Closed gbarter closed 3 years ago
Garrett, thanks for lauching the project. Yes, I am very glad to offer my help. Based on your request, I've made the following four updates in the latest version of the code:
(1) The input files of all cases have been re-checked, and modifications are made wherever the input parameters are not reasonable. In particular, the damping matrices and restoring matrices of the DeepCwind and the HywindSpar cases are defined according to the official "Definition of..." documents released by NREL.
(2) The HywindSpar case is made computationally faster by changing some settings in the control file.
(3) A new functionality of the external quadratic damping has been added to the motion equation to account for viscous drag forces, as required by some cases (e.g., the DeepCwind in the CertTest).
(4) Benchmark results (run by myself) are included as a set of "truth" values (which will not be changed by users' run) to help in the subsequent regression test.
Please do not hesitate to ask me whenever you are in need.
Thank you, Yingyi! I will submit another PR with expanded test coverage and checks, but it might be a bit before I get to it.
Don't worry, Garrett, just take your time.
Implements a continuous integration workflow using Github Actions (addresses #7 )
Occurrence
On any commit or pull request to the
master
branchTest matrix
Linux & mac, gfortran 9 & 10 I tried a couple of times to get a Windows build working, but didn't succeed. I didn't worry too much about it though as Yingyi already provides a compiled Windows executable.
Tests
Right now, this just ensures that the code compiles and runs the available example cases (except HywindSpar and Moonpool because they take too long to run).
Future improvements
(input is welcome to help prioritize next steps)