7c771647ad
adjustments for easier access to output data
2019-04-18 11:55:50 +02:00
c4784e6673
better readable
2019-04-13 15:34:51 +02:00
89679147e8
leaner group structure, centrally handled
2019-04-13 09:47:56 +02:00
52555d8c3c
key words don't contain small letters
...
moved more variables from numerics to crystallite
2019-04-11 11:27:03 +02:00
8f7239b75d
repetition not needed
...
- implicit none at the beginning of the module is enough
2019-04-11 07:35:58 +02:00
67eb39255a
keep connected data together
...
- avoids dependencies
- easier to read and modify
2019-04-11 07:24:04 +02:00
c3925b3497
small polishing
2019-04-07 14:47:21 +02:00
0b70f01e04
polishing
2019-04-07 14:28:08 +02:00
5075e1c2fb
constituent (ex crystallite) results are stored in HDF5
...
currently, not the best code but new structure for crystallite data will
fix that
output of orientations still missing
2019-04-06 12:08:25 +02:00
bfb6ad557f
WIP: crystallite HDF5 results
...
will be stored according to the phase sections
2019-04-06 12:08:25 +02:00
453eb538f7
preparing for PGI compiler
2019-04-03 16:24:15 +00:00
9759d3d041
forall is deprecated
...
- performance-wise, there should be no difference for the small loops we
have
- still, the on-liner syntax was much nicer
2019-04-03 13:25:01 +02:00
e33807aab3
fixed indentation
2019-04-03 12:54:07 +02:00
1d31c5c2db
pInt not needed
2019-04-03 12:32:30 +02:00
4604e65a42
use matmul instead of hand-written functions
...
- performance is the same
- leaner code
- matmul works (was buggy a few years ago)
2019-04-03 08:22:04 +02:00
60f4f9f39c
norm2 for PGI
2019-03-10 10:05:47 +01:00
2394880741
Tstar renamed to S (following the DAMASK paper)
2019-03-09 11:11:40 +01:00
05eb80d38c
pLongInt was not used
2019-03-06 15:49:31 +01:00
6e9b745ca1
shorterned line to avoid compile time truncation warning
2019-02-26 21:22:14 -05:00
ff5ac56cfb
shortened too long lines
2019-02-26 15:31:47 -05:00
cfb2770b93
merged and added correction to FreeSurface.config example
2019-02-26 13:56:49 -05:00
cb9543dff7
more elaborate debug labelling
2019-02-26 13:47:46 -05:00
56e2c1264b
Merge branch 'development' into New-Thermal
2019-02-26 07:54:45 +01:00
6fab99bca5
Merge branch 'development' of magit1.mpie.de:damask/DAMASK into development
2019-02-25 20:10:38 +01:00
d83847446d
Merge branch '32_NewStyleNonlocal-4' into development
2019-02-25 20:06:20 +01:00
7115382729
not used anymore
2019-02-24 10:08:09 +01:00
93ae5cbd07
Merge remote-tracking branch 'origin/development' into 10-consistent-orientation-conversions-3
2019-02-22 23:05:36 +01:00
194824fd0f
WIP: cleaned
...
no file reading
getting rid of a number of obsolete dependencies
2019-02-22 20:37:41 +01:00
435dce220c
move stress conversion one level up
...
should be totally avoided
2019-02-18 07:24:56 +01:00
cf32e7d1f5
use parameter structure and avoid conversion 33<->6
2019-02-17 12:15:46 +01:00
dcd16dda70
variables from mesh object
2019-02-02 12:15:05 +01:00
3a5a50cb03
use variables from theMesh
2019-02-02 10:53:55 +01:00
7a8d98d135
using theMesh (object oriented mesh description)
2019-02-02 10:35:10 +01:00
9a4e9e62b6
using new rotation class
2019-02-01 20:28:51 +01:00
407f94082f
no need for orientation class at the moment
...
implement only if we need symmetry aware operations
2019-02-01 16:52:42 +01:00
9d25d677e6
using new orientation class
...
not sure if transpose is needed for initialization
2019-02-01 10:11:46 +01:00
aabd98bee9
no need to repeat the same code
2019-01-31 09:14:02 +01:00
cbeb3dcff0
use the same formulation for convergence every where
2019-01-31 09:12:44 +01:00
5eaeb37ea4
just polishing
2019-01-31 06:04:49 +01:00
3b13a1af63
calculated convergence criteria wrongly
2019-01-30 17:04:58 +01:00
e1c2747393
logic error for nonlocal
2019-01-30 16:06:14 +01:00
fe88e5bf9c
[skip ci] cleaning
2019-01-30 14:52:12 +01:00
1d88057ce4
avoid superflous variables
2019-01-30 13:24:35 +01:00
64b89484d2
logic better visible
2019-01-30 13:19:05 +01:00
39e766bba0
improved readability
2019-01-30 12:36:02 +01:00
3dd21177a0
no need to store relative residual pointwise
2019-01-30 11:28:41 +01:00
fd069a96cd
unifying name
2019-01-30 10:51:24 +01:00
4ec0fd70a2
only one variable needed
2019-01-30 10:48:59 +01:00
0876787e3c
avoid loops
2019-01-30 10:46:53 +01:00
72c4f2b25f
same names everywhere if possible
2019-01-30 10:37:18 +01:00
0745d7ebc2
convergence flag is set only later
2019-01-30 10:33:57 +01:00
46be595ea8
no need to store relative residual for all points
2019-01-30 10:28:47 +01:00
31906e3ebd
no need for 2 loops
2019-01-30 09:21:50 +01:00
df6ec59f76
use "s" for source
2019-01-30 09:21:50 +01:00
ca7c105f36
only one loop needed
2019-01-30 09:21:50 +01:00
462b1b7c18
sorted according to importance
2019-01-30 06:47:36 +01:00
5908e3fd34
wrong tolerance selection
2019-01-30 06:44:26 +01:00
77f1f45c23
just figured out that RK4 integrator is totally broken
...
readable code helps ;)
2019-01-30 00:17:04 +01:00
a09036ff48
on-the-fly initialization
2019-01-30 00:11:10 +01:00
1a66f976b7
common variable name
2019-01-30 00:01:40 +01:00
6a3dac1df2
still improving readability
2019-01-29 23:45:41 +01:00
bdd193fbd7
now readable (kind of)
2019-01-29 23:31:26 +01:00
eade54a68f
consistent variable names
2019-01-29 23:04:50 +01:00
1408d66c0c
s is used for source
2019-01-29 23:02:59 +01:00
0be05b3ee1
one variable is enough
2019-01-29 22:46:21 +01:00
b4afc303be
clearer logic
2019-01-29 22:40:18 +01:00
73f39136c4
taking over from old branch
2019-01-29 22:19:38 +01:00
38d8e429ff
layout adjustments
2019-01-29 22:07:31 +01:00
066c598203
wrong dot product in state damper
2019-01-29 10:52:00 +01:00
34f3c15552
no need for temp variables
2019-01-29 07:24:06 +01:00
41832fb554
no need for two variables
...
only resulted in confusing code
2019-01-29 07:24:06 +01:00
a24d8b86bf
convergence of plastic state can be done earlier
2019-01-29 07:24:06 +01:00
4a69032637
better readable
2019-01-29 07:24:06 +01:00
1e4da6fbdb
nonlocal convergence check in function
2019-01-29 07:24:06 +01:00
3fdf8e19bb
further simplifications
2019-01-29 07:24:06 +01:00
9892da717a
bugfix: missing initialization
2019-01-29 07:06:16 +01:00
ee586dfa0c
avoid code duplication
2019-01-29 00:46:57 +01:00
2cf44f4060
shorter
2019-01-29 00:39:44 +01:00
2f9a571b96
no need for 2 variables
2019-01-29 00:38:18 +01:00
b62232022b
polishing
2019-01-29 00:27:58 +01:00
95cb404f81
further cleaning
2019-01-28 11:49:24 +01:00
8c2d6400b1
cleaning
2019-01-28 11:28:46 +01:00
f2882f195a
fuction for convergence check
...
avoid code repetition
2019-01-25 07:20:05 +01:00
f4fef6448d
stress integration for all points in one function
2019-01-24 17:59:38 +01:00
fcdab21565
avoid flush of full array
...
more clear logic
2019-01-24 14:15:26 +01:00
30dc8b4831
delta state update for all points
...
replaced stateJump, which works only on one point
2019-01-24 11:33:04 +01:00
c3b48c3484
WIP: update_deltaState
2019-01-24 07:34:30 +01:00
a8a5c8eec0
preparing function for deltaState
...
essential a "stateJump" over all points
2019-01-24 07:12:20 +01:00
a458dc831b
prepare for consistent use of full tensor representation
2019-01-24 06:56:43 +01:00
1cccd761cd
variables were not used
2019-01-23 23:18:14 +01:00
ae931c49a1
more "building block" separation
2019-01-23 23:02:21 +01:00
c60bb2edd3
function for update of dependent state
...
not introduced everywhere
2019-01-23 18:04:19 +01:00
0a6bcadafe
using a function for state update
...
avoids a lot of code repetition
2019-01-23 11:51:43 +01:00
b1522b1b9d
common function to update dot state
2019-01-23 06:14:19 +01:00
ac9d49f6be
state = subState0 + doState * dt
...
this is the usual state update used in all other integrators. also
in-line with logic in crystallite_stress
2019-01-22 23:23:48 +01:00
443519be72
cleaning
...
no reason to assume that the math functions are not thread safe
2019-01-19 10:02:04 +01:00
1be4426dc5
not needed
2019-01-19 09:54:37 +01:00
6fe099b978
[skip ci] re-ordered functions
...
internal/private functions at the end
2019-01-19 09:35:45 +01:00
30f28c9f4e
do concurrent causes problems on some Intel compilers
...
use forall instead.
Mandel/Plain prefixes not needed any more
2019-01-18 23:20:44 +01:00
695b331db0
takeover from old 46-XXX branch
2019-01-18 23:09:46 +01:00
221c587362
using separate functions for stress and tangent
...
extensively tested in 46-simplification-of-crystallite-f90-NEW3 already
2019-01-18 15:30:50 +01:00
406a2cc542
further separation
...
still using old combined function
2019-01-18 14:42:44 +01:00
e433aea193
preparing for separation of stress calculation and tangent calculatin
2019-01-18 12:18:13 +01:00
6049e292c1
no need to store converged tangent
2019-01-15 08:04:50 +01:00
80dca6d304
also not needed
...
was a remainder from time syncinc
2019-01-15 07:52:01 +01:00
daaa7cc2ba
internal (private) functions at the end
...
ordered state integrators according to their id
2019-01-15 04:27:57 +01:00
43f9d043d2
removed time syncing
2019-01-14 12:56:46 +01:00
3f40eeacf9
disorientation was never use
...
not even for nonlocal, but it slows down calculation a lot
2019-01-14 07:44:36 +01:00
d5cf7d2144
Merge branch '19-NewStylePhenopowerlaw' into 'development'
...
Resolve "New coding style for phenopowerlaw"
See merge request damask/DAMASK!39
2018-10-01 16:08:05 +02:00
fc016bbc6e
Merge branch 'development' of magit1.mpie.de:damask/DAMASK into development
2018-09-20 07:16:03 +02:00
310ea62964
only print out the essential information
2018-09-20 06:58:31 +02:00
fcff6b908a
can be easily computed during post processing
2018-09-20 06:40:23 +02:00
901355d2ae
don't use unnecessarily long names
2018-09-20 06:27:53 +02:00
df0464c31b
use (import) debug variables only when needed
2018-09-20 06:24:03 +02:00
6aa4dd842a
define debug variables only if needed
2018-09-20 06:09:02 +02:00
1623a33b48
cleaning (mainly OMP FLUSh)
2018-09-19 21:46:26 +02:00
11d4c28d88
flushes not needed + further cleaning
2018-09-19 19:45:57 +02:00
c313dc1675
only read access
2018-09-19 17:04:12 +02:00
0bf64645a1
should be done by the plasticity laws (for the moment)
2018-09-19 14:22:35 +02:00
a8fb7d7ade
not needed
...
but I'm under the impression that the compiler removes such things anyway
2018-09-19 14:21:10 +02:00
20f0bee459
fallback dPdF not needed
...
save a lot of memory
2018-09-19 06:19:40 +02:00
e6fa3f3d35
correct stress in postResults
2018-09-16 22:57:50 +02:00
f98243e4ac
Merge branch 'development' into 43-wrong-stress-in-dotstate-and-deltastate
2018-09-14 05:39:15 +02:00
ce6e6679d5
causes array access out of bounds, needs further checks
2018-09-11 12:20:05 +02:00
9be2c084e4
Merge branch '39-simplify-obscure-numerics-integration-mode' into 'development'
...
Resolve "simplify obscure numerics integration mode"
Closes #39
See merge request damask/DAMASK!36
2018-09-07 19:34:35 +02:00
a16454f1f9
fixed missing propagation of dependent state variables
2018-09-06 16:48:28 -04:00
b884349e7b
only renaming
...
3333 not needed for dX_dY if X and Y are 3x3 tensors
PK2 stress is S not T according to the DAMASK paper
2018-08-29 13:16:37 +02:00
fc3ce54667
return more than one tangent
2018-08-28 15:07:39 +02:00
4c14f988a3
rename according to paper
2018-08-28 14:58:17 +02:00
b8b5bac684
dotState and deltaState parse Mstar instead of Tstar
...
requires to parse in Fi
2018-08-25 15:59:34 +02:00
ab45818d51
seems to work now
...
anyway, nicer code
2018-08-22 14:30:51 +02:00
52088691d1
improved linked list and fixed solution for strange bug
...
Bug: Using automated LHS re-allocation for a string array that with global scope seems to cause trouble
Hence, "parse_file" works with a local string and assings only once to it
Linked_List: Now storing data in the list head also and last element is always empty.
Finalize allows simple handling of deallocation
2018-08-22 11:51:23 +02:00
3b3e0bc068
forgotten use
2018-08-18 17:11:50 +02:00
2618593568
:Merge branch 'development' into Integration-Mode-Confusion
2018-08-18 16:37:25 +02:00
b64b5b82de
integrationMode was always 1, very obscure code
...
was most likely the leftover of old functionality
2018-08-16 23:33:39 +02:00
7f05bf9c0a
leftover variables not needed anymore
2018-07-16 11:54:46 +02:00
771b8f00ec
leftover from performance reporting/debugging
2018-07-07 13:50:39 +02:00
3c1012ef38
cut out more now useless pieces
2018-07-03 22:12:25 +02:00
a308b2130a
Merge branch 'development' into 21_removeperformanceprofiling
2018-06-28 13:09:11 +02:00
fc83a76318
consistent naming
2018-06-26 20:54:54 +02:00
bb415e8a02
using array size not to rely on correctly set variable, cleaning crystallite data after use
2018-06-26 20:33:41 +02:00
1add0f3d84
gfortran does not recognice an empty array as 'present'
2018-06-22 08:03:22 +02:00
f61f22924a
empty string list as default value simplifies logic
2018-06-21 22:38:06 +02:00
5302782dad
need to search for cumulative tag
2018-06-19 23:42:15 +02:00
9aa211605f
having no output in crystallite caused trouble
2018-06-19 22:58:46 +02:00
9c12ce5539
more general name (should include parsing of debug and numerics)
2018-06-14 06:39:49 +02:00
91d9c11612
material.config is read in centrally
...
moving data from material to config_material. use statements need to
change. All aspects of reading from file will be removed from the
individual modules
2018-06-10 18:01:52 +02:00
e93e380ad3
output tag was empty:
2018-06-02 21:00:47 +02:00
142ce51c9c
homogenization and crystallite use new structure
...
math_transpose33 has no advantage over transpose intrinsic
2018-06-02 19:28:08 +02:00
c45d9f9ee9
Deleted performance profiling code
2018-05-17 16:33:35 +02:00
Martin Diehl