Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
Arnt Karlsen wrote: > ..well, all good jokes can't come up with a potential like the > http://gpgpu.org; your average recent GPU chews code 6 times > faster than your average CPU. So, we can use part of the GPU > to show pretty pictures, and the remainder, say "half", to say, > triple FG framerates? ;-) Well... it's not quite that great. If GPUs were better CPUs than CPUs, guess what we'd all be running for our performance-critical operations? Specifically, find me an implementation of string search (the fundamental operation in gene sequencing) that works on a GPU. Or index building (database stuff); or a big if()-tree based on complicated state (server-side business apps)... GPUs have dedicated hardware for the purpose of 3D rendering. Most importantly this includes: + Parallel, pipelined, single-precision 4D vector operations. A typical CPU has 2 or 3 floating point pipelines. A high end GPU can have dozens. + 1- 2- and 3D lookup tables (i.e. textures), with built-in interpolation/extrapolation (i.e. mipmapping & filtering). The memory bandwidth they can bring to bear on this stuff just crushes what a CPU can do. It also includes less generically useful stuff, like perspective-correct interpolation of vertex constants across a polygon. And it lacks a lot of stuff that CPU users take for granted: + High clock speeds. A CPU can dispatch single instructions 10x faster than the fastest GPU. This means that anything requiring quick decision making or iteration belong on the CPU. + Registers and cache. CPUs get special bonuses for iterating on the same data they were just using. GPUs get almost nothing (there's a little bit of caching that goes on within a single polygon's texture mapping, and a vertex cache of recently transformed vertices), and have to fetch it from main memory every single time it's used. + Turing completeness. Other than recent NVidia hardware, GPUs simply can't decide a lot of problems without CPU intervention anyway. + Double precision floating point math. Sometimes, single just ain't enough. And lots of others. Basically, this is all just a long winded way of saying "use the right tool for the job". Andy ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
Arnt Karlsen wrote: On Thu, 13 Jan 2005 15:05:26 +0100, Steven wrote in message <[EMAIL PROTECTED]>: Citeren "Ampere K. Hardraade" <[EMAIL PROTECTED]>: On January 13, 2005 07:28 am, Arnt Karlsen wrote: On Wed, 12 Jan 2005 21:30:23 -0500, Ampere wrote in message <[EMAIL PROTECTED]>: On January 12, 2005 08:18 pm, Arnt Karlsen wrote: ..these guys use GPU's as math engines? Why not? It makes sense. As a classmate of mine pointed out: the GPU is just a chip for matrix manipulations. ..aye. I'm just wondering how much work we can yank off the Control Process Unit and dump onto these General Purpose processing Unit_s_. ;-) I was referring to Graphics processing units. ..me 2. ;-) Am I missing something here? :-\ A joke ;-) (see the ";-)" smiley of Arnt) ..well, all good jokes can't come up with a potential like the http://gpgpu.org; your average recent GPU chews code 6 times faster than your average CPU. So, we can use part of the GPU to show pretty pictures, and the remainder, say "half", to say, triple FG framerates? ;-) Are there any decent books about those Navier-Stokes equations and how ..chk these: http://download.nvidia.com/developer/SDK/Individual_Samples/DEMOS\OpenGL\src\gpgpu_fluid\docs\GPU_Gems_Fluids_Chapter.pdf (Yeah, those back-slashes worked in Konqueror!!! 8o)) http://www.gpgpu.org/cgi-bin/blosxom.cgi/Miscellaneous/Books/ http://www.gpgpu.org/cgi-bin/blosxom.cgi/Miscellaneous/Courses/ http://www.gpgpu.org/cgi-bin/blosxom.cgi/Miscellaneous/ http://download.nvidia.com/developer/SDK/Individual_Samples/samples.html#gpgpu_fluid http://www.strangebunny.com/techdemo_stokes.php (arrrgh bummer, should have been done in OpenGL) http://www.eet.com/in_focus/silicon_engineering/showArticle.jhtml?articleID=55300900 http://www.eet.com/showArticle.jhtml?articleID=55300904 ..and you might wanna compare Microsofts take with the other guys: http://www.cs.unc.edu/Events/Conferences/GP2/program.shtml ;-) Thanks for the links! I'll look deeper into them after the exams. to implement them in C or java? ..nonono, FG is C++ and will likely remain so for all sortsa good reasons I was buried under when I asked "Why not C like in the kernel?", an extensive discussion of the wisdom of writing Java code can be found over at http://Groklaw.net/, hints on "follow the money", also points towards Redmond and Lindon, Utah. I knew that FG was written in C++ :-p. Problem is that I haven't been properly educated to program in C++ (yet). Steven ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
On January 13, 2005 09:27 am, Christian Mayer wrote: > The general wing geometry (i.e. the stuff you get from an 3-view) is one > thing. I don't use 3-view. > The the real profile of the wing is crucial here - and it's > AFAIK kept as an trade secret. That depends on how old the plane is: http://www.aae.uiuc.edu/m-selig/ads/coord_database.html Alternately, if you managed to get your hands on some technical documents on the aircraft, you may be able to get the information you need regarding the wing ribs' geometry. > Your models also won't have enough detail (= polygons) to reflect that > geometry in the needed detail (or they would be bad models for > visualisation...) Those models put into FlightGear definately don't have the detail required. But If the author keeps the original spline cage of the aircraft, it won't be difficult for he/she to come up with another model that does have enough polygons. However, it will be time consuming to make sure the model is completely one piece. Anyway, this may be useful for an engineering project of mine this term -- create a model glider. Will it be appropiate for me to ask for advice here? Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
On January 13, 2005 02:12 pm, Lee Elliott wrote: > Any particular reason you think the problem is just due to > laziness? No no, I was just trying to say that individual model can be made to be "watertight" if anyone decide to put one through those aerodynamic programs, and he/she will have to close the model manually instead of relying on some automated functions. No offense was intended. Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
On Thursday 13 January 2005 14:27, Christian Mayer wrote: > Ampere K. Hardraade schrieb: > > On January 12, 2005 06:07 pm, Christian Mayer wrote: > >>I see more problems with the correct shape of the wings. The > >> models won't get it right and using just some NACA profiles > >> won't work with the higly optimized profiles of modern > >> aircrafts (like those from Airbus). > > > > I am pretty confident that my models can make it through the > > program, since wing geometry is the first thing I look for. > > I don't know about others though. > > The general wing geometry (i.e. the stuff you get from an > 3-view) is one thing. The the real profile of the wing is > crucial here - and it's AFAIK kept as an trade secret. > > On modern aircraft (like all Airbus modells and the newer > Boeing ones) these are extremely optimized to be able to fly > at high speeds efficently. > > Your models also won't have enough detail (= polygons) to > reflect that geometry in the needed detail (or they would be > bad models for visualisation...) > > CU, > Christian I've thought about this for a few years now and although I had similar concerns about the accuracy of the aerofoil at first I don't think it's important now. If you supply a highly accurate (read high resolution) aerofoil definition to an FDM solver/engine it's going to result in a correspondly high processing requirement. As the aerofoil would only be required to establish the characteristics, or properties, of the flight surface, it might be better to just skip the aerofoil stuff and either supply those co-efficients directly, or generate them from other known characteristics instead. LeeE ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
On Thursday 13 January 2005 12:28, Arnt Karlsen wrote: > On Wed, 12 Jan 2005 21:30:23 -0500, Ampere wrote in message > > <[EMAIL PROTECTED]>: > > On January 12, 2005 08:18 pm, Arnt Karlsen wrote: > > > ..these guys use GPU's as math engines? > > > > Why not? It makes sense. > > > > As a classmate of mine pointed out: the GPU is just a chip > > for matrix manipulations. > > ..aye. I'm just wondering how much work we can yank off the > Control Process Unit and dump onto these General Purpose > processing Unit_s_. ;-) This reminds me of some stuff I read about dsp processors, about ten years ago. Nothing in depth but iirc the effective processing speeds were of the order of several GHz then, with corresponding flop rates. It certainly caused a raised an eyebrow when I read it:) LeeE ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
On Thursday 13 January 2005 02:22, Ampere K. Hardraade wrote: > On January 12, 2005 04:10 pm, Wolfram Kuss wrote: > > However, CFD programs need a "watertight" geometry. I would > > guess that far in excess of 90% of models are not. For > > example, each edge needs to have two neighbour faces. > > The models can be made to be "watertight". People just need > to get off their lazy behind and start creating/merging parts > poly by poly, vertex by vertex. > > > > Ampere I'm afraid that all model objects cannot be made 'water -tight', i.e turned into closed surfaces, at least not without adding lots of redundant faces, which would then need to be set to 100% transparency. This would increase both the surface culling overhead and probably cause further problems with transparency ordering. Examples would include canopies and virtual cockpits. How would you make a closed surface cockpit canopy without adding redundant surfaces to close it? I guess you could actually give it some thickness, and close it that way, but it would increase the face count x 3 (outside faces, edge faces & inside faces). It's similar with virtual cockpits i.e. the window frames and panel. After you've made the basic interior surfaces you'd then have add additional surfaces to close them at the back, but you would never want to see them. Any particular reason you think the problem is just due to laziness? LeeE ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
On Thu, 13 Jan 2005 15:05:26 +0100, Steven wrote in message <[EMAIL PROTECTED]>: > Citeren "Ampere K. Hardraade" <[EMAIL PROTECTED]>: > > > On January 13, 2005 07:28 am, Arnt Karlsen wrote: > > > On Wed, 12 Jan 2005 21:30:23 -0500, Ampere wrote in message > > > > > > <[EMAIL PROTECTED]>: > > > > On January 12, 2005 08:18 pm, Arnt Karlsen wrote: > > > > > ..these guys use GPU's as math engines? > > > > > > > > Why not? It makes sense. > > > > > > > > As a classmate of mine pointed out: the GPU is just a chip for > > > > matrix manipulations. > > > > > > ..aye. I'm just wondering how much work we can yank off the > > > Control Process Unit and dump onto these General Purpose > > > processing Unit_s_. ;-) > > I was referring to Graphics processing units. ..me 2. ;-) > > Am I missing something here? :-\ > > > > A joke ;-) (see the ";-)" smiley of Arnt) ..well, all good jokes can't come up with a potential like the http://gpgpu.org; your average recent GPU chews code 6 times faster than your average CPU. So, we can use part of the GPU to show pretty pictures, and the remainder, say "half", to say, triple FG framerates? ;-) > Are there any decent books about those Navier-Stokes equations and how ..chk these: http://download.nvidia.com/developer/SDK/Individual_Samples/DEMOS\OpenGL\src\gpgpu_fluid\docs\GPU_Gems_Fluids_Chapter.pdf (Yeah, those back-slashes worked in Konqueror!!! 8o)) http://www.gpgpu.org/cgi-bin/blosxom.cgi/Miscellaneous/Books/ http://www.gpgpu.org/cgi-bin/blosxom.cgi/Miscellaneous/Courses/ http://www.gpgpu.org/cgi-bin/blosxom.cgi/Miscellaneous/ http://download.nvidia.com/developer/SDK/Individual_Samples/samples.html#gpgpu_fluid http://www.strangebunny.com/techdemo_stokes.php (arrrgh bummer, should have been done in OpenGL) http://www.eet.com/in_focus/silicon_engineering/showArticle.jhtml?articleID=55300900 http://www.eet.com/showArticle.jhtml?articleID=55300904 ..and you might wanna compare Microsofts take with the other guys: http://www.cs.unc.edu/Events/Conferences/GP2/program.shtml ;-) > to implement them in C or java? ..nonono, FG is C++ and will likely remain so for all sortsa good reasons I was buried under when I asked "Why not C like in the kernel?", an extensive discussion of the wisdom of writing Java code can be found over at http://Groklaw.net/, hints on "follow the money", also points towards Redmond and Lindon, Utah. -- ..med vennlig hilsen = with Kind Regards from Arnt... ;-) ...with a number of polar bear hunters in his ancestry... Scenarios always come in sets of three: best case, worst case, and just in case. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
Andy Ross writes I'm not saying this stuff is impossible; people have been designing aircraft using CFD models for almost 20 years, and CPU cycles have never been cheaper. But it's a lot more work than just feeding c172.ac into a program and getting a working FDM configuration out the other side. Heaven help us if my 3D models are going to be depended on to produce the FDM.Then you will realy have the uncontrolable aircraft. :-) Andy Cheers Innis ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Steven Beeckman schrieb: > Are there any decent books about those Navier-Stokes equations and how > to implement them in C or java? One description of Navier Stokes are at: http://en.wikipedia.org/wiki/Navier-Stokes_equations there are also many many other webpages out there that tell you the equations. To solve them you need a good understanding of solving partial differential equations (PDE). If you can solve them analytical you'll be awarded with a big amount of money and extreme amounts of fame... So you can only solve them (generally) by numeric methods. So just grab any book about solving PDEs numerically and you can solve the Navier Stokes (at least in theory...) CU, Christian -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.0 (MingW32) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFB5pbWlhWtxOxWNFcRAqE1AJ9vU38NF4wLNQOZyTdnXozft0kregCguY4m M+nyYQeJDusZjajHuTRiV2Y= =ugBg -END PGP SIGNATURE- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
Wolfram Kuss wrote: > However, CFD programs need a "watertight" geometry. I would > guess that far in excess of 90% of models are not. For example, > each edge needs to have two neighbour faces. It's even worse than that. Real world aircraft performance is sensitive to all sorts of details that are simply unattainable for an amateur 3D model. Things like exact wing section shapes aren't available unless you have the original design plans and/or an actual aircraft to digitize. And I don't even want to think about the polygon counts involved in an accurate model. :) And remember that the aero modelling is still only part of the problem. You still need to get the mass distribution from somewhere, because it doesn't appear in the photographs. I'm not saying this stuff is impossible; people have been designing aircraft using CFD models for almost 20 years, and CPU cycles have never been cheaper. But it's a lot more work than just feeding c172.ac into a program and getting a working FDM configuration out the other side. Andy ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Ampere K. Hardraade schrieb: > On January 12, 2005 06:07 pm, Christian Mayer wrote: > >>I see more problems with the correct shape of the wings. The models >>won't get it right and using just some NACA profiles won't work with the >>higly optimized profiles of modern aircrafts (like those from Airbus). > > I am pretty confident that my models can make it through the program, since > wing geometry is the first thing I look for. I don't know about others > though. The general wing geometry (i.e. the stuff you get from an 3-view) is one thing. The the real profile of the wing is crucial here - and it's AFAIK kept as an trade secret. On modern aircraft (like all Airbus modells and the newer Boeing ones) these are extremely optimized to be able to fly at high speeds efficently. Your models also won't have enough detail (= polygons) to reflect that geometry in the needed detail (or they would be bad models for visualisation...) CU, Christian -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.0 (MingW32) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFB5oVTlhWtxOxWNFcRArHTAKCqyCcfujqd3y0auAvd4SLaqi4DpQCgvP3H JtvJHKHwhJ2qLJh2rpq2GDk= =gzG2 -END PGP SIGNATURE- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
Citeren "Ampere K. Hardraade" <[EMAIL PROTECTED]>: > On January 13, 2005 08:37 am, Ampere K. Hardraade wrote: > > I was referring to Graphics processing units. > > > > Am I missing something here? :-\ > > > > Ampere > Nevermind. I got confused again. > And I responded before I read this mail :/, so sorry for the first line of my previous reply ;-). Steven ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
Citeren "Ampere K. Hardraade" <[EMAIL PROTECTED]>: > On January 13, 2005 07:28 am, Arnt Karlsen wrote: > > On Wed, 12 Jan 2005 21:30:23 -0500, Ampere wrote in message > > > > <[EMAIL PROTECTED]>: > > > On January 12, 2005 08:18 pm, Arnt Karlsen wrote: > > > > ..these guys use GPU's as math engines? > > > > > > Why not? It makes sense. > > > > > > As a classmate of mine pointed out: the GPU is just a chip for > matrix > > > manipulations. > > > > ..aye. I'm just wondering how much work we can yank off the > > Control Process Unit and dump onto these General Purpose > > processing Unit_s_. ;-) > I was referring to Graphics processing units. > > Am I missing something here? :-\ > A joke ;-) (see the ";-)" smiley of Arnt) Are there any decent books about those Navier-Stokes equations and how to implement them in C or java? Greets, Steven ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
On January 13, 2005 08:37 am, Ampere K. Hardraade wrote: > I was referring to Graphics processing units. > > Am I missing something here? :-\ > > Ampere Nevermind. I got confused again. Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
On January 13, 2005 07:28 am, Arnt Karlsen wrote: > On Wed, 12 Jan 2005 21:30:23 -0500, Ampere wrote in message > > <[EMAIL PROTECTED]>: > > On January 12, 2005 08:18 pm, Arnt Karlsen wrote: > > > ..these guys use GPU's as math engines? > > > > Why not? It makes sense. > > > > As a classmate of mine pointed out: the GPU is just a chip for matrix > > manipulations. > > ..aye. I'm just wondering how much work we can yank off the > Control Process Unit and dump onto these General Purpose > processing Unit_s_. ;-) I was referring to Graphics processing units. Am I missing something here? :-\ Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
On Wed, 12 Jan 2005 21:30:23 -0500, Ampere wrote in message <[EMAIL PROTECTED]>: > On January 12, 2005 08:18 pm, Arnt Karlsen wrote: > > ..these guys use GPU's as math engines? > Why not? It makes sense. > > As a classmate of mine pointed out: the GPU is just a chip for matrix > manipulations. ..aye. I'm just wondering how much work we can yank off the Control Process Unit and dump onto these General Purpose processing Unit_s_. ;-) -- ..med vennlig hilsen = with Kind Regards from Arnt... ;-) ...with a number of polar bear hunters in his ancestry... Scenarios always come in sets of three: best case, worst case, and just in case. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
On January 12, 2005 08:18 pm, Arnt Karlsen wrote: > ..these guys use GPU's as math engines? Why not? It makes sense. As a classmate of mine pointed out: the GPU is just a chip for matrix manipulations. Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
On January 12, 2005 06:07 pm, Christian Mayer wrote: > I see more problems with the correct shape of the wings. The models > won't get it right and using just some NACA profiles won't work with the > higly optimized profiles of modern aircrafts (like those from Airbus). I am pretty confident that my models can make it through the program, since wing geometry is the first thing I look for. I don't know about others though. Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
On January 12, 2005 04:10 pm, Wolfram Kuss wrote: > However, CFD programs need a "watertight" geometry. I would guess that > far in excess of 90% of models are not. For example, each edge needs > to have two neighbour faces. The models can be made to be "watertight". People just need to get off their lazy behind and start creating/merging parts poly by poly, vertex by vertex. Ampere ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
On Wed, 12 Jan 2005 02:06:12 +0100, Oliver wrote in message <[EMAIL PROTECTED]>: > In a german news page (http://www.pro-linux.de/news/2005/7690.html) > i found an article about a software called OpenFOAM which was put > under the GPL license a few days ago and can do the following: > > "The OpenFOAM (Field Operation and Manipulation) software package can > simulate anything from complex fluid flows involving chemical > reactions, turbulence and heat transfer, to solid dynamics, > electromagnetics and the pricing of financial options." > > I read the word "turbulence" and thought that perhaps > this could be usefull somehow for flightgear or jsbsim but i am not > shure about that, so i mention it here maybe you know it better if > this could be somehow usefull for flightgear/jsbsim. > > Here's the website of that software: > http://www.opencfd.co.uk/openfoam/index.html ..dude, I can use this to model thermochemical gasification. ..these guys use GPU's as math engines? -- ..med vennlig hilsen = with Kind Regards from Arnt... ;-) ...with a number of polar bear hunters in his ancestry... Scenarios always come in sets of three: best case, worst case, and just in case. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
On Wed, 12 Jan 2005 16:01:59 + (UTC), Martin wrote in message <[EMAIL PROTECTED]>: > Christian Mayer wrote: > > > Nowadays you can solve navier stokes in real time in your graphics > > hardware... > > as long as you don't need your graphics hardware for other > purposes I think - like displaying FlightGear ;-) > > Martin. ..sure we need all of our gpu irons for that? ;-) On Tue, 11 Jan 2005 01:49:37 +0100, Arnt wrote in message <[EMAIL PROTECTED]>: > > > Hi > > > > ..another way to run code: http://gpgpu.org/ , for a wee quick > > intro, chk out: > > http://www.ece.ucdavis.edu/~jowens/talks/owens-hpec04-gpgpu.pdf > > > > ..note how they waaail for killer apps. ;-) > > > > .."formation flight": http://wwwx.cs.unc.edu/~tgamblin/gpgp/ ;-) > > ..this is well worth a look. > > > ...more gory details: http://gamma.cs.unc.edu/GPGP/ > > http://www.daimi.au.dk/~mosegard/GPGPU_E04 > > > > ..and, ahem, an app: http://flightgear.org/ ;-) > > > > ..so, if I cheat, by stuffing in 5 pci nVidea "math" cards beside > > my new 1xAGP 9250, my trusty 5 yr old AMD K6-2 450MHz can run > > in circles around anything on the market for another 5 years? ;-) > > > > ..since then, I found these (some links span 2 lines and have > spaces in their directory or link file names) to spoonfeed you: > http://www.eet.com/showArticle.jhtml?articleID=55300904 > http://gpgpu.org/cgi-bin/blosxom.cgi/ > Audio and Signal Processing/index.html > http://gpgpu.org/cgi-bin/blosxom.cgi/ > Advanced Rendering/index.html > http://gpgpu.org/cgi-bin/blosxom.cgi/ > Scientific Computing/index.html > http://gpgpu.org/cgi-bin/blosxom.cgi/ > High-Level Languages/index.html > http://gpgpu.org/cgi-bin/blosxom.cgi/Miscellaneous/ > Developer Resources/index.html > http://gpgpu.org/cgi-bin/blosxom.cgi/Miscellaneous/index.html > http://openvidia.sourceforge.net/ > http://www.strangebunny.com/techdemo_stokes.php > > ..note that the latter is useless to us as it requires DirectX 9. > > ..I would guess that if these tricks can pulled off on these cards, > stereo video can be done too, regardless of how Microsoft or > anyone else might feel about it: http://gamma.cs.unc.edu/GPGP/ > -- ..med vennlig hilsen = with Kind Regards from Arnt... ;-) ...with a number of polar bear hunters in his ancestry... Scenarios always come in sets of three: best case, worst case, and just in case. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Wolfram Kuss schrieb: > Erik wrote: > > >>This still might be useful if you can get all the moments and >>coefficients from it. Then you would be able to create a JSBSim >>configuration file from the model geometry. > > > The idea of using the gfx model you need to do anyone (or one of the > thousands or ten thousands you find on the internet) and automatically > get the config file. It would not matter if it takes over night or > even if it takes a week. > > However, CFD programs need a "watertight" geometry. I would guess that > far in excess of 90% of models are not. For example, each edge needs > to have two neighbour faces. As you are only interested in the full shape of the plane (and not it's single parts) you could help yourself by automatically closing these holes. I see more problems with the correct shape of the wings. The models won't get it right and using just some NACA profiles won't work with the higly optimized profiles of modern aircrafts (like those from Airbus). CU, Christian -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.0 (MingW32) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFB5a3NlhWtxOxWNFcRAg9MAJ9Lo8XRiNpaNSGmdU9XHC0TUIWzsACcColJ ejqhd5nYnsk1zpPOO4EbLVU= =ixWC -END PGP SIGNATURE- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
On Wed, 12 Jan 2005 22:10:47 +0100 [EMAIL PROTECTED] (Wolfram Kuss) wrote: Erik wrote: This still might be useful if you can get all the moments and coefficients from it. Then you would be able to create a JSBSim configuration file from the model geometry. The idea of using the gfx model you need to do anyone (or one of the thousands or ten thousands you find on the internet) and automatically get the config file. It would not matter if it takes over night or even if it takes a week. However, CFD programs need a "watertight" geometry. I would guess that far in excess of 90% of models are not. For example, each edge needs to have two neighbour faces. There's always DATCOM+: www.holycows.net/datcom Jon ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
Erik wrote: >This still might be useful if you can get all the moments and >coefficients from it. Then you would be able to create a JSBSim >configuration file from the model geometry. The idea of using the gfx model you need to do anyone (or one of the thousands or ten thousands you find on the internet) and automatically get the config file. It would not matter if it takes over night or even if it takes a week. However, CFD programs need a "watertight" geometry. I would guess that far in excess of 90% of models are not. For example, each edge needs to have two neighbour faces. >Erik Bye bye, Wolfram. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
Christian Mayer wrote: > Nowadays you can solve navier stokes in real time in your graphics > hardware... as long as you don't need your graphics hardware for other purposes I think - like displaying FlightGear ;-) Martin. -- Unix _IS_ user friendly - it's just selective about who its friends are ! -- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Martin Spott schrieb: > "Oliver C." wrote: > > >>I read the word "turbulence" and thought that perhaps >>this could be usefull somehow for flightgear or jsbsim but i am not shure >>about that, so i mention it here maybe you know it better if this could be >>somehow usefull for flightgear/jsbsim. > > > Simulation of fluid dynamices the way I got to know it (in iron casting > processes) typically means that you define a model plus the fluid > environment, start the solver and come back the next day. This is not > necessarily the usual application with FlightGear - but still nice to > know about that project. Nowadays you can solve navier stokes in real time in your graphics hardware... It's really impressive! But the precision isn't good enough for real engineering tasks - yet. CU, Christian -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.0 (MingW32) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFB5TyjlhWtxOxWNFcRAkwIAJ97KQW5UKbBVqAX7aOVq7fwfQOGrgCdFijY 1fXpZl8TC6csUx9PeTpHKWE= =2g0b -END PGP SIGNATURE- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
Martin Spott wrote: "Oliver C." wrote: I read the word "turbulence" and thought that perhaps this could be usefull somehow for flightgear or jsbsim but i am not shure about that, so i mention it here maybe you know it better if this could be somehow usefull for flightgear/jsbsim. Simulation of fluid dynamices the way I got to know it (in iron casting processes) typically means that you define a model plus the fluid environment, start the solver and come back the next day. This is not necessarily the usual application with FlightGear - but still nice to know about that project. This still might be useful if you can get all the moments and coefficients from it. Then you would be able to create a JSBSim configuration file from the model geometry. Erik ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Is this usefull for flightgear/jsbsim?
"Oliver C." wrote: > I read the word "turbulence" and thought that perhaps > this could be usefull somehow for flightgear or jsbsim but i am not shure > about that, so i mention it here maybe you know it better if this could be > somehow usefull for flightgear/jsbsim. Simulation of fluid dynamices the way I got to know it (in iron casting processes) typically means that you define a model plus the fluid environment, start the solver and come back the next day. This is not necessarily the usual application with FlightGear - but still nice to know about that project. Thanks, Martin. -- Unix _IS_ user friendly - it's just selective about who its friends are ! -- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
[Flightgear-devel] Is this usefull for flightgear/jsbsim?
In a german news page (http://www.pro-linux.de/news/2005/7690.html) i found an article about a software called OpenFOAM which was put under the GPL license a few days ago and can do the following: "The OpenFOAM (Field Operation and Manipulation) software package can simulate anything from complex fluid flows involving chemical reactions, turbulence and heat transfer, to solid dynamics, electromagnetics and the pricing of financial options." I read the word "turbulence" and thought that perhaps this could be usefull somehow for flightgear or jsbsim but i am not shure about that, so i mention it here maybe you know it better if this could be somehow usefull for flightgear/jsbsim. Here's the website of that software: http://www.opencfd.co.uk/openfoam/index.html Best Regards, Oliver C. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d