~reducedmodelling/fluidity/ROM_Non-intrusive-ann : changes

~reducedmodelling/fluidity/ROM_Non-intrusive-ann » Changes from revision 4037

From Revision 3301 to 3282

Rev	Summary	Authors	Date
3301	Fixes to the option mangling in sediments Fixes to the option mangling in sediments	jhill1	13 years ago
3300	Add a get_diagnostic, to fetch the previously computed value... Add a get_diagnostic, to fetch the previously computed value of a diagnostic, so that you can do things like add to it. Use this to compute the accumulated sum of all the components of the functional computed so far.	pef	13 years ago
3299	If you have more than one functional, the dump_no should be ... If you have more than one functional, the dump_no should be re-incremented, as we're dumping out multiple vtus for the same timestep. This stops the dump_no wrapping around, and giving you dump numbers like 2147483646.	pef	13 years ago
3298	The adjointer adds a timestep at the end of the run as a dum... The adjointer adds a timestep at the end of the run as a dummy container for the last equation registered. This timestep didn't really happen, and so shouldn't have a functional value associated with it. Take care of that.	pef	13 years ago
3297	Change the adjoint functional derivative callback to match t... Change the adjoint functional derivative callback to match the new interface. Also, change the variables passed through to the derivative computation.	pef	13 years ago
3296	change "void" to "empty" (in the "conserve volume" wetting a... change "void" to "empty" (in the "conserve volume" wetting and drying option) to get rid of the diamond validation warning. thanks jhill for spotting this!	sf1409	13 years ago
3295	Tidy up of some sediment options and adding additional check... Tidy up of some sediment options and adding additional checks to make sure the user isn't stupid. Also running spud-preprocess produced more documentation in the radiation.rng file, so I'm committing that too.	jhill1	13 years ago
3294	Fix the AD for the case where there is more than one dependi... Fix the AD for the case where there is more than one depending timestep	pef	13 years ago
3293	Some reshuffling of how timesteps are conceived of in libadj... Some reshuffling of how timesteps are conceived of in libadjoint.	pef	13 years ago
3292	Add a warning when the user returns an invalid dependency Add a warning when the user returns an invalid dependency	pef	13 years ago
3291	Incorrect comment in Momentum_Equation.F90 fixed. Incorrect comment in Momentum_Equation.F90 fixed.	avdis01	13 years ago
3290	some updates for the shallow water adjoint callbacks. some updates for the shallow water adjoint callbacks.	sf1409	13 years ago
3289	In Adjoint_Main_Loop.F90: In Adjoint_Main_Loop.F90: Fixed usage of uninitialised variable and some comments	sf1409	13 years ago
3288	Small correction to theta_divergence: old_u is necessary if ... Small correction to theta_divergence: old_u is necessary if 'use_theta_divergence', not if 'use_theta_pg'. Also small correction to the recent rotation on the sphere fix, just in case you happen to run something compressible on the sphere.	skramer	13 years ago
3287	Remove from the radiation populate state the energy group se... Remove from the radiation populate state the energy group set fields that where used for the production+scatter+delayed source, aswell as the corresponding group aliased versions Change the radiation group assemble solve to take as an optional argument the scalar field extra_discretised_source, which represents the production+scatter+delayed source Also in the call from radiation to the fluids assemble/Advection_Diffusion_CG.F90 main procedure solve_field_equation_cg include two extra optional arguments that have been added. These are the scalar field extra_discretised_source and an integer iterations_taken. Add a procedure to assemble/Advection_Diffusion_CG.F90 that adds the extra_discretised_source to the rhs if present and they have the same mesh type This is necessary for radiation problems as the coupling source fields between solution fields (ie. energy groups) are discretised with shape_shape, rather than shape_rhs. The former negates the need for a galerkin projection (unless different coordinate mesh ...). Also the radiation extra_discretised_source involves multiple material coeff fields. This is passed in as optional argument rather than inserted/extracted from state as a magic name on the recommendation of Stephan.	btollit	13 years ago
3286	Adding legacy_reservoir_prototype to the make clean list to ... Adding legacy_reservoir_prototype to the make clean list to avoid problems if the main fluidity library is updated but the multiphase code is not.	saunde01	13 years ago
3285	Test-cases: Test-cases: (a) BL_test_10elements_1d: Solving the standard 1D Buckley–Leverett equation. Added analytical solution for comparison. (b) BL_test_gravit and BL_test_gravity_unstable: Solving 1D Buckley–Leverett equation with positive and negative gravity term, respectively. Added analytical solution for comparison. (c) CompModel_1d_2Phases_2Components: Solving multiphase and multicomponent system of equations. In this case, 2 components and 2 phases initialised under equilibrium conditions (i.e., gas and liquid phases co-existing in equilibrium with parameterised EOS -- K-value) with injection of gas. (d) CompModel_1d_2Phases_2Components_NonEquilibrium: Solving multiphase and multicomponent system of equations. In this case, 2 components and 2 phases initialised under non-equilibrium conditions (i.e., just liquid phase) with injection of gas. Added analytical solution for comparison. (e) CompModel_1d_2Phases_3Components_Equilibrium: Solving multiphase and multicomponent system of equations. In this case, 3 components and 2 phases initialised under equilibrium conditions (i.e., gas and liquid phases co-existing in equilibrium with parameterised EOS -- K-value) with injection of gas. (f) CompModel_1d_2Phases_3Components_NonEquilibrium: Solving multiphase and multicomponent system of equations. In this case, 3 components and 2 phases initialised under non-equilibrium conditions (i.e., just liquid phase) with injection of gas. Added analytical solution for comparison. (g) CompModel_1d_2Phases_3Components_LVI_NonEquilibrium: Solving multiphase and multicomponent system of equations. In this case, 3 components and 2 phases initialised under non-equilibrium conditions (i.e., just liquid phase) with injection of gas. The intermediate component has K-Value smaller than one and has been injected along with the first component. Added analytical solution for comparison. (h) CompModel_1d_2Phases_3Components_LVI_NonEquilibrium_DG: Same as (g), but using DG. (i) BL_test_10elements_1d_DG: Same as (a), but using DG. Flml and xml files for some tests still need to be added.	jgomes	13 years ago
3284	I did not mean to submit my debugging lines I did not mean to submit my debugging lines	sf1409	13 years ago
3283	Updating the prototype code. All major developments are fini... Updating the prototype code. All major developments are finished. - Print statements were removed -- using ewrite now - Compositional model is working with good agreement against analytical and semi-analytical solutions. Special constraint for saturation and mass volume fraction were added to ensure the summation is equal to one and all phases and components are in equilibrium conditions.	jgomes	13 years ago
3282	remove the registration of the deprecated manifold projectio... remove the registration of the deprecated manifold projections.	sf1409	13 years ago