Commit Graph

6630 Commits

Author SHA1 Message Date
Martin Diehl 44e5644e78 fixed random Gaussian sampling
sampling needs to be performed from unfiform misorientation, NOT uniformly distributed rotations
for Fiber, compute uniform tilt of Fiber axis
2018-03-13 03:36:24 +01:00
Test User fb710dda0a [skip ci] updated version information after successful test of v2.0.1-1110-g52c9954 2018-03-12 22:56:52 +01:00
Pratheek Shanthraj 52c9954ca4 Revert "more work to get testing to work"
This reverts commit 63c104b374.
2018-03-12 17:44:37 +01:00
Pratheek Shanthraj 63c104b374 more work to get testing to work 2018-03-12 17:35:54 +01:00
Zhuowen Zhao 12adcec41e updated PRIVATE submodule to new state 2018-03-09 18:36:34 -05:00
Zhuowen Zhao 361cf680a5 merged more recent development 2018-03-09 18:19:41 -05:00
Zhuowen Zhao f26fd1d1dc Fixed a physics issue: sense change with respect to difference between resolved stress and backstress 2018-03-09 18:01:09 -05:00
Martin Diehl 8fa8427c99 simple test, only ensures that scatter is not bigger than 2 FWHM (hard max) 2018-03-08 17:35:59 +01:00
Martin Diehl 0ccce7facc fixed missing angle initialization, simplified and commented 2018-03-08 17:33:54 +01:00
Test User 48fce3a49e [skip ci] updated version information after successful test of v2.0.1-1107-g2c3eae6 2018-03-08 10:15:17 +01:00
Philip Eisenlohr 2c3eae6bf0 fixed tests and bug for very small unit cube fractions when grid is also small 2018-03-07 18:59:27 -05:00
Philip Eisenlohr 32c34457c2 direct selection of valid seed coordinates within subfraction of unit cube instead of check and repeat 2018-03-07 17:45:22 -05:00
Test User 03c643c433 [skip ci] updated version information after successful test of v2.0.1-1104-g88c46d5 2018-03-01 06:25:54 +01:00
Philip Eisenlohr 88c46d500e Merge branch '13-higher-order-thermal-expansion' into 'development'
Resolve "higher-order thermal expansion"

Closes #13

See merge request damask/DAMASK!19
2018-02-28 21:08:01 +01:00
Test User f18f076a09 [skip ci] Merge branch 'master' into development 2018-02-27 19:18:42 +01:00
Test User 35e4244887 Merge commit 'v2.0.1-1098-g5c2aeed' 2018-02-27 19:18:18 +01:00
Test User 5c2aeedb73 [skip ci] updated version information after successful test of v2.0.1-1097-gcb073fe 2018-02-27 19:18:18 +01:00
Test User f5f5791b39 [skip ci] Merge branch 'master' into development 2018-02-27 10:01:09 +01:00
Test User 3a4506c5ae [skip ci] updated version information after successful test of v2.0.1-1088-gf35a88d 2018-02-27 10:01:02 +01:00
Philip Eisenlohr cb073fe127 Merge branch '14_LoadCaseRotationPerformance' into 'development'
14 load case rotation performance

See merge request damask/DAMASK!20
2018-02-27 00:41:20 +01:00
Martin Diehl f35a88d2c4 it's 2018 already 2018-02-26 23:51:23 +01:00
Martin Diehl 1bbaa7cefe reverted last commit, scipy.linalg is different from numpy.linalg 2018-02-26 23:50:44 +01:00
Martin Diehl 6046a7ae3d there is no scipy.linalg (at least on my system)
numpy.linalg is automatically imported together with numpy
2018-02-26 20:22:57 +01:00
Martin Diehl e716441787 indexing now follows description 2018-02-26 20:18:38 +01:00
Martin Diehl cd4115bada now using the correct rotation formula 2018-02-26 11:14:02 +01:00
Test User 30d66c2d67 [skip ci] updated version information after successful test of v2.0.1-1084-gb078234 2018-02-26 09:50:06 +01:00
Martin Diehl e1c4c7e6e8 wrong indices in matrix rotation
fixing this resolves issue 14: "Poor performance after load case rotation"
now inline with http://www.continuummechanics.org/rotationmatrix.html
C'_mnop = R_mi R_nj R_ok R_pl C_ijkl
2018-02-25 21:01:23 +01:00
Philip Eisenlohr b078234644 added scipy and scipy.linalg as known modules for calculations 2018-02-25 14:19:22 -05:00
Philip Eisenlohr bc998846ab thermal expansion coefficient alpha = a + b(T-T_0) + c(T-T_0)^2 2018-02-25 14:14:03 -05:00
Martin Diehl c2ced6a274 not needed anymore 2018-02-25 14:45:35 +01:00
Martin Diehl d5d71d2321 math_transpose33 has no benefit here 2018-02-25 13:42:02 +01:00
Martin Diehl 42500eabff simplified (math_transpose33 does not provide any benefit here) 2018-02-25 12:48:58 +01:00
Martin Diehl eedec74a6a correct debug reporting 2018-02-25 12:31:23 +01:00
Martin Diehl 86768c38d9 source code should not be executable 2018-02-25 10:16:27 +01:00
Martin Diehl 4f4fa5daf8 simplified 2018-02-22 14:54:49 +01:00
Martin Diehl e51de7ffd8 explicitly select halton bases in call 2018-02-22 14:16:36 +01:00
Martin Diehl c6c66bb653 using proper Gauss sampling also for the fiber components
function most probably still contains a bu
2018-02-22 00:42:06 +01:00
Martin Diehl 5c908e44ec correct scaling for FWHM
FWHM was wrong by a factor of approx 2.
Analytic expression can be found in https://en.wikipedia.org/wiki/Gaussian_function
2018-02-22 00:10:38 +01:00
Martin Diehl 9173d12d14 correct algorithm for sampling of uniform orientations and fix for Halton series
Halton series gives strange results for large prime numbers, now always starting with 2 for first dimension, 3 for second etc.
Consecutive Halton numbers for rejection sampling seem to cause problems (i.e. introduce patterns).
Algorithm for uniformly distributed orientations with FWHM specified is taken from https://math.stackexchange.com/questions/131336.
WIP: Gauss filtering is currently not implemented!
2018-02-21 20:32:52 +01:00
Martin Diehl ae27660e86 simplified halto procedure (still needs testing) 2018-02-21 18:47:39 +01:00
Martin Diehl 98df2d1427 better description and names 2018-02-21 16:49:29 +01:00
Martin Diehl a52f54a9a0 fixed typo in prime number list and extended to 1600 values
source https://people.sc.fsu.edu/~jburkardt/f_src/halton/halton.f90
2018-02-21 14:37:02 +01:00
Martin Diehl 6b1c1af609 no early return required here 2018-02-21 08:55:33 +01:00
Martin Diehl 051f3592ea Merge branch 'development' of magit1.mpie.de:damask/DAMASK into development 2018-02-21 07:53:00 +01:00
Test User a901a81ba8 [skip ci] updated version information after successful test of v2.0.1-1082-g8d81e12 2018-02-21 01:41:57 +01:00
Philip Eisenlohr 8d81e12c96 Merge branch '11-restart-from-instead-of-restart-at' into 'development'
Resolve "Restart FROM instead of restart AT"

Closes #11

See merge request damask/DAMASK!18
2018-02-20 16:22:37 +01:00
Philip Eisenlohr f5177b6d2d [skip ci] added option for non-periodic geometries 2018-02-20 10:20:35 -05:00
Test User 328c35a099 [skip ci] updated version information after successful test of v2.0.1-1078-g8949d3c 2018-02-17 02:26:51 +01:00
Philip Eisenlohr ad1179117d modified meaning of restart increment to "restart from" instead of "restart at" 2018-02-16 16:41:07 -05:00
Philip Eisenlohr 8949d3cb7c use consistency tests with increased itmin 2018-02-16 10:59:50 -05:00