Hi, everyone,

i'm implementing standard k-omega model proposed by Wilcox based on unstructured mesh by FEM method. BC at wall for Omega is 60.0*dynamic visc/(beta*dy^2), which was proposed by mentor and for k is equal to 0. I use k-l result as initial soultion for k-omega. To my suprise, after first newton-raphson interation, all k is negative inside the mesh domain. i checked many times that the equtions are correct for sure, but i couldn't undersatnd why it's like that. Pls. give me your suggestion and comments. Thanks for your help.

It could be the way you are writing your source terms. Have a look at the following link to see how to make sure that source term gives positive values.

http://personalpages.umist.ac.uk/sta...r/turbcfd2.pdf