Open speth opened 4 months ago
How interesting! Is the line the same for other fluids like O2?
Is there any connection to #1157?
That is strange because up in the supercritical region there, the root finder shouldn't have any issues converging like it would potentially do in a subcritical case with a liquid/vapor region to contend with.
Problem description
For certain combinations of T and P, the method
MixtureFugacityTP::solveCubic
fails, issuing a warning like:and leaving the phase in an invalid state with pressure and density at
inf
and-inf
(or vice versa), which in turn results in all thermodynamic properties evaluating toNaN
.Steps to reproduce
Conditions where this error occurs are relatively rare, but occur on a set of related points.
Behavior
These are points where the solver fails and issues a warning:
The error occurs along a line of points, approximately defined by the line $P=157285712-506093T$. The failures become less common as temperature goes up.
System information
main
branch at d37a76bf6Additional context
Originally reported on the Users' Group