Problem converging with a set particle distribution
 

Hi,

I am running a transient evaporating spray simulation in CFX 11 on a 8 mil mesh using a 8gb ram intel quadcore processor and I have successfully simulated the particle injection using the cone with primary breakup model however when i try to change the method of injection I am having tremendous trouble converging even for one time step. The problem is that when using the cone with primary breakup model all the particles injected were of size 500 micrometers however when using the rosin rammler the representative diameter is 60.7 micrometers so a lot of very small particles are ejected. Is there any way in which I can accelerate the convergence without prohibitively increasing the computational cost?

I also tried using a normal distribution with the mean equal to Sauter mean diameter which 43.7 microns and encountered a similar problem I guess due to a similar reason