cifre Posted February 7, 2021 Share Posted February 7, 2021 This is my first post in the forum, so first things first: Hello everyone!! Down to business: I'm trying to create a custom 3d vector field from a set of imported lines so I can drop some particle tracers on it. I managed to convert the imported lines to vectors using the orient along curve node, and I also created some voxels through bound>volume (ideally one for each of my imported lines, although I'm not sure I succeded). Now, what I want to do, is to link each one of my custom vectors to it's correspondent voxel, so that when a particle is inside that voxel is affected only by that particular vector. This is where I get lost, I don't know how to link my vectors to my voxels... I tried importing my "vec" point attribute inside my volumevop but for the life of me I can't seem to figure it out. I come from Rhino+Grasshopper, so I thought I'd be a bit familiar with the node workflow and all but... wow... Switching to Houdini feels more like stepping off a meek pony and jumping right on top of an untamed pissed off stallion. I'm attaching my files. Any help taming this beast will be much appreciated (Probeta_02_vF_S.igs is the file containing the custom lines) Custom_vField.zip Quote Link to comment Share on other sites More sharing options...
Nicholas Ralabate Posted February 8, 2021 Share Posted February 8, 2021 (edited) Maybe give Volume Velocity from Curves SOP a shot! That will rasterize the curves into a volume for you, then you can just sample that volume with a POP Advect by Volumes DOP in your particle system. This is most often done with velocity vectors, which heads up is sometimes confusingly called v in SOPs land and vel in DOPs land. If you want to instead continue with your current approach, Volume Rasterize Attributes SOP is how you rasterize point/line geometry -- either velocity, or actually, any attribute -- into a volume: https://www.sidefx.com/docs/houdini/nodes/sop/volumevelocityfromcurves.html https://www.sidefx.com/docs/houdini/nodes/dop/popadvectbyvolumes.html https://www.sidefx.com/docs/houdini/nodes/sop/volumerasterizeattributes.html Edited February 8, 2021 by Nicholas Ralabate Quote Link to comment Share on other sites More sharing options...
Librarian Posted February 8, 2021 Share Posted February 8, 2021 Maybe it Helps @cifre FODPAR.hiplc Quote Link to comment Share on other sites More sharing options...
cifre Posted February 9, 2021 Author Share Posted February 9, 2021 Thanks a lot @Nicholas Ralabate! I tried the Volume Rasterize Attributes aproach and I think I got it working. I will have to do some debugging to make sure the resultant field accurately follows my imported lines, but at first glance everything looks good. I also tried the Volume Velocity from Curves SOP approach but I couldn't figure out how to make it work, although it looks like the correct and easiest way of doing what I'm doing, so I will definetely look more into that. Thanks for the file @Librarian, downloading right now to study it. I got a say I was super surprised of how fast you both answered, I thought it would take me at least a week to hear back from anyone. Thanks! Quote Link to comment Share on other sites More sharing options...
Librarian Posted February 9, 2021 Share Posted February 9, 2021 @cifre Just play with the File Have Fun ...Maybe it Helps who Knows Quote Link to comment Share on other sites More sharing options...
cifre Posted February 17, 2022 Author Share Posted February 17, 2022 Hey @Librarian and @Nicholas Ralabate, I just realized I never got back to you on this. How rude of me! Anyway, I thought you'd like to see some of the results that you helped me create! The craziest thing about these is that the lines that I used to create the vector fields in Houdini were the result of a structural stress calculation algorithm using finite element method in Rhinoceros, so the particles in Houdini are actually tracing the stress lines inside the geometry for a given structural load. Which in plain english means that the crazy geometry that you see in the images actually makes sense structurally, is supposed to be the most efficient yet resistant expression of the object. God I sound like a pretentious twat. Well, just wanted to thank you for your help! Good karma to you both ❤ 2 Quote Link to comment Share on other sites More sharing options...
Librarian Posted February 18, 2022 Share Posted February 18, 2022 (edited) @cifre you have tutos on that Topic with entagma(make groups on Pin and Force-rest- (than solver-inside you have to play with ) transfer stress to prim after just smooth ..+ Asia Siggraph 2018 ++you have files on Indie Asia .some of them .Have Fun... Edited February 18, 2022 by Librarian Quote Link to comment Share on other sites More sharing options...
sebkaine Posted February 18, 2022 Share Posted February 18, 2022 i did a small test, i try to keep it simple. for each might not be elegant, but it give a simple setup. hope that help. cheers E trailField_simple.hipnc Quote Link to comment Share on other sites More sharing options...
cifre Posted February 18, 2022 Author Share Posted February 18, 2022 @Librarian Looks super interesting. Can you please post links? I can't seem to find any of the resources you were mentioning. @sebkaine I get an error in your File Cache node: Unable to read file "C:/Users/Cifre/Desktop/geo/trailField_simple.ptc_cache.bgeo.sc". Maybe you also need to attach that file? Quote Link to comment Share on other sites More sharing options...
sebkaine Posted February 18, 2022 Share Posted February 18, 2022 (edited) you just have to regenerate the cache by hiting save to disk / or uncheck the toggle load from disk. you may also need to reset the DOP cache. i have modify the file to invert vel and add fading tail which looks better imo. trailField_simple_fade.hipnc Edited February 19, 2022 by sebkaine Quote Link to comment Share on other sites More sharing options...
cifre Posted February 19, 2022 Author Share Posted February 19, 2022 WOOOAAAAAAH! That looks f... awesome Sebkaine!! Quote Link to comment Share on other sites More sharing options...
cifre Posted February 21, 2022 Author Share Posted February 21, 2022 I just bumped into this picture, and while my approach has the theoretical benefit of topology optimization, when compared to this one it kinda feels chaotic and disorganized. If anyone has any input on how they'd go about achieving a result a bit more similar to the reference picture I'd really appreciate it. Quote Link to comment Share on other sites More sharing options...
sebkaine Posted February 21, 2022 Share Posted February 21, 2022 (edited) https://deskriptiv.com/ https://deskriptiv.com/publications https://oxman.com/projects/data-driven-material-modeling check this they have some in depth pdf Edited February 21, 2022 by sebkaine 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.