Calculus/Further integration techniques

From testwiki
Jump to navigation Jump to search

We have already considered two major techniques we can use to solve integration problems, that of the method of substitution, and integration by parts.

Here we make use of these techniques to help us integrate complicated integrals.


Partial fractions

First, an example. Suppose we want to find 3x+1x2+xdx. One way to do this is to simplify the integrand by finding constants A and B so that

3x+1x2+x=3x+1x(x+1)=Ax+Bx+1.

This can be done by cross multiplying the fraction which gives 3x+1x(x+1)=A(x+1)+Bxx(x+1). As both sides have the same denominator we must have 3x+1=A(x+1)+Bx. This is an equation for x so must hold whatever value x is. If we put in x=0 we get 1=A and putting x=1 gives 2=B so B=2. So we see that

3x+1x2+x=1x+2x+1 

Returning to the original integral

3x+1x2+xdx = dxx+2x+1dx
= ln|x|+2ln|x+1|+C

Rewriting the integrand as a sum of simpler fractions has allowed us to reduce the initial integral to a sum of simpler integrals. In fact this method works to integrate any rational function.

Method of Partial Fractions:

  • Step 1 Use long division to ensure that the degree of P(x) less than the degree of Q(x).
  • Step 2 Factor Q(x) as far as possible.
  • Step 3 Write down the correct form for the partial fraction decomposition (see below) and cross multiply to find the constants.

To factor Q(x) we have to write it as a product of linear factors (of the form ax+b) and irreducible quadratic factors (of the form ax2+bx+c with b24ac<0).

Some of the factors could be repeated. For instance if Q(x)=x36x2+9x we factor Q(x) as

Q(x)=x(x26x+9)=x(x3)(x3)=x(x3)2.

It is important that in each quadratic factor we have b24ac<0, otherwise it is possible to factor that quadratic piece further. For example if Q(x)=x33x22x then we can write

Q(x)=x(x23x+2)=x(x1)(x+2)


We will now show how to write P(x)/Q(x) as a sum of terms of the form

A(ax+b)k and Ax+B(ax2+bx+c)k.

Exactly how to do this depends on the factorization of Q(x) and we now give four cases that can occur.

Case (a) Q(x) is a product of linear factors with no repeats.

This means that Q(x)=(a1x+b1)(a2x+b2)...(anx+bn) where no factor is repeated and no factor is a multiple of another.

For each linear term we write down something of the form A(ax+b), so in total we write

P(x)Q(x)=A1(a1x+b1)+A2(a2x+b2)++An(anx+bn)

Example 1

Find 1+x2(x+3)(x+5)(x+7)dx

Here we have P(x)=1+x2,Q(x)=(x+3)(x+5)(x+7) and Q(x) is a product of linear factors. So we write

1+x2(x+3)(x+5)(x+7)=Ax+3+Bx+5+Cx+7

Multiply both sides by the denominator

1+x2=A(x+5)(x+7)+B(x+3)(x+7)+C(x+3)(x+5)

Substitute in three values of x to get three equations for the unknown constants,

x=31+32=24Ax=51+52=22Bx=71+72=(4)(2)C

so A=5/4,B=13/2,C=25/4, and

1+x2(x+3)(x+5)(x+7)=54x+12132x+10+254x+28

We can now integrate the left hand side.

1+x2dx(x+3)(x+5)(x+7)=54ln|x+3|132ln|x+5|+254ln|x+7|+C


Case (b) Q(x) is a product of linear factors some of which are repeated.

If (ax+b) appears in the factorisation of Q(x) k-times. Then instead of writing the piece A(ax+b) we use the more complicated expression

A1ax+b+A2(ax+b)2+A3(ax+b)3++Ak(ax+b)k

Example 2

Find 1(x+1)(x+2)2dx

Here P(x)=1" and "Q(x)=(x+1)(x+2)2 We write

1(x+1)(x+2)2=Ax+1+Bx+2+C(x+2)2

Multiply both sides by the denominator 1=A(x+2)2+B(x+1)(x+2)+C(x+1)

Substitute in three values of x to get 3 equations for the unknown constants,

x=01=22A+2B+Cx=11=Ax=21=C

so A=1, B=-1, C=-1, and

1(x+1)(x+2)2=1x+11x+21(x+2)2

We can now integrate the left hand side. 1(x+1)(x+2)2dx=lnx+1x+2+1x+2+C


Case (c) Q(x) contains some quadratic pieces which are not repeated.

If (ax2+bx+c) appears we use Ax+B(ax2+bx+c).

Case (d) Q(x) contains some repeated quadratic factors.

If (ax2+bx+c) appears k-times then use

A1x+B1(ax2+bx+c)+A2x+B2(ax2+bx+c)2+A3x+B3(ax2+bx+c)3++Akx+Bk(ax2+bx+c)k

Trigonometric Substitution

If the integrand contains a single factor of one of the forms a2x2 or a2+x2 or x2a2 we can try a trigonometric substitution.

  • If the integrand contains a2x2 let x=asinθ and use the identity 1sin2θ=cos2θ.
  • If the integrand contains a2+x2 let x=atanθ and use the identity 1+tan2θ=sec2θ.
  • If the integrand contains x2a2 let x=asecθ and use the identity sec2θ1=tan2θ.

Sine substitution

This substitution is easily derived from a triangle, using the Pythagorean Theorem.

If the integrand contains a piece of the form a2x2 we use the substitution

x=asinθdx=acosθdθ

This will transform the integrand to a trigonometic function. If the new integrand can't be integrated on sight then the tan-half-angle substitution described below will generally transform it into a more tractable algebraic integrand.

Eg, if the integrand is √(1-x2),

011x2dx=0π/21sin2θcosθdθ=0π/2cos2θdθ=120π/21+cos2θdθ=π4

If the integrand is √(1+x)/√(1-x), we can rewrite it as

1+x1x=1+x1+x1+x1x=1+x1x2

Then we can make the substitution

0a1+x1x2dx=0α1+sinθcosθcosθdθ0<a<1=0α1+sinθdθα=sin1a=α+[cosθ]0α=α+1cosα=1+sin1a1a2

Tangent substitution

This substitution is easily derived from a triangle, using the Pythagorean Theorem.

When the integrand contains a piece of the form a2+x2 we use the substitution

x=atanθx2+a2=asecθdx=asec2θdθ

E.g, if the integrand is (x2+a2)-3/2 then on making this substitution we find

0z(x2+a2)32dx=a20αcosθdθz>0=a2[sinθ]0αα=tan1(z/a)=a2sinα=a2z/a1+z2/a2=1a2za2+z2

If the integral is

I=0zx2+a2z>0

then on making this substitution we find

I=a20αsec3θdθα=tan1(z/a)=a20αsecθdtanθ=a2[secθtanθ]0αa20αsecθtan2θdθ=a2secαtanαa20αsec3θdθ+a20αsecθdθ=a2secαtanαI+a20αsecθdθ

After integrating by parts, and using trigonometric identities, we've ended up with an expression involving the original integral. In cases like this we must now rearrange the equation so that the original integral is on one side only

I=12a2secαtanα+12a20αsecθdθ=12a2secαtanα+12a2[ln(secθ+tanθ)]0α=12a2secαtanα+12a2ln(secα+tanα)=12a2(1+z2a2)za+12a2ln(1+z2a2+za)=12zz2+a2+12a2ln(za+1+z2a2)

As we would expect from the integrand, this is approximately z2/2 for large z.

Secant substitution

This substitution is easily derived from a triangle, using the Pythagorean Theorem.

If the integrand contains a factor of the form x2a2 we use the substitution

x=asecθdx=asecθtanθdθx2a2=tanθ.

Example 1

Find 1zx21xdx.

1zx21xdx=1αtanθsecθsecθtanθdθz>1=0αtan2θdθα=sec1z=[tanθθ]0αtanα=sec2α1=tanααtanα=z21=z21sec1z

Example 2

Find 1zx21x2dx.

1zx21x2dx=1αtanθsec2θsecθtanθdθz>1=0αsin2θcosθdθα=sec1z

We can now integrate by parts

1zx21x2dx=[tanθcosθ]0α+0αsecθdθ=sinα+[ln(secθ+tanθ)]0α=ln(secα+tanα)sinα=ln(z+z21)z21z

Trigonometric integrals

Powers of Sine and Cosine

We will give a general method to solve generally integrands of the form cosm (x)sinn(x). First let us work through an example.

(cos3x)(sin2x)dx

Notice that the integrand contains an odd power of cos. So rewrite it as

(cos2x)(sin2x)cosxdx

We can solve this by making the substitution u = sin(x) so du = cos(x) dx. Then we can write the whole integrand in terms of u by using the identity

cos(x)2 = 1 - sin2(x)=1-u2.

So

(cos3x)(sin2x)dx=(cos2x)(sin2x)cosxdx=(1u2)u2du=u2duu4du=13u3+15u5+C=13sin3x15sin5x+C.

This method works whenever there is an odd power of sine or cosine.

To evaluate (cosmx)(sinnx)dx when either m or n is odd.

  • If m is odd substitute u=sin x and use the identity cos2x = 1 - sin2x=1-u2.
  • If n is odd substitute u=cos x and use the identity sin2x = 1 - cos2x=1-u2.

Example

Find 0π/2cos40(x)sin3(x)dx.

As there is an odd power of sin we let u=cosx so du = - sin(x)dx. Notice that when x=0 we have u=cos(0)=1 and when x=π/2 we have u=cos(π/2)=0.

0π/2cos40(x)sin3(x)dx=0π/2cos40(x)sin2(x)sin(x)dx=10u40(1u2)du=01u40(1u2)du=01u40u42du=[141u41143u43]01=141143.



When both m and n are even things get a little more complicated.

To evaluate (cosmx)(sinnx)dx when both m and n are even.


Use the identities sin2x = 1/2 (1- cos 2x) and cos2x = 1/2 (1+ cos 2x).

Example

Find sin2xcos4xdx.


As sin2x = 1/2 (1- cos 2x) and cos2x = 1/2 (1+ cos 2x) we have

sin2xcos4xdx=(12(1cos2x))(12(1+cos2x))2dx,

and expanding, the integrand becomes

18(1cos22x+cos2xcos32xdx).

Using the multiple angle identities

I=18(1dxcos22xdx+cos2xdxcos32xdx)=18(x12(1+cos4x)dx+12sin2xcos22xcos2xdx)=116(x+sin2x+cos4xdx2(1sin22x)cos2xdx)

then we obtain on evaluating

I=x16sin4x64+sin32x48+C

Powers of Tan and Secant

To evaluate (tanmx)(secnx)dx.

  1. If n is even and n2 then substitute u=tan x and use the identity sec2x = 1 + tan2x.
  2. If n and m are both odd then substitute u=sec x and use the identity tan2x = sec2x-1.
  3. If n is odd and m is even then use the identity tan2x = sec2x-1 and apply a reduction formula to integrate secjxdx.

Example 1

Find sec2xdx.

There is an even power of secx. Substituting u=tanx gives du=sec2xdx so

sec2xdx=du=u+C=tanx+C.


Example 2

Find tanxdx.

Let u=cosx so du=sinxdx. Then

tanxdx=sinxcosxdx=1udu=ln|u|+C=ln|cosx|+C=ln|secx|+C.


Example 3

Find secxdx.

The trick to do this is to multiply and divide by the same thing like this:

secxdx=secxsecx+tanxsecx+tanxdx=sec2x+secxtanxsecx+tanx.

Making the substitution u=secx+tanx so du=secxtanx+sec2xdx,

secxdx=1udu=ln|u|+C=ln|secx+tanx|+C.

More trigonometric combinations

For the integrals sinnxcosmxdx or sinnxsinmxdx or cosnxcosmxdx use the identities

  • sinacosb=12(sin(a+b)+sin(ab))
  • sinasinb=12(cos(ab)cos(a+b))
  • cosacosb=12(cos(ab)+cos(a+b))

Example 1

Find sin3xcos5xdx.

We can use the fact that sin a cos b=(1/2)(sin(a+b)+sin(a-b)), so

sin3xcos5x=(sin8x+sin(2x))/2

Now use the oddness property of sin(x) to simplify

sin3xcos5x=(sin8xsin2x)/2

And now we can integrate

sin3xcos5xdx=12sin8xsin2xdx=12(18cos8x+12cos2x)+C

Example 2

Find:sinxsin2xdx.

Using the identities

sinxsin2x=12(cos(x)cos(3x))=12(cosxcos3x).

Then

sinxsin2xdx=12(cosxcos3x)dx=12(sinx13sin3x)+C

Reduction formula

A reduction formula is one that enables us to solve an integral problem by reducing it to a problem of solving an easier integral problem, and then reducing that to the problem of solving an easier problem, and so on.

For example, if we let

In=xnexdx

Integration by parts allows us to simplify this to

In=xnexnxn1exdx=
In=xnexnIn1

which is our desired reduction formula. Note that we stop at

I0=ex.

Similarly, if we let

In=0αsecnθdθ

then integration by parts lets us simplify this to

In=secn2αtanα(n2)0αsecn2θtan2θdθ

Using the trigonometric identity, tan2=sec2-1, we can now write

In=secn2αtanα+(n2)(0αsecn2θdθ0αsecnθdθ)=secn2αtanα+(n2)(In2In)

Rearranging, we get

In=1n1secn2αtanα+n2n1In2

Note that we stop at n=1 or 2 if n is odd or even respectively.

As in these two examples, integrating by parts when the integrand contains a power often results in a reduction formula.

Rational functions in Sine and Cosine

The "tan half angle" substitution

Another useful change of variables is

t=tan(x/2)

With this transformation, using the double-angle trig identities,

sinx=2t1+t2cosx=1t21+t2tanx=2t1t2dx=2dt1+t2

This transforms a trigonometric integral into a algebraic integral, which may be easier to integrate.

For example, if the integrand is 1/(1 + sin x ) then

0π/2dx1+sinx=012dt(1+t)2=[21+t]01=1

This method can be used to further simplify trigonometric integrals produced by the changes of variables described earlier.

For example, if we are considering the integral

I=111x21+x2dx

we can first use the substition x= sin θ, which gives

I=π/2π/2cos2θ1+sin2θdθ

then use the tan-half-angle substition to obtain

I=11(1t2)21+6t2+t42dt1+t2

In effect, we've removed the square root from the original integrand. We could do this with a single change of variables, but doing it in two steps gives us the opportunity of doing the trigonometric integral another way.

Having done this, we can split the new integrand into partial fractions, and integrate.

I=1122t2+38dt+112+2t2+3+8dt1121+t2dt=4838tan1(3+8)+4+83+8tan1(38)π

This result can be further simplified by use of the identities

3±8=(2±1)2tan(2±1)=(14±18)π

ultimately leading to

I=(21)π

In principle, this approach will work with any integrand which is the square root of a quadratic multiplied by the ratio of two polynomials. However, it should not be applied automatically.

E.g, in this last example, once we deduced

I=π/2π/2cos2θ1+sin2θdθ

we could have used the double angle formulae, since this contains only even powers of cos and sin. Doing that gives

I=π/2π/21+cos2θ3cos2θdθ=12ππ1+cosϕ3cosϕdϕ

Using tan-half-angle on this new, simpler, integrand gives

I=11+2t2dt1+t2=2dt1+2t2dt1+t2

This can be integrated on sight to give

I=42π22π2=(21)π

This is the same result as before, but obtained with less algebra, which shows why it is best to look for the most straightforward methods at every stage.

A more direct way of evaluating the integral I is to substitute t = tan θ right from the start, which will directly bring us to the line

I=11+2t2dt1+t2

above. More generally, the substitution t = tan x gives us

sinx=t1+t2cosx=11+t2dx=dt1+t2

so this substitution is the preferable one to use if the integrand is such that all the square roots would disappear after substitution, as is the case in the above integral.

Alternate Method

In general, to evaluate integrals of the form

A+Bcosx+Csinxa+bcosx+csinxdx,

it is extremely tedious to use the aforementioned "tan half angle" substitution directly, as one easily ends up with a rational function with a 4th degree denominator. Instead, we may first write the numerator as

A+Bcosx+Csinxp(a+bcosx+csinx)+qddx(a+bcosx+csinx)+r.

Then the integral can be written as

(p+qddx(a+bcosx+csinx)a+bcosx+csinx+ra+bcosx+csinx)dx

which can be evaluated much more easily.

Example

Evaluate cosx+2cosx+sinxdx.

Let

cosx+2p(cosx+sinx)+qddx(cosx+sinx)+r.

Then

cosx+2p(cosx+sinx)+q(sinx+cosx)+r
cosx+2(p+q)cosx+(pq)sinx+r.

Comparing coefficients of cos x, sin x and the constants on both sides, we obtain

{p+q=1pq=0r=2

yielding p = q = 1/2, r = 2. Substituting back into the integrand,

cosx+2cosx+sinxdx=12dx+12d(cosx+sinx)cosx+sinx+2cosx+sinxdx.

The last integral can now be evaluated using the "tan half angle" substitution described above, and we obtain

2cosx+sinxdx=2ln|tanx21+2tanx212|+C.

The original integral is thus

cosx+2cosx+sinxdx=x2+12ln|cosx+sinx|+2ln|tanx21+2tanx212|+C.

Irrational functions

Integration of irrational functions is more difficult than rational functions, and many cannot be done. However, there are some particular types that can be reduced to rational forms by suitable substitutions.

Type 1

Integrand contains ax+bcx+dn

Use the substitution u=ax+bcx+dn.

Example

Find 1x1xxdx.

Type 2

Integral is of the form Px+Qax2+bx+cdx

Write Px+Q as Px+Q=pddx(ax2+bx+c)+q.

Example

Find 4x154xx2dx.

Type 3

Integrand contains a2x2, a2+x2 or x2a2

This was discussed in "trigonometric substitutions above". Here is a summary:

  1. For a2x2, use x=asinθ.
  2. For a2+x2, use x=atanθ.
  3. For x2a2, use x=asecθ.

Type 4

Integral is of the form 1(px+q)ax2+bx+cdx

Use the substitution u=1px+q.

Example

Find 1(1+x)3+6x+x2dx.

Type 5

Other rational expressions with the irrational function ax2+bx+c

  1. If a>0, we can use u=ax2+bx+c±ax.
  2. If c>0, we can use u=ax2+bx+c±cx.
  3. If ax2+bx+c can be factorised as a(xα)(xβ), we can use u=a(xα)xβ.
  4. If a<0 and ax2+bx+c can be factorised as a(αx)(xβ), we can use x=αcos2θ+βsin2θ,/theta+β.

Template:Calculus:TOC