nasa polynomials

gardian · March 9, 2011, 05:32

Hello

I m trying to simulate a helium 4 gas in a closed tube 10 m long.
the ends are warm like 120K while the rest of the tube is at 2K.
I wanted to simulate the gas as real so using the peng-robinson method
and the nasa polynomials for the pressure i have inside the tube.
in the CFX-help it is written that in this case all i want is the specific heat , the enthalpy and entropy.
i found them from a software and I have all the above values.
for CFX Cp/R= a1+ a2*T + a3*T^2/2 + a4*T^3/3 +a5*T^4/4
similar equations are given for Ho/R and So/R.

I made a fit for Cp,T and extracted the coefficients . likewise i found a6 and a7 from enthalpy and entropy.

In my case because of the Helium 4 lambda point i divided the interval of Temperatures in the lambda point because there is a discontinuity .

that is around 2K .

So I have the upper coefficients for the range 1.8-130K and the lower one for the range 1.2-1.8

but trying to run this (while I have done it in the past for Helium 3 with success ) the solver crashes with exit code error 255.

can you see something wrong in my steps i followed ??

thank you

** i also receive this error +--------------------------------------------------------------------+
| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| c_fpx_handler: Floating point exception: Invalid Operand |
| |
|

i m using a coefficient for the fit that is like 1.2e-19 !!!!!!
is it too much for cfx?

ghorrocks · March 9, 2011, 19:45

The coefficient of 1.2e-19 just means it is almost zero. Should be fine.

Almost certainly there is a problem with your EOS. Does this run OK with an ideal gas with similar properties?

gardian · March 10, 2011, 05:53

yes . i have also tried the zero's in the very small numbers and is fine.
also the ideal gas case runs properly and gives nice results .
the problem seems to be in the lower temperature interval that is 1.2 -1.8 K.
Is it correct for ansys to divide the intervals like that (1.2-1.8) and (1.8-130) ?
in this low interval the enthalply as i got it from HEPAK software for Helium 4 gives enthalpies that are very small.
PRESSURE TEMP DENSITY CP ENTROPY ENTHALPY QUALITY
[Pa] [K] [Kg/m3] [J/Kg-K] [J/Kg-K] [J/Kg] [-]
1200. 1.200 145.2 317.7 50.96 59.47 -2.000

is it ok ?
thank you

ghorrocks · March 10, 2011, 07:29

Am I correct in reading you have one EOS for above 1.8K and another for below? This type of discontinuity is always difficult to converge. Not only do you need to get the function value smooth over the 1.8K transition, but you also need to minimise the gradient discontinuity (Or preferably eliminate it) at the transition point. Often this means you need to "bend" your EOS a bit at the transition point so the gradients are also continuous.

March 9, 2011, 05:32	nasa polynomials	#1
gardian New Member anton.gardi Join Date: Aug 2010 Posts: 21 Rep Power: 16	Hello I m trying to simulate a helium 4 gas in a closed tube 10 m long. the ends are warm like 120K while the rest of the tube is at 2K. I wanted to simulate the gas as real so using the peng-robinson method and the nasa polynomials for the pressure i have inside the tube. in the CFX-help it is written that in this case all i want is the specific heat , the enthalpy and entropy. i found them from a software and I have all the above values. for CFX Cp/R= a1+ a2T + a3T^2/2 + a4T^3/3 +a5T^4/4 similar equations are given for Ho/R and So/R. I made a fit for Cp,T and extracted the coefficients . likewise i found a6 and a7 from enthalpy and entropy. In my case because of the Helium 4 lambda point i divided the interval of Temperatures in the lambda point because there is a discontinuity . that is around 2K . So I have the upper coefficients for the range 1.8-130K and the lower one for the range 1.2-1.8 but trying to run this (while I have done it in the past for Helium 3 with success ) the solver crashes with exit code error 255. can you see something wrong in my steps i followed ?? thank you ** i also receive this error +--------------------------------------------------------------------+ \| ERROR #001100279 has occurred in subroutine ErrAction. \| \| Message: \| \| c_fpx_handler: Floating point exception: Invalid Operand \| \| \| \| i m using a coefficient for the fit that is like 1.2e-19 !!!!!! is it too much for cfx? Last edited by gardian; March 9, 2011 at 06:54.

March 10, 2011, 05:53	nasa polynomials	#3
gardian New Member anton.gardi Join Date: Aug 2010 Posts: 21 Rep Power: 16	yes . i have also tried the zero's in the very small numbers and is fine. also the ideal gas case runs properly and gives nice results . the problem seems to be in the lower temperature interval that is 1.2 -1.8 K. Is it correct for ansys to divide the intervals like that (1.2-1.8) and (1.8-130) ? in this low interval the enthalply as i got it from HEPAK software for Helium 4 gives enthalpies that are very small. PRESSURE TEMP DENSITY CP ENTROPY ENTHALPY QUALITY [Pa] [K] [Kg/m3] [J/Kg-K] [J/Kg-K] [J/Kg] [-] 1200. 1.200 145.2 317.7 50.96 59.47 -2.000 is it ok ? thank you

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
NASA Format	Stone	CFX	3	August 11, 2021 03:16
therm.dat NASA polynomials	mystix	OpenFOAM	5	May 31, 2018 04:17
Segmentation fault in running alternateSteadyReactingFoam,why?	NewKid	OpenFOAM	18	January 20, 2011 17:55
NASA polynomials, thermo.dat tables	Tomislav Sencic	Main CFD Forum	8	December 15, 2009 11:09
How to Read Nasa Format	Stone	CFX	6	March 20, 2006 03:03

March 9, 2011, 19:45		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	The coefficient of 1.2e-19 just means it is almost zero. Should be fine. Almost certainly there is a problem with your EOS. Does this run OK with an ideal gas with similar properties?

March 10, 2011, 07:29		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	Am I correct in reading you have one EOS for above 1.8K and another for below? This type of discontinuity is always difficult to converge. Not only do you need to get the function value smooth over the 1.8K transition, but you also need to minimise the gradient discontinuity (Or preferably eliminate it) at the transition point. Often this means you need to "bend" your EOS a bit at the transition point so the gradients are also continuous.