You are on page 1of 4

Refresher on Integration Techniques

MA 221

To solve separable equations


dx
= g(t)h(x)
dt
We first separate, then integrate. Remember, any non-linear first order differential equation that you are asked to
solve explicitly or implicitly, must be separable.
To do this successfully, we must be able to quickly and accurately integrate. You are expected to be able to
integrate polynomials (power rule), exponentials and basic trigonometric functions (may require u-substitution,
and/or some algebraic clean-up such as recognizing csc t = 1/ sin t).
We also need to be able to use the following techniques of integration:
1. u-substitution
2. Integration by parts
3. Partial fractions
There are some tricks to help us remember which technique of integration is appropriate.
Before you do any integration, first see if you can do any algebraic simplification.
Example 1. The following integral appears quite complicated at first glance
t2 − 1
Z
e5 ln t dt.
t+1
By using properties of logarithms and exponentials, we can greatly simplify the first term:
5
e5 ln t = eln(t )
= t5 , for t > 0

For the second term, we note that by factoring the numerator, we can see
t2 − 1 (t − 1)(t + 1)
= = t − 1, for t 6= −1.
t+1 t+1
So, our complicated integral can be greatly simplified to the power rule:
2
5 ln t t − 1
Z Z Z
1 1
e| {z } dt = t (t − 1) dt = t6 − t5 dt = t7 − t6 + C.
5
t+1 7 6
t5 | {z }
t−1

Integration by Substitution
Integration by substitution is very useful whenever your integral involves a composition of functions.
Example 2. The integral Z
6
t5 et dt

contains the composition f ◦ g(t) = f (g(t)) with f (t) = et and g(t) = t6 . To make this work, we choose u = t6 , so
6
that then du = 6t5 dt. This allows us to replace et with eu , and t5 dt = 61 du.
We can now integrate a much simpler integral in u:
Z Z
5 t6 1 u 1
t e dt = e du = eu + C
|{z} |{z}
u
6 6
1
du e
6

1
To finish this, we need to return to t, substituting back in u = t6 , we see
Z
6 1 1 6
t5 et dt = eu + C = et + C.
6 6
Example 3. The integral Z
e2t cos(e2t ) dt

contains a function composition since e2t is inside the cosine. This suggests we take u = e2t , then du = 2e2t dt.
Z Z
2t 2t 1 1 1
e cos(e ) dt = cos(u) du = sin(u) + C = sin(e2t ) + C.
|{z} | {z } 2 2 2
1
2 du cos(u)

Big Ideas: Whenever your integrand contains a composition of functions, u-substitution is a good idea. Try
picking u to be the “inner function” in the composition. The rest of the integrand will often be a multiple of du.
This is usually the first integration technique to try.

Integration by Parts
Integration by parts is very useful when you have different types of functions (polynomials, exponentials, trigono-
metric functions or logarithms) multiplying each other. We then use the rule
Z Z
u dv = uv − v du
R
to try to make the integral v du simpler than the one we started with. There are two basic rules to follow:
1. Pick u so that du is simpler than u. (Or, at least not more complicated.)
2. You have to pick dv to be something you can integrate.
Example 4. The integral Z
te3t dt

is best done with integration by parts. We can see this since there is no reasonable u-substitution that can simplify
this, and we have a polynomial, t, multiplying an exponential e3t .
We choose u = t since du = 1 dt is simpler. This leaves dv = e3t , which we can integrate to get v = 13 e3t . So, we
have Z Z
3t 1 3t 1 3t 1 1
t e dt = te − e dt = te3t − e3t + C.
|3 {z } | 3 3 9
|{z} |{z}
u dv {z }
uv R
− v du

Example 5. The integral Z


ln(t) dt

doesn’t look like it’s the product of two functions. We need to view ln(t) = 1 · ln(t). We need to choose u = ln(t)
since we know how to find du = 1t dt and we don’t know how to integrate ln(t), yet. This leaves us with dv = 1, so
v = t. Then Z Z Z
1
1 ln(t) dt = t ln(t) − t dt
ln(t) dt = |{z}
|{z} | {z } t
dv u uv | R{z }
− v du
R
We can simplify the last integral to be 1 dt = t, so we have
Z
ln(t) dt = t ln(t) − t + C.

2
Big Ideas: Whenever your integrand contains a product of different kinds of functions, integration by parts is a
good idea. It is a good idea to (quickly) see if a u-substitution would work. For example,
Z
2
tet dt

initially looks like a product of two different types of functions, but the subsitution u = t2 simplifies it to a
much simpler integral. We want to pick u so that du is simpler. The mnemonic ‘LIATE’ is a good guideline. In
order of preference for u is Logarithm, Inverse trig functions, Algebraic (polynomials), Trigonometric functions,
Exponentials.

Partial Fractions Decomposition


Partial fractions helps us to integrate rational functions. Remember a rational function is of the form
P (t)
R(t) =
Q(t)
where P (t) and Q(t) are polynomials.
The technique of partial fraction decomposition allows us to break up a complicated rational function into
easier to deal with pieces. (Algebraically, we are reversing the common practice of getting terms over a common
denominator.) It does require us to be able to factor Q(t).
Example 6. Consider the integral Z
2t + 1
dt
t2 +t+1
This is a rational function, P (t) = 2t + 1 and Q(t) = t2 + t + 1. However, partial fraction decomposition is not
advised here. The first thing we should always check is a u-substitution.
Selecting u to be the denominator, we have u = t2 + t + 1 and du = 2t + 1 is the numerator.
Z Z
2t + 1 du
2
dt = = ln |u| + C = ln |t2 + t + 1| + C.
t +t+1 u
Note: Always check a u-substitution with u the denominator first. If the numerator is a multiple of du, this will
be a much faster and easier to compute integral.
Example 7. Consider the integral Z
7t + 25
dt.
t2 + 8t + 7
A quick check of u-substitution shows u = t2 + 8t + 7 would require du = 2t + 8. Since the numerator is 7t + 25,
and 72 du = 7t + 28 doesn’t match.
First, we factor the denominator t2 + 8t + 7 = (t + 1)(t + 7). The partial fractions decomposition is then
7t + 25 7t + 25 A B
= = + .
t2 + 8t + 7 (t + 1)(t + 7) t+1 t+7
We now need to solve for A and B. Solving equations with rational can be difficult, instead we will multiply both
sides of the equation by (t + 1)(t + 7) to reduce it to an equation involving only polynomials,

7t + 25 = A(t + 7) + B(t + 1).

Since this is one equation with two unknowns, we need to make two equations to uniquely solve for A and B. We
can do this by counting up the different functions on both sides – here we need to have the same number of t’s on
both sides, and the same number of 1’s on both sides.

7t + 25 = (A + B)t + (7A + B)

We now have two equations:

t equation: 7 = A + B
1 equation: 25 = 7A + B

3
By substracting the t equation from the 1’s equation, we have 18 = 6A, so A = 3. Then, substituting back into the
t equation, we have 7 = 3 + B, so B = 4. We now have
7t + 25 3 4
= + .
t2 + 8t + 7 t+1 t+7
These pieces we can integrate. So,
Z Z
7t + 25 3 4
2
dt = + dt = 3 ln |t + 1| + 4 ln |t + 7| + C.
t + 8t + 7 t+1 t+7
In general, we can do this process for any rational function for which we can factor the denominator into simpler
pieces. We then can break up a complicated rational function into simpler rational functions, which are usually
well set-up for either a quick integration as above, or work well with u-substitution.

Big Ideas: Whenever your integrand contains a rational function, partial fractions decomposition is a strong
possibility. Before you do the partial fractions decomposition, first check to see if u-substitution works. Then,
you’ll need to factor the denominator and solve an algebra problem to break the rational function up into easier to
integrate pieces.
Problem 1. Z
t2 sin(5t3 ) dt

Problem 2. Z
sin(t)
dt
tan(t)
Problem 3. Z
3
+2t2 +2t+17
(3t2 + 4t + 2)et dt

Problem 4. Z
7t cos(t − 1) dt

Problem 5. Z
1
dt
t2 + 4t + 3
Problem 6. Z
t+2
dt
t2 + 4t + 3
Problem 7. Z
4t ln t dt

Problem 8.
t3 + 4t2 + 3t
Z
dt
t2 + 4t + 3
Problem 9. Z
3t + 1
dt
t2 − 1
Problem 10. Z
sin(2t)
eln(t+3) dt
e− ln(t−4) (t2 − t − 12)
Problem 11. Z
t+4
dt
(t + 2)(t + 6)
Problem 12. Z
2t cos(7t) dt

You might also like