During the CFX 10.0 compressor simulation of one passage of an 8+8 splitter and diffuser, a periodic residual convergence is experienced. This is more evident approaching the stall margin. Is there an explanation for this?



This could be due to transient nature of the flow in certain locations that can be avoided by increasing the timestep to bypass the timescale of the transients. On approaching the stall margin, this flow may become more transient, so it is advisable to check isosurfaces of residuals in the impeller and diffuser to understand the flow behaviour. Also, there could be significant component interation effects. You should check the mesh density in these areas of high residuals as these may be influential. Also, restarting a run for the next point to stall margin from a previous result file is generally a good idea, it can be the case that some flow phenomena may be adversely predetermined, so you should try starting the run from a basic initial guess.





Show Form
No comments yet. Be the first to add a comment!