how critical is to have rms residues below 10exp-5 in CFX solver for a CFD simulation. If a problem is set up as a steady state but acutal flow shows transient affets due to reverse flow. Is is likely that residues will not come down due to transient affects. In a simlation i am having very good balances (mass energy , momentum etc) but rms residues are about 10 higher than set up for convergence criteria (10exp-5). some residues like turbultent ke are even higher when using RNGS ke model. Can results be reliable qualiataively and quantatively. can problem be considered as converged.
This issue gets asked so much we should put up a FAQ about it.... Any volunteers?
This is discussed in the CFX manual. Look at the important output parameters for you model. it could be lift, drag, flow rate, free surface level, whatever. If the output parameter of interest does not change to your desired level of accuracy as the thing tries to converge then it is highly likely you are OK. If it moves around you are not OK. Also read http://www.cfd-online.com/Wiki/Ansys...gence_criteria
|All times are GMT -4. The time now is 03:44.|