HS_INPUTS in tabular interpolation

[Marcoliveti97]

Hi everybody!
Do you know why the enthalpy-entropy pair of input is not supported for the tabular backend?
It one of the most common pair for many problems, thanks to the total enthalpy conservation, but unfortunately it seems to be unavailable. Do you think it could be soon implemented?

Thank you for your work

[ibell]

Because H,S, are always the most complicated inputs to handle. Neither input is an independent variable of the EOS and as a consequence, you need to iterate. That is even true in the case of the tabular interpolation. Someone could add the iteration, but it won't be me. Pull requests welcome.

[Marcoliveti97]

Hi Ian and thanks for your reply.
I could help implementing the routine, but my lack of proficiency in C++ would make it extremely inefficient.
If you know somebody among the developer interested in the topic I could work with him in this implementation.
Let me know!

[ibell]

Maybe @msaitta-mpr ?

[msaitta-mpr]

Is your idea to make a new h/s table, or perform a 1D solver on the existing p/h table?

[Marcoliveti97]

I don't really know how the tables are originally created, but I'd say that if the table is generated through p and h, a double interpolation should be enough to get a value from a given h and s. Am I incorrect?

[msaitta-mpr]

It should be possible to simply iterate on s, but there will be a reduction in speed. Often the reason that people use the tables is for the increased property lookup speed. I wasn't sure what you had in mind.

I can take a look into implementing an h/s input pair routine that uses such a routine.

[Marcoliveti97]

That's true, but I don't expect it to be slower than calling the EOS with h and s as inputs.
This would however allow consistency in codes where look-up tables are used, avoiding to call EOS (or PropsSI) whenever hs are used as inputs.

[Marcoliveti97]

Hello!
Is there any update on the issue? Do you think is it something that will be added to the code?

Thank you :)

[msaitta-mpr]

I have been playing around with the new code. However, I am still working to implement a solver for the saturation curves. Currently, they are tabulated by P/T, so a solver needs to be implemented to determine which phase (or two-phase) the state point is in. There are similar methods in the HEOS module, but I will need some more time to move one of them over and make it compatible.

[Marcoliveti97]

Excellent! Thank you for your time and dedication and for considering this request. Let me know if I can be of help in any way or if there is any news!

Cheers