It worked well when I used water’s properties (rho = 1000 and viscosity = 1e-3). However, now I want to simulate the soil fluidization by gas (rho = 1.2 and viscosity = 1.8e-5), there is an issue. After running a few step, the particle is washed out and disappeared (File attached) . I tried to change the timestep, make sure the Mach number is under limit, but it does not help.
Can anyone suggest me a solution for this issue? Thank you
I do not know the details of the code, but if with “viscosity” you mean dynamic viscosity and you are not changing other parameters, it seems that your physical problem stays the same because of the order of magnitude of the kinematic viscosity is the same (i.e. same Stokes number and same Re).
So, my first guess is that the problem is purely numerical. I would check first of all the maximal LBM velocity uLB (should be < 0.04) and the BGK relaxation time tau (< 2). If this doesn’t help probably the problem is in the coupling with LIGGGHTS.
I suggest you work directly with non-dimensional units fixing, for instance, Re and Stk and then find the best lattice parameters for your case.
yes, but \nu is the kinematic viscosity: nu = mu/rho. So, nu should stay on the same order of magnitude. But, I apologize, you are right, the Re does not stay the same, because Stk number changes: if you maintain the same density of the particles in the two simulations you change the density ratio rho_particle/rho_fluid. If Stk is different in the two cases, you cannot guaranty the same dynamics of particles and thus the same characteristic velocity => Re changes.
Still, I confirm that you should
if these parameters are right, then the problem is that the ratio of the characteristic times in fluid and solid changed, both physically and numerically. And so, I guess you should check how to tune the LIGGGHTS timestep and/or change the time integration method (eg using a Runge Kutta integrator in place of an explicit Euler). But these are just feelings, I cannot help in the specific, because I do not know the code and the methodology. I hope somebody else can help you…
Actually, a test you can do is to run the second simulation keeping the same rho_f/rho_particle and same uLB and tau of the first simulation (with water) and check that it runs without problems. This to be sure that your code is not sensitive to physical units but to the dynamics of physical problem you are solving.
Hi mars;
Thank you for your kindness, you guys in this forum have helped me a lot. I following what you told me, changing the LIGGGHTS timestep and Palabos timestep, the simulation ran longer, but after that it corrupted and the error is " free(): invalid next size (fast) "
Do you know how to solve this problem ?
Thank you very much for your time.
