Add MIT yaw correction model (3rd pass)

[jaimeliew1]

Add MIT yaw correction model (3rd pass)

This pull request implements an interpretation of the yawed actuator disk as described in Heck et al. (2023):
https://www.cambridge.org/core/journals/journal-of-fluid-mechanics/article/modelling-the-induction-thrust-and-power-of-a-yawmisaligned-actuator-disk/3A34FC48A6BC52A78B6D221C13F4FC3A

The method assumes that $C_T'$ is invariant to yaw misalignment (an assumption still to be verified), and preforms a wind speed reduction based on the model equations.

Related issue

The structure of the code changes resembles those in the previous pull request (#638) , but now compatible with Floris V4.

Another older pull request can be found here (#624)

Impacted areas of the software

A new operation model is added mit_loss (name to be decided).

I added the code used to calculate the MIT operational model to floris/core/mit_models, but please move this to wherever you think is best.

Additional supporting information

An example has been modified to show the operational models at different controller operation points. The added model is able to transition from below-rated to above-rated wind speeds in a realistic manner as shown in the following figure. I.e. The power output is "clipped" near rated wind speed, and is invariant to yaw far above rated wind speed.

Test results, if applicable

Relevant tests are still to be added.

[paulf81]

Thank you @jaimeliew1 for this exciting contribution! We'll start working on the review/merge soon!

[rafmudaf]

A summary of changes I've made:

• Consolidated all of the model implementation to one file at floris/core/turbine/mit_turbine.py
• Arranged that module in the following order:
  1. Turbine operation module functions; these are the hooks to incorporate this model into the power, thrust, and axial_induction functions defined in the BaseOperationModel interface
  2. Iteration functions
  3. The BaseOperationModel subclass for the MIT Turbine model
  4. Functions implementing the model as described in the paper
• Added a brief description of the model, some equivalent figures from the paper, and a link to the paper in the docs

Notes / suggestions:

• I propose to change the name from "mit-loss" / "MITTurbine" to something like "YawCorrectedADM" or "YawAwareADM"; to a user who has not seen this pull request, the option for a "mit-loss" model will not have much meaning, so it would be helpful to introduce a more descriptive name
• The implementation is elegantly implemented and easy to map back to the model described in the paper. However, the programming paradigm is different than the rest of FLORIS and it may be challenging to maintain in the future. I don't recommend to change the implementation, but this difference is worth noting for our future reference. If there are any maintenance challenges, it might be worthwhile to refactor to match the rest of the FLORIS paradigm.

[rafmudaf]

@jaimeliew1 I wasn't able to understand where mit_rotor_axial_induction and mit_rotor_velocity_yaw_correction are described in the paper. Could you explain this portion a bit more?