3.5 - Integration by substitution
A change of variable can turn an unfamiliar integral into a familiar one. In this part of 3.5, you will use a supplied substitution to change every part of the integral, including its differential and, when appropriate, its limits. Algebraic and trigonometric examples will show why keeping the old and new variables separate matters.
Changing variables
What happens to if we call by the shorter name ? The square root becomes , and differentiation gives . Thus the whole combination becomes . The new integral is simply .
Integration by substitution
Integration by substitution changes the variable of an integral using a stated relationship. The relationship changes both the integrand and the differential, so the new integral contains only the new variable.
The method reverses the chain rule. If , then differentiating gives . This explains why the differential factor is essential.
Changing the variable
If , then , so
For an indefinite result, substitute back to the original variable and include . This substitution rule is not listed separately in the formula booklet: reconstruct it from the chain rule.
The differential notation records the derivative relationship; it is not permission to change to without changing the rest of the integral. Use the given substitution, work out its derivative, then rewrite all remaining old-variable factors.
Worked example 1: a square-root power
Question. Use to find .
Plan. Replace by and the matching combination by .
Since ,
Consequently
Answer. , valid for every real .
Check. Differentiation gives . The inside expression is always positive.
Keep the substitution and change only the square root to its reciprocal. The differential conversion is unchanged; only the power of changes.
Worked example 2: a reciprocal square root
Question. Use to find .
Plan. Use the same differential relationship, then integrate the new power .
Answer. , valid for every real .
Check. Its derivative is . At , both this derivative and the original integrand are zero.
We replaced the combination directly, so we did not need to divide by or invert . This avoids introducing an unnecessary restriction at . The derivative checks establish the displayed antiderivatives on their original domains.
Watch out
From , the line is wrong: it retains the factor already absorbed by . A transformed integral must have the correct differential factor and a single variable.
Use a new inner constant and power.
Transforming limits
For a definite integral, the old limits describe values of . Convert them into values of the new variable before evaluating a transformed antiderivative. A small table keeps the correspondence visible.
The substitution may also require you to express itself in terms of . With , we have and . The root is the non-negative root, so ; this sign condition is part of the substitution.
Substitution with limits
For a valid change on the interval,
Use the transformed limits with a -antiderivative. Alternatively, return the antiderivative to and use the original limits. Do not mix these two routes.
The definite substitution rule is not listed separately in the formula booklet: it follows from the chain rule and definite integration.
Changing variables changes the integrand's height and the interval's width together. The two shaded regions below represent the same integral, although their graphs and horizontal scales differ.
[DIAGRAM: asset_name: The same integral in two variables; asset_slug: substitution-limits-area; caption: The substitution changes both the function and the limits while preserving the integral; description: The left coordinate plot shades y equals x times the square root of x plus one from x equals zero to three. The right shades y equals two times u to the fourth minus two times u squared from u equals one to two. Both areas are labelled 116 over 15. A linking label states u equals the square root of x plus one, x equals u squared minus one, and dx equals two u du.]

Worked example 3: converting every old-variable factor
Question. Use to find the exact value of .
Plan. Express , the square root and in terms of , then convert both limits.
The limits become
Therefore
Answer. .
Check. Rewriting directly gives the antiderivative . At and , its difference is also .
Continue this lesson.
You’re previewing NovaLearn without an account. Create a free account to read the rest.
Create free account