[Student CFD]

want to simulate the motion of a discrete phase in a dirert-flow cyclone. The goal is to obtain a contour of the DPM concentration. The task is stationary. At first, I wanted to obtain a contour of the velocity in the apparatus and used the k-e RNG and Realizeable models for this. But the calculation does not converge at any settings, an error appears "floating point error". The gas velocity in the apparatus flies away to 10^5 m/s. When using the k-w SST model, the velocity calculation is complete.
When turning on the stationary DPM model, the calculation does not converge either when using the k-e models or when using the k-w SST model. When turning on the k-w model, the calculation did not complete even after 500 iterations, I stopped it manually (the velocity is again 10^4 m/s).
1. Why does the k-e models not simulate the velocity profile in the apparatus?
2. How to set the parameters of the DPM model so that the calculation converges when using the k-w model? How many iterations might this require?
I can send you a file of the solid or parasolid model.
Thanks for your help, this is my first task in Ansys
[IMG]
https://i.ibb.co/2KzRCbB/image.jpg
https://i.ibb.co/x3W612t/kw-vel-contour.jpg
https://i.ibb.co/D8NXrNT/kw.jpg
https://i.ibb.co/B37WKHm/ke-vel-real-countour.jpg
https://i.ibb.co/LtpYdYQ/ke-real.jpg
https://i.ibb.co/fvVqZmC/kw-with-par...el-contour.jpg
https://i.ibb.co/Ng01S01/kw-with-particle.jpg
[/IMG]