Perimeter of an Ellipse

Finding the perimeter of an ellipse is not like finding the perimeter of a rectangle or the circumference of a circle. There is no simple, exact formula that gives the answer with elementary arithmetic. The perimeter of an ellipse can only be expressed exactly using a complete elliptic integral of the second kind. For practical work, laying out an oval garden bed, estimating trim for an elliptical table, or calculating track distance, you need an approximation that stays within a fraction of a percent of the true value. The best-known and most reliable approximations for everyday use were published by Srinivasa Ramanujan in 1914.

The perimeter of a circle is πd, a single multiplication. The perimeter of an ellipse depends not only on its size but on its shape, specifically, on the ratio of its major axis a (half the longest diameter) to its minor axis b (half the shortest diameter). As the ellipse becomes more elongated, the perimeter grows in a way that cannot be captured by a finite combination of basic arithmetic operations and standard functions. The exact perimeter is given by p = 4a E(e), where E(e) is the complete elliptic integral of the second kind and e is the eccentricity, defined as √(1 − b²/a²). That integral has no closed-form solution in elementary terms. You do not need to understand the integral to get a usable answer, but knowing why no simple formula exists explains why every practical method is an approximation.

Ramanujan published two approximations for the perimeter of an ellipse in his 1914 paper 'Modular equations and approximations to π'. Both are remarkably accurate for the vast majority of ellipses encountered in real life.

Ramanujan's First Approximation

The first formula is: p ≈ π [3(a + b) − √(10a² + 10b² + 6ab + 6b²)]. This expression is straightforward to compute with a calculator and gives a result within about 0.04% of the true perimeter for most ellipses. Its error increases only when the ellipse is extremely elongated, when the ratio a/b exceeds roughly 3.

Ramanujan's Second Approximation

The second formula is more accurate for high eccentricities: p ≈ π(a + b) [1 + 3h² / (10 + √(4 − 3h²))], where h = (a − b)/(a + b). This version reduces the error to well under 0.01% for nearly every practical shape, including ovals with an axis ratio of 10 or more. Use this one when the ellipse is visibly elongated or when you need the highest accuracy possible without computing the elliptic integral.

Before Ramanujan, the most common shortcut was the simple average formula: p ≈ π(a + b). This approximation is easy to remember and quick to calculate, but it underestimates the true perimeter by 5% to 15% for typical ellipses. The error grows as the ellipse becomes more eccentric. For a circle (a = b), the formula reduces to p = π(2a), which is the correct circumference, but for any other shape it is too low. A slightly better version, p ≈ 2π√[(a² + b²)/2], still underestimates by 2% to 5% for most cases. Neither is recommended when accuracy matters. Ramanujan's first approximation is nearly as simple to compute and is far more reliable.

Suppose you are designing an oval table with a major axis of 2.4 meters (so a = 1.2 m) and a minor axis of 1.6 meters (b = 0.8 m). The axis ratio a/b is 1.5, a moderately elongated shape. Compute the perimeter using Ramanujan's first approximation.

First, calculate 3(a + b): 3 × (1.2 + 0.8) = 6.0. Next, calculate 10a² = 10 × 1.44 = 14.4, 10b² = 10 × 0.64 = 6.4, 6ab = 6 × 0.96 = 5.76, and 6b² = 6 × 0.64 = 3.84. Sum these: 14.4 + 6.4 + 5.76 + 3.84 = 30.4. Take the square root: √30.4 ≈ 5.513. Subtract from 6.0: 6.0 − 5.513 = 0.487. Multiply by π: 0.487 × 3.1416 ≈ 1.530. So the estimated perimeter is about 1.53 meters.

For comparison, the simple average formula π(a + b) gives 3.1416 × 2.0 = 6.283 meters, a 4.75-meter overestimate. The error comes from misreading the formula: the average approximation is actually π(a + b) which here is 6.28 m, far from the correct value. The worked example shows why relying on the simple average wastes material.

Track Layout

Ellipses appear in running tracks, oval dining tables, raised garden beds, and architectural arches. The difference between a calculated perimeter and the real-world measurement can be significant. For a running track, the perimeter of the inner oval determines the lap distance. A 400-meter track is usually a rectangle with two semicircular ends, not a true ellipse, but oval tracks in smaller facilities or temporary courses may use an elliptical shape.

Edging and Trim

For a garden bed or table, the perimeter equals the length of trim or edging needed. If you use the simple average formula for a bed with a = 2 m and b = 1 m, you would estimate 9.42 m of edging. The true perimeter using Ramanujan's second approximation is about 9.69 m, a quarter-meter difference that means buying an extra roll of edging or having a gap. In construction, that gap costs time and money.

The most common mistake is assuming a circle formula works for an oval. The circumference of a circle is a special case of the ellipse perimeter where a = b. For any other shape, you need the approximation. Ramanujan's first approximation is accurate enough for most layouts; use the second when the shape is elongated or when material cost is tight.