Looking for Guidance on TFD Workflow Optimisation for Complex Simulations

[Joy_Nicholes]

Hello Everyone ,

I’ve been using TurbulenceFD (TFD) for a while now, and I adore the features it provides. But as I get farther into the simulations, I’m finding it harder and harder to strike a compromise between performance optimisation and producing high-quality findings.

Large-scale fire and smoke simulations are a part of the project I’m working on right now, and I’m realising that my process could use some tweaking to handle these bigger situations more effectively.

Here are some particular places where I could benefit from guidance:

Cache Management: The size of the cache files might increase significantly when working with high-resolution simulations. What are some practical methods for controlling cache size without making too many quality compromises? Do you advocate any particular TFD settings or compression methods?

Grid Optimisation for Voxels: I’ve read that performance can be greatly affected by voxel grid optimisation. How can one strike the ideal balance between simulation detail and grid resolution? Any advice on how to properly set up adaptable grids?

Hardware Points to Remember: Although my system, which runs TFD, is fairly powerful, I’m wondering whether there are any particular hardware choices or upgrades that could improve performance even further. When selecting GPUs, CPUs, or storage options for TFD, is there anything specific I should be looking for?

Tab button not working on Macos

Hello! I’ve had this issue for a long while on the Mac, on different versions of Macos and C4D. Was hoping that the latest TFD update for Macos would solve the issue, but it hasn’t. I use tab all the time to move from one parameter box to the next to change values. But for some reason it doesn’t work for TFD parameters. So for example: I can type in a voxel size value, the click tab to get to the next parameter (grid size X). But when I type anything, I just get a warning sound, and I can’t c…

General Workflow Advice: When handling intricate TFD simulations, are there any other general guidelines or best practices that you’ve discovered to be beneficial? I’m open to recommendations for settings, plugins, gcp, or outside tools.

Thank you in advance.

[Paul_Selhi]

For a start on GPU choice you want one that has at least 12 G and memory throughput is the bottleneck…get a card with the fastest mem throughput.
For emitters do not use your full objects but proxies where the emission will come from, maybe even simply single poly planes, discs etc. On fast moving objects increase your subsets Only cache velocity if you need it for motion blur or particle advection or upresing. Sim and test at low res to get an overall feel for your sim and then either increase the res which will change the sim or upres. In R26 you can avoid conversion to vdb by using Arnold direct on the TFD container…this does not seem to want to work in 2023. Have a look at my tutorial set that uses the standard advanced render TurbulenceFD (TFD) For Cinema 4D Fundamentals | Software | Training

[Paul_Selhi]

Also don’t do what I have a habit of doing and that is push the sim to the limit of my GPU RAM…more often than not the level of detail is not needed especially when using motion blur.