CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)

 Anthony_impeller October 6, 2011 01:15

Hello Forum-goers,

I've been doing turbomachinery CFD for the last year and have ran into so many problems that I want no one else to hit. So rather than asking a question I will post my solution.

The most recent one was specifying a "user-specified" peng robinson fluid.

My fluid of interest was R245fa and I had obtained my critical pressure, density (volume) and temperature, boiling temperature, accentric factor and molecular mass. This is easy to grab off NIST Refprop if you are looking, I think it is free.

So, the problem was in specifying the five coefficients for the fourth order polynomial of specific heat at constant pressure. After researching ANSYS help files and thermodynamic brain racking I found that is was in fact a zero pressure polynomial. I found the expected order of magnitude by looking at ANSYS' peng-rob data files for r134a.

To determine those specific to R245fa I plotted a bunch of points (C_p) from REFPROP for zero pressure and varying temperature in EXCEL. There was a lower and higher limit to where REFPROP was useful but the middle ground was enough to fit a polynomial to. When the fitted polynomial coefficients were input into CFX my solution would fail before any iteration. The error was:

"Error detected by routine MAKDAT" Blah blah blah.

That means I was making the specific heat capacity at constant volume C_v go negative.

The coefficient values obtained were much smaller than those ANSYS had for R134a. On second look I noted that I had not divided by the fluids specific gas constant (R with a bar in kJ/kgK). On rectifying this I had values on the same order of magnitude.

These values worked fine in ANSYS CFX and my solution solved like a charm (aka I had to work through convergence problems too but that is covered in FAQs and things).

So if you are having this problem now you can fix it.

Good Show,

Anthony.

 enuano June 15, 2013 06:57

R245fa

Hi!
I'm writting my master's thesis and have some problems when modeling an expantion for R245fa. Which kind of pure substance do you use? Peng Robinson WET refrigerant? because my problem is that pressure at outlet is ok, but temperature is too high and enthalpy and entropy make no sense. Also, do you still have the values for the coefficients you used in your simulation? THank you very much
Quote:
 Originally Posted by Anthony_impeller (Post 326855) Hello Forum-goers, I've been doing turbomachinery CFD for the last year and have ran into so many problems that I want no one else to hit. So rather than asking a question I will post my solution. The most recent one was specifying a "user-specified" peng robinson fluid. My fluid of interest was R245fa and I had obtained my critical pressure, density (volume) and temperature, boiling temperature, accentric factor and molecular mass. This is easy to grab off NIST Refprop if you are looking, I think it is free. So, the problem was in specifying the five coefficients for the fourth order polynomial of specific heat at constant pressure. After researching ANSYS help files and thermodynamic brain racking I found that is was in fact a zero pressure polynomial. I found the expected order of magnitude by looking at ANSYS' peng-rob data files for r134a. To determine those specific to R245fa I plotted a bunch of points (C_p) from REFPROP for zero pressure and varying temperature in EXCEL. There was a lower and higher limit to where REFPROP was useful but the middle ground was enough to fit a polynomial to. When the fitted polynomial coefficients were input into CFX my solution would fail before any iteration. The error was: "Error detected by routine MAKDAT" Blah blah blah. That means I was making the specific heat capacity at constant volume C_v go negative. The coefficient values obtained were much smaller than those ANSYS had for R134a. On second look I noted that I had not divided by the fluids specific gas constant (R with a bar in kJ/kgK). On rectifying this I had values on the same order of magnitude. These values worked fine in ANSYS CFX and my solution solved like a charm (aka I had to work through convergence problems too but that is covered in FAQs and things). So if you are having this problem now you can fix it. Good Show, Anthony.

 All times are GMT -4. The time now is 06:22.