criteria for a simulation
Hello everyone :)
I am runng simulation for 2D airfoil with using FLUENT for the calculation of lift and drag coeffcient for different AoA (0 to 16 degree) scince my main objective is to compare the lift and drag coefficient curve for different AoA and different tubulence models, i would be very thankful if anyone can explain me the excat number of iterations to run this simulation/ what is the excat criteria to run this simulation is it okay to stop the solution as soon as the convergence is achieved or do i have to wait for them to stable upto what point for example, i am running for Spalart-Allmaras and my residuals covergence is set to Xe-4 and its achieved, but my coefficients are not converged or i can say that they are still going down. do i have to lower my residuals and continue till they become constant? or is it okay to stop there as i am comparing the coeffcients for different AoA's? thanks in advance |
In your case there are a lot of causes,such as your mesh quality which is not good enough, unreasonable boundary conditions.
|
Quote:
Thanks for the reply. in my case i am using hybrid mesh(unstructured near the airfoil and structured away from it). i would like to know the criteria in general for any simulation. suppose in my case the solution is converged by 300iterations and my coefficients doesn't become constant, so do i have to run another set of iterations till they(coefficeints) become constant. |
Forget the convergence criterium. It doesn't tell you anything.
|
Hello Bharatt,
Since you are interested in the force coefficients (lift and drag), the residuals should not be your only convergence criteria. Of course, it's good that your residuals decrease by at least 2 or 3 orders of magnitude, but you also have to monitor your lift and drag coefficients and check that they converge (for example, I know that in the aeronautical industry, the drag prediction should be within 1 dc or drag count which corresponds to 10^-4, but in your case I don't know the precision you wish to reach). In the Residual Monitors window, you can disable Check Convergence for all your residuals. By doing this, your simulation won't stop before the end of the number of iterations you set. Hope this helps Regards, Jeremie |
Quote:
Thanks for your reply, that made pretty much clear that i should concentrate on the objectives not on the residuals :) like in my case the decimal point for drag is upto 3. thanks again :) bharat |
All times are GMT -4. The time now is 07:40. |