Thread Tools
Nov 19, 2019, 02:42 PM
Sink stinks
Montag DP's Avatar
Thread OP
The easiest way to fix that would be to either put inputs.txt in the win64 directory where you are trying to run it, or run it from the directory where inputs.txt resides.
Sign up now
to remove ads between posts
Nov 19, 2019, 03:20 PM
Registered User
Thank you, it works.

Now I need to figure out where to get coordinates for the various Kline-Fogleman airfoils, how to put the data into xoptfoil (Is it even possible to get some plausible results?) and what settings I have to use in order to achieve something plausible.

Why!? Because nobody ever made a simulation of the KFM airfoil at low reynoldsnumbers, and I think we reached the end of what's possible with conventionell airfoils.

BTW. the link mentioned under 7 is no longer valid: https://www.continuum.io/downloads.
Nov 19, 2019, 03:45 PM
Sink stinks
Montag DP's Avatar
Thread OP
I don't think Xfoil is a good tool to analyze Kline-Fogleman airfoils. It is not designed to model the separation over the step correctly.
Nov 19, 2019, 03:50 PM
Registered User
Quote:
Originally Posted by Montag DP
I don't think Xfoil is a good tool to analyze Kline-Fogleman airfoils. It is not designed to model the separation over the step correctly.
Any suggestion for a suitable program ?
Nov 20, 2019, 09:56 AM
Sink stinks
Montag DP's Avatar
Thread OP
Quote:
Originally Posted by stereodreieck86
Any suggestion for a suitable program ?
You would need an actual Navier Stokes / RANS solver at least. It should probably have a decent transition model too if you are interested in model-scale Reynolds numbers. OpenFOAM is one option for a free open source RANS solver, but I don't know if it has a suitable transition modeling capability. Either way, it is going to be much slower and require a lot more expertise than running something like Xfoil.
Nov 26, 2019, 12:07 PM
Registered User

On optimizing the optimizer


Hi there,

after some time being a passive user of Xoptfoil, watching with fascination the work of the particle swarm, I started to step more and more a little deeper into Xoptfoil ... finally doing modifications of the program code. My target was to get my "perfect airfoil" right out of Xoptfoil which is ready for use also from a more advanced perspective.

During the last months I dealed with some different aspects of the airfoil optimzation. Always when I thought "Now I got it" it turned out, that this was just the beginning of other issues and I had to re-dive into the code. Up to now I tried to cover and implement following aspects :
  • Aerodynamic target values
  • Geometric target values
  • Assessment of the surface of an airfoil
  • Smoothing of the surface
  • ... and some other minor things

When describing these modifications I realized, that it would become a little longer and also some pictures were needed. So I put it all in the attached pdf-file.

Compared to the impressing expertise and capabilities of Dan I'm still an advanced junior in software development as well in aerodynamics or mathematics. So I kindly ask for forbearance if some things should be wrong. I would be happy to get some feedback on my findings - and maybe the one or other approach could be suitable for Xoptfoil 2.0 ...

Regards

Jochen

Dan, would you recommend to load the modifications as a branch into Sourceforge or Github?
Nov 26, 2019, 12:41 PM
Sink stinks
Montag DP's Avatar
Thread OP
Jochen,

Thanks very much for that report. You have done some great work here, and I definitely want to bring all or most of them into the main codebase. I have been away from Xoptfoil development for several months because I was busy getting into FPV flight and then designing and building an airplane (you know, "real" RC modeling stuff, as some people have told me ). Now that I'm done with that, I will get back to working on Xoptfoil 2 soon. I think it would be best if you could make a fork of the project on Github and add in your changes. Then you can just tell me which fork is yours, as there are already several existing (I'm not sure exactly why, as no one has pushed changes back to date). Since the code is currently being rewritten, doing a pull request will not be really productive, but as I am writing the new code I will be sure to incorporate your improvements.

Edit: oh, I should mention, it would be best if you implement each change in a single commit or a set of commits, rather than dumping all the modifications into one commit. That way I can see exactly what was changed for what purpose.

Also, regarding modifying the code, the new version will be mostly written in Python, with a little bit in C, so modifications should be much easier. For most things, it will no longer be necessary to recompile.

Dan
Last edited by Montag DP; Nov 26, 2019 at 12:49 PM.
Nov 28, 2019, 12:52 AM
Registered User

Xoptfoil-JX


Hi Dan,

thanks for the recognition

The sources are now on my very first Github repo being a fork of Xoptfoil. I couldn't make it to divide all the modifications into different commits but I tried to tag the reason of a modification with keywords in the comments e.g. "! jx-mod Geo targets". I'm sure you are able to read all the modifications like a book ...

Forgot to say that I switched off multi threading in the make file because I a had strange results from time to time (also my mini PC is happy to cool down a little during optimization runs).

I really hope that the "bad weather season" started now also at the place you are living - so you do not have to go outside for flying and can enjoy working on 2.0 ...

regards

Jochen
Nov 28, 2019, 12:16 PM
Sink stinks
Montag DP's Avatar
Thread OP
Jochen,

Thanks again. Your changes are easy enough to follow. I will be sure to incorporate them in version 2.

Dan


Quick Reply
Message:

Thread Tools