About latticeU

Although I have calculated several cases, I’m still not very clear about the lattice Units.

Re = PhyL * PhyU / PhyNu = Resolution * LatticeU / LatticeNu
Usually we use LatticeU= PhyU / M.

  1. Must LatticeU be proportionable with PhyU? Can it be any value? And according to the comments of the code, it should be “proportional to Mach number”. I’ve no idea what it means.

  2. In some of my cases of High Re with smagorinsky model, If I make my deltaX halved, I found that the original LatticeU can not be used in the new case for the simulating divergent. In this situation, I have to make my latticeU smaller. Why this happened?

  3. I found that if the M is different in the same case (means LatticeU different), there are some differences in the results velocity. Was it right? If so, how should we determine the value of latticeU?

  4. During the iteration, we use averaged rho to monitor the simulation. Usually it should be drifted around 1. But if the latticeU is not suitable, the ave rho will continually grow up, even from 1 to 4. Why will this happen? As we know that rho is the sum of the distribution functions, if it grow up, does it suit the mass conservation? And is this kind of result correct?

Maybe some of the questions are so primitive,thank you for your focus.

with best wishes,