Number of elements x double precision
I'm running a simulation of wind tunnel and have a question related to the number of elements and double precision. On the computer that I'm running it's possible to run (without double precision) a mesh with 3 million elements (maximum). I'm worried about the results (because of single precision) so I decreased the mesh to 1.5 million to run it with double precision. It's always good to perform a simulation with high number of elements? Double precision is always important or it's better to use a higher mesh and run with single precision?
What is your solver? Mesher? Application?
Solver: ANSYS CFX Solver
Mesher: ANSYS (Advanced meshing)
Application: I'm need to validate experimental results (academic purposes).
|All times are GMT -4. The time now is 23:14.|