Added a main routine to make calls to 'solve_fe'. Plots can now be generated and saved to file.
Added stabilization to the 2D fe code. Tests with various element sizes, shapes and convection vector directions show good stabilization performance.
Added a small 2D Finite Element code for testing purposes
Regarding the dependency on the reference node by using a regular triangle as the reference element, new code is added including a transformation from the regular to a equilateral reference element which defines a equilateral mesh element with edge length 1 as being the reference with size equal to 1, the 'one-element'.
Added routines to dosplay an animation on the rotation of either gamma or the element to the testing routines for nu_art computation
renamed the file 'coord_transform' to 'test_nu_art_computation' to indicate that the actual testing of computation of artificial diffusion tensor in the element space 'nu_art' is done in this routine including a conditional cll to plotting
added some missed comments
Implementation of python scripts to test and play around with a computation of a artificial diffusion tensor in the reference element spaye and the efects of the gamma direction and the element shape
First implementation (Testing Stage) of direct Stabilization without Pe and Da computation by means of the element Jacobi matrix. Additional examples for testing
Implementation of Peclet and Damkoehler Numbers calculation trhough inverse of the Element Matrix Jacobian. Implementation only, tests outstanding.
Implementation of distinct Peclet and Dahmkoehler computation algorithms either as integral of coefficient quotients or as quotients of coefficient integrals (IOQ, QOI))
Added an initial implementation of Pe & Da tensors
Construction scripts for four homogeneous meshes added to the 2D-DI Project
Addition of 0.1u thick meshes to the 2D-DI example simulation
Example Simulation pseudo 2D Semiconductor Resistor diode 2D-DI now working
Updates to the 2D-DI Example
Minor cleanup and bug fixes. Adding new pseudo 2D semiconductor simulation.
Updates to carrier life times in the Semicunductor simulation
Updated lauch JSON files for Linux and MacOS
Update fieldquantity3D.f90
Update launch_mac.json
Integration of separated build for Debug and Release
Added stab_pe and stab_damk to the list of variables for the OpenMP case
In the Example of the Eikonel Equation fixed the Dirichlet-Boundary ambiguity at light entrance, adjusted script to fix missing beta terms
Minor re-naming and fixes
Corrected errors in the implementation of '3D_DIFF_ADVECT_REACT' Physics Mode
Corrected 'beta' Jacobian Terms
Minor Fixes
Merge branch 'master' of https://git.code.sf.net/p/polyde-fem/git
reactivated memory release
Added a net Python plotting routine that does need FEMProject.json File, but instead receives the right Probes-Path as argument
Bug Fixes
Initial implementation of Eikonal Study Project (including extending the 3D_DIFF_ADVECT_REACT Physics Mode with additional terms)
Minor updates
Ported to MacOS and adapted to work without Intel MKL. Tested with Examples.
Added sf_stab-control to both stabilization parameters.
Mostly cosmetic changes, minor corrections
Reactivation of the Semiconductor Natural-Formulation 3DSEMICONDUCTOR_NAT physics mode
Addition of first Examples for 3D Simulations, more to add.
Corrected formulas for calculation of Pe and Da numbers
Additional Print of Convergence data at Iteration origin
Implementation of Two-Parameter Stabilization. Tests outstanding.
Re-Implementation of the VTK-Solution Export Feature
Implementation of a central modules directory for the building process
Code improvements due to new interface to fieldquantity3d
Swap23 Feature implemented
Improvements to the Ray-Tracing Feature Wavelength is a (outside) physics parameter now
Minor Modifications needed for the Physical Scaling Feature
Minor Modifications needed for the PROBES Feature
Implementation of new linear Solvers and Re-implementation of MKL Pardiso with gfortran
Jacobian Stabilization Implementation
Probes Feature Data Type
Non-Linear Solver Re-Worked
Physical Scaling Features
Implementation of the PROBES Feature in PolyDE
Minor Improvement
Updated according to Intel OneAPI 2023
Minor Improvements and Cleanup
Addition of PSBLAS and Intel MKL Folders
Cleanup Packages
Implementation of 3D EMWAVE physics modes
Implementation of the scaling factors
Implemented MKL PARDISO Libraries with gfortran compiler
New Settings related to Load limits and Jacobian Stabilization control
Fix for what seems to be a gfortran c-preprocessor bug
Added the PSBLAS Package
Resolved the ambuguity with four diferent files for the linear solver routine.
Added MUMPS CMake Files, The Solver is not integrated yet
Cleaned mgmres
updated the physics mode
Minor fixes and clean up
Re-activation of Parametric solver study for the 3DMTSO Physics mode including optical carrier generation
minor additions
Implementation of Ray-Tracing
Addition of 'SP' - single precision matrix mode to the PARDISO solver
Auxiliarry and helping routines that support implementation of the ray-tracing and new options in cunjunction with the 3DMTSO physics mode
Test implementation of Jacobian stabilization
Single-Precision PARDISO solver interface
Update of the CMakeList
Updates to the CMakeLists
Minor changes and clean-up
Implementation of 3DMTSO - Mechanics Thermal Semiconductor Optical Simulation
New field quantities for evaluation of Jacobian Peclet Numbers: first implementation
A Testing implementation of a subroutine to detect active dofs on the global dof basis
Variable 'IDAMAX' interfeared with a same-name function of UMF solver. changed the naming to 'IDA_MAX'
minor tweaks, modifications, bug_fixes, cleanup
additional flag for linux
Implemented an additional global variable that controlls biasing conditions at the boundary conditions without affecting the sf_load which applies to the whole RHS. Modified loaddescent-nonlinear solver makes use of it.
Call to 'tabdat2mesh' moved to solver3d
added new global variable sf_bias to the list and changed the default values for sf_load and sf_relax
get_jacpec is now parallel
started '3DMTSO' implementation - not the final version yet
Minor tweaks, Feminterface changed because implementation of user-definable BC-Biasing required to do so.
A SUPG Stabilization - first implementation. Still in testing phase.
updated settings
Minor corrections and some clean-up