I was hoping some people on this forum might have some suggestions on this sort of simulation.

Imagine a rotating cylinder (cooled) where the id is is heated by a radiation beam (similar to a flashlight beam where there are different levels of intensity as it diverges from the center).

The radiation heating (as I received it) is in W/M2 (a.k.a. heater flux). But it rotates around the ID of the cylinder at a certain RPM.

I need to both verify that the cooling is adequate, and then do a fatigue study.

I have done static studies like this before. However I am at a loss of how to do it rotating.

No mater what type of simulation I select, I don't have the option to both rotate a part and add heat flux.

I have tried a transient thermal analysis (questions on that I have posted before about time curves), but it never really heats up (maybe a problem with my time curve?).

Because the beam is large, I divided it up into 10 equal (36 degrees) spacings, and as the first spacing was full on, the second one was heating up. as the first one was cooling down, the second one was full one, and the third was warming on etc... This doesn't seem to work. The entire system with 1/2 a megawatt of power only heated up 9 degrees C after 40 minutes with minimal cooling.

Right now, I am running a simulation where the radiation is just on one section and constant (not rotating) of the cylinder. At least then I can figure out what coolant requirements will be needed to remove the heat (a.k.a heat going in = heat needing to go out).

I would still need to rotate the stress around to do a fatigue study.

Any suggestions?

David