2

I am currently working on axisymmetric bodies like fuselages/droplets etc. For this, I am interested in the point of flow separation (if there is) and velocity components at a certain point in space.

I know that XFoil is capable of doing this in a short fraction of time/effort compared to CFD for 2D sections.

What is the validity of taking this approach towards axisymmetric bodies? With this I mean: using the cross section of the axisymmetric body and simulating this 2D cross section in XFoil, or any other 2D airfoil simulation programme.

Or: can I change the software in some way such that it can represent the physics better?

vipers120
  • 23
  • 3
  • You should edit the title to be more specific (and to transform it in a question as we are on a Q&A website). Maybe "how to determine the point of flow separation?" or "what are the alternative to XFoil to find point of flow separation?" – Manu H Sep 25 '19 at 12:53
  • Edited title and text – vipers120 Sep 25 '19 at 13:01

2 Answers2

3

Xfoil analyzes airfoils, not wings, and as such does not account for wingtip effects so its results are invalid for bodies of revolution or even low aspect ratio wings.

More generally, you can reference this answer about the derivation of wing lift from airfoil data. It boils down to the fact that aspect ratio has a massive impact on the 3D flow around the wing. For high AR wings the flow at most sections is close to what can be predicted with a 2D method, but this condition does not hold for slender bodies or near wingtips.

To phrase it differently, we can only apply the results from 2D simulations to wings by avoiding 3D effects. In the case of slender bodies, the entire flowfield is impacted by 3D effects.

Since flow separation is dependent on the pressure field around the body, you need to have an accurate idea of the 3D flow to reliably predict separation points.

AEhere supports Monica
  • 8,581
  • 1
  • 35
  • 67
  • Yes, I understand that XFoil does not account for 3D effects as wingtip etc. Following your URL, I also understand that this 2D simulation does not compute the correct lift/drag caused by not accounting for this. What I am, however, interested in is the point of separation and the velocity field surrounding the body. I do not see how these get affected significantly by having a body instead of a 2D cross-section. – vipers120 Sep 25 '19 at 13:14
  • I'm glad you found this useful, but if you don't accept this answer just yet you might get some better ones from other time zones. – AEhere supports Monica Sep 25 '19 at 13:29
0

You can use Xfoil to do this kind of work: you simply scale your axisymmetric thickness form by 2.3 (IIRC). Maybe 2.4 if you are more interested in the transition location on your body of revolution. The reason for the thickness scaling factor is that a body of revolution appears as a much less 'thick' object to the 3D inviscid flow.

Tim
  • 1