Hello,
I am trying to implement the Shepard filter for a system with boundaries defined by Morris ghost particles. The host particles have an associated mass equal to the mass of the SPH particles insuring the no-slip or free slip conditions.
At the beginnong of the simulation all gas and ghost particles are arranged in a regular lattice.
Is it required to take into account this particles for the Shepard corrrection?
Best regards,
Cristian Achim
http://cfd.mace.manchester.ac.uk/sph/SPH_PhDs/2008/crespo_thesis.
Not necessary....just give common mass inertia
Thanks for the reply. I will check the thesis. the link does not work, but this should be ok http://cfd.mace.manchester.ac.uk/sph/SPH_PhDs/2008/crespo_thesis.pdf
Since the initial configuration is a regular periodic, I am giving the same mass to all particles ghost or real. Can you indicate some pages where I can see it?
Best regards,
Cristian Achim
Shepard correction is typically used to improve the density calculation on particles close to free boundaries (well, the consistency of the resulting mass and momentum system is another big story).
The ghost particles, which are used for the simulation of rigid boundaries, provide by default an additional support for the internal, real fluid, particles. Thus, the Shepard correction is redundant for real fluid particles, while I am not sure if there is any good reason to use increased density accuracy for the ghost particles. It is the mass of the ghost particles that real fluid particles "see", not their density.
Regarding the initialization and assuming that you are treating monophase flow, yes, you should give the same mass to all particles.
- Badges
- Science topic
- Similar topics
- Engineering
- Thermal Engineering
- Thermofluid
- Fluid Mechanics
- Fluid Dynamics