hy, what can be changed or improved in mesh generation or boundaries conditions in order to increase the time step for the transient calculations? the biggest time step for which the calculations are not divergent after few iterations is 0.001s... thanx! lelia

Can you please try to decrease the mesh size and see whether it works or else you have to check the boundary conditions again.. that is the best if you have to increase the time step.

Sachin

hy! thankx for the advice...but can u be more explicit...by decreasing the mesh size u mean the number of cells or the cells dimensions? because i have around 200.000 cells but the smallest one has 0.0006m. then, for the boundaries conditions, i have an inlet and an outlet...i'am confused a bit...i appreciate if u give a more detalied explanation. thankx! regards,lelia

You have to decrease the mesh size. Refine your mesh. But this will cause increase in number of cells which will take more PC run time. Regarding boundary conditions, do check them again by plotting each and every boundary conditions. Even if you miss small thing it will diverge. Good luck.

Sachin

thankx, just that i can't go bellow 0.0006m which is the smallest cell dimension, because i already have 270.000 cells...and for the boundaries conditions, i cant find the problem, if there any...im trying to solve the fluid flow in steady state and then just the particles ( because i have a 2 phase problem) in transient hopping that it will take less time and that the errors induced are acceptable.. regards, lelia

It is ok if you can not reduce the mesh size for some of the cells. Try to refine the mesh whereever you can. I hope this helps.

Regards, Sachin

If u are using ccm+ turn off volume sources, and hope for the best by refining your surface mesh

no, i'am working under Star-Cd 3.26...but is there any chance to simulate in the same time the fluid flow in steady state and particles motion in transient?

No, it should be done in two steps. If you do uncoupled simulation (i.e. particle do not affect the flow) then you can first get a steady state solution for the flow without particles, afterwards add particles, turn off flow field solution and make a restart. In this case at the second step you'll have to solve only particle motion equations.

hi Anton! you always have answers for my problems?! thankx a lot! in fact that's what i have done. first i considered just the fluid flow in steady state, then i have restarted the simulation using the steady state results and adding the particles and this time in transient state. i also turned off the solvers for fluid ( equation behavior - primary variables - u-mom,v-mom...). that's the idea, isn't it? thankx! regards, lelia