added preprocessor directive DEBUG_BLADED_INTERFACE to output ugly degubbing files when calling Bladed DLLs (this is to fix a problem caused by the MPI task not making separate copies of the dynamically loaded library) added STATIC_DLL_LOAD preprocessor directive for OpenFOAM to load the DLL on program start regardless of if it is being used added HubRad as InitOutput for FAST surface visualization added ED_UserSubs.f90, which is basically the ED subroutines from ServoDyn's UserSubs.f90 file, now in a module (so Doxygen doesn't get so confused) moved routines for file I/O and data validation from ElastoDyn.f90 into ElastoDyn_IO.f90. split part of ElastoDyn.f90 into ElastoDyn_IO.f90 updated ElastoDyn blade input meshes to receive point loads instead of distributed loads because previous mesh ignored any values input on the blade tip and blade root changed ElastoDyn tower input mesh from line2 to point because it didn't handle any loads transferred to the tower top or tower base correctly. added initialization output for tower base position and added accelerations for tower top and tower base (which involved defining several arrays starting at 0)
fixed type difference in UserSubs.f90/UserTeet() that prevented code from compiling converted the output array to single-precision before writing to text files to prevent asteriks in text files in double-precision Instead I increase the size of the time channel and print a warning about some NREL postprocessors. removed the fatal error when trying to output text files of long simulations (because it doesn't fit in the 10-character limit on the time channel). in file name) when the code aborts on error when visualization is requested, all meshes that currently exist are written out (with.
Added code to output VTK graphics files Code is based on work from Tom Brown (visiting professional from Eastern New Mexico University, summer 2015) added check so BD can't be in static analysis when coupled to FAST fixed issue where Orca data wasn't destroyed at end of simulation (caused issues in Simulink simulations). fixed copy of OtherStates in Orca, which we wouldn't notice because it doesn't use other states. changed intent(out) to intent(inout) on FAST_RestoreFromCheckpoint_T() because gfortran (gcc v4.6.2 from mingw32) ran out of memory when compiling it otherwise.
updated many comments for doxygen and reordered subroutines for better readability split FAST_Subs.f90 into FAST_Subs.f90 and FAST_Solver.f90 updated all modules except MAP++ to use new MiscVars type added check to compute jacobian for first 5 steps of simulations with OrcaFlex because Orcina's newest release (10.0?) didn't work when computing just the first 2 added Matlab function to read (text) linearization output files generated by FAST 8 added matlab script to convert FAST 8.15 input files to 8.16 modified registry command in compile scripts for IfW_UniformWind so it doesn't use -noextrap anymore updated FAST2Matlab.m and Matlab2FAST.m so that variables can contain the word "OutList" in their comments without it thinking it is the OutList section. Scripts were already set to ignore *writing* columns that didn't exist. This is useful for converting AD15 airfoil files, which may or may not contain the cm column. modified FAST2MATLAB/FAST2Matlab.m so that if it *reads* a table with more header columns than columns in the table, it ignores the headers for the missing columns. replaced SimulationToolbox with MATLAB_Toolbox Added Test 14, linearization and eigenanalysis fixed potential memory leak in CreateCheckpoint if it encountered an error fixed visualization rates specified in FAST summary file (times displayed were multiplied by DT_out instead of DT) fixed issue in complex FFT that caused it to crash with large arrays These two are particularly bad without preprocessor directives. changed case on BladedInterface.F90 and OrcaFlex/OrcaFlexInterface.F90 because Doxygen won't preprocess files unless they have uppercase file extensions. modified several modules to get their ModName from FAST instead of appending it inside the module fixed seg fault that occurred when you asked for VTK stick-figure files when using BeamDyn without AeroDyn * added new FAST_Lin.f90 source file and updated all compiling files with this change
* updated AeroDyn v15 input files with new FrozenWake input (also changed SumPrint to true because CertTest expected AD summary file) * added linearization section to primary FAST input files and documented it in ReadMe Added Linearization features for ED, AD, SrvD, and IfW: