Skip to content

Conversation

@akeeste
Copy link
Contributor

@akeeste akeeste commented Aug 15, 2024

This PR was intended to bring #1274 from main into dev so that #1275 can be tested appropriately. However apparently we haven't brough out bugfixes in consistently so it's looking complicated.

There are a lot of commits confusing the history of main, I'm guessing from bug fix PRs that were not squashed and merged. @kmruehl Thoughts on how to approach this?

dforbush2 and others added 30 commits March 15, 2022 13:53
When MoorDyn is used in WEC-Sim design, it is called as an external DLL executable library, and it runs on a separate console.exe under MATLAB's main process, with its separate process-ID(PID) in the windows task manager. The issue is related to the following scenarios on the Windows platform:
1>> if the lines data text file has an error in it.
2>> MoorDyn fails to converge.
3>> MoorDyn is launched, but the model has some other error.
Under such a scenario, MATLAB fails to terminate the MoorDyn Console and hangs instead, and either crash immediately or crashes if we press the close button on the MoorDyn console window. Either way, there seems to be no way the MATLAB session can recover from MoorDyn failure without crashing. It becomes a major issue if we are trying to tune mooring lines and want to try different lines configurations and stiffnesses, etc., to find an optimal mooring system because some configurations will not work, which will keep crashing MATLAB sessions during the test and trial.
The goal is for MoorDyn's failure not to crash the MATLAB session but instead MoorDyn dll execution to close and the MoorDyn console to close. I've found a solution to that.
For scenarios 1 and 2, the following Fix#01 seems to work on the Windows platform. And to cover scenario 3, an additional Fix#02 is needed.
Fix#01
In the MoorDyn block's 'InitFcn' call, making the following suggested changes check for MoorDyn Initialization failure and terminates the MoorDyn.dll console without crashing the MATLAB Session.
Fix#02
Suppose the MoorDyn console has been launched, but there is an error in the Simulink model. In that case, MATLAB throws an error and terminates the Simulink execution without terminating the MoorDyn console window. A try-catch statement in the wecsim file can address this issue. If a Simulink model error occurs, we can print the error and terminate the MoorDyn Console process.
* updates_docs_on_master

* incorporating_kelleys_minor_edits
Fixes function parsing error
Change the subscript of the pitch dof from 3,3 to 5,5 in ystring.
* Update readNEMOH.m for compatibility with NEMOH v3.0

* Removing line 374

This capability is within Dev, but not currently pulled over to Master. A separate PR was submitted to Dev for this fix.
* Port dev testing PR to master

* Update out of date actions versions

* Re-include R2020b and R2021a in Windows tests

* Use file to capture logging to support R2020b issues
kmruehl and others added 23 commits February 9, 2024 12:40
adding community code of conduct
fix to #1217, nonlinFK and body block changed
Changes:
+ Updated to MATLAB Actions V2
+ Support for R2020b on Windows was removed
+ Updated to checkout V4
+ Removed logfile option from run-command as it has no effect
Bug fixes for WEC-Sim GUI and Run From Simulink features
* update CI to use main branch
* update simulationClass to reference git commits on main
* fix cubes.LIS file with merge conflicts
* update repository tree references in docs from master to main
* update readme and github workflow to run-tests-main.yml
* update redirect in gh-pages-redirect.html
* Add second redirect for old /mastet/index.html route
* Build main branch rather than master
* update other references to documentation

---------

Co-authored-by: Mathew Topper <dataonlygreater@gmail.com>
As part of #1235, a sphinx option was incorrectly renamed from master_doc to main_doc. This PR corrects this using the root_doc option, available in Sphinx >= 4.
Fix sign bug in the Generator Equivalent Circuit Block in PTO-Sim - Rebase PR to main
this is a sign correction on damping and inertia in the pto-sim/simple direct drive PTO block
* fix on pDis function call

* preliminary fix for issue #1288
readCapytaine - fix reading multiple bodies with <6 dofs
@akeeste akeeste requested a review from kmruehl August 15, 2024 14:16
@kmruehl kmruehl self-assigned this Aug 21, 2024
@kmruehl
Copy link
Collaborator

kmruehl commented Aug 21, 2024

@akeeste it's looking like this is pulling in commits from 2 years ago, which certainly exist within dev, because dev was created from main after our last release. I'm going to go through this and see if I can detangle the branches. We may need to manually pull the PRs over.

Copy link
Collaborator

@kmruehl kmruehl left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The majority of these file revisions are due to the conversion from master --> main.

@kmruehl kmruehl mentioned this pull request Aug 21, 2024
@kmruehl
Copy link
Collaborator

kmruehl commented Aug 21, 2024

@akeeste in order to resolve these merge conflicts, I would have to make commits to main, which I would like to avoid. Instead, I've migrated this PR to a branch on my fork and re-opened it here: #1297

@kmruehl kmruehl closed this Aug 21, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.