[bassamelsayed]

Hi everyone,
i have an axial flow compressor model(IGV-Rotor-Stator) undergoes a rain ingestion flow.
i done a dry run (without rain phase) and every thing was successful. when i tried to include the rain phase, i had some troubles.
firstly, i use Eurlian-Eurlian flow simulation with Air as continuous flow and rain as polydispersed flow (MUSIG).
the flow inlet condition is total pressure with Pt=1.13 atm and total temperature Tt=288K.
the outlet condition is totoal mass flow rate mdot=77Kg/s
all other conditions are as per pictures attached.
i was run this simulation since the last Oct. 2022 and most of the runs i face the following issues:
1- Both the two phases has a back flow and the cfx put a wall in % (starting from 100% in decreased with simulation) to prevent the flow either from return to the model in the outlet or escape from the model in the inlet. although this phenomenon did not raised during the dry run.
2-after about (48-52 iterations) the program exit with "Fatal overflow in linear solver."

Kindly if any one can help in this issue either by supporting me with any reference or any solution.

[ghorrocks]

Why are you using a complex multiphase model (MUSIG)? Why isn't a simpler multiphase model appropriate?

What are you trying to get out of this model? Why are you doing it? Of course ingesting rain into a compressor is going to reduce the performance a bit - but what aspect of this are you trying to investigate?

[bassamelsayed]

Thanks for your feedback, i am doing this simulation in order to get the effect of the deterioration of the compressor due to water ingestion and what is the shifting of the operating line in the compressor map looks like.

The second thing is why is using MUSIG is that i have a several group sizes of water droplets and i can not use the dispersed type because i can not use the mean diameter when i have the actual diameters of the group sizes.

[Opaque]

Originally Posted by bassamelsayed

Thanks for your feedback, i am doing this simulation in order to get the effect of the deterioration of the compressor due to water ingestion and what is the shifting of the operating line in the compressor map looks like.

The second thing is why is using MUSIG is that i have a several group sizes of water droplets and i can not use the dispersed type because i can not use the mean diameter when i have the actual diameters of the group sizes.

Theoretically, you are correct; however, since you have no experience running multiphase flow, you must take the lowest approximation, i.e. assume the same diameter for all the water droplets.

Once you have fully evaluated the impact of uniform droplets on your models, and you still have a need for a more complex model, you have already gained some experience (with useful results in between) before proceeding to INCREMENTALLY refine your model.