Prove that the improper integral ∫₀^∞ 1/(1 + x² sin²x) dx is divergent.
15 marks · directive: Prove
The core of a good answer
On each interval [kπ, (k + 1)π], x² sin²x ≤ (k + 1)²π² sin²x, so the integral over it is at least ∫₀^π dt/(1 + (k + 1)²π² sin²t) = π/√(1 + (k + 1)²π²) ≥ 1/(2(k + 1)); summing over k gives a harmonic-type series, so ∫₀^∞ diverges.
An outline written by upsc.help, not an official UPSC answer; every fact checked by a separate review. How we work
What the question asks
The integrand is positive but does not tend to 0 (it equals 1 at every x = kπ), and comparison with a simple function fails because it dips to about 1/x² between the multiples of π. The method is to split into periods, bound each piece from below using the standard integral ∫₀^π dt/(1 + c sin²t) = π/√(1 + c), and compare with the harmonic series.
Answer outline
Introduction
The integrand f(x) = 1/(1 + x² sin²x) is continuous and positive on [0, ∞), so ∫₀^∞ f converges iff the partial integrals ∫₀^X f are bounded. Split [0, ∞) into Iₖ = [kπ, (k + 1)π], k = 0, 1, 2, ...
Step 1: Bound on each period
- On Iₖ, 0 ≤ x ≤ (k + 1)π, so x² sin²x ≤ (k + 1)²π² sin²x.
- Hence f(x) ≥ 1/(1 + cₖ sin²x), cₖ = (k + 1)²π².
Step 2: The standard integral
- sin² has period π, so ∫_{Iₖ} dx/(1 + c sin²x) = ∫₀^π dt/(1 + c sin²t).
- Using t ↦ tan t: ∫₀^π dt/(1 + c sin²t) = 2∫₀^{π/2} sec²t dt/(1 + (1 + c)tan²t) = 2 · (1/√(1 + c)) · (π/2) = π/√(1 + c).
Step 3: Lower bound for each piece
- ∫_{Iₖ} f ≥ π/√(1 + (k + 1)²π²).
- Since 1 ≤ (k + 1)π, √(1 + (k + 1)²π²) ≤ √2 (k + 1)π ≤ 2(k + 1)π.
- So ∫_{Iₖ} f ≥ 1/(2(k + 1)).
Step 4: Divergence
- ∫₀^{Nπ} f = Σ_{k=0}^{N-1} ∫_{Iₖ} f ≥ (1/2)Σ_{k=1}^{N} 1/k → ∞ as N → ∞.
- The partial integrals are unbounded, so ∫₀^∞ dx/(1 + x² sin²x) diverges (to +∞).
Conclusion
The integral diverges, since each period contributes about 1/k and the partial integrals grow roughly like ln X. Numerical check: the integral from 0 to 10, 100 and 1000 is about 2.97, 5.18 and 7.49; the last tenfold increase of X adds about 2.31, close to ln 10 ≈ 2.30, which is logarithmic growth with no finite limit.
Facts to remember
- For a non-negative integrand, ∫₀^∞ f converges iff ∫₀^X f is bounded in X.
- ∫₀^π dt/(1 + c sin²t) = π/√(1 + c) for c > -1.
- The harmonic series Σ 1/k diverges.
- Comparison test for improper integrals: 0 ≤ g ≤ f and ∫ g = ∞ imply ∫ f = ∞.
- An integrand need not tend to 0 for convergence, nor does tending to 0 between peaks guarantee it.