1

I am doing some research on ellipsoids. I am not sure where the formula for the surface area of a prolate ellipsoid comes from. Can anyone please help me with how to derive the formula. I have the formula below

the formula

  • Where comes form this formula? See: https://en.wikipedia.org/wiki/Ellipsoid#Surface_area – Emilio Novati Dec 08 '16 at 20:50
  • @EmilioNovati It does not show how the formula is derived – user99820 Dec 08 '16 at 21:12
  • For prolate (or oblate) ellipsoid, use surface of revolution. – Ng Chung Tak Dec 08 '16 at 22:17
  • @NgChungTak Is there a way you can explain to me how that works ? Thank you – user99820 Dec 09 '16 at 07:31
  • see (https://en.wikipedia.org/wiki/Pappus%27s_centroid_theorem) – Jean Marie Dec 09 '16 at 10:58
  • @NgChungTak Is there a way the formula can be derived from the formula in the link below. (http://tutorial.math.lamar.edu/Classes/CalcIII/QuadricSurfaces_files/eq0003MP.gif ) – user99820 Dec 10 '16 at 22:36
  • Your formula is approximation only, it's a bit empirical. For the formula I mentioned in my comments below, please refer to the journal here – Ng Chung Tak Dec 11 '16 at 07:56
  • @NgChungTak Is it possible for you to explain the formula. Because I don't understand some things on the formula . Thank you – user99820 Dec 14 '16 at 06:28
  • What's your background? For the approximate formula you mentioned in your question, I haven't seen any literature about it. While elliptic integrals are rarely taught in university undergraduate courses. Basically, I self-studied elliptic integrals by reading plenty of books in university library. I suggest you to consult your professor or instructor for further information. – Ng Chung Tak Dec 14 '16 at 10:28

1 Answers1

7

The method is very standard and appears in most calculus texts.

Let $\dfrac{x^2}{a^2}+\dfrac{y^2}{b^2}=1$ be the ellipse such that $a>b$.

  • For prolate spheroid, it rotates about the $x$-axis.

\begin{align*} y &= \frac{b}{a} \sqrt{a^2-x^2} \\ \frac{dy}{dx} &= -\frac{bx}{a\sqrt{a^2-x^2}} \\ ds &= \sqrt{1+\left( \frac{dy}{dx} \right)^2} \, dx \\ &= \sqrt{1+\frac{b^2x^2}{a^2(a^2-x^2)}} \, dx \\ &= a\frac{\sqrt{1-\left( 1-\frac{b^2}{a^2} \right) \frac{x^2}{a^2}}} {\sqrt{a^2-x^2}} \, dx \\ S &= \int_{-a}^{a} 2\pi y \, ds \\ &= 4b\pi \int_{0}^{a} \sqrt{1-\left( 1-\frac{b^2}{a^2} \right)\frac{x^2}{a^2}} \, dx \\ &= 4b\pi \left[ \frac{x}{2} \sqrt{1-\left( 1-\frac{b^2}{a^2} \right) \frac{x^2}{a^2}}+ \frac{a^2}{2\sqrt{a^2-b^2}} \sin^{-1} \frac{x\sqrt{a^2-b^2}}{a^2} \right]_{0}^{a} \\ &= 2\pi b \left( b+\frac{a^2}{\sqrt{a^2-b^2}} \sin^{-1} \frac{\sqrt{a^2-b^2}}{a} \right) \\ \end{align*}

  • For oblate spheroid, it rotates about the $y$-axis.

\begin{align*} x &= \frac{a}{b} \sqrt{b^2-y^2} \\ \frac{dx}{dy} &= -\frac{ay}{b\sqrt{b^2-y^2}} \\ ds &= \sqrt{1+\left( \frac{dx}{dy} \right)^2} \, dy \\ &= \sqrt{1+\frac{a^2y^2}{b^2(b^2-y^2)}} \, dy \\ &= b\frac{\sqrt{1+\left( \frac{a^2}{b^2}-1 \right) \frac{y^2}{b^2}}} {\sqrt{b^2-y^2}} \, dy \\ S &= \int_{-b}^{b} 2\pi x \, ds \\ &= 4a\pi \int_{0}^{b} \sqrt{1+\left( \frac{a^2}{b^2}-1 \right)\frac{y^2}{b^2}} \, dy \\ &= 4a\pi \left[ \frac{y}{2} \sqrt{1+\left( \frac{a^2}{b^2}-1 \right) \frac{y^2}{b^2}}+ \frac{b^2}{2\sqrt{a^2-b^2}} \sinh^{-1} \frac{y\sqrt{a^2-b^2}}{b^2} \right]_{0}^{b} \\ &= 2\pi a \left( a+\frac{b^2}{\sqrt{a^2-b^2}} \sinh^{-1} \frac{\sqrt{a^2-b^2}}{b} \right) \\ \end{align*}

The two cases are interchangeable by flipping the roles of $a$ and $b$ together with $\sinh iz=i\sin z$

Ng Chung Tak
  • 18,990
  • [+1} Note that $\sqrt{a^2-b^2}=c$ (distance from center to focii) and even more, for the first case $c/a=e$ (eccentricity) giving simpler looking formulas. – Jean Marie Dec 09 '16 at 14:59
  • Is there a connection with the formula given by the OP ? – Jean Marie Dec 09 '16 at 15:01
  • $$\dfrac{x^2}{a^2}+\dfrac{y^2}{b^2}+\dfrac{z^2}{c^2}=1$$

    where $a\ge b\ge c$.

    $$S=2\pi \left[ c^{2}+\frac{bc^{2}}{\sqrt{a^{2}-c^{2}}}, F\left( \cos^{-1} \frac{c}{a}, k \right)+b\sqrt{a^{2}-c^{2}}, E\left( \cos^{-1} \frac{c}{a}, k \right) \right]$$

    $$k=\frac{a}{b}\sqrt{\frac{b^{2}-c^{2}}{a^{2}-c^{2}}}$$

    For prolate, $$b=c\implies k=0$$

    For oblate, $$a=b\implies k=1$$

    – Ng Chung Tak Dec 09 '16 at 15:13
  • Very interesting. Thank you. – Jean Marie Dec 09 '16 at 15:19
  • ... where $E$ and $F$ are "incomplete elliptic integrals" – Jean Marie Dec 09 '16 at 15:26
  • @JeanMarie Yes, of course. – Ng Chung Tak Dec 09 '16 at 15:31
  • I just added this information for people unaware of the meaning of E and F. Their meaning cannot be retrieved unless they have at least the keywords "elliptic integrals". – Jean Marie Dec 09 '16 at 15:36
  • Here is a site where one can find the formula given by the OP (http://planetcalc.com/149/). It is an approximate formula with a precise value of $p$... – Jean Marie Dec 09 '16 at 15:39
  • An interesting reference: (www.citr.auckland.ac.nz/researchreports/CITR-TR-165.pdf) – Jean Marie Dec 09 '16 at 15:57
  • 1
    I do not have the reputation to comment on his answer, but there is a typo in at least one line of Ng Chung Tak's answer above: the term before the asinh() expression should be ${ b^2 \over 2 \sqrt{a^2 - b^2}}$. As printed now, the numerator is "b" instead of "$b^2$". [converted to comment by a moderator] – Rudi May 13 '18 at 18:51
  • @Rudi, Amended, thanks for your response. – Ng Chung Tak May 14 '18 at 00:33