About speed up calculations by modifying model geometry

I want to reduce the number of particles by modifying the initial settings of the model, so as to achieve the effect of speeding up the calculation, but I found that the calculation results are completely incorrect. I would like to ask, does this DualSPHysics code support this idea? The detail is as following:

First of all, I simulated the settlement of a sphere in water through normal modeling, and the calculation result is no problem, using the solid ball simulated by the SPH rigid body. And then I modified the modeling of the solid sphere part, nothing else has changed.

Changed: Changed the solid sphere to a hollow sphere (the mass of the sphere has not changed, it is still the same mass as before)

Result: After the change, the sphere hardly sinks (a total of 0.2mm sinks, the initial height of the sphere is 4cm, and the sphere sinks to the bottom normally when no changes are made)

Initial model(Modified and Unmodified)

https://forums.dual.sphysics.org/uploads/268/AWTB68BICRKB.png There was an error displaying this embed.


Partial enlargement:

https://forums.dual.sphysics.org/uploads/827/WJL41RXNTC1R.png There was an error displaying this embed.


result:

https://forums.dual.sphysics.org/uploads/040/2IGIEJ4YSNP1.png There was an error displaying this embed.


I also tried the coupling method of DualSPHysics and Chrono library, still the same problem, where is the problem of my operation (is it still necessary to modify the total quality?), or the code does not currently support this aspect of acceleration?

Sign In or Register to comment.