Particles removed due to Rhop_out
Greetings everybody,
I have a problem related to particle removal due to rhop_out.
My basic setup is a straight cylinder, a cylindrical "plug" at the entrance that is moving at constant velocity forcing fluid through the cylinder (i.e. piston in 1 direction). I'm working under Re=50 and water. I'm as well using delta-SPH (0.1) to stabilize my simulation. The walls are only 1 particle thick but no fluid particles escape through the boundary and the velocity is mostly fine near the walls.
When the simulation is executed, after some time particles start being removed from the centre of the cylinder due to too low or too high density (Rho). The rest of the fluid is perfectly fine. I know SPH deals with instabilities and to prevent "blowing-up" rhop_out is used to remove "bad" particles. However, when I widen my bounds (from 700 & 1300 to 500 & 1500) less particles are removed and the result is better.
So my question/problem is:
1. Why are particles spiking/dipping so much in density, and can I somehow prevent it?
2. Why does widening the bounds improve the result and not destabilize it.
Disclaimer: I know inlet & outlet combination exists, but I got access to it a bit too late in my project thus cannot switch to it.
Regards,
Bladomas
I have a problem related to particle removal due to rhop_out.
My basic setup is a straight cylinder, a cylindrical "plug" at the entrance that is moving at constant velocity forcing fluid through the cylinder (i.e. piston in 1 direction). I'm working under Re=50 and water. I'm as well using delta-SPH (0.1) to stabilize my simulation. The walls are only 1 particle thick but no fluid particles escape through the boundary and the velocity is mostly fine near the walls.
When the simulation is executed, after some time particles start being removed from the centre of the cylinder due to too low or too high density (Rho). The rest of the fluid is perfectly fine. I know SPH deals with instabilities and to prevent "blowing-up" rhop_out is used to remove "bad" particles. However, when I widen my bounds (from 700 & 1300 to 500 & 1500) less particles are removed and the result is better.
So my question/problem is:
1. Why are particles spiking/dipping so much in density, and can I somehow prevent it?
2. Why does widening the bounds improve the result and not destabilize it.
Disclaimer: I know inlet & outlet combination exists, but I got access to it a bit too late in my project thus cannot switch to it.
Regards,
Bladomas
Comments
If you want, you can send to us your case (XML) to dualsphysics@gmail.com and we will check it when we have time
Regards
The increase in density could be because of a "squeezing" of particles, so they clump together in search a way that numerically their density increases and get kicked out of the sim. This has at least been my experience earlier. A way to avoid this is try checking what happens if you both decrease and increase particle sizing.
Kind regards