MixtureFugacityTP::solveCubic -- root didn't converge

[jamie]

Hello,

I'm trying the evaluate hydrogen thermodynamics with the Redlich-Kwong EOS and I'm running into warnings that the root does not converge. I get the following error message 

MixtureFugacityTP::solveCubic(T = 288.15, p =11416262): WARNING root didn't converge V = -6108.801148602612

I only get this warning near 11.4 MPa for T=288.15. And if the warning appears, then evaluating properties like density are inf. Attached are files to reproduce. The H2 thermo is copied from GRI30. The RK parameters are calculated from the critical properties. Running on python 3.10.13, cantera 3.0.0.  Is there a way to tighten the root finder to get conversion?

Thanks,

Jamie

[Ray Speth]

Hi Jamie,

Thanks for reporting this. I’ve investigated a little bit more and created Issue #1699 on our tracker. I don’t know of any workarounds for this, other than perhaps trying a nearby point if you can handle the amount of error this introduces in the state.

Regards,
Ray

​

[jamie]

Hi Ray, 

Thanks for creating the GtiHub issue. Interesting to see how linear the error points are. I'm currently perturbating the pressure + and - some amount and then taking the average density for now. 

Jamie