divergence issue when updating dpm sources
 

I have two similar cases using k-omega SST turbulence model and DDPM multiphase model:

case 1. dpm iteration interval is greater than max iteration per time step, so the DPM source is not updated within each time step (urf DPM =1)

case 2. dpm iteration interval is set 2 and the reporting interval is 4 so at least 2 dpm source update occurs for each time step (urf DPM 0.9)

the first case is fine and does not have any divergence issue, but the second one divergence after a while (with turbulent viscosity ratio warning)

in both cases, I track overall DPM sources and there in no sharp change from one iteration to another and from one time step to another.

as I understood, to ensure that DPM sources are calculated accurately at least on update must occur in each time step so that is why I moved from case 1 to case 2.

what can be the cause of such divergence?