Hey, hope someone can help, I am trying to model the flow of a 2d aerofoil in water at 7m/s (High Reynolds flow) and i am having a bit of trouble with my grid optimisation.

Basically i am using a steady flow, k-epsilon model with 1% turbulance and 0.01 length scales.

Now this has been working fine up until now however i decided to increase the mesh in the streamwise direction with the aim to get a squarer elements around the aerofoil, particually at the tip and trailing edge. Now for some reason this simple change has thrown my epsion residuals into an occilating unstable mess, has sent my continuity rising sky high, and every other residual is following them up.

As I would like to converge sometime in the next millenium this is a problem.

Any ideas on what i have done wrong??

I believe the problem is related to the increase in mesh density?? can this cause instablities

THIS IS THE THIRD REPLY TO THIS PROBLEM BUT BY HOW MUCH DID YOU INCREASE THE MESH DENSITY?

If continuity shoots away up then your model is probably unrealistic compared with physical experiments and SIMPLY DOESN'T SATISFY THE GOVERNING EQUATIONS. You must have something in the model that you don't know about or the boundary conditions, etc are wrong.

Phil

PS. I think I read somewhere that anything above 2 million cells becomes very hard to solve and great care and experience is required so don't go above this.