Open andrew-platt opened 5 years ago
@jjonkman may have thoughts on this given similar changes implemented in FAST.farm
Thanks, @andrew-platt. I would prefer the InflowWind-based solution because (1) there is no need to specify a point on the ground in TurbSim where the wind velocity is always zero and (2) in floating offshore wind turbines, the tower may heave downward and I'd prefer that any aerodynamic analysis nodes that pass below the ground be given zero velocity rather than fatally aborting OpenFAST.
We don't use the tower points when using TurbSim to generate wind inflow for FAST.Farm, but we have opened up two issues here pertinent to wind generation for use in FAST.Farm (and OpenFAST): https://github.com/OpenFAST/openfast/issues/199 https://github.com/OpenFAST/openfast/issues/256
Thanks,
Background
When a course grid is used in TurbSim, the tower points may be disabled when the bottom of the grid is too close to the ground to allow for a point on the tower with the same spacing. Then in the simulation, the aero for the tower and tower potential must be turned off to avoid errors. This can create some confusion when doing studies based on grid resolution (@ewquon ran into this).
InflowWind based solution
After a brief chat with @bjonkman, rather than change anything in TurbSim, we can change InflowWind to interpolate between the bottom of the grid and the ground when tower points are requested (set
p%InterpTower
to true if the bottom of the grid is less thandz
from the ground). Any points below ground would be set to zero (Envision ran into a few cases where this was needed -- slight numerical rounding issues if I recall correctly).The only concern with this is that the linear interpolation will be rather crude, but this may not be much of a concern when only a very small number of grid points (i.e. 5 on a 140 m grid) since the wind field is very course to start with.
Aternatives
Modifying TurbSim could be problematic due to the the factorization, so that is likely not a good approach.