Small scale FEM

[Jordan Walsh]

Just looking at the FEM solver in 15

I get nice results on a 1x1x1m cube from the shelf but I'm wanting to do some fx that are more of the scale of 1-2 cm cubes.

Im having some issues getting similar results when scaling down the geo. Anyone know how to sort this out? 

Thanks!

 

[Farmfield]

Yeah, this is probably one of those situations where simulating at the correct scale might be impossible and you have to sim it at a larger scale, tweak the time factor to get it looking the right scale... You know the drill, in the end it's all about how you/director/producer want it to look, not how it would look if it was real.

Edited December 9, 2015 by Farmfield

[ch3]

On a similar topic, I did a few tests a while ago, trying out the different scales. Even if you change the units in the hip file options, prior to creating any DOP nodes, you may still have to adjust a few things to get the correct dynamics.

For example, going from meters to centimeters, you need to divide the density of any preset by 1,000,000, as well as the tangent drag. You probably also need more substeps or change the tolerance levels to make the simulation stable.

 

The simple test I am always doing to make sure my scale settings are set correctly, is to drop something from a certain height and measure how long it takes to hit the ground without any air drag or resistance.

 

Based on the acceleration formula, that's how long it should take for an object to hit the ground:

 

1m        0.45s = 10.8f @24fps
10m      1.43s = 34.3f @24fps
100m    4.52s =108.5f @24fps

 

 

I am still trying to work out some of these things my self, so it would be nice to let us know any of your findings.

 

thanks

georgios

Edited December 16, 2015 by ch3