3Dblade

dan

hi, I'm simulating flow and heat transfer through a gas turbine blade passage. I'm using a dual core with 4GB RAM. With the 3D (1.5E06 nodes) mesh I'm using now, and using a Density-Based solver, I can run the simuation only partitioning the mesh. one partition for each cpu. The problem is that x and y velocity residuals at some point drop several orders of magnitude, then increase again or oscillate and I have the following error:

Error: > (greater-than): invalid argument [2]: wrong type [not a number]

The previous mesh I used (1.3E06 nodes), worked well with the Pressure-Based Solver without partitioning the mesh. Any suggestions? Thanks