Explore the distinction between absolute and conditional convergence for improper integrals, master the Dirichlet and Abel tests for oscillating integrands, and analyze classic examples like ∫sin(x)/x dx.
When dealing with improper integrals of oscillating functions, we encounter a fascinating phenomenon: some integrals converge due to cancellation effects, even though the absolute value of the integrand doesn't have a finite integral.
Classic Example
Converges (conditionally)
But its absolute value...
Diverges!
This distinction leads to two fundamentally different types of convergence with important implications for analysis.
The concepts of absolute and conditional convergence for integrals mirror those for infinite series:
An improper integral is said to converge absolutely if:
That is, the integral of the absolute value of the function converges.
An improper integral is said to converge conditionally if:
The convergence relies on cancellation between positive and negative parts.
If converges, then converges.
That is: Absolute convergence ⟹ Convergence
Given: converges
To Show: converges
Proof:
Note that , so:
By comparison test, converges.
Since and both integrals on the right converge:
converges by linearity. ∎
Important:
Convergence does NOT imply absolute convergence.
Example: converges, but diverges.
Problem: Show converges absolutely.
Solution:
We have:
Since converges (p = 2 > 1):
By comparison test, converges.
Therefore the original integral converges absolutely.
Problem: Classify
Solution:
Check absolute convergence:
Since converges (p = 3/2 > 1):
The integral converges absolutely.
Let and be defined on . The integral converges if:
Idea: Use integration by parts and show the boundary term vanishes while the remaining integral converges.
Let . Then and for all x.
As : since and is bounded.
The integral converges absolutely since and . ∎
The Dirichlet test is perfect for integrals involving:
Problem: Show converges.
Solution (Dirichlet Test):
Let and .
Check conditions:
By Dirichlet test, the integral converges.
Problem: Show converges.
Solution:
Let and .
By Dirichlet test, the integral converges.
The integral converges if:
Dirichlet Test
Requires: bounded partial integrals of f, g → 0 monotonically
Abel Test
Requires: ∫f converges, g bounded and monotonic
Problem: Show converges.
Solution:
Let and .
By Abel test, the integral converges.
The integral is the canonical example of conditional convergence. Let's prove both parts: convergence and non-absolute convergence.
The integral converges.
Apply the Dirichlet test with and :
All conditions satisfied. ∎
The integral diverges.
Use the inequality :
The first integral .
The second integral converges (by Dirichlet test).
So . ∎
Therefore, converges conditionally.
For :
If and both converge, then converges absolutely for any constants .
Warning:
Unlike absolutely convergent integrals, conditionally convergent ones:
Problem: Classify
Solution:
Substitute , so :
This is a Fresnel-type integral. Since converges conditionally:
The original integral converges conditionally.
Problem: Determine convergence of
Solution:
Use identity:
Substitute :
This converges conditionally (same as sin(x)/x).
Problem: Evaluate
Solution:
Problem: Evaluate
Solution:
Substitute , :
Problem: Evaluate
Solution:
Note as , so this is Type II.
Since (by L'Hôpital):
Problem: Evaluate
Solution:
Singularities at both and .
Near :
Near :
Both are p = 1/2 < 1, so both parts converge.
Using substitution : the integral equals .
Problem: Classify
Solution:
Check absolute convergence:
Since p = 3 > 1, converges.
The integral converges absolutely.
Problem: Classify
Solution:
Use identity:
First integral diverges (p=1). Second converges (Dirichlet).
Since one part diverges, the integral diverges.
Problem: Classify
Solution:
Check absolute convergence:
The Gaussian integral converges.
The integral converges absolutely.
Problem: Classify
Solution:
Check absolute convergence:
Since converges (p = 1/2 < 1):
The integral converges absolutely.
Problem: Classify
Solution:
Let , .
Check Dirichlet conditions:
By Dirichlet test, the integral converges conditionally.
Wrong: Thinking that if converges, then must also converge.
Right: Convergence does NOT imply absolute convergence. Example: converges but diverges.
Wrong: Trying to apply comparison test directly to .
Right: Comparison test requires non-negative functions. For oscillating integrands, use Dirichlet or Abel test.
Wrong: Applying Dirichlet test when partial integrals are not bounded.
Right: Verify ALL conditions: (1) bounded partial integrals of f, (2) g monotonically decreasing, (3) g → 0.
Wrong: Thinking Abel test requires g → 0.
Right: Abel requires to converge + g bounded monotonic. Dirichlet requires bounded partials of f + g → 0 monotonically.
Wrong: Assuming converges absolutely.
Right: For : absolute convergence only when p > 1. Here p = 0.5 < 1, so it's conditionally convergent.
Wrong: Concluding conditional convergence after showing converges without checking .
Right: You must verify BOTH that converges AND that diverges to conclude conditional convergence.
Wrong: Thinking .
Right: , but they're not equal. Use for analysis.
Problem: Evaluate
Solution:
Use integration by parts twice, or recognize this as the real part of:
Taking the real part:
Problem: Evaluate
Solution:
This is related to the Gamma function. Using integration by parts twice:
The boundary term vanishes. For the remaining integral:
So
Problem: Analyze
Solution:
Singularities at both and .
Near : → p = 1/2 < 1 ✓
Near : → p = 1/2 < 1 ✓
Both singularities are integrable. Using substitution : the integral equals .
Problem: Evaluate
Solution:
Factor:
Partial fractions:
Problem: Evaluate
Solution:
Note
Substitute , :
Before evaluating any integral, scan the integrand for points where it might blow up. Check denominators, square roots, and logarithms.
: needs p > 1. : needs p < 1. Opposite!
Many improper integrals require L'Hôpital's rule or known limits like .
For mixed integrals, choose splitting points that separate singularities and infinite limits. Often x = 1 works well.
Substitutions can convert Type I to Type II or vice versa. For example, swaps 0 and ∞.
Many convergence tests for improper integrals parallel those for series. This helps with intuition.
| Integral | Type | Converges? | Value |
|---|---|---|---|
| Type I | Yes (p=2>1) | 1 | |
| Type I | No (p=1) | — | |
| Type II | Yes (p=½<1) | 2 | |
| Type II | No (p=1) | — | |
| Type I | Yes | 1 | |
| Type I (both) | Yes | π | |
| Type II | Yes | −1 | |
| Mixed | No | — |
Challenge 1
Determine convergence and evaluate if convergent:
Substitute u = 1 + x².
Challenge 2
For what values of α does the following converge?
Near x = 1, use L'Hôpital. Near x = 0, compare with x^α.
Challenge 3
Evaluate the following integral:
Split at x = 1 and use the substitution u = 1/x on one part.
Challenge 4
Prove that the following integral converges:
Find the singularities and check the exponent at each.
| Integral | Classification | Reason |
|---|---|---|
| Conditional | Dirichlet test; diverges | |
| Absolute | , p-test | |
| Conditional | Dirichlet test; p = 1/2 < 1 | |
| Absolute | p = 1.5 > 1, comparison test | |
| Absolute | ||
| Conditional | Fresnel-type, substitution u=x² |
Check if ∫|f| converges
Use comparison, p-test, or other standard tests on |f(x)|
If ∫|f| converges → Absolutely Convergent
Done! The integral converges and is well-behaved.
If ∫|f| diverges, check if ∫f converges
Use Dirichlet or Abel test for oscillating integrands
If ∫f converges but ∫|f| diverges → Conditionally Convergent
The convergence depends on cancellation effects.
If both diverge → Divergent
The integral does not converge.
Problem: Classify
Solution:
Step 1: Check absolute convergence.
Since diverges (logarithmic p-test with p=1):
Step 2: Apply Dirichlet test.
Let (bounded partials), (decreasing to 0).
The integral converges conditionally.
Problem: Classify
Solution:
Check absolute convergence:
Since converges:
The integral converges absolutely.
Problem: Classify
Solution:
Substitute , so :
This integral diverges (oscillates without decay).
Absolute: ∫|f| converges. Conditional: ∫f converges but ∫|f| diverges. Absolute convergence is stronger and implies regular convergence.
Use Dirichlet when you have a product f(x)g(x) where g(x) → 0 monotonically and the partial integrals of f are bounded (like sin x or cos x).
If ∫f converges and g is bounded and monotonic, then ∫fg converges. It's useful when g doesn't go to zero but is well-behaved.
The oscillations of sin(x) cause cancellation that makes the integral converge. But |sin(x)| ≥ sin²(x), and ∫sin²(x)/x dx ~ ∫1/(2x) dx which diverges.
First check if ∫|f| converges using comparison or p-tests. If yes, it's absolutely convergent. If ∫|f| diverges but ∫f converges (often by Dirichlet/Abel), it's conditionally convergent.
Oscillating functions like sin(x), cos(x), sin(ax+b), cos(ax+b), and bounded periodic functions. The key is that positive and negative areas cancel out.
Not directly. Comparison test requires non-negative functions. For oscillating integrands, use Dirichlet or Abel tests instead.
For p > 1: absolutely convergent. For 0 < p ≤ 1: conditionally convergent. For p ≤ 0: divergent. The threshold is p = 1.
Absolutely convergent integrals can be rearranged, split arbitrarily, and composed more freely. Conditionally convergent integrals are 'fragile' - their value can change with rearrangement.
Fresnel integrals ∫sin(x²)dx and ∫cos(x²)dx converge conditionally. They arise in optics and are related to ∫sin(x)/x through substitution.
converges → stronger, well-behaved
converges but diverges → fragile
Bounded partials of f + g → 0 monotonically
converges + g bounded monotonic
For : Absolute if p > 1, Conditional if 0 < p ≤ 1, Divergent if p ≤ 0