Guest xionmark Posted November 9, 2005 Share Posted November 9, 2005 Hi, Wasn't there a SIM_CFD class in the beta H8 HDK? I can't seem to find it now ... has it been replaced? If so, by what? --Mark Quote Link to comment Share on other sites More sharing options...
Wolfwood Posted November 9, 2005 Share Posted November 9, 2005 Hi,Wasn't there a SIM_CFD class in the beta H8 HDK? I can't seem to find it now ... has it been replaced? If so, by what? --Mark 22479[/snapback] I remember there being a SIM_CFD class as a sample in $HT/samples/SIM but it was so out dated that it got removed. In its place they've got a Snow and Hair Solver as examples. Quote Link to comment Share on other sites More sharing options...
Guest xionmark Posted November 9, 2005 Share Posted November 9, 2005 I remember there being a SIM_CFD class as a sample in $HT/samples/SIM but it was so out dated that it got removed. In its place they've got a Snow and Hair Solver as examples. 22480[/snapback] Oh? Was the SIM_CFD class derived from something else (already in the HDK)? Hmm, I'll give the hair and snow examples a quick look. Thanks, Mark Quote Link to comment Share on other sites More sharing options...
Jason Posted November 9, 2005 Share Posted November 9, 2005 Oh? Was the SIM_CFD class derived from something else (already in the HDK)? Hmm, I'll give the hair and snow examples a quick look.Thanks, Mark 22481[/snapback] I remember the 2D Jos Stam solver was part of the examples. Quote Link to comment Share on other sites More sharing options...
Guest xionmark Posted November 9, 2005 Share Posted November 9, 2005 I remember the 2D Jos Stam solver was part of the examples. 22483[/snapback] Really? I was just starting on writing that myself, except in 3D. I suppose I should look through the earlier beta distros ... Thanks Jason. --Mark Quote Link to comment Share on other sites More sharing options...
mtucker Posted November 9, 2005 Share Posted November 9, 2005 Yes, all the real work in SIM_CFD was done by code written by Jos Stam, copied and completely unmodified from a Siggraph paper he published ages ago. It was just a proof of concept and an attempt to show how easy it could be to take an existing solver (written in straight C no less) and turn it into a DOP solver. It was removed before we released to the public because we didn't feel right shipping someone else's code with Houdini (even if it was code in the public domain). Besides, it was just a 2D solver, didn't respond to forces, didn't let you control density birthing, didn't respond to other objects, etc. So it wasn't terribly useful. I very much look forward to hearing how your effort goes, and to seeing the code on the odforce sourceforge site if it's not proprietary... Mark Quote Link to comment Share on other sites More sharing options...
Jason Posted November 9, 2005 Share Posted November 9, 2005 I still maintain that having the 2D solver exsistent in a 3D context is still very useful. I'd love to see that Jos Stam solver that responds to (the 2D projection of) 3D forces and allows density to be added during the simulation and so on. I keep picturing a vase or a glass with 2d bump-mapped droplets (the density) dripping down the sides. IIRC, 2d simulation was used pretty effectively as displacement mapped ripples on puddles on Chicken Little. Nice:) Quote Link to comment Share on other sites More sharing options...
mtucker Posted November 11, 2005 Share Posted November 11, 2005 For anyone who is interested, I just uploaded the source (and a compiled version for Windows) of the old-timey 2D CFD solver to the Houdini Exchange (http://www.sidefx.com/index.php?option=com_wrapper&Itemid=8). So now you can all see just how bad it really is But you are encouraged to make it better... I tried adding it to the odforce sourceforge page, but I couldn't connect through CVS to add the source. So for now, if anyone has any improvements they want to share, I guess they can be posted here or to the exchange. Mark Quote Link to comment Share on other sites More sharing options...
Jason Posted November 11, 2005 Share Posted November 11, 2005 Thanks Mr. Tucker! Quote Link to comment Share on other sites More sharing options...
mtucker Posted November 11, 2005 Share Posted November 11, 2005 And if anyone would like to try something similar, try downloading http://www.cs.ubc.ca/~rbridson/download/simple_flip2d.tar.gz and make a DOP Solver out of that. It took me about 15 minutes to compile the test application, convert the output files to .chan format, and use a File CHOP and Channel SOP to visualize the simulation in Houdini. Making it a DOP Solver should only take a few hours... Mark Quote Link to comment Share on other sites More sharing options...
Guest xionmark Posted November 11, 2005 Share Posted November 11, 2005 And if anyone would like to try something similar, try downloading http://www.cs.ubc.ca/~rbridson/download/simple_flip2d.tar.gz and make a DOP Solver out of that. It took me about 15 minutes to compile the test application, convert the output files to .chan format, and use a File CHOP and Channel SOP to visualize the simulation in Houdini. Making it a DOP Solver should only take a few hours...Mark 22550[/snapback] Hey Mark, Thanks for posting this! --Mark Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.