Flight Question What if Orbiter simulated aerodynamics based on the shape of your vehicle?

[Pipcard]

Because right now, you can only define the flight model separately and manually in your .dll or spacecraft3 .cfg.

 

[SolarLiner]

Even X-Plane does not simulate aerodynamism based on the shape of the model. Because that would be too expensive to calculate.
Instead (IIRC), you povide a simplified model with only the most important parts (fuselage, wings, ailerons), and then define what's what in the configuration files.

So technically you'd still have to configure stuff, and basically with Orbiter when defining your ship's aerodynamics you are reproducing that basic aircraft model, defining the fuselage and the wings, as well as the moving surfaces.

 

[N_Molson]

Keep this for Orbiter 2100 Quantum Physics edition :lol:

 

[Pipcard]

I'm not currently developing the cancelled Japanese "Fuji" space capsule, but if someone wanted to, how would someone define the aerodynamics for it? Because it doesn't have wings or anything.

"...a very flat shape that resembled an Apollo Command Module that had somehow got itself run over by a steamroller. It’s also reminiscent of the lenticular (read: saucer-shaped) re-entry vehicles considered in the early days of American spaceflight... The strange shape was driven by one of Fuji’s design goals... By re-entering at a high angle of attack the Core Module could generate lift..."

 

[Urwumpe]

Also, even X-Plane does not simulate the aerodynamics accuratly - it is OK below Mach 0.5, but gets increasingly inaccurate until Mach .95... something. From transonic flight on, it collapses and pretends to provide accurate values.

A space shuttle in X-plane would not work out well.

If aerodynamics would be really that simple... we would not need wind tunnels and number crunchers for it.

 

[SolarLiner]

A space shuttle in X-plane would not work out well.

And yet, Endeavour is there, ready to be piloted. :lol:

 

[Urwumpe]

And yet, Endeavour is there, ready to be piloted. :lol:

Yes, but it will likely fare as well during a realistic reentry scenario as a Space Shuttle in Orbiter... I still remember why both SSU and Shuttle Fleet got their aerodynamics upgraded. :lol:

 

[boogabooga]

Essentially, you are asking Orbiter to be computational fluid dynamics software.

It is a lot more complicated than you think. The Navier-Stokes partial differential equations do not have known closed-form solutions, so they are solved numerically using various finite difference techniques (including finite volume and less commonly finite element).

Long story short, it can takes hours to days to get a converged solution on a healthy system. Nevermind trying to render a 3D environment at the same time. It is not something Orbiter can do on the fly.

You can learn to use such software, if you wanted. (I think CFD is fun.) Either shell out over $1000 for a professional level commercial suit (Fluent and CFX are the most common that I am aware of), or try to get free software on the internet (OpenFoam, but you will need to have Linux). Essentially, you would use the data from the CFD simulations to get the numbers to use in Orbiter's config files.

Also, don't think that you can take a 3D model from Orbiter and use it for the CFD simulations. The grid needs to be entirely different. For one it needs to be the negative space- the space around your object and not the object itself. Also, special structured or unstructured grids need to be used and the quality and resolution will greatly affect the accuracy of the simulation.

It might be possible to take a 3D model from Orbiter and use it to generate the CFD grid. I know for sure that you can use Solidworks models to generate CFD grids, but I'm not sure about Orbiter's file types, which are gaming models and not engineering models. Plus it may be difficult to find the software for free.

The thought of doing CFD on Orbiter craft is intriguing, though.

 

[Dantassii]

Essentially, you are asking Orbiter to be computational fluid dynamics software.

It is a lot more complicated than you think. The Navier-Stokes partial differential equations do not have known closed-form solutions, so they are solved numerically using various finite difference techniques (including finite volume and less commonly finite element).

Long story short, it can takes hours to days to get a converged solution on a healthy system. Nevermind trying to render a 3D environment at the same time. It is not something Orbiter can do on the fly.

You can learn to use such software, if you wanted. (I think CFD is fun.) Either shell out over $1000 for a professional level commercial suit (Fluent and CFX are the most common that I am aware of), or try to get free software on the internet (OpenFoam, but you will need to have Linux). Essentially, you would use the data from the CFD simulations to get the numbers to use in Orbiter's config files.

Also, don't think that you can take a 3D model from Orbiter and use it for the CFD simulations. The grid needs to be entirely different. For one it needs to be the negative space- the space around your object and not the object itself. Also, special structured or unstructured grids need to be used and the quality and resolution will greatly affect the accuracy of the simulation.

It might be possible to take a 3D model from Orbiter and use it to generate the CFD grid. I know for sure that you can use Solidworks models to generate CFD grids, but I'm not sure about Orbiter's file types, which are gaming models and not engineering models. Plus it may be difficult to find the software for free.

The thought of doing CFD on Orbiter craft is intriguing, though.

I also have done some CFD work while I was working on my Ph.D. in Aerospace Engineering. One of my projects at the time was trying to find a way to reduce the number of grids required to model the flow over the nose part of an F/A-18 fighter jet. Not the whole jet, just the nose section up to the edge of the cockpit bubble.

The 'current' technology used at that time (ca 1995) required over 10 billion grids, about 100 Gb of RAM, and about 2-3 weeks of run time on a CRAY-XMP super computer to get a steady state solution for a single moment in time that was even remotely close to the experimental data obtained from a heavily instrumented F/A-18 model in a wind tunnel.

Although computers today are much faster, the size of the grid mesh has not gone down at all. My Ph.D. Project (which I never got to finish) was an attempt to reduce the number of grids required down by at least 5 orders of magnitude and possibly as many as 8 orders of magnitude using an unusual (at that time) algorithm. Sadly, I was booted out of the program before I could complete my work due to a Comprehensive Math exam that very few USA engineering students ever passed. I later found out that 3 Chinese grad students tried to pick up where I left off, but were unable to apply the algorithm correctly. After another 3 years trying to figure out what I had done, they gave up on trying to get it to work.

In any case, real-time use of CFD solutions for aerodynamics over a vehicle in a flight simulator program (especially in atmospheres other than Earth's) will probably have to wait until the Quantum computer technology works its way into home computers.

Dantassii
HUMONGOUS IMS shipbuilder

 

[Urwumpe]

The 'current' technology used at that time (ca 1995) required over 10 billion grids, about 100 Gb of RAM

Did not get much better - a tiny CFD simulation here has 60 million volume elements.

But still, it depends on the accuracy. You would also get a solution if you only use maybe 250,000 elements. It just would have a severe lack in accuracy, not something that you would want to enjoy in real life.

But even 25 million elements would still be in the range that you can calculate a steady-state solution on a modern computer (Xeon + OpenCL) in around 6 hours. A full set of sample data for creating a lift function for Orbiter would be possible in a month.

And a satellite aerodynamics solution (for high Mach numbers and altitude > 100 km), you could even do it in real-time - these aerodynamics are already accurate if you simulate tiny exosphere atoms bouncing off the hull of the spacecraft, no volume needed in first or second order accuracy.

 

[jedidia]

What if Orbiter simulated aerodynamics based on the shape of your vehicle?

... Most of our add-ons would crash and burn?

 

[C3PO]

On the last STS mission NASA still had special tiles installed to disturb the boundary layer in order to test their CFD model. Hypersonic CFD is not 100% accurate yet.

 

[malisle]

Would it be possible to precompute the parameters based on the shape of the aircraft (including variations of AOA, speed, angles of control surfaces etc.) and then use those parameters is .dll?

Few days of computing would probably be acceptable to most addon developers that want somewhat realistic aerodynamics.

 

[Urwumpe]

Would it be possible to precompute the parameters based on the shape of the aircraft (including variations of AOA, speed, angles of control surfaces etc.) and then use those parameters is .dll?

Few days of computing would probably be acceptable to most addon developers that want somewhat realistic aerodynamics.

Sure. You could even write a program that starts the simulation jobs, analyses the results and generates C++ code for Orbiter.

But that's a lot of software behind.

 

[Cizurator]

Huh, just imagine how complicated and maybe even expensive would development of addons become!

 

[Urwumpe]

Huh, just imagine how complicated and maybe even expensive would development of addons become!

Well, don't you think it is a bit annoying that everything in Orbiter flies like a DG? :lol:

A simplified aerodynamics editor would maybe be really a nice tool. Not a CFD simulation... but maybe some "combine some shapes and calculate the coarse aerodynamic profile" style editor.

 

[boogabooga]

Well, don't you think it is a bit annoying that everything in Orbiter flies like a DG? :lol:

A simplified aerodynamics editor would maybe be really a nice tool. Not a CFD simulation... but maybe some "combine some shapes and calculate the coarse aerodynamic profile" style editor.

So, Kerbal Space Program?

 

[BruceJohnJennerLawso]

Well, don't you think it is a bit annoying that everything in Orbiter flies like a DG? :lol:

A simplified aerodynamics editor would maybe be really a nice tool. Not a CFD simulation... but maybe some "combine some shapes and calculate the coarse aerodynamic profile" style editor.

What exactly is the current aerodynamics model? Does Orbiter just model vessels as spheres, then augment it with aerodynamic surfaces if they are defined, or is it more subtle.

 

[Thunder Chicken]

Orbiter simply takes the projected areas of your objects and applies a drag coefficient that is usually a function of Mach number. It does the same with wings, applying lift and moment coefficients. These coefficients are all empirical, you have to go dig through literature to find test data for a body similar in shape to your vessel.

In Orbiter, you can actually take a mesh in the shape of a brick and write a dll to give it the flight qualities of a glider. You'll find that a lot of add-ons fly exactly like the DeltaGlider, because they were made by someone with great graphics skills but not such a great understanding of the aerodynamic coding requirements.

I recall a couple of years ago, maybe back on M6, that someone was working on a compressible flow solver that would accept a *.msh file, mesh it and determine the lift, drag, and moment coefficients for use in .dlls. I forget who it was, but I worked with them about the numerics. They got it running after a fashion, but it was not polished by any means.

I used to do computational fluid modeling when I was in industry. It really is computationally intensive, and the code really needs to be run by someone who has an understanding of the underlying numerical methods and fluid physics or else you will get wrong answers (that can look correct!) or NaNs. Compressibility, turbulence modeling, heat transfer, variable densities...these are all modeled differently and one needs to know in which situation which model is appropriate.

Much of my consulting work involved walking into a low-budget engineering firms and yanking the CFD software away from the fresh-out-of-college engineer assigned to use it who couldn't solve anything with it. Many companies think because they have a CFD license that they know what they are doing. There are many companies bleeding R&D money because they truly don't know what they don't know.

 

[Dantassii]

...There are many companies bleeding R&D money because they truly don't know what they don't know.

Actually, its impossible for them to have a clue what they don't know because they think they know everything. They also believe that anyone trying to tell them otherwise obviously doesn't have any idea what they are talking about.



Dantassii
HUMONGOUS IMS shipbuilder