Lines of Code
[root]/Packages/Uintah/CCA/Components/MPM/HeatConduction
| Author | Changes | Lines of Code | Lines per Change |
|---|---|---|---|
| Totals | 19 (100.0%) | 118 (100.0%) | 6.2 |
| guilkey | 6 (31.6%) | 54 (45.8%) | 9.0 |
| luitjens | 8 (42.1%) | 31 (26.3%) | 3.8 |
| jas | 4 (21.1%) | 28 (23.7%) | 7.0 |
| harman | 1 (5.3%) | 5 (4.2%) | 5.0 |
Switched the order of time integration from updating stress at the
top of the cycle to updating it at the bottom based on evidence that
for GIMP this provides significantly better accuracy and stability.
This will change the answers for everything.
42 lines of code changed in 2 files:
Make heat conduction in the explicit code optional, but on
by default. Turning it off via an input file flag results
in a substantial speedup. See bigbar.ups for usage.
2 lines of code changed in 1 file:
Deleted some commented out code, and commented out some unused code.
A few formatting and general cleanup changes as well.
0 lines of code changed in 2 files:
Merge from r38647 and r38648
10 lines of code changed in 1 file:
new->scinew (more coming tomorrow)
2 lines of code changed in 1 file:
Change the way interpolators are specified. Previously, this was done
with the "8or27" variable, but the new 3rd order bspline interpolator also
uses 27 nodes, so this is no longer sufficiently descriptive. The
new method uses a designation called "interpolator", which currently can
be linear or gimp (for 8 and 27 respectively) with new possibilities
in the works. Note that the d_8or27 variable is still part of MPMFlags,
but now instead of coming directly from the input file, it's value is
set based on which interpolator is specified. 8or27 has been removed
from all input files, and for most of them (except for some long unused
ones which were using the default value anyway)) it has been replaced with
the new specification. Bulletproofing has been inserted into MPMFlags
to exit and inform the user of this change is the old specification is found.
Cross your fingers.
10 lines of code changed in 1 file:
on the path to implicit heat conduction with AMR.
8 lines of code changed in 1 file:
Don't pass the number of nodes into the solvers, instead use the solvers
internal state to determine the number of rows.
PetscSolver.cc
Added some debugging code.
1 lines of code changed in 1 file:
Backed out revision 37716 and implemented a much simpler method of
allocating the diagonal.
0 lines of code changed in 1 file:
Force allocation of the diagonal for the ImplicitHeatConduction matrix.
Petsc's preallocation information is lost once a matrix is assembled and
thus any future calls to MatSetValues can cause mallocs. The implicit
heat code adds 1's to the diagonals at a later point since the diagonals
were previously zero this means that mallocs will be called.
2 lines of code changed in 1 file:
Fixed the preallocation code.
The coordinates of the nodes need to be mapped from global coordinates
into local coordiantes and then mapped back to a local index. The offset
to map to the local index is equal to the low index of the first patch
on each processor. Previously nothing was subtracted (which will work
on a single processor run) and before that the low index of the current
patch was subtracted (which will work in a single patch case).
5 lines of code changed in 1 file:
Fix for multi patch MPM performance issues.
Call createLocalToGlobalMapping once per patch set instead of once per patch.
Do not subtract the patches low corner from the local to global mapping.
3 lines of code changed in 1 file:
Performance improvement. Will not change results
-Changed list of boundary cells vector<IntVector> to vector<IntVector> *
-Nuked tabs
-changed variable names
bound -> bound_ptr
nbound -> nbound_ptr
sfcx -> sfcx_ptr
sfcy -> sfcy_ptr
sfcz -> sfcz_ptr
5 lines of code changed in 1 file:
Add further refinements for getting heat flux polynomial data to the particles.
28 lines of code changed in 4 files: