FAST + OrcaFlex Simulation Crash

Dear Jason,

I have an interesting question relating to coupled analysis in OrcaFlex with respect to the vessel lateral degree of freedom.

Using the OC3 Hywind model as an example I created my own model of the floater and I successfully performed the analysis when I only had Wind and Waves acting on the vessel. Now I have added Current to the model as well. My forces are uni-directional in the x-direction. The displacement of the vessel in the x-direction is much larger now than it used to be (understandably so). The problem is that my simulation crashes 112s after the start giving either: a supersonic blade error or that my wind field has been exhausted.

When I looked at the visualisation of the 112s simulation in OrcaFlex, I noticed that my model proceeds to drift in the x-direction up to 29.8m and at roughly 30m away from the origin it appears to be literally yanked back to the origin instantly, which I assume gives the supersonic error for the blades. Conversely, when my turbine drifts away out of the range of the wind field, I assume it is considered exhausted.

The “face” or front plane of my turbine is located in the YZ plane. That is the plane in OrcaFlex where the wind is being applied to the hub in the x-direction

My question to you is: Does AeroDyn or any other of the FAST modules have a specific distance restriction? Given that we can specify the width and height of the front plane, in which the wind field is acting, via the grid, can we specify the depth of the field, i.e. if my SPAR drifts too far back, how can I tell the software not to realign the vessel at the origin?

The reason I suspect that the issue lies with AeroDyn is because I checked the Wind field (which I have been running successfully for the past 3 months with different alterations), I checked the controller and then I adjusted some of the values in the AeroDyn module, which delayed the crash by 40s, from 78s to 112s into the simulation.

Have you encountered such an issue before? Where does the fault lie in your opinion? Is it even possible to specify a high current and wave force without the crash occurring? When I say “high” I mean a 50-year extreme return case.

Dear Evgeniy,

Regarding the wind field, the basic limit is set by the InflowWind module i.e. that no analysis node where ambient wind is used (distributed along the blades and tower) can lie outside of the rectangular box defining the wind field. The box is defined by the width and height of the TurbSim wind field. The length of the box is effectively infinite for periodic wind files or derived from the time range of the wind file times the mean wind speed for wind files that are not created as periodic within TurbSim. For finite-length boxes, the wind field will start at half the grid width downstream of the FAST origin (0,0,0). The orientation box is determined by the wind propagation direction within the InflowWind module. Regardless, you can increase the domain of the box by changing the grid width, grid height, and useable time length when generating the wind files in TurbSim.

Regarding the supersonic error, I’m not sure what is causing the wind turbine to suddenly jump back to the origin. I have not heard of that problem before and cannot really comment on it.

I hope that helps.

Best regards,

Dear Jason,

Thank you for the answer as it did help me locate the error.
I finally discovered the flaw in my design which was in the controller. I do have one final question.

If I get an error that the wind angle of attack lies at Z=453m above my rectangular wind field box, what module is responsible for this issue?

Given that I have specified my grid points set as 31/31 and height as 170.5m and am using a generic TurbSim input file for a .bts output. The same wind file ran perfectly fine when I fixed the controller issue. So I assume the controller is responsible, but what Line of Code (Command) would govern this error?


Dear Evgeniy,

How exactly is the error message worded? I can’t find “wind angle of attack” or “rectangular wind field box” in the FAST source code.

Best regards,