# convergence problem

 Register Blogs Members List Search Today's Posts Mark Forums Read

 November 4, 2004, 10:49 convergence problem #1 limseokmin Guest   Posts: n/a Hi every one. I'm facing with serious problem that problem dosen't converge. If somebody help me, greatly appreciated. Problem description) Large tank with small inlet & ourlet, contain wall which has heat flux to inner fluid. The fluid is liquid nitrogen. It's the cryostat. My solving approach) Use turbulent model(k-e model) - if do not use, the temperature near hear flux wall is much high. not phisical. Use varying properties(k, rho, C etc.) - polynomial interpolation (4th order). It's steady. It include enegy equation. Use default setting(relaxation factor, descretization scheme) The difficulty is hard to obtain convergence. Especially the continuity convergence condition is main cause. How do I can improve my problem solving conditions and reach the complete convergence. Please help me to follow up.

 November 4, 2004, 11:53 Re: convergence problem #2 Ertan Guest   Posts: n/a Hi, Why dont you try first laminar case (although it is not laminar), after the solution stabilized turn it into turbulent and continue to simulate. It may work...

 November 5, 2004, 00:43 Re: thanks for advice. but... #3 limseokmin Guest   Posts: n/a As you say, I try laminar case to solve that. but in the convergence grafical window, continuity value is maintained about 1 and it is not converge(infinite calculation). What is the problem, you think. I use tetrahedral automesh(mesh quarlity is not good, but it satisfies min,max required range). Is it effect calculation condition largely and shows poor result?

 November 14, 2004, 13:43 Re: thanks for advice. but... #4 H Guest   Posts: n/a lower urfs will give you better convergence. does you model have any porous zone? also good quality mesh goes a long way in improving convergence. 80% of any CFD problem efforts should be devoted to creating a good mesh