First Year Engineering is coming soon

We're still writing and reviewing the notes for every subject, and we'll open it when every subject is ready. Meanwhile you can practise aptitude, ask Aria about any subject, or see the branches that are ready.

Reduction formulae

Reduction formulae: systematic recurrence relations for families of integrals (powers of trig, products with exponentials/polynomials/logs) that let you reduce an integral with a large exponent/power to one with a smaller exponent until it is elementary.

Drafted with Aria, reviewed by the AiCanCode.org team · Updated . How we write and check our notes. Spotted an error? Use Give Feedback at the bottom of the page.

Why it matters

Engineers often integrate families of similar functions repeatedly: powers of sine/cosine in signal processing, products like x^n e^{ax} in control and communications, or x^m (ln x)^n in data transforms. Reduction formulae give a mechanical, step-by-step way to simplify such integrals, crucial for hand-computation, derivations, and for verifying computer algebra outputs.

Key ideas

  • Trig identities used: sin^2 x + cos^2 x = 1; sin^2 x = (1 − cos 2x)/2; cos^2 x = (1 + cos 2x)/2; tan^2 x + 1 = sec^2 x; sec^2 x − 1 = tan^2 x.

  • Standard reduction patterns (with outlines of derivation steps)

    1. Powers of sine:

      • Define I_n = ∫ sin^n x dx, n ≥ 1.
      • For n ≥ 2, write sin^n x = sin^{n−1} x · sin x.
      • Substitute u = cos x so du = −sin x dx, or more commonly: use integration by parts with u = sin^{n−1} x, dv = sin x dx.
      • Derivation sketch:
        1. Let u = sin^{n−1} x ⇒ du = (n−1) sin^{n−2} x cos x dx; dv = sin x dx ⇒ v = −cos x.
        2. I_n = ∫ u dv = −sin^{n−1} x cos x + (n−1) ∫ sin^{n−2} x cos^2 x dx.
        3. Replace cos^2 x by 1 − sin^2 x and split integrals to get I_n in terms of I_{n−2} and a boundary term −sin^{n−1} x cos x.
      • Resulting reduction formula (see next section) relates I_n and I_{n−2}.
    2. Powers of cosine:

      • Analogous steps starting with J_n = ∫ cos^n x dx, using u = cos^{n−1} x, dv = cos x dx, and identity sin^2 x = 1 − cos^2 x.
    3. Even/odd powers over symmetric limits:

      • For definite integrals on [0, π/2], boundary terms from (1) and (2) vanish because sin 0 = 0, cos(π/2) = 0. This yields clean recurrences for ∫_0^{π/2} sin^n x dx and ∫_0^{π/2} cos^n x dx.
    4. Powers of secant:

      • K_n = ∫ sec^n x dx, n ≥ 2.
      • For even n, peel off sec^2 x and use tan substitution; for general n ≥ 2, an integration-by-parts reduction with u = sec^{n−2} x, dv = sec^2 x dx is standard.
      • Use identities d/dx(sec x) = sec x tan x, d/dx(tan x) = sec^2 x, and sec^2 x = 1 + tan^2 x to reduce the power by 2.
    5. Powers of tangent:

      • T_n = ∫ tan^n x dx, n ≥ 2.
      • Write tan^n x = tan^{n−2} x (sec^2 x − 1) to get T_n = ∫ tan^{n−2} x sec^2 x dx − T_{n−2}.
      • Substitute u = tan x to handle the first term and reduce the power by 2.
    6. Polynomial times exponential:

      • I_n(a) = ∫ x^n e^{a x} dx, n ≥ 0, a ≠ 0 typically for compact expressions.
      • Use integration by parts: choose u = x^n (reduces degree), dv = e^{a x} dx (easy to integrate). This yields a simple recurrence reducing n by 1. Repeat until n = 0.
    7. Polynomial times sine/cosine:

      • S_n(a) = ∫ x^n sin(a x) dx, C_n(a) = ∫ x^n cos(a x) dx, n ≥ 0.
      • Use integration by parts with u = x^n, dv = sin(ax) dx or cos(ax) dx. Each step reduces the power of x by 1 and alternates between sine and cosine terms; results contain 1/a factors.
    8. Powers of x times logarithm powers:

      • L_{m,n} = ∫ x^m (ln x)^n dx, m > −1 for convergence on (0, b], n ≥ 1 integer.
      • Use integration by parts: u = (ln x)^n, dv = x^m dx. Each step lowers the log power n → n−1 and introduces factors 1/(m+1).
  • Diagram note: This topic is algebraic; no diagram is usually required. If an examiner asks for a “flow” of reduction, draw a simple box-arrow chain labeled I_n → I_{n−2} → … → I_0 or I_1, and mark the identities used (e.g., cos^2 x = 1 − sin^2 x).

Formulas

I_n = ∫ sin^n x dx = −(sin^{n−1} x · cos x)/n + ((n−1)/n) ∫ sin^{n−2} x dx, where n ≥ 2; x in radians (dimensionless); applies for indefinite integral over any interval where sine is defined.

J_n = ∫ cos^n x dx = (cos^{n−1} x · sin x)/n + ((n−1)/n) ∫ cos^{n−2} x dx, where n ≥ 2; x in radians; applies for indefinite integral.

∫_0^{π/2} sin^n x dx = ((n−1)/n) ∫_0^{π/2} sin^{n−2} x dx, n ≥ 2; angle in radians; boundary terms vanish at 0 and π/2.

∫_0^{π/2} cos^n x dx = ((n−1)/n) ∫_0^{π/2} cos^{n−2} x dx, n ≥ 2; angle in radians; boundary terms vanish.

K_n = ∫ sec^n x dx = (sec^{n−2} x · tan x)/(n−1) + ((n−2)/(n−1)) ∫ sec^{n−2} x dx, n ≥ 2; x in radians; holds where sec x and tan x are defined.

T_n = ∫ tan^n x dx = (tan^{n−1} x)/(n−1) − ∫ tan^{n−2} x dx, n ≥ 2; x in radians; holds where tan x is defined.

I_n(a) = ∫ x^n e^{a x} dx = (x^n e^{a x})/a − (n/a) ∫ x^{n−1} e^{a x} dx, n ≥ 1; a real nonzero constant; x has units inverse to a so that ax is dimensionless.

S_n(a) = ∫ x^n sin(a x) dx = −(x^n cos(ax))/a + (n/a) ∫ x^{n−1} cos(a x) dx, n ≥ 1; a ≠ 0; x in units so that ax is dimensionless (radians).

C_n(a) = ∫ x^n cos(a x) dx = (x^n sin(ax))/a − (n/a) ∫ x^{n−1} sin(a x) dx, n ≥ 1; a ≠ 0; same unit note as above.

L_{m,n} = ∫ x^m (ln x)^n dx = (x^{m+1} (ln x)^n)/(m+1) − (n/(m+1)) ∫ x^m (ln x)^{n−1} dx, m > −1, n ≥ 1; x > 0 (for ln x); dimensionless ln argument.

Worked examples

  1. Derive the reduction formula for ∫ sin^n x dx and evaluate ∫_0^{π/2} sin^5 x dx.

    • Derivation (indefinite):
      1. Let I_n = ∫ sin^n x dx, n ≥ 2. Choose u = sin^{n−1} x, dv = sin x dx.
      2. Then du = (n−1) sin^{n−2} x cos x dx, v = −cos x.
      3. I_n = ∫ u dv = −sin^{n−1} x cos x + (n−1) ∫ sin^{n−2} x cos^2 x dx.
      4. Replace cos^2 x = 1 − sin^2 x: I_n = −sin^{n−1} x cos x + (n−1) ∫ sin^{n−2} x dx − (n−1) ∫ sin^n x dx.
      5. Bring the last term to the left: I_n + (n−1) I_n = −sin^{n−1} x cos x + (n−1) I_{n−2}.
      6. So n I_n = −sin^{n−1} x cos x + (n−1) I_{n−2} ⇒ I_n = −(sin^{n−1} x cos x)/n + ((n−1)/n) I_{n−2}.
    • Definite evaluation: ∫_0^{π/2} sin^5 x dx.
      1. Apply the definite reduction without boundary term: For n ≥ 2, ∫_0^{π/2} sin^n x dx = (n−1)/n ∫_0^{π/2} sin^{n−2} x dx.
      2. Step down: For n = 5: ∫_0^{π/2} sin^5 x dx = (4/5) ∫_0^{π/2} sin^3 x dx. For n = 3: ∫_0^{π/2} sin^3 x dx = (2/3) ∫_0^{π/2} sin x dx.
      3. Compute base integral: ∫_0^{π/2} sin x dx = [−cos x]_0^{π/2} = (−0) − (−1) = 1.
      4. Back-substitute: ∫_0^{π/2} sin^3 x dx = (2/3)·1 = 2/3. Then ∫_0^{π/2} sin^5 x dx = (4/5)·(2/3) = 8/15.
      5. Final answer: ∫_0^{π/2} sin^5 x dx = 8/15 (dimensionless).
      • Check: 8/15 ≈ 0.5333; consistent with sin^5 averaging below 1 over [0, π/2].
  2. Use the reduction formula for ∫ sec^n x dx to evaluate ∫ sec^6 x dx.

    • Reduction formula recall: For n ≥ 2, ∫ sec^n x dx = (sec^{n−2} x tan x)/(n−1) + ((n−2)/(n−1)) ∫ sec^{n−2} x dx.
    • Steps:
      1. For n = 6: ∫ sec^6 x dx = (sec^4 x tan x)/5 + (4/5) ∫ sec^4 x dx.
      2. Apply again with n = 4: ∫ sec^4 x dx = (sec^2 x tan x)/3 + (2/3) ∫ sec^2 x dx.
      3. And ∫ sec^2 x dx = tan x + C.
      4. Assemble: ∫ sec^4 x dx = (sec^2 x tan x)/3 + (2/3) tan x + C.
      5. Substitute back: ∫ sec^6 x dx = (sec^4 x tan x)/5 + (4/5)[(sec^2 x tan x)/3 + (2/3) tan x] + C.
      6. Simplify coefficients: (4/5)·(1/3) = 4/15; (4/5)·(2/3) = 8/15.
      7. Final antiderivative: ∫ sec^6 x dx = (sec^4 x tan x)/5 + (4/15) sec^2 x tan x + (8/15) tan x + C.
      • Optional check: Differentiate term-by-term using product rule and identities to recover sec^6 x.
  3. Find ∫ x^4 e^{2x} dx using reduction.

    • Use I_n(a) = ∫ x^n e^{a x} dx = (x^n e^{a x})/a − (n/a) ∫ x^{n−1} e^{a x} dx.
    • Here, n = 4, a = 2. Proceed step-by-step, carrying units notionally so that ax is dimensionless.
      1. I_4(2) = (x^4 e^{2x})/2 − (4/2) I_3(2) = (x^4 e^{2x})/2 − 2 I_3(2).
      2. I_3(2) = (x^3 e^{2x})/2 − (3/2) I_2(2).
      3. I_2(2) = (x^2 e^{2x})/2 − (2/2) I_1(2) = (x^2 e^{2x})/2 − I_1(2).
      4. I_1(2) = (x e^{2x})/2 − (1/2) I_0(2).
      5. I_0(2) = ∫ e^{2x} dx = (e^{2x})/2.
    • Back-substitute from I_0 upward:
      1. I_1(2) = (x e^{2x})/2 − (1/2)·(e^{2x}/2) = e^{2x}(x/2 − 1/4).
      2. I_2(2) = (x^2 e^{2x})/2 − e^{2x}(x/2 − 1/4) = e^{2x}(x^2/2 − x/2 + 1/4).
      3. I_3(2) = (x^3 e^{2x})/2 − (3/2)·e^{2x}(x^2/2 − x/2 + 1/4) = e^{2x}[x^3/2 − (3/4)x^2 + (3/4)x − 3/8].
      4. I_4(2) = (x^4 e^{2x})/2 − 2·e^{2x}[x^3/2 − (3/4)x^2 + (3/4)x − 3/8] = e^{2x}[x^4/2 − x^3 + (3/2)x^2 − (3/2)x + 3/4] + C.
    • Final answer: ∫ x^4 e^{2x} dx = e^{2x}(x^4/2 − x^3 + 3x^2/2 − 3x/2 + 3/4) + C.
    • Quick derivative check (outline): Differentiate e^{2x}P(x) to get e^{2x}(2P + P′). Verify that 2P + P′ = x^4.
  4. Derive and use the reduction formula for ∫ x^m (ln x)^n dx to compute ∫ x^3 (ln x)^2 dx (x > 0).

    • Reduction derivation:
      1. Let L_{m,n} = ∫ x^m (ln x)^n dx, with m > −1, n ≥ 1.
      2. Choose u = (ln x)^n ⇒ du = n(ln x)^{n−1}·(1/x) dx; dv = x^m dx ⇒ v = x^{m+1}/(m+1).
      3. By parts: L_{m,n} = (x^{m+1}(ln x)^n)/(m+1) − ∫ [x^{m+1}/(m+1)]·n(ln x)^{n−1}·(1/x) dx.
      4. Simplify: L_{m,n} = (x^{m+1}(ln x)^n)/(m+1) − (n/(m+1)) ∫ x^m (ln x)^{n−1} dx.
    • Using it for m = 3, n = 2:
      1. L_{3,2} = ∫ x^3 (ln x)^2 dx = (x^4 (ln x)^2)/4 − (2/4) ∫ x^3 (ln x) dx.
      2. Reduce L_{3,1}: ∫ x^3 (ln x) dx = (x^4 ln x)/4 − (1/4) ∫ x^3 dx = (x^4 ln x)/4 − (1/16) x^4.
      3. Substitute back: L_{3,2} = (x^4 (ln x)^2)/4 − (1/2)[(x^4 ln x)/4 − x^4/16] = (x^4 (ln x)^2)/4 − (x^4 ln x)/8 + x^4/32 + C.
    • Final answer: ∫ x^3 (ln x)^2 dx = x^4[(ln x)^2/4 − (ln x)/8 + 1/32] + C, for x > 0.

Common mistakes

  • Sign errors in integration by parts, especially with sine/cosine alternation and the minus from v = −cos(ax)/a.
  • Misapplying identities: mixing tan^2 x = sec^2 x − 1 vs sec^2 x = 1 + tan^2 x. A wrong identity ruins the recurrence.
  • Stopping one step too early: After reduction, remember to evaluate the final base case (e.g., I_0, I_1, ∫ sec^2 x dx) and substitute back fully.
  • For ∫ x^n e^{ax} dx, choosing u = e^{ax} and dv = x^n dx makes the integral harder; the correct choice is u = x^n.
  • For ∫ x^m (ln x)^n dx, forgetting domain x > 0 (ln x undefined for x ≤ 0 in real analysis).

For your exam

  • Definite integrals: Use the reduced form to compute ∫_0^{π/2} sin^n x dx or cos^n x dx for a given n. Show each reduction step and the base case value (usually 1 for odd powers or π/2 for even powers).
  • Short notes: Explain when boundary terms vanish (e.g., on [0, π/2]) and the even/odd patterns for powers of sine/cosine.
  • Numericals: Evaluate an integral with a medium power (e.g., ∫ sec^6 x dx, ∫ x^5 e^{3x} dx) showing step-by-step reductions and final simplified answer.
  • Presentation: Keep the recurrence boxed (clearly written), label each step, and state domain assumptions (x in radians; x > 0 for logs; a ≠ 0 when needed). A full-marks answer is clean, logically ordered, arrives at a correct reduced form, evaluates the base case correctly, and includes +C for indefinite integrals.

Quick check

  1. Write the reduction formula for I_n = ∫ sin^n x dx in terms of I_{n−2}.
  2. Using reduction, express ∫ tan^6 x dx in terms of ∫ tan^4 x dx.
  3. State the reduction step for S_n(a) = ∫ x^n sin(ax) dx.

Answers: (1) I_n = −(sin^{n−1} x cos x)/n + ((n−1)/n) I_{n−2}. (2) ∫ tan^6 x dx = (tan^5 x)/5 − ∫ tan^4 x dx. (3) S_n(a) = −(x^n cos(ax))/a + (n/a) ∫ x^{n−1} cos(ax) dx; a ≠ 0.

Finished this topic? Mark it so your progress, study plan and readiness keep up.

Stuck on something here?