Integrals over infinite intervals, or of functions that blow up, defined carefully as limits. The recurring question is always the same: does the limit exist? The tests here are a direct rehearsal for the series tests two units from now.
Many vital integrals run to infinity or have a singularity — probability distributions, decaying signals, gravitational potentials. Improper integrals give them a rigorous meaning by taking a limit of ordinary, proper integrals.
The Gaussian integral $\int_{-\infty}^{\infty}e^{-x^2}\,dx=\sqrt{\pi}$ normalizes the bell curve; Laplace and Fourier transforms over infinite domains power signal processing and control engineering; the exponential waiting-time distribution in queueing theory is defined by one.
Torricelli's "Gabriel's Horn" (1640s) has finite volume but infinite surface area — fillable with a finite amount of paint, yet impossible to paint. A vivid early warning that infinity defies intuition, and a direct ancestor of the $p$-test below.
$\displaystyle\int_a^\infty f=\lim_{t\to\infty}\int_a^t f$. If $f$ is unbounded near $b$, $\displaystyle\int_a^b f=\lim_{t\to b^-}\int_a^t f$. The integral converges exactly when this limit exists (as a finite number); otherwise it diverges. An integral improper at both ends is split at any interior point, and both pieces must converge.
$\displaystyle\int_1^\infty x^{-p}\,dx$ converges iff $p>1$; $\displaystyle\int_0^1 x^{-p}\,dx$ converges iff $p<1$. The two cases are mirror images: near infinity, large $p$ decays fast enough; near $0$, only small $p$ keeps the blow-up mild enough.
If $0\le f\le g$ on $[a,\infty)$ and $\int_a^\infty g$ converges, so does $\int_a^\infty f$; if $\int_a^\infty f$ diverges, so does $\int_a^\infty g$. (Squeeze a nonnegative function between $0$ and something integrable, or below something that blows up.)
If $f,g\ge0$ and $\displaystyle\lim_{x\to\infty}\dfrac{f(x)}{g(x)}=L$ with $0<L<\infty$, then $\int_a^\infty f$ and $\int_a^\infty g$ either both converge or both diverge. Useful when direct comparison is awkward to set up.
$\int_1^\infty\dfrac{dx}{x^2+1}$ converges: $0\le\dfrac1{x^2+1}\le \dfrac1{x^2}$ and $\int_1^\infty x^{-2}dx$ converges ($p=2>1$).
$\int_1^\infty\dfrac{x}{\sqrt{x^5+1}}\,dx$: compare with $g(x)=x^{-3/2}$.
• Always rewrite as a limit; never substitute $\infty$ or the singular point
directly into the antiderivative.
• Watch for a hidden singularity inside the interval, not just at the
endpoints — split the integral there.
• Comparison (direct or limit) needs eventually nonnegative functions.
• Limit comparison requires the limit to be a positive finite number — $L=0$ or
$L=\infty$ only give one-directional information.
Decide convergence before you try to evaluate — often you don't need the antiderivative at all.
$\int_0^\infty e^{-x}\,dx$.
Does $\int_1^\infty\dfrac{dx}{x}$ converge?
Converge or diverge: $\int_1^\infty\dfrac{2+\sin x}{x^2}\,dx$?
Does $\int_0^1\dfrac{dx}{\sqrt x}$ converge? Evaluate it.
Does $\int_0^\infty\dfrac{dx}{\sqrt x(1+x)}$ converge?
Does $\int_2^\infty\dfrac{\ln x}{x^2}\,dx$ converge?
Show $\int_1^\infty\dfrac{\cos x}{x^2}\,dx$ converges.
The horn formed by rotating $y=1/x$ for $x\ge1$ has volume $\int_1^\infty\pi/x^2\,dx$ and surface area at least $\int_1^\infty2\pi/x\,dx$. Show the volume is finite but the surface area is infinite.
Course coverage. This unit is part of the complete MAT159H1 topic map. It should cover the following items before you move on:
Use the current UofT MAT159H1 calendar topics as the coverage spine, and pair lecture notes with a Spivak-style proof-calculus text for extra reading. The exact assigned textbook should still be checked against the instructor syllabus.
The improper-integral spine is: locate every improper point, split the integral there, and require every resulting limit to converge.
Use these after the main practice set. They are designed to force definition-level reasoning, theorem selection, and mixed application rather than routine calculation only.
Decide convergence of $\int_2^\infty dx/(x(\ln x)^2)$.
Determine convergence of $\int_1^\infty (3x^2+1)/(x^5+x)dx$.
Analyze $\int_0^1 dx/x$.
If $\int |f|$ converges, what can you conclude about $\int f$?