Physics Based sims?

[MCarlton]

Hello,

I wondered, are there any sims out there which run on a "true physics" model, in that they accurately model airflow?

If so, my next question is, is it possible to import a model into any of them, based on inputting actual parameters from a real model?

Most I have seen seem to use a "virtual" aerodynamics model in which the various factors (drag/lift coefficient, induced drag, thrust, pitch speed etc) are tweaked until the model flies how the author wants it to.

What I feel would be better is a sim which can generate those values by modelling airflow over a 3D model. So, you could "build" a model in the sim, using something like 3DS/Gmax/Metaseqoia etc and the sim would use simulated airflow around that as the physical model.

Prop type, size etc could be input also, as could other variables.

The sim could use something akin to Motocalc in its modelling to acheive some of this perhaps?

Any thoughts?

[skirtz]

The current ClearView physics is built that way. The model is considered a collection of liftsurfaces and props. All physics forces are calculated based on this information - see how to create a new plane in ClearView here:



ClearView models - Introduction (http://docs.google.com/Doc?id=dc9rgzk7_19cjf6wgcw)

Plane models by example - Part 1 (http://docs.google.com/Doc?id=dc9rgzk7_23n75shdg2)
Plane models by example - Part 2 ("http://docs.google.com/Doc?id=dc9rgzk7_50d96sjdfp)
Plane models by example - Part 3 ("http://docs.google.com/Doc?id=dc9rgzk7_63dmnm2sg4)
Helicopter models by example ("http://docs.google.com/Doc?id=dc9rgzk7_73hhr8c9f3)


Stefan
http://rcflightsim.com

[slipstick]

I wondered, are there any sims out there which run on a "true physics" model, in that they accurately model airflow?

That's how all modern sims work....with some variation in the value of "accurately" ;).

If you're asking if any of them are perfect then the answer is no. Even the multi-million dollar "sims" used by Boeing, Airbus etc still require flight testing and modifications after all the numbers have been crunched.

Steve

[Montag DP]

Hello,

I wondered, are there any sims out there which run on a "true physics" model, in that they accurately model airflow?

If so, my next question is, is it possible to import a model into any of them, based on inputting actual parameters from a real model?

Most I have seen seem to use a "virtual" aerodynamics model in which the various factors (drag/lift coefficient, induced drag, thrust, pitch speed etc) are tweaked until the model flies how the author wants it to.

What I feel would be better is a sim which can generate those values by modelling airflow over a 3D model. So, you could "build" a model in the sim, using something like 3DS/Gmax/Metaseqoia etc and the sim would use simulated airflow around that as the physical model.

Prop type, size etc could be input also, as could other variables.

The sim could use something akin to Motocalc in its modelling to acheive some of this perhaps?

Any thoughts?The only sim I know of that really does that (meaning, you can create a model with defined geometry and it will actually use aerodynamic theory to predict flow physics around the geometry) is X-Plane, but that's not a radio-controlled model simulator.

I think with all RC simulators you will need to do some tweaking of parameters in order to get it to fly realistically. Frankly, I think this is how it should be for this type of thing...as long as the game has a good physics engine that produces realistic behavior I don't mind tweaking some parameters. It's necessary due to the wide variety in sizes and types of models.

[skirtz]

I have doubts how good is X-Plane simulation is if we take into account the wider flight envelope the rc model fly.

With model simulators, you actually see how the model perform in a way that is directly comparable to a real rc model. Well trained RC pilots will very easily detect if a model in RC simulator does not perform as expected. RC model is used to do stunt flying and the flight envelope is very wide, including variety stall conditions. The simulated model is tested against all these conditions. Therefore, the criteria for "being close" is very high.

The simulators for real planes like X-Plane are tested and probably perform well in the limited envelope of taking off, flying with certain speed well above stall and landing. Any attempt to simulate rc plane in Microsoft flight simulator or X-Plane and do real world 3d flying will show how limited their simulation really is, because the simulated rc model will not fly at all as an rc model in most of the stall conditions or 3d flight maneuvers. Therefore, if the criteria for flight physics is the ability for a model to perform in all situations (like an rc model) and not in the limited flight envelope of a real plane, the flight physics for these simulators is rather poor.


What I try to say is, that the models in RC simulators actually fly very close to how the real models fly and they are A LOT BETTER AT THAT than the simulators for real planes like X-Plane. Obviously, X-Plane excels in many other areas that are not important for rc planes.

My 2 cents.

Stefan

[Montag DP]

I have doubts how good is X-Plane simulation is if we take into account the wider flight envelope the rc model fly.

With model simulators, you actually see how the model perform in a way that is directly comparable to a real rc model. Well trained RC pilots will very easily detect if a model in RC simulator does not perform as expected. RC model is used to do stunt flying and the flight envelope is very wide, including variety stall conditions. The simulated model is tested against all these conditions. Therefore, the criteria for "being close" is very high.

The simulators for real planes like X-Plane are tested and probably perform well in the limited envelope of taking off, flying with certain speed well above stall and landing. Any attempt to simulate rc plane in Microsoft flight simulator or X-Plane and do real world 3d flying will show how limited their simulation really is, because the simulated rc model will not fly at all as an rc model in most of the stall conditions or 3d flight maneuvers. Therefore, if the criteria for flight physics is the ability for a model to perform in all situations (like an rc model) and not in the limited flight envelope of a real plane, the flight physics for these simulators is rather poor.


What I try to say is, that the models in RC simulators actually fly very close to how the real models fly and they are A LOT BETTER AT THAT than the simulators for real planes like X-Plane. Obviously, X-Plane excels in many other areas that are not important for rc planes.

My 2 cents.

StefanI agree with you completely. X-Plane actually comes with a couple model-size planes and they don't fly realistically at all, even for simple takeoff, circuits, and landing. I assume that's because X-Plane's algorithms don't work very well for low reynolds numbers. Even if they did, I doubt it would be accurate across the wide flight envelope of models.