Calculus II Section 3.7 Exhaustive Study Notes on Improper Integrals and Comparison Tests
Overview and Foundations of Improper Integrals
In standard Calculus I definite integration, integrals are interpreted geometrically as the net area under a curve. These proper integrals require two fundamental conditions:
The domain of integration must be a closed, bounded interval [a,b] of finite length.
The integrand function f(x) must remain bounded on the entire interval [a,b].
Improper integrals extend integration and geometric area calculations to scenarios where these standard conditions are relaxed, specifically:
When the domain of integration is unbounded (extending to \n\infty, −∞, or both).
When the integrand function f(x) has one or more infinite discontinuities (vertical asymptotes) on or within the interval of integration.
Type 1: Improper Integrals over Unbounded Intervals
An improper integral over an unbounded interval is defined rigorously as the limit of proper definite integrals over bounded subintervals as an endpoint approaches infinity.
Upper Limit Unbounded ([a,∞)):
For a fixed real number a∈R:
∫a∞f(x)dx=limb→∞∫abf(x)dx
Convergence Condition: The improper integral converges if the limit exists as a finite real number.
Divergence Condition: The improper integral diverges if the limit does not exist, or equals ∞ or −∞
Lower Limit Unbounded ((−∞,b]):
For a fixed real number b∈R:
∫−∞bf(x)dx=lima→−∞∫abf(x)dx
Both Limits Unbounded ((−∞,∞)):
For integration across the entire real line:
∫−∞∞f(x)dx=lima→−∞∫acf(x)dx+limb→∞∫cbf(x)dx
where c is any arbitrarily chosen real number (commonly chosen as c=0).
Crucial Rule: The integral ∫−∞∞f(x)dx converges if and only if both individual limits on the right-hand side converge independently. If either limit diverges, the entire integral diverges.
p-Integrals on Unbounded Intervals
Integrals of the form ∫a∞xp1dx (where a>0) are known as p-integrals. They represent a fundamental benchmark in calculus for convergence testing.
General Theorem for Unbounded p-Integrals:∫a∞xp1dx={p−1a1−pdivergesp>1,p≤1,for any a>0
Detailed Proof/Derivation by Cases (taking a=1):
Case 1 (p>1):∫1∞xp1dx=limb→∞∫1bx−pdx=limb→∞[1−px1−p]1b=limb→∞(1−pb1−p−1−p11−p)
Because p>1, the exponent 1−p<0, meaning limb→∞b1−p=0. Thus:
∫1∞xp1dx=0−1−p1=p−11
This yields a finite value, so the integral converges.
Case 2 (p=1):∫1∞x1dx=limb→∞∫1bx1dx=limb→∞[ln(x)]1b=limb→∞(ln(b)−ln(1))=∞
The limit is infinite, so the integral diverges.
Case 3 (p<1):∫1∞xp1dx=limb→∞[1−px1−p]1b
Because p<1, the exponent 1−p>0, meaning limb→∞b1−p=∞. Thus, the integral diverges.
Type 2: Improper Integrals with Discontinuous Integrands
A Type 2 improper integral occurs when the domain is bounded [a,b], but the integrand f(x) approaches ±∞ at one or more points in the domain.
Discontinuity at Left Endpoint (x=c):
If limx→c+f(x)=±∞ for a real number c:
∫cbf(x)dx=limr→c+∫rbf(x)dx
Discontinuity at Right Endpoint (x=c):
If limx→c−f(x)=±∞ for a real number c:
∫acf(x)dx=limr→c−∫arf(x)dx
Discontinuity at an Interior Point (x=c where a<c<b):
If f(x) has a vertical asymptote at c∈(a,b), the integral must be split into two separate integrals:
∫abf(x)dx=∫acf(x)dx+∫cbf(x)dx=limr→c−∫arf(x)dx+lims→c+∫sbf(x)dx
Both limits must exist and be finite for the overall integral to converge.
p-Integrals on Bounded Intervals
When considering the bounded interval (0,a] with an infinite discontinuity at the origin x=0, the behavior of p-integrals reverses compared to the unbounded case.
General Theorem for Bounded p-Integrals:∫0axp1dx={1−pa1−pdivergesp<1,p≥1,for any a>0
Summary Comparison of p-Integral Rules:
Domain [a,∞) (Unbounded region): Converges for p>1, Diverges for p≤1.
Domain (0,a] (Bounded interval with singularity at 0): Converges for p<1, Diverges for p≥1.
Boundary Case p=1: ∫x1dxdiverges in both domains because ln(x) grows without bound both as x→∞ and as x→0+.
Non-Elementary Functions and Motivation for Testing Convergence
Antiderivatives cannot always be expressed using elementary functions (polynomials, rational functions, algebraic expressions, trigonometric functions, inverse trigonometric functions, logarithms, and exponentials).
Key Examples of Functions Without Elementary Antiderivatives:
xsin(x) (Sinc function / Sine Integral)
e−x2 (Gaussian distribution function)
1−x4 (Elliptic integral integrand)
sin(x2) (Fresnel sine integral)
In practical applications, evaluating the precise numerical value of an integral is often unnecessary or impossible analytically. Determining whether the improper integral converges to a finite value is sufficient.
The Comparison Test for Improper Integrals
The Comparison Test allows us to determine the convergence or divergence of an improper integral with a complex or non-elementary integrand by comparing it to a simpler function with known convergence properties (such as a p-integral or exponential function).
Theorem 0.1 (Comparison Test):
Fix a∈R and suppose f and g are integrable functions satisfying:
f(x)≥g(x)≥0for all x∈(a,∞)
Then their improper integrals preserve this inequality:
∫a∞f(x)dx≥∫a∞g(x)dx
Direct Comparison Rules:
Convergence Rule: If the larger integral ∫a∞f(x)dx converges, then the smaller integral ∫a∞g(x)dx must also converge.
Divergence Rule: If the smaller integral ∫a∞g(x)dx diverges, then the larger integral ∫a∞f(x)dx must also diverge.