SIM Groups (different results when using them)

[Alexx31]

I have a scene, where I want to split the flow randomly into three SIM-groups (1,2,3)
and I spent again 2 hours trying to understand the unpredictabe behaviour of SIM groups

So now I have a logical problem, that the simulation differs extremely, when doing exactly the same with using SIM groups or doing just the same without them.

In the scene is only a "Find Target" Operators with Filtering enabled or disabled for SIM groups, and all3 setups should give the same result, but they look very different:
- the first is my simulation without any usage of sim groups
- the second screenshot is my simulation setting the flow to SIM1 and enable Filter for SIM 1 (which is everything in the flow)
and should not change anything, because like before everything is addressed
- the third screenshot is my (original) setup, where I wanted to split the flow randomly into three SIM groups 1,2 and 3.
And again it looks even more different, despite it is the same flow with everything.

What's here the reason?
Thanks for explanation!

p.s. Also to be sure to understand correctly:
I understand that "Simulation groups > change > set" means to overwrite everything existing.
e.g. if I have set it to SIM 1 and send this out this into next Particle Groups operator "SIM Change > Set 2" so it will also remove automatically SIM 1.
But to be very sure I also removed the SIM1 with another operator, when I wanted it to be only SIM2

   

   

   

[tyFlow]

Did you consult the docs and also look at the official particle groups sample scene that explains how they affect PhysX sims? I can see from your screenshot why the simulations would be different...

Sim groups work the same as collision flags in Unity, if you are familiar with that. Particle groups must overlap in order to collide. It's an easy way to filter out specific collisions for particles.

[Alexx31]

Thanks, okay, so they do not collide with each other... this explains a lot.

I am looking of course always in the docs (but I'm of course not aware of everything and sometimes there is a small language barriere to catch every detail as non-native English reader... ).

I was not aware that they must overlap to collide. I thought of them to assist purely as "filters".
=> "Overlap" means they need share at least one Sim number, as I understand, but not all numbers.

So it seems it's more important to go through all official example scenes.

---

p.s.
But for me one interesting take-away from this is, that the simulation looks much better and much easier to control, when I split it into 3 different SIM groups.
This wasn't intended, but is maybe a trick to prevent too extreme collisions when using many particles.
Without those SIM groups it needs extreme values to direct the flying particles.

So at least I learned this now...