Aerodynamic Analysis - is it worth it?

Question

Aerodynamic Analysis - is it worth it?

phoenix221 (Computer)

(OP)

17 Sep 04 20:36

Hello all,

I am relatively new to this field and my latest question has to do with computer aided aerodynamic analysis. I know that it takes weeks/months worth of investment in learning any given packages so I would like to have a basis for confidence in a given code before making such a committment.

I was wondering what state the current tools are in, that is, given an accurate CAD model (meshed for FEA) is it possible to do a realistic aerodynamic analysis of the airplane using any of today's software packages? If not, how close to real-world behavior will such an analysis be?

I would be interested what the issues may arise with such an undertaking as well as what software is recommended?

Thanks,

Sparweb (Aerospace) · Answer 1 · 2004-09-19 03:24:57

The accuracy of your results is dependent on the assumptions you build into the analysis, the amount of detail you desire, and the quantity of computing power (read: "money") you intend to throw at it.  The FEA model of the object doesn't have to be accurate, so much as it has to be broken down into elements that equal the level of detail you're looking for.

CFD analysis can be done using different sets of principles.  For example, you can ignore compressibility if you are considering speeds lower than 30%Mach.  Are you going to consider viscosity?  You may not need it for some projects.

Designing a CFD analysis that produces practical results is an art form unto itself, and I'm far from an expert.  But also consider that CFD only replaces another art form: design of wind tunnel analysis, which is also tricky, in that the results don't easily scale up to full size.

Steven Fahey, CET
"Simplicate, and add more lightness" - Bill Stout

phoenix221 (Computer) · Answer 2 · 2004-09-22 00:39:20

SparWeb,

Thanks for your post...

Is there a good reference for CFD that does not require a Phd in Math? I would like to understand the "lay of the land" as it were, i.e. the principles you are talking about which underlie a CFD analysis.

The "art form" comment is a bit unsettling since "art" implies a certain amount of "undefinable" quality to the hole thing...

GraviMan (Automotive) · Answer 3 · 2004-09-22 08:03:15

Is there a basic CFD package that could provide a good intro, without huge cost? Limited elements or time works fine for me...

Mart

Sparweb (Aerospace) · Answer 4 · 2004-09-22 10:59:13

I'm not trying to promote anything - and believe me I haven't tried this software so I don't know anything about it, but:

http://www.aerodyn.org/
http://www.desktopaero.com/
http://www.hanleyinnovations.com/

Is this the kind of stuff you are looking for?

If either of you are good with computer programming languages and took a few aerodynamics courses in school, you probably could bang together your own program, especially if you can simplify the problem to a level that fits within your programming abilities. The book Foundations of Aerodynamics, by Kuethe and Chow, could help at this task (it seems to be written for people doing this).

Steven Fahey, CET
"Simplicate, and add more lightness" - Bill Stout

GraviMan (Automotive) · Answer 5 · 2004-09-23 07:52:00

Hey this is great stuff, thanks Sparweb!

I can handle simple Navier-Stokes stuff, and am reading some notes on high speed flows. There seem to be enough CFD packages out there, and I doubt I could get anywhere near...

Mart

IRstuff (Aerospace) · Answer 6 · 2004-09-23 11:42:04

THere is a also a program called MicroTunnel, which is a CFD written specifically around a wind tunnel environment. Sells for $99 http://www.microcfd.com/

Seems to do its thing, but it's a bit slow and awkward to use. There is a free trial available for download

TTFN

GregLocock (Automotive) · Answer 7 · 2004-09-23 18:13:53

The free trial lasts for 3 days

Make sure you have a fast computer, it really does take hours to work through a problem.

The most annoying thing is that it uses the pixels to represent the airfoil, so if you are trying to model a sail, as I was, rescaling the section led to holes in it, which was not very helpful.

But I have to say I found it a very easy program to use.

Cheers

Greg Locock

pja (Aerospace) · Answer 8 · 2004-09-23 22:08:39

Simple Navier-Stokes stuff?? That's the equivalent of saying
you ran in an easy marathon... : )

phoenix221 (Computer) · Answer 9 · 2004-09-24 18:51:42

>THere is a also a program called MicroTunnel, which is a CFD
>written specifically around a wind tunnel environment. Sells
>for $99 http://www.microcfd.com/

It looks this is for 2D only... is there an equivalent for 3D analysis?

GraviMan (Automotive) · Answer 10 · 2004-09-29 07:50:29

"Simple Navier-Stokes stuff??"

Ah well, the trick is to cheat! You use Bernoulli for most of the calc (eg lift estimation), then use Navier-Stokes in the boundary layer to estimate drag. I can see why the civilised world moved on to finite difference methods though...

Mart

Come Join Us!

Posting Guidelines

Aerodynamic Analysis - is it worth it?