* Update fmt submodule to 10.0.0.
* Update format wrapper, add format_as overloads.
- Every enum that is formatted needs to define a format_as overload in
the same namespace.
- Add a wrapper for sprintf for fmt versions without ArgList.
* Update index.rst
- Updating the PropiedadesDeFluidos link
- Adding a new app 'CuadernoDeProblemas' that uses CoolProp
* Update indext.rst with English description
PropiedadesDeFluidos and CuadernoDeProblemas link changed with short English description
* Update index.rst with English translation
Added English translation for the Spanish apps
* Liquid sodium properties added in incompressible liquids from range 400 K to 2500 K in steps of 100 K.
np.float changed to float in SecCoolFluids.py (Deprecated alias)
json file generated and added
* a bug in liquid sodium viscosity value corrected
* specific heat capacity fit was not proper for liquid sodium.
after removing two data points at the high temperature end, the fit is reasonable and is retained as of now.
* liquid sodium json file updated
* specific heat fit for liq sodium was not satisfactory for full range 400 K to 2300 K.
hence the range of values for specific heat of liq sodium reduced to 400 K to 1000 K and new coefficients updated
The rhoV auxilliary for n-Hexane had accidentally been set to use
reduced tau. This commit fixes the typo. This prevents bad guesses
for a number of property evaluations which use saturated states.
* Remove R1233zd(E) hL ancillary
The R1233zd(E) hL ancillary performs poorly. This can result in a
failure to converge when provided inputs (p/h) lie between the
ancillary saturation curve and the true saturation curve. Removing
the ancillary makes CoolProp fall back to a robust saturation
solver.
* Add 0.01 K cushion to Tmax in HSU_P_flash
Within HSU_P_flash, if an ancillary is used it can have a small
error. If the point of interest is near that saturation curve, this
error can prevent the solver from converging. Adding a 0.01 K margin
account for small ancillary errors and allow the solver to converge.
* Correct 0.01K cushion
The 0.01K adjustment should increase the potential range of the
solver. Previously, the minimum was increased, but it should be
decreased.
Verify that mole fractions are set before using them by implementing mole fractions as an internal object. This provides a more robust way to check if values have been set.
* Provide better feedback for bad DQ inputs
DQ is poorly defined (not unique) for densities above critical
density. Provide the user better feedback when they try to do this,
but do not give them a non-unique answer that might give them
trouble later.
* Prevent DQ warning for saturation curves
The saturation curves (Q=0, Q=1) are monotonic and invertible, so we
should not tell the user otherwise.
* Add small tolerance to DQ quality checking
Exact checking of floating point values is bad practice, as
iterative calculations can introduce small errors. This commit
ensures that when the DQ flash routine checks for the saturation
curves that it allows for some slop.
* Prevent DQ evaluation when D>Dcrit
The combination of density and quality inputs are not valid when the
density is greater than the critical density and the quality is
greater than 0. A small tolerance is added to allow for numeric
slop.
* Add discussion of Q to DQ warning
