a)}{b-a}=\dfrac1{b-a}\displaystyle\int_a^bv\,dt\); average speed \(=\dfrac1{b-a}\displaystyle\int_a^b\lvert v\rvert\,dt\).
Careful: speed is \(\lvert v\rvert\); the speed at \(t\) increases when \(v\) and \(a\) have the same sign (Lesson 4.2).
Worked example · Position, distance, and speed
A particle moves on a line with acceleration \(a(t)=6t-12\) m/s\(^2\), \(v(0)=9\) m/s and \(s(0)=0\). Find \(v(t)\), \(s(t)\), the displacement and the total distance for \(0\le t\le4\).
Displacement: \(s(4)-s(0)=64-96+36=4\) m. Direction changes at \(t=1\) and \(t=3\): \(s(1)=4\), \(s(3)=0\), \(s(4)=4\). Total distance \(=\lvert4-0\rvert+\lvert0-4\rvert+\lvert4-0\rvert=12\) m. (The displacement is only 4 m because the particle backtracks.)
Worked example · From a velocity graph
A car's velocity (m/s) is \(v(t)=10-2t\) for \(0\le t\le8\). The car starts at position 5. Find its position at \(t=8\) and the distance it travels.
\(s(8)=5+\displaystyle\int_0^8(10-2t)\,dt=5+\Big[10t-t^2\Big]_0^8=5+(80-64)=21\). \(v=0\) at \(t=5\): \(\displaystyle\int_0^5=25\), \(\displaystyle\int_5^8(10-2t)\,dt=\Big[10t-t^2\Big]_5^8=16-25=-9\). Distance \(=25+9=34\) m.
Worked example · Particle with a trigonometric velocity
\(v(t)=3\cos t\) for \(0\le t\le\pi\), with \(s(0)=2\). \(s(t)=2+3\sin t\). Displacement over \([0,\pi]\): \(s(\pi)-s(0)=0\). Total distance: \(v\gt0\) on \((0,\pi/2)\) and \(v\lt0\) on \((\pi/2,\pi)\), so \(\displaystyle\int_0^\pi\lvert3\cos t\rvert\,dt=2\int_0^{\pi/2}3\cos t\,dt=6\).
Worked example · Speed from a table
Velocity of a particle (m/s): \(v(0)=3\), \(v(2)=1\), \(v(4)=-2\), \(v(6)=-1\). Estimate the displacement over \([0,6]\) with a trapezoidal sum, and explain why the total distance can't be found accurately from this table.
Displacement \(\approx\tfrac{3+1}{2}(2)+\tfrac{1-2}{2}(2)+\tfrac{-2-1}{2}(2)=4-1-3=0\). Velocity changes sign between \(t=2\) and \(t=4\), but we do not know exactly when, so \(\int\lvert v\rvert\) can only be estimated (with assumptions such as linear behavior between data points).
Applying a calculator
For velocity or acceleration given by a complicated formula, use the calculator's numerical integral with the correct limits and absolute value for distance: \(\displaystyle\int_a^b\lvert v(t)\rvert\,dt\). Write the integral in your answer.
For the same particle, find the total distance on \([0,5]\).
Show answer
\(v=0\) at \(t=3\). \(s(0)=1\), \(s(3)=-8\), \(s(5)=-4\). Distance \(=\lvert-8-1\rvert+\lvert-4-(-8)\rvert=9+4=13\).
A particle has acceleration \(a(t)=4\) and \(v(0)=-8\). When does it change direction, and what is the displacement from \(t=0\) to \(t=4\)?
Show answer
\(v=-8+4t=0\) at \(t=2\). \(s(4)-s(0)=\displaystyle\int_0^4(4t-8)\,dt=\Big[2t^2-8t\Big]_0^4=0\).
Velocity is \(v(t)=t^2-4\) for \(0\le t\le3\). Write an integral for total distance and evaluate it.
Show answer
\(v\lt0\) on \((0,2)\), \(\gt0\) on \((2,3)\). \(\displaystyle\int_0^2(4-t^2)\,dt+\int_2^3(t^2-4)\,dt=\left(8-\tfrac83\right)+\left(9-12-\tfrac83+8\right)=\tfrac{16}3+\tfrac73=\tfrac{23}{3}\).
A student computes total distance as \(\displaystyle\int_0^4v(t)\,dt\) for \(v(t)=3-t\). What is the mistake? Compute both quantities.
Show answer
The integral without absolute value is displacement: \(\Big[3t-\tfrac{t^2}{2}\Big]_0^4=12-8=4\). Since \(v\lt0\) for \(t\gt3\), the total distance is \(\displaystyle\int_0^3(3-t)\,dt+\int_3^4(t-3)\,dt=4.5+0.5=5\).
The acceleration of a particle is given in a table. What can you say about the velocity change from \(t=0\) to \(t=4\) if \(a(0)=2\), \(a(2)=1\), \(a(4)=-1\)? Use a trapezoidal sum.
Show answer
\(\tfrac{2+1}{2}(2)+\tfrac{1-1}2(2)=3+0=3\). The velocity increased by about 3 units from \(t=0\) to \(t=4\).
Integrals answer a particular kind of question in the real world: starting from a known amount, and given how fast things come in and go out, how much is there now? This lesson collects the applied setups: flows, populations, costs, and any context where a rate is known.
CHA-4.D. Interpret the meaning of a definite integral in accumulation problems.
CHA-4.E. Determine net change using definite integrals in applied contexts.
Essential knowledge.
CHA-4.D.1. A function defined as an integral represents an accumulation of a rate of change.
CHA-4.D.2. The definite integral of the rate of change of a quantity over an interval gives the net change of that quantity over that interval.
CHA-4.E.1. The definite integral can be used to express information about accumulation and net change in many applied contexts.
The accumulation model
If \(Q(t)\) is an amount and \(Q'(t)\) its rate of change, then for any time \(t\) \[Q(t)=Q(a)+\int_a^tQ'(u)\,du.\] When a quantity has an inflow rate \(R_{\text{in}}(t)\) and an outflow rate \(R_{\text{out}}(t)\), then \(Q'(t)=R_{\text{in}}-R_{\text{out}}\), and \[Q(t)=Q(0)+\int_0^t\big(R_{\text{in}}(u)-R_{\text{out}}(u)\big)\,du.\] Interpreting the integral: \(\displaystyle\int_a^bQ'(t)\,dt\) is the net change in \(Q\) over \([a,b]\), measured in units of \(Q\).
Finding when the amount is largest or smallest
\(Q\) increases when \(Q'\gt0\) (inflow exceeds outflow) and decreases when \(Q'\lt0\). A maximum of \(Q\) on an interval occurs where \(Q'\) changes from positive to negative, or at an endpoint. Use the candidates test: find where \(Q'=0\), then compare \(Q\) at those times and at the endpoints, using the accumulation formula.
Worked example · A tank with inflow and outflow
A tank holds 100 gallons at \(t=0\). Water flows in at \(R_{\text{in}}(t)=30e^{-0.2t}\) gallons per minute and out at a constant 10 gallons per minute. Find the amount of water at time \(t\), the time when the amount is greatest, and the greatest amount, for \(t\ge0\).
\(Q'(t)=30e^{-0.2t}-10\). Then \(Q(t)=100+\displaystyle\int_0^t\big(30e^{-0.2u}-10\big)\,du=100+\Big[-150e^{-0.2u}-10u\Big]_0^t=250-150e^{-0.2t}-10t\).
\(Q'(t)=0\): \(e^{-0.2t}=\tfrac13\), \(t=5\ln3\approx5.493\). For \(t\lt5.493\), \(Q'\gt0\), and for larger \(t\), \(Q'\lt0\), so \(Q\) is greatest then. \(Q(5.493)=250-150\left(\tfrac13\right)-10(5.493)=250-50-54.93\approx145.07\) gallons.
Worked example · Interpreting an integral
\(\displaystyle\int_2^6R_{\text{in}}(t)\,dt\) is the total amount of water that flows in between \(t=2\) and \(t=6\) minutes, in gallons. For the inflow above, it equals \(\Big[-150e^{-0.2t}\Big]_2^6=150\left(e^{-0.4}-e^{-1.2}\right)\approx150(0.6703-0.3012)=55.4\) gallons.
Worked example · A table of rates
The rate at which people enter a stadium is \(E(t)\) people per minute and the rate they leave is \(X(t)\), for \(0\le t\le 60\), measured as follows.
\(t\) (min)
0
20
40
60
\(E(t)\)
50
80
60
20
\(X(t)\)
10
20
40
70
There are 200 people at \(t=0\). Estimate the number at \(t=60\) using a trapezoidal sum.
Net rate \(N=E-X\): \(40,\ 60,\ 20,\ -50\). Trapezoids with width 20: \(\tfrac{40+60}{2}(20)+\tfrac{60+20}2(20)+\tfrac{20-50}{2}(20)=1000+800-300=1500\). Estimate: \(200+1500=1700\) people.
Worked example · Population with immigration
The rate of change of a town's population is \(P'(t)=400+30t-2t^2\) people per year for \(0\le t\le10\), with \(P(0)=5000\). Find \(P(10)\) and the time when the population is greatest.
\(P(10)=5000+\Big[400t+15t^2-\tfrac23t^3\Big]_0^{10}=5000+4000+1500-666.67\approx9833.3\). \(P'=0\): \(2t^2-30t-400=0\), \(t^2-15t-200=0\), \(t=\dfrac{15+\sqrt{1025}}{2}\approx23.5\) (the other root is negative), outside the interval. On \([0,10]\), \(P'\gt0\) throughout (for example \(P'(10)=400+300-200=500\)), so \(P\) is increasing and the greatest population is at \(t=10\).
Units and sentences
The answer to an applied integral question is a sentence with units, such as "There are about 1700 people in the stadium at 60 minutes." If asked to interpret \(\displaystyle\int_a^bf\), say what it measures in the context: "the net change in ... from time \(a\) to time \(b\)," with units.
Exam focus
\(Q(t)=Q(a)+\int_a^tQ'\); net rate = in minus out.
Maximum or minimum of \(Q\): where \(Q'\) changes sign, compared with endpoints.
From a table: trapezoidal or Riemann sums for \(\int Q'\).
Interpret integrals with units and context.
The integral of a rate gives the net change, not the amount (add the initial amount).
Practice
Water flows into a pool at \(r(t)=20-t\) gallons per minute for \(0\le t\le20\). The pool has 50 gallons at \(t=0\). How much water is in the pool at \(t=10\)?
Show answer
The rate of growth of a tree's height is \(h'(t)=\dfrac{12}{\sqrt{t+1}}\) inches per year. By how much does it grow from \(t=0\) to \(t=8\)?
Show answer
A factory's cost rate is \(C'(x)=0.3x+5\) dollars per unit when \(x\) units have been made. Find the extra cost to raise production from 100 to 200 units.
Show answer
Sand falls onto a pile at \(R(t)=10\sqrt t\) kg per hour and is removed at 12 kg per hour. When is the pile growing, and at what time is its mass greatest on \([0,9]\)?
Show answer
\(10\sqrt t-12\gt0\) when \(\sqrt t\gt1.2\), \(t\gt1.44\). The pile shrinks on \((0,1.44)\) and grows on \((1.44,9)\). The mass at \(t=0\) versus \(t=9\): net change on \([0,9]\) is \(\Big[\tfrac{20}{3}t^{3/2}-12t\Big]_0^9=180-108=72\gt0\), so the greatest mass is at \(t=9\).
Explain what \(\displaystyle\int_0^{24}\big(R_{\text{in}}(t)-R_{\text{out}}(t)\big)\,dt\) means for a reservoir (\(t\) in hours, rates in thousand gallons per hour).
Show answer
The net change, in thousand gallons, in the amount of water in the reservoir over the 24 hours: positive means the reservoir gained water overall.
A student says the amount at time \(t\) is \(\int_0^tR'(u)\,du\). What is missing?
Show answer
The starting amount: \(Q(t)=Q(0)+\displaystyle\int_0^tR'(u)\,du\). The integral alone is only the change.
Finding the area between curves expressed as functions of x
Area under a curve is an integral. The area between two curves is an integral too, of a difference. This lesson uses vertical strips: slice the region with thin vertical rectangles, find the height of each, and add.
Required content · CHA-5.A.1
Learning objective.
CHA-5.A. Calculate areas in the plane using the definite integral.
Essential knowledge.
CHA-5.A.1. Areas of regions in the plane can be calculated with definite integrals.
The formula
If \(f(x)\ge g(x)\) on \([a,b]\), the area of the region between the graphs of \(f\) (top) and \(g\) (bottom) and the lines \(x=a\) and \(x=b\) is \[A=\int_a^b\big[f(x)-g(x)\big]\,dx.\] Each thin strip has width \(dx\) and height top minus bottom, so its area is \((f(x)-g(x))\,dx\). The formula works whether the curves are above or below the \(x\)-axis, since the difference is still the strip height.
Figure 8.2. The region between \(y=2x\) and \(y=x^2\) has area \(\displaystyle\int_0^2(2x-x^2)\,dx=\tfrac43\) using vertical strips (top minus bottom), and the same \(\displaystyle\int_0^4\left(\sqrt y-\tfrac y2\right)dy=\tfrac43\) using horizontal strips (right minus left).
Procedure
Sketch the region and find which curve is on top.
Find the intersection points (set \(f(x)=g(x)\)) if the limits are not given.
Write the integral of top minus bottom.
Evaluate it (by antiderivatives or with the calculator).
Worked example · Bounded by two curves
Find the area of the region bounded by \(y=2x\) and \(y=x^2\).
Intersections: \(2x=x^2\) gives \(x=0\) and \(x=2\). On \((0,2)\), \(2x\ge x^2\) (e.g., at \(x=1\): \(2\gt1\)). \(A=\displaystyle\int_0^2(2x-x^2)\,dx=\Big[x^2-\tfrac{x^3}{3}\Big]_0^2=4-\tfrac83=\tfrac43\).
Worked example · Given vertical boundaries
Find the area between \(y=e^x\) and \(y=x\) for \(0\le x\le1\).
Find the area between \(y=\sin x\) and \(y=\cos x\) from \(x=\dfrac\pi4\) to \(x=\dfrac{5\pi}{4}\).
They cross at both endpoints (\(\sin x=\cos x\) when \(\tan x=1\)). On the interval, \(\sin x\ge\cos x\). \(A=\displaystyle\int_{\pi/4}^{5\pi/4}(\sin x-\cos x)\,dx=\Big[-\cos x-\sin x\Big]_{\pi/4}^{5\pi/4}=\left(\tfrac{\sqrt2}2+\tfrac{\sqrt2}2\right)-\left(-\tfrac{\sqrt2}2-\tfrac{\sqrt2}2\right)=2\sqrt2\approx2.828\).
Worked example · A region partly below the \(x\)-axis
Find the area between \(y=x^2-4\) and \(y=-x^2+4\).
The intersections: \(x^2-4=-x^2+4\) gives \(x=\pm2\). On \((-2,2)\) the top is \(4-x^2\). \(A=\displaystyle\int_{-2}^{2}\big[(4-x^2)-(x^2-4)\big]dx=\int_{-2}^2(8-2x^2)\,dx=\Big[8x-\tfrac{2x^3}3\Big]_{-2}^2=2\left(16-\tfrac{16}3\right)=\tfrac{64}3\).
Using the calculator
For curves like \(y=x^2\) and \(y=\cos x\), the intersection points need a calculator. Find them (or store them as \(a\) and \(b\)), then compute \(\displaystyle\int_a^b\big[\text{top}-\text{bottom}\big]dx\) numerically. On free-response questions you must still write the integral setup with the limits you found.
Exam focus
\(A=\int_a^b(\text{top}-\text{bottom})\,dx\) with vertical strips.
Find the intersections to get the limits; confirm which curve is on top.
The strip height is top minus bottom even when parts are below the axis.
Use the calculator for intersections and the integral; write the setup.
Area is positive; if you get a negative number, you reversed top and bottom.
Practice
Find the area between \(y=x\) and \(y=x^2\) from \(x=0\) to \(x=1\).
Show answer
Write (do not evaluate) the integral for the area between \(y=x^2\) and \(y=\cos x\), with intersection points at \(x=\pm c\) where \(c\approx0.824\).
Show answer
\(\displaystyle\int_{-0.824}^{0.824}\big(\cos x-x^2\big)\,dx\) (cosine is on top near 0). This evaluates to about 1.095.
Find the area between \(y=\dfrac1x\) and \(y=x\) for \(1\le x\le2\). First decide which is on top.
Show answer
At \(x=1.5\): \(\dfrac1x=0.67\lt x=1.5\), so \(y=x\) is on top. \(\displaystyle\int_1^2\left(x-\tfrac1x\right)dx=\left[\tfrac{x^2}2-\ln x\right]_1^2=(2-\ln2)-\tfrac12=\tfrac32-\ln2\approx0.807\).
A student computes the area between \(y=x^2-4\) and \(y=0\) from \(-2\) to \(2\) as \(\displaystyle\int_{-2}^{2}(x^2-4)\,dx\) and gets \(-\tfrac{32}3\). What is the mistake?
Show answer
The curve is below the axis, so top minus bottom is \(0-(x^2-4)=4-x^2\). The area is \(\tfrac{32}{3}\). A negative area is a sign that top and bottom were reversed.
Area between curves in y and with several crossings
Two situations make the vertical-strip method awkward: a region whose boundaries are naturally written as \(x=\) a function of \(y\), and curves that cross each other more than twice. Both are handled by the same idea: choose the direction of the strips to fit the region, and split the region where the top (or right) curve changes.
Required content · CHA-5.A.2, CHA-5.A.3
Learning objective.
CHA-5.A. Calculate areas in the plane using the definite integral.
Essential knowledge.
CHA-5.A.2. Areas of regions in the plane can be calculated using functions of either \(x\) or \(y\).
CHA-5.A.3. Areas of certain regions in the plane may be calculated using a sum of two or more definite integrals or by evaluating a definite integral of the absolute value of the difference of two functions.
Horizontal strips: integrating with respect to \(y\)
If the region is described by \(x=R(y)\) on the right and \(x=L(y)\) on the left for \(c\le y\le d\), slice it with thin horizontal strips of height \(dy\). Each has length right minus left: \[A=\int_c^d\big[R(y)-L(y)\big]\,dy.\] Horizontal strips are best when the boundaries are easier to write as functions of \(y\), or when vertical strips would force you to split the region into pieces (the right boundary changes formula).
Worked example · The same region two ways
Find the area between \(y=2x\) and \(y=x^2\) using horizontal strips.
Solve for \(x\): \(x=\dfrac y2\) (the line) and \(x=\sqrt y\) (the parabola, \(x\ge0\)). The intersection points are at \(y=0\) and \(y=4\). For \(0\le y\le4\), \(\sqrt y\ge\dfrac y2\) (at \(y=1\): \(1\ge0.5\)), so right \(=\sqrt y\) and left \(=\dfrac y2\): \[A=\int_0^4\left(\sqrt y-\frac y2\right)dy=\Big[\tfrac23y^{3/2}-\tfrac{y^2}4\Big]_0^4=\tfrac{16}3-4=\tfrac43.\] This agrees with Lesson 8.4 (Figure 8.2).
Worked example · A parabola opening sideways
Find the area bounded by \(x=y^2\) and \(y=x-2\).
Write the line as \(x=y+2\). Intersections: \(y^2=y+2\), \((y-2)(y+1)=0\), so \(y=-1\) and \(y=2\) (points \((1,-1)\) and \((4,2)\)). For \(-1\lt y\lt2\), the line is to the right (e.g., \(y=0\): \(x=2\gt0\)). \(A=\displaystyle\int_{-1}^2\big[(y+2)-y^2\big]\,dy=\Big[\tfrac{y^2}2+2y-\tfrac{y^3}{3}\Big]_{-1}^2=\left(2+4-\tfrac83\right)-\left(\tfrac12-2+\tfrac13\right)=\tfrac{10}3+\tfrac76=\tfrac92\).
With vertical strips this region needs two integrals (the lower boundary changes at \(x=1\)); with horizontal strips, one.
Curves that intersect more than twice
If the top and bottom curves swap places, the integral of \(f-g\) over the whole interval subtracts the pieces where \(g\) is on top and gives the wrong area. Find all the intersection points, determine which curve is on top on each subinterval, and add the separate integrals (or integrate \(\lvert f-g\rvert\)).
Figure 8.3. The curves \(y=x^3\) and \(y=x\) meet at \(x=-1,0,1\). On \((-1,0)\) the cubic is on top; on \((0,1)\) the line is on top. The area is \(\displaystyle\int_{-1}^0(x^3-x)\,dx+\int_0^1(x-x^3)\,dx=\tfrac14+\tfrac14=\tfrac12\); one integral from \(-1\) to \(1\) of \(x-x^3\) would give 0, the wrong answer.
Worked example · The cubic and the line
Find the area bounded by \(y=x^3\) and \(y=x\).
Intersections: \(x^3=x\), \(x(x-1)(x+1)=0\): \(x=-1,0,1\). On \((-1,0)\), \(x^3\ge x\) (at \(x=-0.5\): \(-0.125\ge-0.5\)); on \((0,1)\), \(x\ge x^3\). \(A=\displaystyle\int_{-1}^0(x^3-x)\,dx+\int_0^1(x-x^3)\,dx=\tfrac14+\tfrac14=\tfrac12\). (A single integral \(\int_{-1}^1(x-x^3)\,dx=0\) would be wrong.)
Worked example · A curve and the \(x\)-axis
Find the area of the region bounded by \(y=x^3-4x\) and the \(x\)-axis.
\(x^3-4x=x(x-2)(x+2)\): zeros at \(-2,0,2\). On \((-2,0)\) the cubic is above the axis; on \((0,2)\), below. \(A=\displaystyle\int_{-2}^0(x^3-4x)\,dx+\int_0^2(4x-x^3)\,dx=\Big[\tfrac{x^4}{4}-2x^2\Big]_{-2}^0+\Big[2x^2-\tfrac{x^4}{4}\Big]_0^2=4+4=8\).
Worked example · With a calculator
Use a calculator to find the area between \(f(x)=e^{-x}\) and \(g(x)=\sin x+0.5\) for \(0\le x\le2\).
Graph both. They intersect once in \([0,2]\), at \(x_1\approx0.268\) (solve \(e^{-x}=\sin x+0.5\) numerically). Before \(x_1\), \(f\) is on top (at \(x=0\): \(1\gt0.5\)); after it, \(g\) is on top. So \[A=\int_0^{0.268}\big(e^{-x}-\sin x-0.5\big)\,dx+\int_{0.268}^{2}\big(\sin x+0.5-e^{-x}\big)\,dx\approx0.065+1.617=1.682.\] One numerical integral of \(\displaystyle\int_0^2\lvert f(x)-g(x)\rvert\,dx\) gives the same value. Always write the integral you used before giving the value.
Exam focus
Horizontal strips: \(A=\int_c^d(\text{right}-\text{left})\,dy\); express the boundaries as \(x=\) functions of \(y\).
Choose the strip direction that avoids splitting the region.
For curves that cross, find every intersection and add the integrals of (top − bottom) on each piece, or integrate the absolute value.
A single integral over an interval where curves swap gives the wrong (net) result.
Check the sign: each piece must be a positive area.
Practice
Find the area between \(x=y^2\) and \(x=4\) using horizontal strips.
Show answer
Find the area bounded by \(x=y^2-2\) and \(x=y\).
Show answer
\(y^2-2=y\): \(y=-1,2\). For \(y\) between, \(x=y\) is to the right. \(\displaystyle\int_{-1}^2\big[y-(y^2-2)\big]dy=\Big[\tfrac{y^2}2-\tfrac{y^3}3+2y\Big]_{-1}^2=\left(2-\tfrac83+4\right)-\left(\tfrac12+\tfrac13-2\right)=\tfrac{10}{3}+\tfrac76=\tfrac92\).
Find the area bounded by \(y=x^3-x\) and \(y=0\).
Show answer
Zeros at \(-1,0,1\). \(\displaystyle\int_{-1}^0(x^3-x)\,dx+\int_0^1(x-x^3)\,dx=\tfrac14+\tfrac14=\tfrac12\).
Find the area of the region between \(y=x^2\) and \(y=8-x^2\).
Show answer
Why can't you just compute \(\displaystyle\int_{0}^{2\pi}(\sin x-\cos x)\,dx\) for the area between the curves on \([0,2\pi]\)?
Show answer
The curves cross at \(x=\tfrac\pi4\) and \(\tfrac{5\pi}4\), so sin is on top only on part of the interval. The integral over \([0,2\pi]\) is \(0\) because positive and negative pieces cancel. Split at the intersections or integrate \(\lvert\sin x-\cos x\rvert\).
For the region bounded by \(y=\sqrt x\), \(y=0\) and \(x=4\), would horizontal or vertical strips be easier?
Show answer
Vertical strips: \(\displaystyle\int_0^4\sqrt x\,dx=\tfrac{16}3\). (Horizontal strips need \(x=y^2\) to \(x=4\): \(\int_0^2(4-y^2)\,dy=\tfrac{16}3\), also fine.)
Integrals do not only find area. If you know the area of every slice of a solid, you can add up the slices to get the volume. The key is a thin slab: its volume is (cross-sectional area) \(\times\) (thickness).
CHA-5.B. Calculate volumes of solids with known cross sections using definite integrals.
Essential knowledge.
CHA-5.B.1. Volumes of solids with square and rectangular cross sections can be found using definite integrals and the area formulas for these shapes.
CHA-5.B.2. Volumes of solids with triangular cross sections can be found using definite integrals and the area formulas for these shapes.
CHA-5.B.3. Volumes of solids with semicircular and other geometrically defined cross sections can be found using definite integrals and the area formulas for these shapes.
The idea
Take a solid that lies along the \(x\)-axis from \(x=a\) to \(x=b\), and suppose the cross section perpendicular to the \(x\)-axis at \(x\) has area \(A(x)\). Slicing into slabs of thickness \(dx\) gives \[V=\int_a^bA(x)\,dx.\] (If the slices are perpendicular to the \(y\)-axis, use \(V=\int_c^dA(y)\,dy\).) The area \(A(x)\) depends on the shape of the cross section and on a length \(s(x)\), usually a distance across the base region between two curves.
Figure 8.4. The base is the region under \(y=\sqrt x\) for \(0\le x\le4\). Each cross section perpendicular to the \(x\)-axis is a square standing on the base, with side \(s=\sqrt x\) and area \(x\). The volume is \(\displaystyle\int_0^4x\,dx=8\).
Areas of the usual cross sections
With \(s\) the length of the base (the side or the diameter, as given):
Cross section
Area \(A\)
square with side \(s\)
\(s^2\)
rectangle with base \(s\) and height \(h\)
\(s\,h\)
equilateral triangle with side \(s\)
\(\dfrac{\sqrt3}{4}s^2\)
isosceles right triangle with leg \(s\)
\(\dfrac12s^2\)
isosceles right triangle with hypotenuse \(s\)
\(\dfrac14s^2\)
semicircle with diameter \(s\)
\(\dfrac{\pi}{8}s^2\)
For a region between curves, \(s(x)=\) top minus bottom.
Worked example · Squares
The base of a solid is the region under \(y=\sqrt x\) for \(0\le x\le4\). Cross sections perpendicular to the \(x\)-axis are squares. Find the volume.
The base is bounded by \(y=x\) and \(y=x^2\). Cross sections perpendicular to the \(x\)-axis are equilateral triangles. Find the volume.
The curves meet at \(x=0\) and \(x=1\); \(s=x-x^2\). \(A=\dfrac{\sqrt3}4(x-x^2)^2\). \(V=\dfrac{\sqrt3}4\displaystyle\int_0^1(x^2-2x^3+x^4)\,dx=\dfrac{\sqrt3}{4}\left(\tfrac13-\tfrac12+\tfrac15\right)=\dfrac{\sqrt3}{4}\cdot\dfrac1{30}=\dfrac{\sqrt3}{120}\approx0.0144\).
Worked example · Rectangles of fixed height
The base is the region between \(y=4-x^2\) and the \(x\)-axis. Cross sections perpendicular to the \(x\)-axis are rectangles of height 3. Find the volume.
Worked example · Slices perpendicular to the \(y\)-axis
The base is the region between \(x=y^2\) and \(x=4\). Cross sections perpendicular to the \(y\)-axis are squares. Find the volume.
\(s=4-y^2\) for \(-2\le y\le2\). \(V=\displaystyle\int_{-2}^2(4-y^2)^2\,dy=2\int_0^2(16-8y^2+y^4)\,dy=2\left(32-\tfrac{64}{3}+\tfrac{32}{5}\right)=\tfrac{512}{15}\approx34.13\).
Checking the setup
Check units (volume is in cubic units, \(\text{area}\times\text{length}\)). Ask: which direction are the slices (so which variable is the integration variable and which limits)? What is \(s\) in terms of that variable? A quick sanity check: the volume should be positive and not exceed the volume of a box that contains the solid.
Exam focus
\(V=\int_a^bA(x)\,dx\), \(A\) = area of a slice perpendicular to the axis of integration.
Express the length \(s(x)\) as top minus bottom (or right minus left) across the base.
The base is the region between \(y=\sqrt x\) and \(y=0\) for \(0\le x\le4\); cross sections are isosceles right triangles with one leg in the base. Find the volume.
Show answer
A solid has base bounded by \(y=x^2\) and \(y=4\). Cross sections perpendicular to the \(y\)-axis are squares. Write the volume as an integral and evaluate.
Show answer
\(s=2\sqrt y\) (the width from \(x=-\sqrt y\) to \(\sqrt y\)). \(A=4y\). \(V=\displaystyle\int_0^44y\,dy=\Big[2y^2\Big]_0^4=32\).
A student says the volume with equilateral triangle cross sections of side \(s\) is \(\int\dfrac{\sqrt3}{2}s\,dx\). What is wrong?
Show answer
The area of an equilateral triangle is \(\tfrac{\sqrt3}{4}s^2\) (it depends on \(s^2\)). The student used something linear in \(s\), which is the height-type quantity, not the area.
Volume with the disc method: revolving around axes
Spin a region around a line and it sweeps out a solid. If the region touches the axis of revolution along its whole length, every slice perpendicular to the axis is a disc, a circle of radius equal to the distance from the axis to the curve. Adding up the discs gives the volume.
Required content · CHA-5.C.1, CHA-5.C.2
Learning objective.
CHA-5.C. Calculate volumes of solids of revolution using definite integrals.
Essential knowledge.
CHA-5.C.1. Volumes of solids of revolution around the \(x\)- or \(y\)-axis may be found by using definite integrals with the disc method.
CHA-5.C.2. Volumes of solids of revolution around any horizontal or vertical line in the plane may be found by using definite integrals with the disc method.
The disc method
Revolve the region between \(y=f(x)\), the \(x\)-axis, and the lines \(x=a\), \(x=b\) about the \(x\)-axis. The slice at \(x\) is a disc of radius \(R=\lvert f(x)\rvert\), with area \(\pi R^2\) and thickness \(dx\): \[V=\pi\int_a^b\big[f(x)\big]^2\,dx.\] To revolve about the \(y\)-axis a region described as \(x=g(y)\), \(c\le y\le d\): \[V=\pi\int_c^d\big[g(y)\big]^2\,dy.\]
Figure 8.5. Left: revolving the region under \(y=\sqrt x\) on \([0,4]\) about the \(x\)-axis gives discs of radius \(\sqrt x\); \(V=\pi\displaystyle\int_0^4x\,dx=8\pi\). Right: revolving the region between \(y=x\) and \(y=x^2\) on \([0,1]\) gives washers with outer radius \(R=x\) and inner radius \(r=x^2\); \(V=\pi\displaystyle\int_0^1\left(x^2-x^4\right)dx=\dfrac{2\pi}{15}\).
The radius is always the distance from the axis to the boundary curve, and the integration variable is measured along the axis: the discs are stacked in that direction.
Worked example · About the \(x\)-axis
Revolve the region under \(y=\sqrt x\) on \([0,4]\) about the \(x\)-axis (the left half of the figure).
Revolve the region bounded by \(y=x^2\), the \(y\)-axis, and \(y=4\) about the \(y\)-axis.
Slices perpendicular to the \(y\)-axis; write \(x=\sqrt y\), radius \(\sqrt y\) for \(0\le y\le4\). \(V=\pi\displaystyle\int_0^4y\,dy=\pi\cdot8=8\pi\).
Worked example · About a horizontal line other than the \(x\)-axis
Revolve the region between \(y=x^2\) and the line \(y=4\), for \(0\le x\le2\), about the line \(y=4\).
The region lies between the curve and the axis of revolution, so there is no hole: a disc with radius equal to the distance from the curve to the line, \(R=4-x^2\). \(V=\pi\displaystyle\int_0^2(4-x^2)^2\,dx=\pi\int_0^2(16-8x^2+x^4)\,dx=\pi\left(32-\tfrac{64}{3}+\tfrac{32}{5}\right)=\dfrac{256\pi}{15}\approx53.62\).
Worked example · About a vertical line other than the \(y\)-axis
Revolve the region between \(x=y^2\) and the line \(x=4\), for \(0\le y\le2\), about the line \(x=4\).
The distance from the curve \(x=y^2\) to the axis \(x=4\) is \(R=4-y^2\). \(V=\pi\displaystyle\int_0^2(4-y^2)^2\,dy=\dfrac{256\pi}{15}\) (the same computation as above, with the roles of the variables exchanged).
Finding the radius
Sketch the region and the axis of revolution.
The radius is the distance from the axis to the curve (\(x\)-axis: \(y\) value; horizontal line \(y=k\): \(\lvert y-k\rvert\); \(y\)-axis: \(x\) value; vertical line \(x=k\): \(\lvert x-k\rvert\)).
Integrate along the axis. Limits are the extreme values of that variable on the region.
Exam focus
Disc method: \(V=\pi\int R^2\,d(\text{variable along the axis})\), radius = distance from axis to curve.
About the \(x\)-axis: \(\pi\int[f(x)]^2dx\); about the \(y\)-axis: \(\pi\int[g(y)]^2dy\).
For a line \(y=k\) or \(x=k\): \(R=\lvert\text{curve}-k\rvert\).
Use discs only when the region touches the axis of revolution (no hole).
Write the integral, then evaluate; answer in cubic units.
Practice
Revolve the region under \(y=x^2\) from \(x=0\) to \(x=2\) about the \(x\)-axis. Find the volume.
Show answer
Revolve the region between \(y=\sqrt x\), the line \(y=2\), and the \(y\)-axis about the line \(y=2\).
Show answer
The curve meets \(y=2\) at \(x=4\). \(R=2-\sqrt x\) for \(0\le x\le4\). \(V=\pi\displaystyle\int_0^4(2-\sqrt x)^2\,dx=\pi\int_0^4(4-4\sqrt x+x)\,dx=\pi\left(16-\tfrac{64}{3}+8\right)=\tfrac{8\pi}{3}\).
A student sets up the volume of the region under \(y=x^2\) (\(0\le x\le2\)) revolved about the \(x\)-axis as \(\pi\displaystyle\int_0^2x^2\,dx\). What is wrong?
Show answer
The radius is \(x^2\), so the integrand is \((x^2)^2=x^4\), not \(x^2\). The student forgot to square the radius.
Set up (do not evaluate) the integral for the region under \(y=e^{x}\) on \([0,1]\) revolved about the \(x\)-axis.
Show answer
\(\pi\displaystyle\int_0^1e^{2x}\,dx\), which equals \(\dfrac\pi2(e^2-1)\approx10.04\).
Volume with the washer method: revolving around axes
When the region does not touch the axis of revolution, the solid has a hole, and each slice is a washer: a disc with a smaller disc removed. The volume of a washer slab is the outer disc minus the inner disc.
Required content · CHA-5.C.3, CHA-5.C.4
Learning objective.
CHA-5.C. Calculate volumes of solids of revolution using definite integrals.
Essential knowledge.
CHA-5.C.3. Volumes of solids of revolution around the \(x\)- or \(y\)-axis whose cross sections are ring shaped may be found using definite integrals with the washer method.
CHA-5.C.4. Volumes of solids of revolution around any horizontal or vertical line whose cross sections are ring shaped may be found using definite integrals with the washer method.
The washer method
If the region lies between an outer curve and an inner curve and is revolved about an axis, the slice perpendicular to the axis has outer radius \(R\) and inner radius \(r\) (both measured as distances from the axis), area \(\pi\left(R^2-r^2\right)\): \[V=\pi\int\big(R^2-r^2\big)\,d(\text{variable along the axis}).\] Note that it is \(R^2-r^2\), not \((R-r)^2\): subtract the areas, not the radii.
For the \(x\)-axis (region between \(y=f(x)\) above and \(y=g(x)\) below, both nonnegative): \(V=\pi\displaystyle\int_a^b\big([f(x)]^2-[g(x)]^2\big)dx\).
Worked example · About the \(x\)-axis
Revolve the region between \(y=x\) and \(y=x^2\) (\(0\le x\le1\)) about the \(x\)-axis (the right half of Figure 8.5).
Outer radius \(R=x\) (the line is farther from the axis), inner radius \(r=x^2\). \(V=\pi\displaystyle\int_0^1\big(x^2-x^4\big)\,dx=\pi\left(\tfrac13-\tfrac15\right)=\dfrac{2\pi}{15}\approx0.419\).
Worked example · About the \(y\)-axis
Revolve the same region about the \(y\)-axis. Use horizontal slices: for \(0\le y\le1\), the region extends from \(x=y\) (the line, left) to \(x=\sqrt y\) (the parabola, right). Distances from the \(y\)-axis: outer \(R=\sqrt y\), inner \(r=y\). \(V=\pi\displaystyle\int_0^1\big(y-y^2\big)\,dy=\pi\left(\tfrac12-\tfrac13\right)=\dfrac\pi6\approx0.524\).
Worked example · About a horizontal line below the region
Revolve the same region about the line \(y=-1\).
Distances from \(y=-1\): to the line \(y=x\): \(R=x+1\); to the parabola: \(r=x^2+1\). \(V=\pi\displaystyle\int_0^1\left[(x+1)^2-(x^2+1)^2\right]dx=\pi\int_0^1\left(-x^4-x^2+2x\right)dx=\pi\left(-\tfrac15-\tfrac13+1\right)=\dfrac{7\pi}{15}\approx1.466\).
Worked example · About a vertical line
Revolve the region about the line \(x=2\).
For \(0\le y\le1\), distances from \(x=2\): to the left boundary \(x=y\): \(R=2-y\); to the right boundary \(x=\sqrt y\): \(r=2-\sqrt y\). \(V=\pi\displaystyle\int_0^1\left[(2-y)^2-(2-\sqrt y)^2\right]dy=\pi\int_0^1\left(-5y+y^2+4\sqrt y\right)dy=\pi\left(-\tfrac52+\tfrac13+\tfrac83\right)=\dfrac\pi2\).
Worked example · A rectangular hole
The region between \(y=4\) and \(y=x^2+1\), \(-\sqrt3\le x\le\sqrt3\), revolved about the \(x\)-axis. Outer radius \(R=4\), inner \(r=x^2+1\). \(V=\pi\displaystyle\int_{-\sqrt3}^{\sqrt3}\big(16-(x^2+1)^2\big)dx=2\pi\int_0^{\sqrt3}\big(15-2x^2-x^4\big)dx=2\pi\left(15\sqrt3-2\sqrt3-\tfrac{9\sqrt3}{5}\right)=\dfrac{112\sqrt3\,\pi}{5}\approx121.9\).
Choosing between discs and washers
Look at the region and the axis: if the region touches the axis everywhere along its length, use discs; if there is a gap between the region and the axis, use washers. The method for the radius is the same in each case: distance to the axis. Often the choice of slicing direction (vertical or horizontal) is dictated by the axis: slices are perpendicular to the axis.
Exam focus
Washer: \(V=\pi\int(R^2-r^2)\), with \(R\) and \(r\) the distances from the axis to the outer and inner curves.
Subtract the squares, not square the difference.
Slices are perpendicular to the axis; integrate along the axis.
Axes other than the coordinate axes: radii are shifted by the axis position.
Check that \(R\ge r\) on the whole interval.
Practice
Revolve the region between \(y=x\) and \(y=\dfrac{x}{2}\) for \(0\le x\le4\) about the \(x\)-axis.
Show answer
Revolve the region between \(y=\sqrt x\) and \(y=x\) (\(0\le x\le1\)) about the \(x\)-axis.
Show answer
Here \(\sqrt x\ge x\), so \(R=\sqrt x\), \(r=x\): \(\pi\displaystyle\int_0^1(x-x^2)\,dx=\tfrac\pi6\).
Revolve the region between \(y=x^2\) and \(y=2x\) about the line \(y=-2\).
Show answer
Intersections \(x=0,2\). Line on top: \(R=2x+2\), \(r=x^2+2\). \(\pi\displaystyle\int_0^2\left[(2x+2)^2-(x^2+2)^2\right]dx=\pi\int_0^2\left(-x^4+8x\right)dx=\pi\left(-\tfrac{32}{5}+16\right)=\tfrac{48\pi}{5}\).
Set up (do not evaluate) the volume when the region between \(y=e^x\) and \(y=1\), for \(0\le x\le1\), is revolved about the \(x\)-axis.
Show answer
A student writes the washer volume as \(\pi\displaystyle\int_0^1(x-x^2)^2\,dx\) for the region between \(y=x\) and \(y=x^2\) revolved about the \(x\)-axis. What is wrong?
Show answer
The student squared the difference of the radii. The correct integrand is \(x^2-(x^2)^2=x^2-x^4\) (outer radius squared minus inner radius squared).
Why does revolving the region between \(y=x\) and \(y=x^2\) about the \(y\)-axis (\(\tfrac\pi6\)) give a different volume from revolving it about the \(x\)-axis (\(\tfrac{2\pi}{15}\))?
Show answer
The radii are distances from the axis of revolution, and the region is at different distances from the two axes, so the washers differ. Each volume must be computed with its own radii; there is no reason for the two solids to have equal volume.
Unit 8 practice · 10 problems
Unit 8 practice: Applications of Integration
Ten problems covering the whole unit. Work each one on paper before
revealing the answer. No calculator needed.
Find the average value of \(f(x) = 6x^2\) on \([0, 2]\).
Water enters a tank at \(12 - t\) gal/min and leaves at 4 gal/min for \(0 \le t \le 12\); the tank holds 50 gal at \(t = 0\). When is the amount of water greatest, and how much is there then?
Show answer
Net rate \(8 - t\), positive then negative, zero at \(t = 8\). Amount \(= 50 + \displaystyle\int_0^8 (8 - t)\,dt = 50 + 32 = 82\) gallons at \(t = 8\).
Find the area between \(y = x^2 - 4\) and \(y = 2x - 1\).
Show answer
Intersections at \(x = -1, 3\); the line is on top. \(\displaystyle\int_{-1}^{3}\big(-x^2 + 2x + 3\big)\,dx = \Big[-\tfrac{x^3}{3} + x^2 + 3x\Big]_{-1}^{3} = 9 + \tfrac{5}{3} = \tfrac{32}{3}\).
Set up and evaluate an integral with respect to \(y\) for the area bounded by \(y = \sqrt{x}\), \(y = 0\), and \(x = 4\).
Show answer
Horizontal strips run from \(x = y^2\) to \(x = 4\) for \(0 \le y \le 2\): \(\displaystyle\int_0^2 (4 - y^2)\,dy = 8 - \tfrac{8}{3} = \tfrac{16}{3}\). (Check: \(\displaystyle\int_0^4\sqrt{x}\,dx = \tfrac{16}{3}\).)
The region bounded by \(y = x^2\), \(y = 0\), and \(x = 2\) is revolved about the \(x\)-axis. Find the volume.
The region between \(y = \sqrt{x}\) and \(y = \dfrac{x}{2}\) is revolved about the \(x\)-axis. Find the volume.
Show answer
They meet at \(x = 0\) and \(4\), with \(\sqrt{x}\) on top. \(\pi\displaystyle\int_0^4\left(x - \tfrac{x^2}{4}\right)dx = \pi\Big[\tfrac{x^2}{2} - \tfrac{x^3}{12}\Big]_0^4 = \pi\left(8 - \tfrac{16}{3}\right) = \dfrac{8\pi}{3}\).
The base of a solid is the region between \(y = x\) and \(y = x^2\). Cross sections perpendicular to the \(x\)-axis are equilateral triangles. Find the volume.
Animated recap with on-screen narration. Turn on Voice to
have it read aloud (uses your device's built-in voice). Pressing play
counts as your one free video.
Free preview complete
That's the end of the free preview.
You've opened five lessons, which is
as much as we can show without a subscription. Everything you've
already opened stays available; use the outline on the left to go
back to it.