2

Given an ellipse with a = semi-major axis and b = semi-minor axis, what is the exact formula for determining the average radius, i.e., the average distance between center and perimeter over the enter ellipse?

I see a few approximation formulas online. For example, the "arithmetic mean radius" is:

$$\frac{2a + b}{3}$$

And the "quadratic mean radius" is:

$$\sqrt{\frac{a^2 + b^2}{2}}$$

But these are just approximations. I want some formula, like an integral or a hypergeometric function, that gives an exact answer. It's easy to find one such formula for the circumference of an ellipse:

$$4a \int_0^{\pi/2} \sqrt{1 - (1 - b^2/a^2) \sin^2(x)} dx$$

Which given $e = \sqrt{1 - b^2/a^2}$, simplifies to:

$$4a \int_0^{\pi/2} \sqrt{1 - e^2 \sin^2(x)} dx$$

But I can't seem to find a working similar one for the mean radius. This answer seems to supply one, but it evaluates to exactly the semi-minor axis, which seems to be by definition not what I'm looking for.

EDIT: I specifically mean in terms of arc-length (thank you Parcly Taxel).

  • Are you averaging over the angle from the centre or over the perimeter or something else? You will get a smaller average over the angle. Neither of these two are appropriate for planetary motion – Henry Feb 26 '23 at 01:55
  • @ParclyTaxel I'm wondering: if one were to draw a line from the center of the ellipse to a random point chosen on the perimeter, what is the expected value of the length of that line. Is that arc-length measure? – Johansson McFleppers Feb 26 '23 at 02:06
  • @Henry I'm interested in this from a geometry perspective, not a planetary motion or astronomical one. – Johansson McFleppers Feb 26 '23 at 02:07
  • @JohanssonMcFleppers In what sense is the point on the perimeter chosen "at random?" In other words, what is the process by which the point is chosen? Do you pick an arbitrary direction/angle from the origin and find the point of intersection of that ray to the ellipse, or do you pick by taking an number uniformly from $[0,1)$ and this represents the proportion along the ellipse's perimeter at which the point is taken? – heropup Feb 26 '23 at 02:14
  • @heropup The second one. Random in terms of proportion along the perimeter, not in terms of angle. – Johansson McFleppers Feb 26 '23 at 02:15
  • @ParclyTaxel The second answer you referenced seems to just say it's the integral of the "arclength integrand". But what actually is the formula? Sorry, I realize I am ignorant about this. – Johansson McFleppers Feb 26 '23 at 02:16
  • Please see the summary of an older post of mine here. – Ng Chung Tak Feb 26 '23 at 04:01
  • @NgChungTak I did already see that post of yours while researching. The only one of those six formulas that seems like it could be what I want is the Hooke's law orbit arclength average one, but when I evaluated it with some test parameters, it did not result in something that made sense for my question (the result was smaller than the semi-minor axis). Either I did it incorrectly, or the discrepancy is due to the fact I am not interested in this in relation to orbital mechanics, but just geometry/calculus. – Johansson McFleppers Feb 26 '23 at 04:13
  • @ParclyTaxel See my response to NgChungTak above. I don't think it does answer it, but I'm not completely confident. – Johansson McFleppers Feb 26 '23 at 05:09
  • @ParclyTaxel OH WAIT. The equation now seems to be giving me results that make sense. So. . .I must have miswrote it before. In which case, thank you! Seems like that is likely the answer after all. – Johansson McFleppers Feb 26 '23 at 05:27
  • @NgChungTak Thank you too of course (see above comment) – Johansson McFleppers Feb 26 '23 at 05:28
  • I most likely will, but I will first wait a few more hours just in case I realize I missed something. – Johansson McFleppers Feb 26 '23 at 05:45
  • 1
    @JohanssonMcFleppers: I just wasted a lot of time reading through the entire question and comments, just to find that the question has been resolved. Please either post the answer and accept it, or delete the question. If you want to wait before you do that, please edit the question to include a warning that you currently believe it's been answered. – joriki Feb 26 '23 at 07:09

0 Answers0