Using WISDEM on user-defined blade

[ariff-alzakri]

Hi,

I am trying to use WISDEM to perform a simple aero optimization of the blade's chord and twist, as of now separately. I have successfully managed to define the input, though I have made assumptions on the structural definition of the blade which does not reflect the actual design to run the simulation.

WISDEM states that the optimization was successful:

Unknown solver options: rhobeg, catol, adaptiveOptimization terminated successfully (Exit mode 0)
Current function value: -0.2852793555503351
Iterations: 14
Function evaluations: 28
Gradient evaluations: 14
Optimization Complete

But an error occurs as follows:
UnboundLocalError: 'wt.wt_rna.rotorse.rc_in' : Error calling compute(), cannot access local variable 'area_root_ss' where it is not associated with a value

Could I ask for help on why this occurs and how can I solve this?

Thank you.

[ptrbortolotti]

hi, @ariff-alzakri , I may need some extra input to help you out. are you starting from one of the examples? if so, is it example 03? those are part of the unit tests and run successfully. what have you changed?

[ariff-alzakri]

Hi @ptrbortolotti ,

Thank you for your reply.

Before I provide more detail on the input I would like to provide further clarification that I have made several other changes from the previous result. I am running WISDEM using windows subsystem for linux, using python to run WISDEM in vscode. The turbine blade is for a tidal turbine blade, hence the smaller blade design. The latest result of the simulation run is as below:

Result

Geometry file: components['blade']

The blade that I am using is different from the example, with a much smaller size. These have been changed as per the blade design including airfoils and its polar definitions.

I have kept all of the 'structure', root, hub and other definitions the same. This could be a point of error as the thicknesses may not align well with the current definitions of the blade.

Geomtry file: control

I have changed the control value for the criteria that I am designing to. The blade is designed as a fixed pitch.

Modelling file

Kept the same

Analysis file

I have changed the TSR to user-defined, and max iter was set to 200. The rest was untouched.

[ptrbortolotti]

hi @ariff-alzakri, I am afraid that you will need to go back to a working example andgo step by step through the changes. the first error that you listed indicated that area_root_ss has no value associated to it. that value is within the blade cost model, which is documented here https://docs.nrel.gov/docs/fy19osti/73585.pdf
the model is implemented here https://github.com/NLRWindSystems/WISDEM/blob/master/wisdem/rotorse/blade_cost.py
That model is quite specific to wind turbine blades, and it probably won't work without some changes for much smaller tidal blades. you should probably start by skipping that model entirely.
in your later messages something else is going wrong, as the beam model frame3dd seems to be failing. that may happen because the cross sectional properties computed by PreComp are unrealistic. debugging that takes some exercise. precomp is called here https://github.com/NLRWindSystems/WISDEM/blob/master/wisdem/rotorse/rotor_elasticity.py and is documented here https://docs.nrel.gov/docs/fy06osti/38929.pdf
good luck!