Improper integrals appear when an interval is infinite or when the function becomes unbounded at a point in the interval. They matter because they decide whether a total accumulated quantity, such as area, probability, charge, or work, is finite or infinite. Instead of evaluating the integral in the usual way, we rewrite it as a limit and test whether that limit exists as a finite number.
Convergence means the infinite process produces a finite result, while divergence means it does not.
Understanding Calculus: Convergence Tests for Improper Integrals
A useful way to think about convergence is to watch a running total. Start at a fixed point and keep extending the interval farther out. The graph of the function may stay above zero forever, yet each new strip can become so small that the total approaches one fixed value.
In that case, the remaining tail becomes negligible. By contrast, a function can approach zero and still add enough area to keep the total growing without bound. Approaching zero is necessary for many positive tails to converge, but it is not enough by itself.
Reciprocal powers mark an important boundary because integration changes the exponent in a predictable way. Far from zero, faster decay gives less area in every later interval. The borderline reciprocal function decreases, though too slowly to produce a finite total.
A slightly larger power makes the decay strong enough. Near zero, the situation reverses. The issue is not a long tail but an increasingly tall spike squeezed into a tiny width.
A reciprocal power that is harmless at infinity can be too large near zero. Students should always identify where the trouble occurs before choosing a test, since the same expression can have different behavior at different locations.
Comparison tests are most useful when an exact antiderivative is difficult or unavailable. For a positive rational function, the highest powers in the numerator and denominator often control its far away behavior. Lower degree terms become relatively unimportant as the input grows.
This lets you compare the function with a simpler reciprocal power. Limit comparison formalizes that idea. If the ratio of two positive functions settles to a finite positive number, they eventually differ only by an approximate constant factor.
They therefore have matching convergence behavior. With direct comparison, the direction matters. A smaller nonnegative function inherits convergence from a larger convergent one.
A larger nonnegative function inherits divergence from a smaller divergent one. The comparison only needs to work after some point, because any finite starting piece contributes only a finite amount.
A vertical asymptote needs equally careful treatment. Splitting at the bad point is essential, even if the function has values on both sides. One side cannot rescue the other side.
This becomes especially important when signs change. Positive and negative infinite contributions may appear to cancel in an informal calculation, but that is not a genuine finite integral unless the required one-sided limits behave properly. In later courses, absolute convergence gives a safer standard by checking whether the integral of the magnitude is finite.
These ideas appear in probability densities, where total probability must be finite, and in physics models with concentrated charge or mass. When practicing, sketch the location of the tail or spike, choose a known benchmark, and state clearly which part of the interval your reasoning addresses.
Key Facts
- Infinite interval definition: ∫_a^∞ f(x) dx = lim_(b→∞) ∫_a^b f(x) dx
- Vertical asymptote definition: ∫_a^b f(x) dx with f unbounded at c equals ∫_a^c f(x) dx + ∫_c^b f(x) dx, and both parts must converge.
- p-integral on [1,∞): ∫_1^∞ 1/x^p dx converges if p > 1 and diverges if p ≤ 1.
- p-integral near 0: ∫_0^1 1/x^p dx converges if p < 1 and diverges if p ≥ 1.
- Direct comparison: If 0 ≤ f(x) ≤ g(x) for large x and ∫ g(x) dx converges, then ∫ f(x) dx converges.
- Limit comparison: If f(x), g(x) ≥ 0 and lim_(x→∞) f(x)/g(x) = L with 0 < L < ∞, then ∫ f(x) dx and ∫ g(x) dx have the same behavior.
Vocabulary
- Improper integral
- An integral with an infinite interval of integration or an integrand that becomes unbounded on the interval.
- Convergence
- The property that an improper integral has a finite value when written as a limit.
- Divergence
- The property that an improper integral does not approach a finite value.
- Comparison test
- A test that proves convergence or divergence by bounding one nonnegative function above or below another.
- Limit comparison test
- A test that compares the long-term size of two nonnegative functions using the limit of their ratio.
Common Mistakes to Avoid
- Forgetting to rewrite the integral as a limit. An improper integral is not evaluated directly over infinity or through an asymptote.
- Using the p-integral rule with the wrong interval. For ∫_1^∞ 1/x^p dx convergence requires p > 1, but for ∫_0^1 1/x^p dx convergence requires p < 1.
- Applying direct comparison in the wrong direction. To prove convergence, the function must be smaller than a known convergent function; to prove divergence, it must be larger than a known divergent function.
- Ignoring a discontinuity inside the interval. If an integrand is unbounded at an interior point, the integral must be split there and every resulting improper integral must converge.
Practice Questions
- 1 Determine whether ∫_1^∞ 5/x^3 dx converges or diverges, and find its value if it converges.
- 2 Use limit comparison to decide whether ∫_2^∞ (3x + 1)/(x^3 - 4) dx converges or diverges.
- 3 Explain why ∫_1^∞ 1/(x + sin x) dx diverges by comparing its long-term behavior to a simpler improper integral.