[averis007]

I'm running a dynamic mesh simulation on Fluent. The overall mesh size is of the order of ~40 mil. In order to prevent the mesh quality becoming too bad due to the movement of one of the walls, I'm regenerating the mesh after every 0.5 seconds, interpolating the data from the previous mesh. But the problem is, I get a huge spike in the coefficients (mainly Cl and Cd), which lasts about 0.1 to 0.15 seconds and then the coefficients return near the expected values. Also, the occurrence of the spike is not the same everytime. My time-step is 0.00005 seconds, so it takes 10000 steps to complete a flow time of 0.5 seconds. Sometimes the spike would start around 1500th step, sometimes 5000th, sometimes 8000.
The turbulence model was RANS (k-w, BSL, DDES ). I tried repeating with DES and LES, but got almost identical spikes both times.


I'm sharing images of the spikes occurring at different flow times.

[averis007]

Forgot to mention that coincident with the spike in the coefficients, there is a spike in the residuals as well. Please see the attached image corresponding to the 2nd image in the original post.