Yes, that is a good list. I have some of those. Just if I ever had time to dig in some of them in more detail.
I have studied mostly DATCOM. We had an idea some years ago to integrate DATCOM into JSBSim so that one could test the designed plane with that simulator. They seem to have done some progress in that area since then.
But I don't like the FGFS simulator.. I think it is a too large heap of amateur programming without any real "brain" in it. For example the heavy pointless usage of XML just makes me mad. They write and write code without ever getting to the actual point... they just keep going around it all the time. Really annoying! The computer gets filled with crap.. Somebody should take over FGFS and really force them to do some meaningfull code.
http://www.crapmachine.com/
:)
X-Plane might also have some design-and-simulate capabilities currently? More professionally done there maybe? Hope so anyway.
I had somewhere a small program that could also simulate a proposed plane parametically. Just where do I have it and what was it's name...
>X-Plane might also have some design->and-simulate capabilities currently? More >professionally done there maybe? Hope so >anyway.
X-plane has the cycle dump, outputs all values for one simulation round. It outputs to a text file, you can find one from one of my previous comments on this blog.
Then there is a function for finding stability derivatives. That gets also outputted to a text file.
One more thing: The little challenge with X-plane is that one has to know what to do when creating a model - the X-plane uses pre-computed airfoil pressure distribution. Usually it seems that people do a model of some airplane X and then they just use wrong data for the airfoil and it really does not behave like the real one. To be accurate you have to put just the airfoil data in that you are investigating that time, e.g. if I am investigating flying at 47000 ft with Re 1000000, I need to convert the simulation output of X-foil to X-plane and use that when the plane flies at 47000 ft in the sim at Re 1000000. To be accurate, one may need to have multiple versions of a model to test different Re conditions.
Many plane designers, don't even approximate and they don't even have a simulated data for the proper airfoil, but for example they use NACA 2412 for a plane that is supposed to have e.g. GU25 airfoil which has completely different behavior. People are just lazy, because the X-plane does not input the airfoil dat file but requires the file to be pre-processed before inputting to X-plane.
Yes, I have had that same thought about XML. There is a lot to write. However, for text configuration files, XML has a lot to offer; the alternatives are not very nice. Ideally, the various XML files would have a GUI editor, so one would never need to edit the actual XML code, itself.
4 comments:
Yes, that is a good list. I have some of those. Just if I ever had time to dig in some of them in more detail.
I have studied mostly DATCOM. We had an idea some years ago to integrate DATCOM into JSBSim so that one could test the designed plane with that simulator. They seem to have done some progress in that area since then.
But I don't like the FGFS simulator.. I think it is a too large heap of amateur programming without any real "brain" in it. For example the heavy pointless usage of XML just makes me mad. They write and write code without ever getting to the actual point... they just keep going around it all the time. Really annoying! The computer gets filled with crap.. Somebody should take over FGFS and really force them to do some meaningfull code.
http://www.crapmachine.com/
:)
X-Plane might also have some design-and-simulate capabilities currently? More professionally done there maybe? Hope so anyway.
I had somewhere a small program that could also simulate a proposed plane parametically. Just where do I have it and what was it's name...
>X-Plane might also have some design->and-simulate capabilities currently? More >professionally done there maybe? Hope so >anyway.
X-plane has the cycle dump, outputs all values for one simulation round. It outputs to a text file, you can find one from one of my previous comments on this blog.
Then there is a function for finding stability derivatives. That gets also outputted to a text file.
One more thing:
The little challenge with X-plane is that one has to know what to do when creating a model - the X-plane uses pre-computed airfoil pressure distribution. Usually it seems that people do a model of some airplane X and then they just use wrong data for the airfoil and it really does not behave like the real one. To be accurate you have to put just the airfoil data in that you are investigating that time, e.g. if I am investigating flying at 47000 ft with Re 1000000, I need to convert the simulation output of X-foil to X-plane and use that when the plane flies at 47000 ft in the sim at Re 1000000. To be accurate, one may need to have multiple versions of a model to test different Re conditions.
Many plane designers, don't even approximate and they don't even have a simulated data for the proper airfoil, but for example they use NACA 2412 for a plane that is supposed to have e.g. GU25 airfoil which has completely different behavior. People are just lazy, because the X-plane does not input the airfoil dat file but requires the file to be pre-processed before inputting to X-plane.
Yes, I have had that same thought about XML. There is a lot to write. However, for text configuration files, XML has a lot to offer; the alternatives are not very nice. Ideally, the various XML files would have a GUI editor, so one would never need to edit the actual XML code, itself.
Good luck!
Jon Berndt
JSBSim Project Coordinator
Post a Comment