-
Notifications
You must be signed in to change notification settings - Fork 2
/
ChangeLog.txt
72 lines (60 loc) · 5.02 KB
/
ChangeLog.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
BeamDyn Change Log
================================================================================
v1.05.00 9-Dec-2016 Envision Energy USA, Ltd.
- added a check that the values for the damping are similar
- bug fix: the P_d portion of the stiffness matrix (Eq 5.43 in the BeamDyn manual) was implemented with incorrect indices, so the (4:6,1:3) entry
was getting overwritten by what should have been the (4:6,4:6) entry of that matrix. This does not seem to affect the 5MW model results; however
the (5,1) entries of the 6x6 submatrices that make up the stiffness matrix reported in the BeamDyn summary file
- cleaned up the code:
+ Created new routines to isolate more of the code in the BD_Init routine.
+ added some comments
+ removed unused variables
+ changed BD_ElasticForce() and BD_DissipativeForce() so that we don't have to create temporary variables that don't get used when they are called when fact=.false.
+ replaced BD_Tilde() with SkewSymMat() and/or cross_product() from the NWTC Library
+ redimensioned several arrays so they can be indexed easier (This appears to be working, but I [adp] have not tested with multiple elements.
There is potential that the multiple element handling is now broken, but I don't have a test case for that right now.)
+ added some local variables to miscVar data type
+ reordered some loops for efficiency
+ removed some routines that were only calling another subroutine without actually doing anything
Bypassed several subroutines that were only passing to another subroutine without actually doing anything.
+ added BD_TrapezoidalPointWeight routine to initialize the trapezoidal quadrature points and weights in the same routine.
+ I also renamed some local variables to match the parameter-type versions of the same variable.
+ Removed many ErrStat/ErrMsg arguments that weren't used.
+ Added parameters for GAUSS or TRAP quadrature instead of always checking for integers 1 or 2
+ Replaced flags in BD_CrvCompose with compile-time parameters for easier reading of code
+ Modified BD_ComputeIniNodalPosition so it uses the same indexing order as SP_Coef, allowing us to get rid of a temporary matrix holding the transpose of SP_Coef
+ BD_ComputeIniCoef was computing and factoring the coefficient matrix 3 more times than necessary so I rearranged some of the code to avoid that.
================================================================================
v1.01.04 26-Jul-2016
- added the full finite-element mass and stiffness matrices of the blade model at time zero to the summary file
- fixed a problem in BeamDyn whereby its velocities were initialized incorrectly when ElastoDyn was initialized
with nonzero deflections/displacements
================================================================================
v1.01.03 12-Apr-2016
- modified validity check to use single-precision variables
================================================================================
v1.01.02 5-Apr-2016
- fixed a bug that would use the wrong node numbers in gaussian quadrature
- fixed some initialization problems and type conversions reported/fixed by Johannes Rieke
- incorporated some changes by Johannes Rieke (precision differences so driver can be compiled in single precision, too;
moved some subroutines from Driver_Beam.f90 to Driver_Beam_Subs.f90)
- fixed makefile so it will compile the sources distributed in archive
- fixed problem in initial configuration (At0 E10) whereby an initially curved beam induces strain and motion.
- updated "xb", "yb" and "zb" text on key point coordinates in input files to match "xr", "yr", "zr" in documentation
- added the MiscVar type (split from OtherStates) to conform to changes to FAST framework
- updated comments to doxygen style
- moved the section of CalcOutput that computes the displacements and velocities for the BldMotion mesh into a subroutine.
explicitly initialized BD output mesh with this subroutine due to coupling with AD in option 2 solve in FAST.
This allows AeroDyn to have a better initial guess.
- reworked some subroutine arguments so that we send parameter and MiscVar types instead of individual components of those types.
moved some of the local variables that were allocated each time step into the miscVar type. This has sped up FAST's Test 26 2-5%.
- removed a few variables the compiler marked as unused
- added validity checks on mass matrix, M, from input file: M(1,1)=M(2,2)=M(3,3) and M(4,4)+M(5,5)=M(6,6) (in input file coordinates) for each input station
- added validity check that kp_zr(1) must be zero.
- added BD_STATIC_ANALYSIS and BD_DYNAMIC_ANALYSIS constants for check (in FAST) that static analysis isn't used when coupled to FAST
- added text to indicate mapping of element/node number to global node number in summary file
- updated the manual
================================================================================
v1.00.00 7-Oct-2015
Initial release
================================================================================