Add multiple reactions in particles

[AntoniaBerger]

This pull request introduces improvements to the reaction modeling infrastructure in the codebase, primarily by integrating a new ReactionSystem abstraction that supports multiple reaction types (liquid, solid, cross-phase) and refactoring the configuration and residual calculation logic to use this new system. The changes also update various model classes to utilize the new reaction interface and extend support for cross-phase reactions.

Reaction system integration and refactoring:

• Added the ReactionSystem abstraction to GeneralRateModel, LumpedRateModelWithPores, and LumpedRateModelWithoutPores, replacing direct use of _dynReactionBulk and related vectors, and providing helper methods for reaction offsets and management. (src/libcadet/model/GeneralRateModel.hpp [1] src/libcadet/model/LumpedRateModelWithPores.hpp [2] src/libcadet/model/LumpedRateModelWithoutPores.hpp [3]
• Refactored configuration logic in LumpedRateModelWithoutPores to support multiple reaction types (liquid, solid, cross_phase) using the new ReactionSystem, and updated workspace management accordingly. (src/libcadet/model/LumpedRateModelWithoutPores.cpp [1] [2] [3]
• Updated the residual and Jacobian calculation logic to use the new reaction system, supporting multiple reactions per type and properly handling cross-phase, liquid, and solid reactions in the computation. (src/libcadet/model/LumpedRateModelWithoutPores.cpp [1] [2]

API and interface updates:

• Changed calls from residualLiquidAdd and analyticJacobianLiquidAdd to the more general residualFluxAdd and analyticJacobianAdd in several model implementations, reflecting the new interface. (src/libcadet/model/GeneralRateModel2D.cpp [1] src/libcadet/model/GeneralRateModelDG.cpp [2] [3] src/libcadet/model/LumpedRateModelWithPoresDG.cpp [4] [5] src/libcadet/model/LumpedRateModelWithPoresDG2D.cpp [6]

Support for cross-phase reactions:

• Registered and integrated MassActionLawReactionCrossPhase as a new reaction type, with corresponding registration in the factory and inclusion in the build system. (src/libcadet/CMakeLists.txt [1] src/libcadet/ReactionModelFactory.cpp [2] [3]

Include and dependency updates:

• Added necessary includes for ReactionSystem in model header files to support the new abstraction. (src/libcadet/model/GeneralRateModel.hpp [1] src/libcadet/model/LumpedRateModelWithPores.hpp [2] src/libcadet/model/LumpedRateModelWithoutPores.hpp [3]

[AntoniaBerger]

The new suggested user interface for cross phase reactions will be similar to the interface for bulk reactions:

model.root.input.model.unit_000.particle_type_000.nreac_cross_phase= 1
model.root.input.model.unit_000.particle_type_000.cross_phase_reaction_000.type = "MASS_ACTION_LAW_CROSS_PHASE"
model.root.input.model.unit_000.particle_type_000.cross_phase_reaction_000.mal_stoich_liquid = [-1, 0]
model.root.input.model.unit_000.particle_type_000.cross_phase_reaction_000.mal_stoich_solid = [0, 1]
model.root.input.model.unit_000.particle_type_000.cross_phase_reaction_000.mal_kfwd_liquid = [1.0]
model.root.input.model.unit_000.particle_type_000.cross_phase_reaction_000.mal_kbwd_solid = [0.5]

Up to discussion is still, how we want to continue with parameters like MAL_EXPONENTS_LIQUID_FWD_MODSOLID.

And for reactions located only in the pore (liquid) phase I purpose:

model.root.input.model.unit_000.particle_type_000.nreac_liquid= 1
model.root.input.model.unit_000.particle_type_000.liquid_reaction_000.type = "MASS_ACTION_LAW"
model.root.input.model.unit_000.particle_type_000.liquid_reaction_000.mal_stoich = [-1, 0]
model.root.input.model.unit_000.particle_type_000.liquid_reaction_000.mal_kfwd = [1.0]
model.root.input.model.unit_000.particle_type_000.liquid_reaction_000.mal_kbwd = [0.5]

To use the same reaction modul for different phases, I gave up the _liquid / _solid description.
For more information about the general plan see #92
For more and about the documentation files.

[jbreue16]

Should we have a different field in the LRM, ie liquid_reaction_000 instead of bulk ? i think this is more descriptive and breaking the interface compared to other unit operations is actually good as it emphasizes that the phase considered int he model is indeed something else.

Also, do we want the names to start with reaction_ or with the description, followed by _reaction ? I dont mind either way but both ways were wrote down in two different comments above.

[AntoniaBerger]

Should we have a different field in the LRM, ie liquid_reaction_000 instead of bulk ? i think this is more descriptive and breaking the interface compared to other unit operations is actually good as it emphasizes that the phase considered int he model is indeed something else.

I agree, it is the most precise name.

Also, do we want the names to start with reaction_ or with the description, followed by _reaction ? I dont mind either way but both ways were wrote down in two different comments above.

I switched it to phase reaction_xxx because, in my opinion, it is more fluid. I need to update the comment, though.

[AntoniaBerger]

This will PR will be continued in #491