According to mathworld, let the sphere have radius $R$, then the surface area a spherical cap of height $h$ and base radius $a$ is given by $$S=2\pi Rh=2\pi(a^2+h^2).$$

What is this value for an n-dimensional hypersphere?
If it helps simplify the problem we can assume $R=1$ and $a=0.5$.

Many thanks.

  • $\begingroup$ The terms "$n$-dimensional" and "hyper<object>" have different meanings to mathematicians than to the public at large. To clarify, by "$n$-dimensional hypersphere" do you mean (1) An $n$-dimensional sphere (i.e., a sphere in $\mathbf{R}^{n+1}$), or (2) A sphere in $\mathbf{R}^{n}$? $\endgroup$ – Andrew D. Hwang Apr 17 '17 at 11:12
  • 1
    $\begingroup$ @AndrewD.Hwang thanks for the clarification, I wanted to mean a sphere in $R^n$ $\endgroup$ – dontloo Apr 17 '17 at 12:02
  • $\begingroup$ There are easier forms for asymptotic approximations of these values, when n is large. For a sphere of radius 1, the spherical cap of height a has volume/surface area (1-a^2)^[n/2 + o(n)]. $\endgroup$ – TMM Apr 17 '17 at 12:12
  • $\begingroup$ @TMM thank you that's very useful, do you have any idea about the asymptotic approximation of the ratio between the cap and the total surface area? $\endgroup$ – dontloo Apr 17 '17 at 12:30

The solution requires special classes of functions, namely regularized incomplete beta functions $I_{\sin^2 \Phi} (\frac{n-1}{2}, \frac{1}{2})$ and Gamma functions. The result and its derivation can e.g. be found in the very nice article by S. Li here. Let $\Phi$ be the angle of the cap, so $a = R\sin \Phi$.

Then the surface area of the cap is $$ A(\Phi) = R^{n-1} \frac{\pi^{n/2}}{\Gamma(n/2)} I_{\sin^2\Phi} (\frac{n-1}{2}, \frac{1}{2}) $$

  • $\begingroup$ Thanks for the answer, I just saw that paper as well, I think one crucial factor is the regularized incomplete beta functions, however after some googling I still don't have a clue what is the function like, especially how it grows wrt n with different $\phi$ values, could you explain a bit? $\endgroup$ – dontloo Apr 17 '17 at 14:15
  • $\begingroup$ It would be great if you could also share some insight on how the ratio between the cap and the total surface area grows (asymptotically) wrt n, thank you :) $\endgroup$ – dontloo Apr 17 '17 at 14:18
  • $\begingroup$ For constant angles $\Phi \in (0, \pi/2)$, the ratio of cap surface area to total area is exponentially small. $\endgroup$ – Chris Jones Mar 29 '19 at 21:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.