[treviusss]

Hello,

One more time I encountered problem while trying to solve a 2D Steady simulation of Ahmed Body. I was doing mesh independece test where I was decreasing element size in refinement zone. I got to a point where element size in refinement zone is 3mm and outside 4mm. I wanted to do one more try with 2mm elements in refinement zone and 4mm outside but when I solved a problem with First Order Upwind for Momentum/Turbulent Kinetic Energy/Disspiration Rate and changed to Second Order residuals started to fluctuate and I have no idea what to do about it.

I'm not sure if I'm doing it right but for First Order I'm monitoring Drag Coefficient and when it seems to level out I change to Second Order.
I did this after around 3000 iterations (see attachment)

Average Skewness for mesh is 5,1202e-002 and Average Orthogonal Quality is 0,96859. Number of elements is 351 941

I attach residuals, drag coefficient plots, contours of velocity magnitude around model where it looks like it failed to predict vortices behind the model and how the mesh looks like.

Is there anything I can do about it? Maybe decreasing Under Relaxation Factors at some point can help converging the solution?

Is it normal that Drag Coefficient seems to go lower everytime I make denser mesh? Because Cd is the variable that I'm monitoring to find out whether the solution is mesh independent or no.
I read in this topic: Mesh Independent Study, that it is really hard to get mesh independent solution for car especially with limited computational resources.

Thanks for any replies in advance.

[Aero_Denis]

Are you consider ground effect? what are the boundary conditions?

May be, if you change to transient, could be better, i think the solver is not geting the vortex formed on the wake...