Clapeyron v0.6.20

Diff since v0.6.19

v0.6.20

New Features

• revamp to Fugacity-based bubble/dew solvers. Improved speed of successive substitution iterations and support for Activity models.
• Activity models: support for second-order Michelsen TP flash. in VLE and LLE equilibria
• Activity models: support for QP/QT flash
• Improved implicit differentiation for all solvers.
• New method: thermodynamic factor (thermodynamic_factor)
• New method: eos_repr, to create parseable julia code capable of reproducing a model
• BACKSAFT: support for multicomponent mixtures.
• crit_pure: improved convergence.
• CoolProp: support for CoolProp 7.2
• Association: removed small static solvers for sizes 2-5 due to compilation slowdown.
• Cubics: improved robustness in the volume solver
• PCSAFT: improved robustness in the volume solver at low temperatures

Method deprecations

• ActivityDewTemperature and ActivityBubbleTemperature were removed, FugBubbleTemperature and FugDewTemperature are now the default for activity methods, with proper support for non-condensables/non-volatiles
• DETPFlash now uses another global optimizer, self-adaptive spherical search algorithm, with the main intention of dropping BlackBoxOptim as a dependency.

Bug fixes

• Convergence failure in Michelsen TP flash when equilibrium = :unknown and LLE was detected.
• Fixes on MultiphaseTPFlash
• Various fixes on saturation_pressure initial points
• Fixed an extra RT division in Obj_de_tp_flash; it now returns the molar reduced Gibbs energy G/(nRT).
• __eval_G_DETPFlash(model::EoSModel, p, T, ni, equilibrium) now returns reduced Gibbs energy, consistent with the PTFlashWrapper overload.
• gammaphi_gibbs now handles unnormalized inputs (mole amounts) correctly and always returns a (g, v) tuple for empty phases.

Merged pull requests:

• Stable computation of log(g_Mie) in SAFTVRMie (#491) (@Garren-H)
• Bump actions/checkout from 5 to 6 (#493) (@dependabot[bot])
• Adds thermodynamic factor (#494) (@se-schmitt)
• Add link to Electrolyte Models documentation (#496) (@se-schmitt)
• Update ref UNIFAC 2.0 (#497) (@se-schmitt)
• Fixes #438 (#499) (@Garren-H)
• improve stability of crit_pure (#501) (@longemen3000)
• Some documentation fixes (#503) (@se-schmitt)
• Implicit differentiation of root finders through implicit function theorem. (#507) (@Garren-H)
• typo in internal_energy_res (#510) (@Sush1090)
• Bump actions/cache from 4 to 5 (#514) (@dependabot[bot])
• Refactor DETPFlash, add SASS backend, and fix GammaPhi (#516) (@abcdvvvv)
• Fix DETPFlash reduced Gibbs scaling for GammaPhi/PTFlashWrapper (#517) (@abcdvvvv)
• Implicit AD: Corrected NestedAD error checking, added error tests and custom error displays (#518) (@Garren-H)
• crit_pure: Fixed solver residual check and AD call (#519) (@Garren-H)
• update NEWS (#520) (@abcdvvvv)

Closed issues:

• bubble point pressre and temperature (#395)
• SAFTgammaMie (and probably other GC methods) fails when using MonomerIdeal and custom compound names (#438)
• LLE pressure (#486)
• ph_flash failure for Multifluid models in the two-phase region (#492)
• IAPWS() Not returning correct water vapor pressure (#495)
• crit_pure returns numeric values even when the solver does not converge (#500)
• ps_flash bug (#505)
• ps_flash fails when phase set to vapour (#506)
• T_scale not defined for Cubic with ref state (#511)
• ReferenceState not setting for nitrogen (PCSAFT) (#512)
• tcRK R1243 saturation curve fails near critical point (#513)
• [Dependency] JSON version conflict caused by BlackBoxOptim; propose updating to master or switching backend (#515)