Intro Task: Added Mass and Buoyancy in Actuator Cylinder Model

[kevmoor]

Here is the intro task: copy added mass and buoyancy equations/logic from the double multiple streamtube model to the actuator cylinder model.

Successful completion will require:

• Basic plots comparing the DMS and AC models’ added mass and buoyancy impacts using the existing RM2 plotting script
• A pull request against OWENSAero
• A list of recommended improvements to OWENS or its docs based on completing this task

Recommended workflow:

1. Copy the added mass and buoyancy equations/logic from the DMS model to the AC model. This is <20 lines of addition/modification.

• Specifically, the streamtube() function of DMS.jl has the needed additions https://github.com/sandialabs/OWENSAero.jl/blob/master/src/DMS.jl
• These additions will go in the radialforce() function in acmultiple.jl https://github.com/sandialabs/OWENSAero.jl/blob/master/src/vawt-ac/src/acmultiple.jl.
• All of the i/o architecture “should” already be there for the AC radialforce() function so it should be nearly copy paste from the DMS model.
• This is the paper that the simple added mass equations came from https://www.sciencedirect.com/science/article/pii/S0960148118304312

2. Install OWENS https://sandialabs.github.io/OWENS.jl/dev/installation/

• There is also a short setup and demo video for OWENS in the readme https://github.com/sandialabs/OWENS.jl.
• Please use mac/linux if you can. If you are trying to install on Windows, don’t try to build the openfast wrappers at this point.
• If you run into issues, we’ll do a quick working meeting.

3. Run the RM2_medium.jl (https://github.com/sandialabs/OWENSAero.jl/blob/master/examples/RM2/RM2_medium.jl) and update to compare the current “DMS” model to the “AC” model with added mass and buoyancy.

• There is a flag currently on “DMS” that you’d just need to change to “AC” at the minimum

4. Create a pull request

[kevmoor]

Should be able to finish in a day or two. Will create draft pull request

[kevmoor]

Read through paper and code is making more sense. Kevin will send other papers on VAWT aero. Some of the non-standard variable names should be updated to be more descriptive. Global variables are confusing.

Next steps: Kevin help finish/guide code modifications, and Carlos start on the documentation updates while these questions and pain points are fresh. Kevin double check on P/T and send over.

[cmichelenstrofer]

Recommended Improvements

• Use a specific style guide. E.g.,
  • Specifically: variable naming convention.
  • Can use tools (e.g.g, Julia Formatter) to make this somewhat automatic
• Use descriptive variable names in source code. Leave symbols for examples/tutorials.
• Allow nslices to be different from ntheta. This is not just desired functionality but would help debug and understand variable flow.
• Avoid Globals! These are confusing and difficult to debug.
• Functions have too many outputs. Consider multiple smaller functions (postprocessing) or a "results" struct
• Have a general added mass function that works for both DMS and AC instead of repeating code in both.
• Not all function inputs are documented.

[kevmoor]

Items to minimally finalize this task:

• Buoyancy
• Check that AC model hasn't changed since before since AC and DMS aren't on top of each other as expected.

Kevin push the csv files

[kevmoor]

Just pushed the CSV files to master