Tests edge cases, boundary conditions, and common misconceptions for integrals.
UT-1: Riemann Sum Identification and Limit Conversion
Question:
A student encounters the limit:
limn→∞∑k=1nn3k24−n2k2
(a) Express this limit as a definite integral. (b) Identify which type of Riemann sum (left, right,
midpoint, or trapezoidal) this represents, or state if it is ambiguous. (c) Evaluate the definite
integral.
Solution:
(a) Factor out n1:
∑k=1nn3k24−n2k2=∑k=1nn1⋅(nk)24−(nk)2
With Δx=n1 and xk=nk (right endpoint of the k-th subinterval on
[0,1]):
limn→∞∑k=1nn1⋅xk24−xk2=∫01x24−x2dx
(b) Since xk=nk uses the right endpoint of each subinterval, this is a right Riemann
sum.
(c) Use the substitution x=2sinθ, dx=2cosθdθ:
∫01x24−x2dx=∫0arcsin(1/2)4sin2θ⋅2cosθ⋅2cosθdθ
=16∫0π/6sin2θcos2θdθ=4∫0π/6sin2(2θ)dθ
Using sin2(2θ)=21−cos(4θ):
=4∫0π/621−cos(4θ)dθ=2[θ−4sin(4θ)]0π/6
=2(6π−4sin(2π/3))=2(6π−43/2)=3π−43
UT-2: Integration by Parts with LIATE Trap
Question:
Evaluate ∫e2xsin(3x)dx.
A student chooses u=sin(3x) and dv=e2xdx for the first application of integration by
parts. Show that this choice works but leads to a longer computation than choosing u=e2x.
Evaluate the integral completely using the more efficient choice and explain why “LIATE” alone does
not settle this choice.
Solution:
LIATE ranks: Logarithmic, Inverse trig, Algebraic, Trigonometric, Exponential.
Both e2x (Exponential) and sin(3x) (Trigonometric) are in the LIATE list. Trigonometric
comes before Exponential, so LIATE suggests u=sin(3x). But for products of exponentials and
trig functions, either choice works, and the key is to apply integration by parts twice and
solve algebraically.
Using u=e2x, dv=sin(3x)dx:
du=2e2xdx, v=−31cos(3x).
∫e2xsin(3x)dx=−31e2xcos(3x)+32∫e2xcos(3x)dx
Apply parts again to the remaining integral. Let u=e2x, dv=cos(3x)dx:
du=2e2xdx, v=31sin(3x).
∫e2xcos(3x)dx=31e2xsin(3x)−32∫e2xsin(3x)dx
Substituting back:
I=−31e2xcos(3x)+32(31e2xsin(3x)−32I)
I=−31e2xcos(3x)+92e2xsin(3x)−94I
913I=e2x(−31cos(3x)+92sin(3x))
I=13e2x(2sin(3x)−3cos(3x))+C
The misconception: students sometimes stop after one application of parts, or they choose a
different assignment for the second integration by parts (switching u and dv), which creates a
circular argument instead of solving for I.
UT-3: Area Between Curves When Functions Cross the x-Axis
Question:
Find the total area of the region bounded by y=x3−4x and y=x2−4.
A student computes ∫−22[(x3−4x)−(x2−4)]dx and gets the
wrong answer. Explain the error and compute the correct total area.
Solution:
First find intersection points: x3−4x=x2−4⟹x3−x2−4x+4=0.
By inspection x=1 is a root: 1−1−4+4=0. Factoring:
(x−1)(x2−4)=(x−1)(x−2)(x+2)=0
Intersection points: x=−2,1,2.
The student’s error: they integrated from −2 to 2 without accounting for the curve crossing at
x=1. On [−2,1]We must determine which curve is on top; on [1,2]The other may be on top.
Test point x=0: x3−4x=0, x2−4=−4. So x3−4x>x2−4 on [−2,1].
Test point x=1.5: x3−4x=3.375−6=−2.625, x2−4=2.25−4=−1.75. So
x2−4>x3−4x on [1,2].
The student’s integral gives ∫−22(x3−x2−4x+4)dx=328
only because the areas happen to be positive in both subintervals when separated correctly. The
student’s single integral from −2 to 2 actually evaluates to:
∫−22(x3−x2−4x+4)dx=[4x4−3x3−2x2+4x]−22
This gives 328 here too, but this is coincidental. The fundamental error is not splitting
at intersection points, which would give wrong answers .
Integration Tests
Tests synthesis of integrals with other topics.
IT-1: Volume of Revolution with Washer and Shell Method Verification (with Derivatives)
Question:
Let R be the region bounded by y=x, y=0And x=4.
(a) Find the volume generated when R is revolved about the line x=6 using the shell
method. (b) Verify your answer using the washer method. (c) A student claims the volume about
x=6 should be the same as the volume about the y-axis because “it’s just a translation.”
Explain why this is false, and compute the volume about the y-axis for comparison.
Solution:
(a) Shell method (parallel to axis of revolution): use horizontal shells.
A shell at height y has radius r=6−x=6−y2 and height h=dy (thin strip). For shells,
we integrate along the axis perpendicular to the axis of revolution.
Since we revolve about x=6 (vertical line), shells are vertical: radius =6−xHeight
=x−0=x.
(b) Washer method: washers perpendicular to x=6So we integrate with respect to y.
Outer radius: R=6−0=6 (from x=6 to the y-axis). More precisely, for washers
perpendicular to the axis x=6: at height yThe region extends from x=y2 to x=4.
Revolved about x=6:
Outer radius: 6−y2 (from axis to the left edge of region at x=y2)
Inner radius: 6−4=2 (from axis to the right edge of region at x=4)
V=π∫02[(6−y2)2−22]dy=π∫02(36−12y2+y4−4)dy
=π∫02(32−12y2+y4)dy=π[32y−4y3+5y5]02
=π(64−32+532)=π(32+532)=5192π
(c) Volume about the y-axis (disk method):
V=π∫04(x)2dx=π∫04xdx=π⋅8=8π
The volumes are different (5192π≈120.6 vs 8π≈25.1). Revolving about
a different axis changes the radius of every point, so the volume changes. The student’s
“translation” argument is wrong because the region itself does not translate — only the axis does,
which changes the distance from every point to the axis.
IT-2: Improper Integral Convergence with Parameter (with Limits)
Question:
For what values of p does the improper integral
∫0∞1+xxp−1dx converge?
Hint: Split the integral at x=1 and analyze the behavior of each part separately using
comparison with p-integrals.
Solution:
Split at x=1:
∫0∞1+xxp−1dx=∫011+xxp−1dx+∫1∞1+xxp−1dx
Near x=0 (first integral): When x∈(0,1), 21<1+x1<1So:
2xp−1<1+xxp−1<xp−1
By comparison with ∫01xp−1dx (converges iff p−1>−1I.e., p>0):
The first integral converges iff p>0.
As x→∞ (second integral): When x>1,
21<1+x1<x1So:
2xp−1<1+xxp−1<xp−2
By comparison with ∫1∞xp−2dx (converges iff p−2<−1I.e.,
p<1):
The second integral converges iff p<1.
Conclusion: The full integral converges iff both parts converge, i.e., 0<p<1.
Note: this integral equals B(p,1−p)=πcsc(πp) (the beta function), which has the same
domain of convergence.
IT-3: FTC Part 1 vs Part 2 Confusion with Piecewise Functions (with Continuity)
Question:
Let
f(x) = \begin{cases} 2x & \text{if 0 \leq x \lt 2 \\ 8 - 2x & \text{if 2 \leq x \leq 4 \end{cases}
Let F(x)=∫0xf(t)dt.
(a) Find and graph F(x). (b) Is F differentiable at x=2? Compute F′(2) from the definition
of the derivative and explain. (c) A student claims that since f has a corner at x=2, F must
also have a corner at x=2. Is this correct?
Since 4=4, Fis differentiable at x=2With F′(2)=4=f(2).
(c) The student is incorrect. Although f has a corner at x=2 (f changes from slope +2
to slope −2), f is continuous at x=2 (f(2−)=4=f(2+)). By FTC part 1, since f is
continuous at x=2, F is differentiable at x=2 with F′(2)=f(2)=4. Integration
“smooths” the corner: F is continuously differentiable even though f is not.
Summary
The key principles covered in this topic are linked in the sub-pages above. Focus on understanding
the definitions, applying the formulas or frameworks, and evaluating strengths and limitations of
each approach.
Worked Examples
Worked examples demonstrating the application of key concepts are covered in the detailed sub-pages
linked above.
Common Pitfalls
Confusing terminology or concepts that appear similar but have distinct meanings.
Overlooking key assumptions or boundary conditions that limit applicability.