Partial fractions and integrating rational functions
Ten rational functions split into partial fractions and integrated: distinct linear factors by the cover-up rule, three factors, a repeated factor, an irreducible quadratic, an improper fraction that needs long division first, and the definite integral of 1/(x² − 1), with worked solutions and the mistakes that choose the wrong form or skip the division.
Before you start
Substitution and parts do not integrate ,x2−11, but splitting it into 21(x−11−x+11) turns it into two logs. Partial fractions is that split, done systematically: the form of the answer is fixed by the factors of the denominator, and only the constants are unknown. The work goes wrong in the set-up more than in the arithmetic: a repeated or quadratic factor given too few terms, an improper fraction decomposed without dividing first, a cover-up evaluated at the wrong root. These ten problems decompose first, then integrate.
A rational function is Q(x)P(x) with P and Q polynomials. It is proper when .degP<degQ. If it is not, divide first: QP=S+QR with ,degR<degQ, and decompose only .QR.
Factor Q completely into linear factors and irreducible quadratics (quadratics with no real roots, such as ).x2+1). Each factor contributes terms to the form:
Factor of Q
Terms in the decomposition
,x−a, once
x−aA
(x−a)2
x−aA+(x−a)2B
,x2+px+q, irreducible
x2+px+qBx+C
To find the constants, multiply both sides by Q and either substitute convenient values of x (the roots of Q first) or match coefficients of each power of .x. The cover-up rule is the substitution done in one step: for a factor x−a that appears once, A is QP with the factor x−a covered up, evaluated at .x=a. For the factor ,x+2,.a=−2.
x(x−1)2x2+3=xA+x−1B+(x−1)2CThe repeated factor (x−1)2 needs a term for each power, x−1 and .(x−1)2. Three unknowns for a cubic denominator.
x2+3=A(x−1)2+Bx(x−1)+CxMultiply both sides by .x(x−1)2.
:x=0:.3=A.:x=1:.4=C.Each root kills every term but one.
Coefficient of :x2:,1=A+B, so .B=−2.No root isolates ,B, so match the x2 terms: A(x−1)2 gives ,Ax2,Bx(x−1) gives ,Bx2, and Cx has none.
x3−x−12+(x−1)24Check at :x=2: the left side is 27 and the right side is .23−2+4=27.
Problem 5
Decompose .(x+1)(x2+1)2x2+3x+3.
(x+1)(x2+1)2x2+3x+3=x+1A+x2+1Bx+Cx2+1 has no real roots, so it stays as a factor, and its numerator is a general linear polynomial ,Bx+C, not a constant.
2x2+3x+3=A(x2+1)+(Bx+C)(x+1)Multiply both sides by .(x+1)(x2+1).
:x=−1:,2−3+3=2A, so .A=1.The root of the linear factor.
:x2:,2=A+B, so .B=1. Constant: ,3=A+C, so .C=2.Match coefficients. .(Bx+C)(x+1)=Bx2+(B+C)x+C.
x+11+x2+1x+2Check the x coefficient, which was not used: ,B+C=3, as required.
Problem 6
Write x2−1x3+x2+x+3 as a polynomial plus partial fractions.
The fraction is improper: degree 3 over degree 2.Divide before decomposing.
x3+x2+x+3=(x+1)(x2−1)+(2x+4)Long division: x⋅(x2−1)=x3−x leaves ;x2+2x+3;1⋅(x2−1) leaves ,2x+4, of degree less than 2. Quotient ,x+1, remainder .2x+4.
x2−1x3+x2+x+3=x+1+(x−1)(x+1)2x+4Step 2 divided by .x2−1.
(x−1)(x+1)2x+4=x−13−x+11Cover-up: at ,x=1,;26=3; at ,x=−1,.−22=−1.
x+1+x−13−x+11Check at :x=2: the left side is 317 and the right side is .3+3−31=317.
Problem 7
Find .∫(x+1)(x−2)5x−1dx.
∫(x+12+x−23)dxThe decomposition from Problem 1.
∫x+12dx=2ln∣x+1∣ and .∫x−23dx=3ln∣x−2∣.,∫x−adx=ln∣x−a∣, with the constant carried along. The absolute value keeps the log defined on both sides of the root.
2ln∣x+1∣+3ln∣x−2∣+CCheck by differentiating: ,x+12+x−23, which is the integrand.
Problem 8
Evaluate ∫23x2−1dx exactly.
(x−1)(x+1)1=x−11/2−x+11/2Cover-up: at ,x=1,;21; at ,x=−1,.−21.
∫x2−1dx=21(ln∣x−1∣−ln∣x+1∣)=21lnx+1x−1Integrate each term, then combine with .lnp−lnq=lnqp.
At :x=3:.21ln42=21ln21. At :x=2:.21ln31.Neither root of x2−1 lies in ,[2,3], so the integrand is continuous there and the antiderivative can be evaluated at the ends.
21ln21−21ln31=21ln1/31/2Upper value minus lower value, then the same log rule.
21ln23≈0.2027Sanity check: the integrand falls from 31 to 81 on an interval of length 1, so the integral lies between 0.125 and .0.333.
Problem 9
Find .∫x(x−1)2x2+3dx.
∫(x3−x−12+(x−1)24)dxThe decomposition from Problem 4.
∫x3dx=3ln∣x∣ and .∫x−1−2dx=−2ln∣x−1∣.The two linear terms give logs.
∫(x−1)24dx=4∫(x−1)−2dx=−x−14The squared term is a power, not a log: ∫u−2du=−u−1 with .u=x−1.
3ln∣x∣−2ln∣x−1∣−x−14+CCheck the last term: .dxd(−4(x−1)−1)=4(x−1)−2.
Problem 10
Find .∫(x+1)(x2+1)2x2+3x+3dx.
∫(x+11+x2+1x+x2+12)dxThe decomposition from Problem 5, with x2+1x+2 split into its x part and its constant part, which integrate differently.
∫x2+1xdx=21ln(x2+1)Substitution ,u=x2+1,:du=2xdx: the numerator is half of .du. No absolute value is needed because .x2+1>0.
∫x2+12dx=2arctanxThe constant part gives the arctangent.
ln∣x+1∣+21ln(x2+1)+2arctanx+CSteps 2 and 3 plus .∫x+1dx=ln∣x+1∣.
Where this goes wrong
1. Decomposing an improper fraction without dividing
The cover-up rule produces numbers for any fraction, proper or not, so nothing seems to go wrong.
(x−1)(x+1)x3+x2+x+3Right so far: Problem 6 with the denominator factored.
“Two linear factors, so ;x−1A+x+1B; cover up to get A=26=3 and .B=−22=−1.”The habit that causes the mistake: going straight to the form without comparing degrees.
x2−1x3+x2+x+3=x−13−x+11The right side goes to 0 as x→∞ while the left side grows like ,x, so they cannot be equal; at x=2 they are 317 and .38. A proper-fraction form can only match a proper fraction. Divide first; the missing piece is the quotient x+1 (Problem 6).
2. One term for a repeated factor
(x−1)2 looks like a single factor, so it gets a single fraction.
x(x−1)2x2+3Right so far: Problem 4.
“One fraction per factor: x and .(x−1)2.”The habit that causes the mistake: counting factors, not powers.
x(x−1)2x2+3=xA+(x−1)2BMultiplying out gives .x2+3=A(x−1)2+Bx. The roots give A=3 and ,B=4, but then the x2 coefficients say :1=A=3: no constants work. A repeated factor needs one term for each power, x−1B+(x−1)2C (Problem 4).
3. A constant over an irreducible quadratic
Every linear factor gets a constant on top, and that pattern gets carried to .x2+1.
(x+1)(x2+1)2x2+3x+3Right so far: Problem 5.
“Put a constant over each factor.”The habit that causes the mistake: the numerator over a quadratic needs one degree less than the quadratic, which is degree 1, not 0.
(x+1)(x2+1)2x2+3x+3=x+1A+x2+1CMatching coefficients of 2x2+3x+3=A(x2+1)+C(x+1) gives A=2 from ,x2,C=3 from ,x, and then 3=A+C=5 from the constants: no solution. With x2+1Bx+C there are three unknowns for three equations (Problem 5).
4. Evaluating the cover-up at the wrong root
The factor is ,x+2, and 2 is the number in it.
,(x−3)(x+2)x+7=x−3A+x+2B, with A=2Right so far: Problem 2, steps 2 and 3.
“Cover up x+2 and put in .x=2.”The habit that causes the mistake: reading the number off the factor without its sign.
B=2−32+7=−9The cover-up works because the covered factor is 0 at the chosen ,x, and x+2 is 0 at ,x=−2, not 2. At :x=−2:.B=−55=−1. A quick check catches it: ,2(x+2)−9(x−3)=−7x+31, not .x+7.
5. Integrating 1/(x − 1)² as a log
Every other term in the decomposition integrates to a log, so this one does too.
∫(x3−x−12+(x−1)24)dxRight so far: Problem 9, step 1.
“∫udu=ln∣u∣ with .u=(x−1)2.”The habit that causes the mistake: ∫udu needs du on top, and du=2(x−1)dx is not there.
∫(x−1)24dx=4ln(x−1)2Differentiating the claimed answer gives ,x−18, not .(x−1)24. The term is the power ,4(x−1)−2, whose integral is −x−14 (Problem 9).
Print this set: partial-fractions.pdf (problems, answers, and worked solutions on separate pages).