Double Integrals over Ellipses in Polar Coordinates
What You're Actually Solving Here
Double integrals over elliptical regions aren't some abstract math exercise. They're the tool you reach for when circular symmetry doesn't fit your problem. Maybe you're calculating moments of inertia for an elliptical plate, or modeling heat distribution in anisotropic materials. Either way, you need to know how to handle this.
Polar coordinates make circular integrals trivial. Ellipses break that simplicity. But there's a transformation that restores it — and once you see it, the whole process clicks.
The Geometry You Need to Understand First
An ellipse centered at the origin has the equation:
x²/a² + y²/b² = 1
Where a is the semi-major axis (horizontal radius) and b is the semi-minor axis (vertical radius). When a = b, this collapses into a circle. That's your baseline.
The standard polar substitution x = r·cos(θ) and y = r·sin(θ) only works directly for circles where r goes from 0 to the radius. For ellipses, you need to account for the different radii in each direction.
The Scaling Trick That Fixes Everything
Here's what actually works: scale the x-direction by 1/a and the y-direction by 1/b. This transforms your ellipse into a unit circle.
Define new variables:
u = x/a and v = y/b
Now u² + v² = 1 is a unit circle. Your ellipse becomes a circle in (u, v) space.
The Coordinate Transformation
For the ellipse region, your radial bounds in standard polar coordinates aren't constant — they depend on θ. The ellipse boundary in polar form is:
r(θ) = (ab) / √((b·cos(θ))² + (a·sin(θ))²)
This formula comes directly from substituting x = r·cos(θ) and y = r·sin(θ) into the ellipse equation and solving for r.
For 0 ≤ θ ≤ 2π, r runs from 0 to r(θ). That's the direct polar approach, but the integral gets messy fast because r(θ) has that square root in the denominator.
The Better Approach: Scaled Polar Coordinates
Use the substitution that maps the ellipse to a circle:
x = a·r·cos(θ)
y = b·r·sin(θ)
Where 0 ≤ r ≤ 1 and 0 ≤ θ ≤ 2π.
This is the key move. You're not using standard polar coordinates — you're using scaled polar coordinates where the radial scaling differs in each direction.
The Jacobian Determinant
The Jacobian accounts for the area distortion when you change variables. This is where most people get sloppy, and that's where errors creep in.
Compute the Jacobian for x = a·r·cos(θ), y = b·r·sin(θ):
|∂(x,y)/∂(r,θ)| = |det| = a·b·r
The absolute value of this determinant is a·b·r.
This means: dA = dx·dy = a·b·r·dr·dθ
The extra factor of r is the standard polar Jacobian. The a·b comes from the scaling. Don't forget either piece.
Setting Up the Integral
Your double integral over the ellipse region D transforms to:
∬_D f(x,y) dA = ∫(θ=0 to 2π) ∫(r=0 to 1) f(a·r·cos(θ), b·r·sin(θ)) · a·b·r dr dθ
The integrand changes to match your scaled coordinates. The Jacobian is a·b·r. The bounds are now constant: r from 0 to 1, θ from 0 to 2π.
This is the payoff. You traded complicated radial bounds for a constant integration region and a modified integrand.
How To Actually Do It
Here's the step-by-step process you can follow:
- Identify your ellipse parameters — find a and b from the ellipse equation x²/a² + y²/b² = 1
- Set up the coordinate transformation — substitute x = a·r·cos(θ) and y = b·r·sin(θ)
- Compute the Jacobian — you'll get a·b·r
- Transform the integrand — replace x and y in f(x,y) with your scaled expressions
- Change the bounds — r goes from 0 to 1, θ goes from 0 to 2π
- Evaluate — work through the inner r-integral, then the outer θ-integral
Example: Area of an Ellipse
The simplest check: find the area of the ellipse x²/a² + y²/b² ≤ 1.
Set f(x,y) = 1. The integral becomes:
∫(θ=0 to 2π) ∫(r=0 to 1) 1 · a·b·r dr dθ
Inner integral: ∫(0 to 1) a·b·r dr = a·b · [r²/2](0 to 1) = a·b/2
Outer integral: ∫(0 to 2π) a·b/2 dθ = a·b/2 · 2π = πab
Area = πab. When a = b = R, this gives πR², the circle area. The formula checks out.
Example: Moment of Inertia
For the moment of inertia about the x-axis of a uniform elliptical plate with density ρ:
I_x = ∬_D y² dA
Using our transformation with f(x,y) = y²:
I_x = ∫(0 to 2π) ∫(0 to 1) (b·r·sin(θ))² · a·b·r dr dθ
= ∫(0 to 2π) ∫(0 to 1) b²·r²·sin²(θ) · a·b·r dr dθ
= a·b³ ∫(0 to 2π) sin²(θ) dθ · ∫(0 to 1) r³ dr
= a·b³ · π · (1/4)
= (πab³)/4
For a unit density plate, that's I_x = πab³/4. The symmetry makes the θ-integral straightforward because sin²(θ) integrates to π over a full period.
Direct Polar vs. Scaled Polar: Which to Use?
You have two paths. Here's how to choose:
| Method | Radial Bounds | Jacobian | Best When |
|---|---|---|---|
| Direct polar | r varies with θ | r | f(x,y) is simple; you want to avoid extra scaling in integrand |
| Scaled polar | r = 0 to 1 constant | a·b·r | Ellipse is explicit; you want simpler integration bounds |
The direct polar method keeps the standard Jacobian but forces you to integrate r(θ) from 0 to that complicated expression. The scaled polar method gives you constant bounds but adds a·b to every integrand.
For most textbook problems, scaled polar is cleaner. For problems where your integrand already has nice circular symmetry, direct polar might avoid polluting the integrand with extra factors.
Common Mistakes to Avoid
- Forgetting the Jacobian entirely — dx·dy ≠ r·dr·dθ for scaled coordinates. You need a·b·r.
- Using wrong radial bounds — r goes to 1 in scaled coordinates, not to the ellipse boundary. The scaling does that work.
- Confusing a and b — a is horizontal, b is vertical. Mix them up and your answer is wrong.
- Dropping the r factor from the Jacobian — this is the classic error. a·b is not enough. You need a·b·r.
When the Ellipse Isn't Centered at the Origin
The methods above assume the ellipse is centered at (0,0). If your ellipse is centered at (h, k), shift first:
x' = x - h
y' = y - k
Then apply the same transformation to x' and y'. The Jacobian doesn't change — translation doesn't affect area scaling. Only the bounds shift.
The Bottom Line
Double integrals over ellipses in polar coordinates come down to one move: scale your polar coordinates so the ellipse becomes a unit circle. The transformation x = a·r·cos(θ), y = b·r·sin(θ) gives you constant bounds, a Jacobian of a·b·r, and a problem you can actually solve without grinding through complicated radial functions.
Work through the area example until the process is automatic. Once the transformation clicks, every other elliptical integral follows the same pattern.