usually one would design the welds so that the parent material fails first so no would be my answer. I would use IGES and just go from Rhino to Abaqus. If you go from SWX I would use IGES as well - it is your best hope with surface models. Given that things are made of steel do you really think anything will shatter? Assuming things are slow why not just do a NL static and push a rigid surface into the wind turbine structure and get a load displacement curve. Fracture is a tougher thing to model - is it needed is what I would ask.
My next try will be surface models yes, at first I will do another try with a solid.
The hullplates will tear, no question about that, Usualy the side has a ballast tank. If that ballast tank is leaking, a ship can counterballast and there is no direct threath, (offcource with bulkcarriers the story is different as the cargo hold will get compromised by the first tear)
The drift speeds taken in a lot of the research I have found have been lower than a lot of the real data I have recieved, and I can state have eyewitnessed this month on the north sea.
The big issue occurs when the cargo area is compromised, and it is of value to know at what force that occurs. From there I can calculate back to speed and windspeed per shipclass.
The point of this research is to figure out at what windspeed a threath occurs that a drifting ship without power leaks cargo or worse sinks if it after a certain amount of time drifing collides with a wind turbine.
With that information I can define a risk area with the current and planned windfarms, together with the responce force available, and the time it takes for them to react. That risk area will be the area in which the responce force will with high probability, not make it in time.
The tear is key, as the important part is the cargo spill. The recent incidents have learned us that a lot of cargo, while at first glance may not look that dangerous, actualy have a big, if not a life threathing impact.