search

ibell / pdsim

Steady-State simulation code for positive displacement machines
MIT License
38 stars 25 forks source link

Modeling mixtures in scroll compressor #56

Open mohsint7 opened 4 years ago

mohsint7 commented 4 years ago

This issue is related to both PDSim and Coolprop.

I am trying to model mixtures (R454B.mix) using existing PDSim scroll core class. I am getting some errors like "DHSU_T_flash does not exist" or need to build the phase envelop.

The complete core class is structured around CoolProp State.update or State.copy. Due to some missing functions for mixtures in Coolprop, I got these errors. Is there any way around these errors?

I also tried calling REFPROP using Coolprop but that is extremely slow and has convergence issues if, during iterations, temperature is in near critical region.

davideziviani commented 4 years ago

@mohsint7: these are well known issues. I am working with R454C.mix with REFPROP10, and by using REFPROP backend it works. It is slow due to the wrapping around REFPROP. Are you intentionally running nearby C.P. ?

mohsint7 commented 4 years ago

Thank you for quick response. I am not intended to run it for near C.P but trying different lump temperature guess values to avoid Pr number error in PDSim code. during iteration, it automatically go near C.P.

using REFPROP, a single iteration is taking hours. is it normal?

davideziviani commented 4 years ago

@mohsint7 nope, it takes just minutes to run. There must be something else for instance geometric parameters.

Inside core/core.py , in the OBJECTIVE_CYCLE( ) you can add a check for Tlumps to avoid unrealistic values.

For multi-lumps:

Check the new Tlumps

                for jj in range(len(Tnew)):
                    if Tnew[jj] > 400 or Tnew[jj] < 273:
                        Tnew[jj] = 340 #[K]
ibell commented 4 years ago

Also, please see the example examples/scroll_compressor_bicubic_with_mixtures.py which shows how to use tabular interpolation with mixtures. That can help with speed. You might also want to provide a better guess value for the lump temperature(s).