Learning Forum

[CFDGer]

Hello,I have trouble finding a solution to implement the NASA 9 coeffcient polynom which is applicable for temperature ranges of 200K-20000K. This polynom has the form:n cp/R = a1/T^2+a2/T+a3+a4*T+a5*T^2+a6*T^3+a7*T^4nnThe 7 coefficient polynom is of course easy to implement but it is only valid up to 6000K which is not n Do you have to write an UDF to be able to implement this kind of polynom or is there a different way in ANSYS FLUENT?.n Thanksn

[Rob]

What are you modelling to need to go up to 20,000 K ? You can have several bands of polynomials which may be more accurate: very high order polynomials tend to have silly coefficients which rarely follow the real data curve when plotted as nothing can hold enough significant figures. n

[CFDGer]

Thanks for your reply Rob.nI am modeling a flow at Ma=7.9 and an ambient temperature of 1113K. With the calorically perfect gas I should get values of ~15000K. With consideration of high temperature effects and chemical reactions I assume to get around 7000K-8000K. So I assume the cp(T) which is fitted for up to 20000K suits this case better. Further I would like to implement it for the academic standard. nThe coefficients I use are from the NASA 9 coefficient polynomials tables. I already plotted them with matlab and they seem to be valid (they calculate the same values in the range of 200K-6000K as the 7 coefficient polynomials).n

[Rob]

OK, as a process engineer anything over about 25m/s tends to have gone a bit wrong, over Mach 1 and it's gone badly wrong! Remember to check the limits in the solver, default is 5000K maximum.nMax for a polynomial looks to be 8 coefficients (so x^7) so you'll need a UDF. Have a look at DEFINE_PROPERTY in the UDF manual. n

[CFDGer]

Yes I have already run plenty of simulations with this real gas setting with chemical reactions and without (so only species transport) and therefore I already adjusted the limiters and for example the positivity rate as it is recommended by Ansys User Guide for high speed flow.nThe species transport simulations converges very nice and the results are reasonable. As soon as I turn on the volumetric reactions, the solution diverges after a few hundred-thousand iterations. To be more precise: I set the temperature limit to 25000K to give the solver a little room to converge. But as I said after a few hundred iterations the Ma- number increases in a few cells to Ma=230 or even more unrealistic values (often directly upstream of the shock wave) and the temperature then decreases in these cells to 1K (limited) and in others to the max limited value. In these simulations I used the 7 coeff. cp(T) polynomials and I thought maybe that could be the reason for the divergence.nIs there any information I could forward you that would maybe give you a hint on why chemical reactions pose such a problem for my case?.

[Amine Ben Hadj Ali]

I recommend using Real Gas Property Table here ( if you have version 20R2) or sticking to Real CUEOS. For the latter you need to define your heat capacity via UDF DEFINE_SPECIFIC_HEAT. Another approach is to UDRGM Real Gas which is more general.nRegarding reactions you probably need to account for stiff chemistry solver and rely on well calibrated reaction mechanism.n