Build integrals from accumulated change

Connect signed sums, integral notation and accumulation-function behavior

What to practice

  • Connect signed sums, integral notation and accumulation-function behavior

Units 2 through 5 started with an amount and asked how fast it changes. Unit 6 reverses the question: given a rate, how much has accumulated? This chapter builds the answer in four steps. A rate times a width gives an amount (topic 6.1). Sums of such products estimate accumulation from tables and graphs (topic 6.2). Their limit defines the definite integral (topic 6.3). An accumulation function then changes at exactly the rate being accumulated, and a graph of that rate describes the accumulation function's behavior (topics 6.4 and 6.5). Chapter 15 turns these ideas into exact evaluation with antiderivatives. Here you work from geometry, sums and derivatives.

Predict first. Water flows into a tank at 30 liters per minute for 4 minutes, then drains at 20 liters per minute for 3 minutes. Is the tank fuller or emptier than at the start, and by how much? Keep your answer for Worked example 14.1.

Rate times width gives accumulated change (topic 6.1)

If a rate is constant, the amount accumulated is rate times elapsed input: 30 liters per minute for 4 minutes gives 120 liters. On a graph of the rate, that product is the area of a rectangle. When the rate varies, the area between the rate graph and the horizontal axis still gives the accumulated change. Units follow the product: (rate unit) times (input unit), so liters per minute times minutes gives liters.

Area below the axis counts as negative. A negative rate means the amount is decreasing, so that region subtracts from the net change. The net change is the signed total. The gross amount, such as the total volume that passed through a pump in either direction, adds the areas without signs.

Worked example 14.1: a pump that reverses

A pump moves water into a tank at 30 liters per minute for 4 minutes, out at 20 liters per minute for the next 3 minutes, then in at 10 liters per minute for 2 minutes. Describe the change in the tank.

Draw the rate as three horizontal segments: height 30 on [0,4][0,4], height −20-20 on [4,7][4,7] and height 10 on [7,9][7,9]. The signed areas are 30(4)=12030(4)=120, −20(3)=−60-20(3)=-60 and 10(2)=2010(2)=20.

  • Net change: 120−60+20=80120-60+20=80 liters. After 9 minutes the tank holds 80 liters more than at the start. After 7 minutes, the moment in the prediction, it holds 120−60=60120-60=60 liters more.
  • Gross amount pumped: 120+60+20=200120+60+20=200 liters of water moved through the pump.
  • When was the tank fullest? The amount rises while the rate is positive and falls while it is negative, so compare the ends of the rising stretches: 120120 liters above the start at t=4t=4 and 8080 at t=9t=9. The fullest moment is t=4t=4.

The third question is a preview of topic 6.5. The rate tells you how the amount moves, and its sign changes mark where the amount turns.

Common error. Answering a net question with a gross amount, or the reverse. "How much more water is in the tank?" asks for the signed total. "How much water passed through the pump?" asks for the total without signs. Read which one the question wants before you add areas.

Approximate accumulation with Riemann sums (topic 6.2)

When the rate is known only at some inputs, split the interval into subintervals, choose one rate value for each, and add (rate value) times (width). Four standard choices appear on the exam.

SumHeight used on each subinterval
Leftthe value at the left end
Rightthe value at the right end
Midpointthe value at the middle, if it is known
Trapezoidalthe average of the two end values

The subintervals need not have equal widths. Use the width of each one. A table normally supports left, right and trapezoidal sums. A midpoint sum needs values at the midpoints, so with a table it usually uses wider subintervals whose midpoints are listed.

Over or under? If the rate is increasing on a subinterval, the left value is the smallest there, so the left sum underestimates and the right sum overestimates. If it is decreasing, the roles swap. This conclusion needs monotonic behavior on every subinterval, stated in the problem or visible on a graph. Table values alone do not show what happens between entries.

Worked example 14.2: sums from unequal time steps

The table gives the rate E(t)E(t), in visitors per hour, at which people enter a county fair tt hours after the gates open. It is constructed practice data.

tt (hours)01347
E(t)E(t) (visitors per hour)120160220200110

The widths are 1, 2, 1 and 3 hours.

  • Left: 1(120)+2(160)+1(220)+3(200)=12601(120)+2(160)+1(220)+3(200)=1260 visitors.
  • Right: 1(160)+2(220)+1(200)+3(110)=11301(160)+2(220)+1(200)+3(110)=1130 visitors.
  • Trapezoidal: 1⋅120+1602+2⋅160+2202+1⋅220+2002+3⋅200+1102=11951\cdot\frac{120+160}{2}+2\cdot\frac{160+220}{2}+1\cdot\frac{220+200}{2}+3\cdot\frac{200+110}{2}=1195 visitors.

The trapezoidal sum is the average of the left and right sums, a useful check. In context: about 1195 visitors entered during the first 7 hours. Figure 14.1 draws the left sum as rectangles under one possible curve through the table. The curve is an illustration; the table does not determine the rate between its entries.

A graph of entry rate E of t, in visitors per hour, against t, time since opening in hours, with tick marks at the table times 0, 1, 3, 4 and 7. Five filled dots mark the table values (0, 120), (1, 160), (3, 220), (4, 200) and (7, 110), joined by a smooth curve labeled one possible rate curve through the table. Four dotted gray rectangles form the left sum: from 0 to 1 with height 120, from 1 to 3 with height 160, from 3 to 4 with height 220 and from 4 to 7 with height 200. Each rectangle is labeled with its width, 1, 2, 1 and 3, and its height. On the first two subintervals the rectangles lie below the curve; on the last two they rise above it.

Figure 14.1. A left Riemann sum with unequal widths. Each rectangle's area is a height from the table times its own width.

Worked example 14.3: diagnose a table estimate

A student estimates the fair's total with a left sum and writes:

"There are 4 subintervals on [0,7][0,7], so Δt=74=1.75\Delta t=\frac{7}{4}=1.75. The estimate is 1.75(120+160+220+200)=12251.75(120+160+220+200)=1225 visitors. Since EE is decreasing at the end, this is an overestimate."

Two errors. First, the formula Δt=b−an\Delta t=\frac{b-a}{n} assumes equal widths, but these widths are 1, 2, 1 and 3. Each product must use its own width, which gives 1260. Second, the over/under claim has no support. The table shows E(4)>E(7)E(4)>E(7), but it does not say EE decreases throughout each subinterval, and the left sum overestimates only where the rate decreases. Figure 14.1 shows why one curve cannot settle it: on the first two subintervals its rectangles fall short, and on the last two they overshoot.

Repair: "The left sum is 1(120)+2(160)+1(220)+3(200)=12601(120)+2(160)+1(220)+3(200)=1260 visitors. Without information that EE is monotonic on each subinterval, the table alone cannot show whether 1260 is an overestimate or an underestimate."

Turn a limit of sums into a definite integral (topic 6.3)

Divide [a,b][a,b] into nn equal subintervals of width Δx=b−an\Delta x=\frac{b-a}{n}, with right endpoints xk=a+kΔxx_k=a+k\Delta x. For a continuous function ff, the right sums approach a single number as nn increases, and that number is the definite integral:

∫abf(x) dx=lim⁡n→∞∑k=1nf(a+kb−an)b−an\int_a^b f(x)\,dx=\lim_{n\to\infty}\sum_{k=1}^{n}f\left(a+k\frac{b-a}{n}\right)\frac{b-a}{n}

Left and midpoint sums have the same limit. The integral is the signed area between the graph of ff and the xx-axis from aa to bb; when ff is a rate, it is the net change.

To read a limit of sums, find three ingredients. The factor multiplying the sum gives Δx\Delta x. The expression for xkx_k gives the starting value aa. The function applied to xkx_k gives ff. Then b=a+nΔxb=a+n\Delta x.

Worked example 14.4: read the sum's ingredients

Write lim⁡n→∞∑k=1n(1+3kn)3n\lim_{n\to\infty}\sum_{k=1}^{n}\left(1+\frac{3k}{n}\right)\frac{3}{n} as a definite integral and evaluate it.

Method choice. There are two valid readings. With Δx=3n\Delta x=\frac{3}{n} and xk=1+3knx_k=1+\frac{3k}{n}, the start is a=1a=1 and the end is b=1+3=4b=1+3=4. The function applied to xkx_k is f(x)=xf(x)=x:

∫14x dx\int_1^4 x\,dx

Alternatively, let xk=3knx_k=\frac{3k}{n} on [0,3][0,3], so the summand is f(xk)=1+xkf(x_k)=1+x_k and the integral is ∫03(1+x) dx\int_0^3 (1+x)\,dx. The two integrals describe the same region shifted one unit, and both equal the area of a trapezoid with parallel sides 1 and 4 and width 3: 1+42(3)=7.5\frac{1+4}{2}(3)=7.5. Choose whichever reading makes the geometry or the next step simplest. Do not mix them: ∫14(1+x) dx\int_1^4 (1+x)\,dx would add the shift twice.

Differentiate an accumulation function (topic 6.4)

For a function ff continuous on an interval containing aa, define

g(x)=∫axf(t) dtg(x)=\int_a^x f(t)\,dt

the signed area from aa to a variable endpoint xx. The Fundamental Theorem of Calculus says g′(x)=f(x)g'(x)=f(x): the accumulated amount grows at the rate being accumulated. When the upper limit is a function u(x)u(x), the chain rule adds a factor:

ddx∫au(x)f(t) dt=f(u(x)) u′(x)\frac{d}{dx}\int_a^{u(x)}f(t)\,dt=f(u(x))\,u'(x)

A constant lower limit contributes nothing to the derivative. If the variable is in the lower limit, reverse the bounds first, which changes the sign: ddx∫x5f(t) dt=−f(x)\frac{d}{dx}\int_x^5 f(t)\,dt=-f(x).

Worked example 14.5: a variable upper limit

Let G(x)=∫1x211+t3 dtG(x)=\int_1^{x^2}\frac{1}{1+t^3}\,dt. Find G′(x)G'(x) and G′(1)G'(1).

With u(x)=x2u(x)=x^2, G′(x)=11+(x2)3⋅2x=2x1+x6G'(x)=\frac{1}{1+\left(x^2\right)^3}\cdot 2x=\frac{2x}{1+x^6}, and G′(1)=22=1G'(1)=\frac{2}{2}=1.

Two wrong answers are common. "G′(x)=11+x6−12G'(x)=\frac{1}{1+x^6}-\frac{1}{2}" subtracts the integrand's value at the lower limit, as if evaluating; but the lower limit is a constant and contributes 0. "G′(x)=11+x6G'(x)=\frac{1}{1+x^6}" substitutes x2x^2 but drops the chain factor 2x2x. Check by asking what happens when xx changes: x2x^2 changes 2x2x times as fast, so area accumulates 2x2x times as fast.

Read an accumulation function from its rate graph (topic 6.5)

Because g′=fg'=f, a graph of ff is a graph of the slopes of gg. Every Chapter 12 test applies, with ff in the role of g′g'.

  • gg increases where f>0f>0 and decreases where f<0f<0.
  • gg has a relative maximum where ff changes from positive to negative, and a relative minimum where it changes from negative to positive.
  • gg is concave up where ff is increasing and concave down where ff is decreasing; points of inflection of gg occur where ff changes between increasing and decreasing.
  • Values of gg come from signed areas, starting from g(a)=0g(a)=0.

Worked example 14.6: from the graph of ff to the behavior of gg

The graph of ff in Figure 14.2A consists of segments joining (0,1)(0,1), (2,3)(2,3), (5,−3)(5,-3) and (8,0)(8,0). Let g(x)=∫0xf(t) dtg(x)=\int_0^x f(t)\,dt for 0≤x≤80\leq x\leq 8.

Values. The region on [0,2][0,2] is a trapezoid of area 1+32(2)=4\frac{1+3}{2}(2)=4. The segment from (2,3)(2,3) to (5,−3)(5,-3) crosses the axis at x=3.5x=3.5, giving triangles of signed area 2.252.25 and −2.25-2.25. The last region is a triangle of signed area −4.5-4.5. So g(2)=4g(2)=4, g(3.5)=6.25g(3.5)=6.25, g(5)=4g(5)=4 and g(8)=−0.5g(8)=-0.5.

Direction. f>0f>0 on [0,3.5)[0,3.5) and f<0f<0 on (3.5,8)(3.5,8), so gg has its absolute maximum, 6.25, at x=3.5x=3.5. For the absolute minimum, compare the candidates g(0)=0g(0)=0, g(3.5)=6.25g(3.5)=6.25 and g(8)=−0.5g(8)=-0.5: the minimum is −0.5-0.5, at the endpoint x=8x=8.

Concavity. ff increases on (0,2)(0,2), decreases on (2,5)(2,5) and increases on (5,8)(5,8). So the graph of gg is concave up, then down, then up, with points of inflection at x=2x=2 and x=5x=5. Figure 14.2B confirms every conclusion, but each one came from Figure 14.2A alone.

Two stacked graphs sharing a horizontal axis from 0 to 8. Panel A, the rate f: a broken line from (0, 1) up to (2, 3), down through (3.5, 0) to (5, negative 3), then up to (8, 0). The regions above the axis are shaded gray and labeled plus 4 from 0 to 2 and plus 2.25 from 2 to 3.5; the regions below are hatched and labeled minus 2.25 from 3.5 to 5 and minus 4.5 from 5 to 8. Panel B, the accumulation function g of x equals the integral from 0 to x of f: a smooth curve from (0, 0) rising through (2, 4), labeled inflection, to a peak at (3.5, 6.25), labeled maximum, then falling through (5, 4), labeled inflection, and crossing the axis near x equals 7 to end at (8, negative 0.5), labeled minimum. Dotted vertical lines at x equals 2, 3.5 and 5 run through both panels.

Figure 14.2. A rate graph and its accumulation function. Sign changes of ff become turning points of gg; turning points of ff become inflection points of gg.

Try it

Try it 14.1. Water enters a tank at a rate r(t)r(t) liters per minute, with r(0)=8r(0)=8, r(4)=12r(4)=12, r(6)=9r(6)=9 and r(10)=5r(10)=5. Use a trapezoidal sum with these subintervals to estimate the water added during the first 10 minutes, with units.

Solution. 4⋅8+122+2⋅12+92+4⋅9+52=40+21+28=894\cdot\frac{8+12}{2}+2\cdot\frac{12+9}{2}+4\cdot\frac{9+5}{2}=40+21+28=89 liters.

Try it 14.2. Write lim⁡n→∞∑k=1n2n e2k/n\lim_{n\to\infty}\sum_{k=1}^{n}\frac{2}{n}\,e^{2k/n} as a definite integral.

Solution. Δx=2n\Delta x=\frac{2}{n} and xk=2knx_k=\frac{2k}{n} on [0,2][0,2], so the limit is ∫02ex dx\int_0^2 e^x\,dx. An equivalent form is ∫13ex−1 dx\int_1^3 e^{x-1}\,dx.

Try it 14.3. Find H′(x)H'(x) for H(x)=∫x3cos⁡(t2)dtH(x)=\int_x^3\cos\left(t^2\right)dt and K′(x)K'(x) for K(x)=∫0sin⁡x1+t2 dtK(x)=\int_0^{\sin x}\sqrt{1+t^2}\,dt.

Solution. Reverse the bounds of HH: H′(x)=−cos⁡(x2)H'(x)=-\cos\left(x^2\right). For KK, the chain rule gives K′(x)=1+sin⁡2x cos⁡xK'(x)=\sqrt{1+\sin^2 x}\,\cos x.

Check your work

  • Do the units of an accumulated amount equal the rate unit times the input unit?
  • Did each Riemann term use its own subinterval width?
  • Does every over/under claim rest on stated monotonic behavior?
  • In a limit of sums, did you read Δx\Delta x, the start aa and ff before writing the bounds?
  • In a derivative of an accumulation function, did you include the chain factor and the sign from a variable lower limit?
  • When reading gg from a graph of ff, did you use the sign of ff for direction and the slope of ff for concavity?

Practice

This chapter's bank items are APCAL-C14-M01 to APCAL-C14-M16, printed after the companion panel as Questions 182–197, and the nine-point written tasks APCAL-C14-F01 and APCAL-C14-F02, printed as Written tasks 19 and 20. Questions 182–191 and Written task 20 are no-calculator work. Questions 192–197 and Written task 19 require a graphing calculator in radian mode; give decimals to the nearest thousandth. Answers, the error behind each wrong choice, rubrics and complete responses are in Answers, explanations and written responses, under Chapter 14: answers and explanations.

Review and retrieve

Wait two days. Then, without notes, recompute the three sums in Worked example 14.2 and redraw Figure 14.2B from Figure 14.2A alone. About a week later, mix this chapter with Chapter 12: for g(x)=∫0x(t2−4t+3)dtg(x)=\int_0^x \left(t^2-4t+3\right)dt, use only the integrand to locate the relative extrema and the point of inflection of gg. Check: a relative maximum at x=1x=1, a relative minimum at x=3x=3, and an inflection point at x=2x=2.

Use this chapter companion

Review the lesson, attempt the original practice, then check the explanation. Your printed study guide remains complete without the website.

Independence

Skill by Skill Prep is independent. SAT® and AP® are trademarks registered by the College Board, which is not affiliated with, and does not endorse, this website or book. PSAT/NMSQT® is a registered trademark of the College Board and the National Merit Scholarship Corporation, which are not affiliated with, and do not endorse, this site. Bluebook™ is a trademark of the College Board. ACT® is a trademark of ACT Education Corp., which is not affiliated with and does not endorse this website or book. Desmos is a trademark of Desmos Studio PBC. Practice results describe these questions, not official scores or admission predictions.

Last updated .