Conversation
- remove unused variables - convert `print*` to `CALL WrScr()` - remove AA version number - efficiency updates for powers of numbers: when exponent is an integer, don't make it a real number (e.g., use `x**2` instead of `x**2.`
- removed version numbers from example input files - fixed some typos in TwrShadow and TwrPotent comments - fixed outdated AD example files in AA documentation
Attempt to fix this issue: OpenFAST#972
This is to attempt to debug an intermittent problem with gfortran and the ProcessComFile function and related routines
Passing this as an array has led to some issues with some gnu compilers in the past: `BL_files(:) = AFI(:)%bl_file`. I've updated the code to pass AFI instead of BL_files as subroutine arguments.
- use parameter `X_BLMethod_Tables` instead of `2` where appropriate - change `call check` to `call check()` for clarity
When the string is empty, we add "\lib" to the path instead of ".\lib", which then causes the scripts to fail to find the library files.
fixed some comments, removed unused variables
FVW doesn't need all of the calculations from that subroutine
Trying not to repeat so many of the same calculations in different subroutines
hub_theta_x_root for FAST.Farm and BPitch outputs are set as in other projection method; note that this projection method isn't actually used in production mode anyway.
also fix some comments and spacing
|
I'm going to double check the failing AD15 driver cases to make sure we don't have an issue with the channel name changes. If we do, it should be a minor update to those r-test cases. |
|
I think the problem is that intermittent issue (likely some memory issue related to the AD input file reading), since some of those files do pass on my fork: https://github.com/bjonkman/openfast/actions/runs/1918669057; also the gfortran11 debug test does pass here, too. |
|
Good to know. I was hoping it was our friendly intermittent failures and not an output channel name change. |
andrew-platt
left a comment
andrew-platt
left a comment
There was a problem hiding this comment.
Looks good to me! Thanks for putting this together!
The test case that is failing are the AD driver cases that occasionally fails -- it passed in Bonnie's branch so it is not due to output channel renaming.
2 participants
Feature or improvement description
Previously, AeroDyn had some undocumented outputs:
B{1:4}Aero{F|M}{x|y|z},B{1:4}AeroPwr,B{1:3}Aero{F|M}{x|y|z}g,RtAero{F|M}{x|y|z}g. This pull request adds those outputs to theOutListParameters.xlsxfile and updates the Matlab-generated parameters in AeroDyn_IO.f90.In addition, there are some minor changes to comments and example input files in the documentation.
Impacted areas of the software
AD, documentation
Test results, if applicable
This should not change any test results.