[B1ngus]

Hello,


for my PhD-Work i am simulating a Rotating piston compressor. It contains 4 inlets and 4 outlets (they are all periodic). Inside i have 2 rotors, which are rotating in the same direction but with a different phase-angle (motion is similar to a wankel-engine). So between the rotors, gases (air and refrigerants) will be compressed. As a result the fluid domain is changing and also rotating dependend of the time.



I have read a lot of older posts (09-18) about the Immersed Solid method and their limitations here in this forum. Is the immersed solid method still suitable for my case or does it stay in conflict with the inaccurate calculation of compressible flows?


If so, would u recommend to simulate this with a moving/deformation mesh (maybe even remeshing?)? Or is there any other possible method outside to simulate rotating compressors?


The focus of the simulation is to achieve the pressure ratio and several diagrams like p,v for this compressor.



Thanks

[ghorrocks]

Have you read the documentation on these models? For instance if you read about immersed solids in the documentation you would see:
* An immersed solid domain cannot model heat transfer
* For transient cases, immersed solids do not interact properly with fluid domains that involve compressibility or multiphase flow.

I don't know what you are trying to achieve with this model, but I suspect these issues on immersed solid mean it is not a suitable model. You will have to use moving mesh. I suspect you won't need remeshing here, just moving mesh should be OK.

Another issue:
If you are trying to get performance curves for this machine I would have thought it would have been dominated by adiabatic compression/expansion and leakage at the seals. The compression/expansion can be worked out analytically and seal leakage requires knowledge of the sealing performance and is not a CFD issue. So I don't see how CFD will help you in this case.

[B1ngus]

Originally Posted by ghorrocks

Another issue:
If you are trying to get performance curves for this machine I would have thought it would have been dominated by adiabatic compression/expansion and leakage at the seals. The compression/expansion can be worked out analytically and seal leakage requires knowledge of the sealing performance and is not a CFD issue. So I don't see how CFD will help you in this case.

Alright, i agree with your tips thanks! So sealing performance could be done another analysis by FEM/FSI, right?

[ghorrocks]

Seal performance is a complex topic. FEM/FSI might be part of it, but there is a lot beyond that. For instance surface roughness, friction, fatigue of sealing elements, wear, pressure history, bedding in and issues like that are critical - and FEM/FSI will not give you useful information on most of that. I was involved in a project where seal design was a key component of it and we did almost no simulation work on the seals as they were too complex. We developed the seals using experiments and clever interpretation of experimental data.