When a simulation is started at 0 m/s, there's a seg fault.

[dzalkind]

Desired use case

Wind ramp starting at 0 m/s

Description

https://forums.nrel.gov/t/unhandled-exception-in-libdiscon-dll-when-debugging-fast/8025/6

Current behavior

Divide-by-zero in AeroDynTorque

Expected behavior

Saturate initial inflow speed. I think this was done in the past and may have been removed?

Code versions

v2.9

[dzalkind]

The divide-by-zero should be caught here:

ROSCO/rosco/controller/src/ControllerBlocks.f90

Line 283 in 122c758

LocalVar%WE%v_m = max(LocalVar%HorWindV, 3.0_DbKi) ! avoid divide by 0 below if HorWindV is 0, which some AMRWind setups create