Running Simulation...

Replies to This Discussion

Permalink Reply by Arthur van der Harten on October 12, 2021 at 7:33pm

Hi Tarek,

I can't  comment on your specific case because I don't know anything about your model, but unrealistically low absorption coefficients will tend to run a model for a very long time. (There may be other issues with your geometry, but this is usually the issue people run into.)

Rays will lose energy much more slowly when a room is very reflective, but such conditions only really exist with polished painted concrete. I don't know how many such rooms you've ever been in. I can count them on one hand - and I've been to acoustic reverberation chambers on 3 continents.

Pachyderm works with a ray decimation algorithm. This means that a ray will be traced out to the cutoff time, unless it has lost enough energy that its further contribution to the impulse response is negligible. At this time, the ray is discontinued. So, the more absorption in the room, the less time it will take to run the simulation. So, more realistic absorption coefficients will allow a model to run in a more reasonable amount of time.

Gypsum board/Plasterboard will typically be between 8 and 11 percent at mid and high frequencies, and higher at low frequencies, for example. Concrete, unless polished and painted, is still a little porous, so you can probably make it 3 to 5 percent.

Regarding sources and receivers, adding more receivers adds almost nothing to calculation time (until the every end, when it extrapolates a pressure response, anyway). This is because the receivers are entered into a spatial partition system, which allows it to be done in an optimized way. Adding multiple sources will effectively multiply the calculation time by the number of sources, because the simulation needs to be repeated for each source individually.

Hopefully this helps you figure out how to run simulations with reasonable run times.

Arthur