refactor: clarify new gravity, clean up kernel flags usage, move CFL computations to FVM solver (#3486)

Co-authored-by: Matteo Cusini <49037133+CusiniM@users.noreply.github.com>

Author: paveltomin

.integrated_tests.yaml

@@ -1,6 +1,6 @@
 baselines:
   bucket: geosx
-  baseline: integratedTests/baseline_integratedTests-pr3479-9362-cffefcc
+  baseline: integratedTests/baseline_integratedTests-pr3486-9492-f0c817c
 allow_fail:
   all: ''
   streak: ''

BASELINE_NOTES.md

@@ -6,6 +6,10 @@ This file is designed to track changes to the integrated test baselines.
 Any developer who updates the baseline ID in the .integrated_tests.yaml file is expected to create an entry in this file with the pull request number, date, and their justification for rebaselining.
 These notes should be in reverse-chronological order, and use the following time format: (YYYY-MM-DD).
 
+PR #3486 (2025-01-06)
+=====================
+useNewGravity became gravityDensityScheme.
+
 PR #3479 (2024-12-15)
 =====================
 Refine inputFiles/compositionalMultiphaseFlow: shift reference pressures to initial pressures, make nonlinear tuning more reasonable, minimize output.

src/coreComponents/finiteVolume/docs/FiniteVolume.rst.txt

@@ -16,17 +16,27 @@ The numerical flux is obtained using the following expression for the mass flux
 .. math::
   F_{KL} = \Upsilon_{KL} \frac{\rho^{upw}}{\mu^{upw}} \big( p_K - p_L - \rho^{avg} g ( d_K - d_L ) \big),
 
-where :math:`p_K` is the pressure of cell :math:`K`, :math:`d_K` is the depth of cell :math:`K`, and :math:`\Upsilon_{KL}` is the standard TPFA transmissibility coefficient at the interface.
+
+where :math:`p_K` is the pressure of cell :math:`K`, :math:`\rho^{avg}` is the average fluid density, :math:`d_K` is the depth of cell :math:`K`, and :math:`\Upsilon_{KL}` is the standard TPFA transmissibility coefficient at the interface.
 The fluid density, :math:`\rho^{upw}`, and the fluid viscosity, :math:`\mu^{upw}`, are upwinded using the sign of the potential difference at the interface. 
 
-This is currently the only available discretization in the :ref:`CompositionalMultiphaseFlow`. 
+For :ref:`CompositionalMultiphaseFlow` there are two options to compute the average density, :math:`\rho^{avg}`. The desired option can be selected using the `gravityDensityScheme` parameter:
+
+#. `ArithmeticAverage`: :math:`\rho^{avg}` is computed using simple arithmetic average: :math:`\rho^{avg} = 0.5 \cdot (rho_K + rho_L)`, where :math:`rho_K` and :math:`rho_K` are densities in the two cells.
+
+#. `PhasePresence`: average phase density is computed using checking for phase presence:
+
+    * :math:`\rho^{avg} = 0.5 \cdot (\rho_K + \rho_L)` if phase is present in both cells :math:`K` and :math:`L`
+
+    * :math:`\rho^{avg} = \rho_K` if phase is present only in cell :math:`K`
+
+    * :math:`\rho^{avg} = \rho_L` if phase is present only in cell :math:`L`
 
 Hybrid FVM
 ~~~~~~~~~~
 
 This discretization scheme overcomes the limitations of the standard TPFA on non K-orthogonal meshes.
 The hybrid finite-volume scheme--equivalent to the well-known hybrid Mimetic Finite Difference (MFD) scheme--remains consistent with the pressure equation even when the mesh does not satisfy the K-orthogonality condition.
-This numerical scheme is currently implemented in the `SinglePhaseHybridFVM` solver.
 
 The hybrid FVM scheme uses both cell-centered and face-centered pressure degrees of freedom.
 The one-sided face flux, :math:`F_{K,f}`, at face :math:`f` of cell :math:`K` is computed as:
@@ -60,5 +70,3 @@ For a given interior face :math:`f` between two neighboring cells :math:`K` and
 We obtain a numerical scheme with :math:`n_{\textit{cells}}` cell-centered degrees of freedom and :math:`n_{\textit{faces}}` face-centered pressure degrees of freedom.
 The system involves :math:`n_{\textit{cells}}` mass conservation equations and :math:`n_{\textit{faces}}` face-based constraints.
 The linear systems can be efficiently solved using the MultiGrid Reduction (MGR) preconditioner implemented in the Hypre linear algebra package.  
-
-The implementation of the hybrid FVM scheme for :ref:`CompositionalMultiphaseFlow` is in progress.

src/coreComponents/physicsSolvers/PhysicsSolverBase.cpp

@@ -458,15 +458,6 @@ real64 PhysicsSolverBase::setNextDtBasedOnNewtonIter( real64 const & currentDt )
   return nextDt;
 }
 
-
-real64 PhysicsSolverBase::setNextDtBasedOnCFL( const geos::real64 & currentDt, geos::DomainPartition & domain )
-{
-  GEOS_UNUSED_VAR( currentDt, domain );
-  return LvArray::NumericLimits< real64 >::max;       // i.e., not implemented
-}
-
-
-
 real64 PhysicsSolverBase::linearImplicitStep( real64 const & time_n,
                                               real64 const & dt,
                                               integer const GEOS_UNUSED_PARAM( cycleNumber ),

src/coreComponents/physicsSolvers/PhysicsSolverBase.hpp

@@ -239,17 +239,6 @@ class PhysicsSolverBase : public ExecutableGroup
   virtual real64 setNextDtBasedOnStateChange( real64 const & currentDt,
                                               DomainPartition & domain );
 
-  /**
-   * @brief function to set the next dt based on state change
-   * @param [in]  currentDt the current time step size
-   * @param[in] domain the domain object
-   * @return the prescribed time step size
-   */
-  virtual real64 setNextDtBasedOnCFL( real64 const & currentDt,
-                                      DomainPartition & domain );
-
-
-
   /**
    * @brief Entry function for an explicit time integration step
    * @param time_n time at the beginning of the step