Understand the idea
An improper integral involves an infinite interval or an unbounded integrand. Its value is defined by a limit of ordinary definite integrals. A finite limit means the improper integral converges; an infinite or nonexistent limit means it diverges. An antiderivative alone does not settle convergence.
Replace an infinite bound with a variable, evaluate the finite integral, and only then take the limit. For an integrand unbounded at a finite endpoint, approach that endpoint from within the interval. The integral can converge even when the function becomes arbitrarily large near an endpoint.
If a singularity lies inside the interval, split the integral there. Both one-sided improper integrals must converge independently. For example, cancellation on symmetric intervals cannot rescue an integral whose two sides separately diverge; that would be a different limiting procedure.
A method you can apply
- Identify every infinite bound or point where the integrand is unbounded.
- Write a separate limit for each improper part, then evaluate its finite integral.
- State explicitly whether every required limit is finite before combining the results.
Worked example
Determine whether converges and find its value.
- Write .
- An antiderivative is , so the finite integral is .
- The term approaches as tends to infinity.
Try it yourself
Does converge? Find its value if it does.
Show solution and reasoning
Use . The unbounded endpoint does not prevent convergence. In contrast, diverges.
