changelog tags manifest rss atom

shortlog for magma/3D/stgMADDs/Magma

navigate: (0) -300 -100 -60 tip

3 years dave 3D ridge model input velocity BCs now map correctly to the re-scaled
3 years dave adding the context as an argument when initialising the _* feVariables. otherwise have a NULL ptr issue
3 years dave need to explicitly set the operators for the sub block pcs else KSPSetUp fails - wasn't a problem for older version of PetscExt...
3 years dave adding some meta files - old ones causing build issues - and removing
3 years dave adding new .meta file - old one creating build issues
3 years dave removing old .met file - causing build issues
3 years dave adding new .meta files - old ones causing build issues
3 years dave removing .meta files - causing build issues
3 years dave fixing up petsc MatTranspose arguments for different PETSc versions
3 years dave added new .meta files - old ones causing build issues
3 years dave removing .meta files, causing build issues
3 years dave adding new .meta files - old ones were creating build issues
3 years dave removing old .meta files - causing build problems
3 years dave adding new .meta terms - old ones causing build problems
3 years dave modified the VecSetOption protype for different PETSc vetsions
3 years dave removing old .meta files - causing build problems
3 years dave replaced the .meta files - the old ones were causing build problems
3 years dave removing the old .meta files - causing build issues
3 years dave modified the .meta files - old ones were giving build problems
3 years dave .meta files causing building issues - replace
3 years www-data Added the v1.2.x branch, backdated to revision 468 (01-10-2008)
3 years dave courant number default of 1.0
3 years dave set the time step as dt = courant_number * mininum_node_seperation / wave_speed for the solitaryWaves model
3 years dave modifying the time step conditions
3 years dave only halve the minimum nodal seperation when using quadratic elements when calculating the time step
3 years dave adding a Courant condition time step for the solitaryWaves problem. still defines the time step by the wave speed by defult, but if this is not defined, then use the Courant condition
3 years dave adding additional meltvelocityZfield for 3D and dumping all the McKenzie eqn fields in the VTK output plugin
3 years dave fixed up the preconditioning for the initialis guess problem
3 years dave adding the 3d ridge model for the isoviscousMcKenzie system
3 years dave removed a dodgy field split call, and set the field split pc type
3 years dave fixing up the MatBlockMergeSubBlocks function to be consistent with the laterst PetscExt/dev interface
3 years dave using FGMRES KSP type for the block Jacobian
3 years dave modified the [vel,dyn] Jacobian block assembly
3 years dave commented out a function for generating the finite difference jacobian that interacted directly with the SNES and caused problems when building with Petsc-dev
3 years dave changing PETSc interfaces for PETSc dev
3 years dave changing PETSc interfaces for PETSc dev
3 years dave setting the isoviscous Mckenzie solver up for field split preconditioning. (note: need to build against petsc-dev and PetscExt/dev for this now, and some of the Petsc function prototypes are different for this (MatTranspose, VecSetOptions) )...
3 years dave removing PETSc wrapper classes and replacing with native PETSc calls
3 years dave depreciating McKenzie_SLE class
3 years dave depreciating McKenzie_SLE class
3 years dave removing PETSC wrapper classes and replacing with native PETSc calls
3 years dave removing PETSc wrapper classes and replacing with native PETSc calls
3 years dave have depreciated the MCKenzie_SLE class
3 years dave using LU preconditioning on each of the sub blocks for the Isoviscous McKenzie
3 years dave using the FeVariables to get the local element velocity and dynamic pressures.
3 years Luke Hodkinson Updating SConscript to operate like the other
3 years DavidLee using native PETSc SNES stuff, not the wrappers now for the SLE solvers
3 years DavidLee using optimised interpolate routines from the StgFEM function suite.
3 years DavidLee type casting issue resolved. not a problem for gcc, but tau_cc freaks out
3 years DavidLee using upper traingular preconditioning, and ilu for all the diagonal blocks (at least until
3 years DavidLee fixed a bug in the neumann flux terms evaluation
3 years DavidLee setting the y_solidus at 0 (just looks nicer and takes out the discontinuity)...
3 years DavidLee modifying viewport title
3 years DavidLee using Marc's initial porosity distribution (cause solver converges with this now)..
3 years DavidLee fixed a bug in the melt fraction evaluation (was using F, not dF/dy)
3 years DavidLee fixed a mug in the melt velocity
3 years DavidLee ...actually i bugged some already debugged assembly terms - fixing this (again)!
3 years DavidLee isoviscousMcKenzie XML:
3 years DavidLee IsoviscousMcKenzieAssembly:
3 years DavidLee ifixing a but in the assembly terms XML

navigate: (0) -300 -100 -60 tip