3 Replies Latest reply on May 16, 2017 5:52 PM by James Lee

    how to set up right boundary conditions to rotating walls

    James Lee

      Hello, I have to admit that I come from experience of Fluent, and am exploring Flow Simulation to simulate rotating flow for electric machine.

      What I am trying to run heat transfer conjugated CFD analysis, it is to simulate an electric machine which will be cool down in the room by natural convection. Please take a look at my attached snapshot for your reference. Here is some basic summary to my settings:

       

      1. internal flow. (I create a large case to put my machine there, thus there will be one surface P1 with pressure in, another surface P2 with pressure out, P1-P1 is a tiny bit....), Rotation is not turned on....

       

      2. the machine's stator will be cooled down by forced convection through channels in its water jacket. thus I set up HTC and its bulk temp through flow simulation Outer wall B.C.

       

      3. rotor assembly will rotate at a high RPM, thus some walls (as I marked out in yellow lines) will rotate. I set up its rotation B.C. through Real Wall B.C. specifying its rotating speed there. --> I am not sure if it is right

       

      4. heat loss to all components, such as stator coppering winding, stator stack, rotor sleeve, rotor magnets, rotor hubs.

       

      I am not sure if I set up all of B.C. all right and I feel like there might be something wrong, but not sure...

       

      Any suggestions?

      thank you.

        • Re: how to set up right boundary conditions to rotating walls
          James Lee

          It was my first time leave notes on this forum, not sure if Solidworks or some one else could give me any suggestions.

           

          thank you.

          • Re: how to set up right boundary conditions to rotating walls
            Amit Katz

            James,

            I have had difficulty in the past attempting to run conjugate simulations involving rotating regions. It seems the software does not like to move the heat transfer boundary. What ends up happening is that the momentum simulation and the energy simulation become decoupled. The energy equations are solved as if there is no rotation.

             

            With that said, I think you might be able to get away with ignoring the rotation in this model. I believe that the heat generated by the fluid motion here will be negligible compared to the waste heat from your electric machine. If I were in your shoes, I would first do a complete simulation with no rotating regions and see how the results look.

              • Re: how to set up right boundary conditions to rotating walls
                James Lee

                Thank you, Amit,

                 

                Not sure how Solidworks can help me to answer this question officially.

                 

                I was not only interested in windage loss calculation, but also heat transfer associated with rotating assembly. if the solver failed to solve energy equation together with momentum equation, it means that viscous force item in the moment equation only plays role of converting force to fluid momentum from its adjacent rotating components, friction force will not be able to convert any heat.

                 

                I am not sure if it will be the case, given that energy equation will have to be solved together with momentum equation in most of solidworks heat transfer conjugated CFD cases.

                 

                My original concern is I may not set up the right rotating B.C. In FLUENT, there is option to use rotating reference frame; basically you can set up fluid region rotating at the speed you want, and set up rotating walls with zero speed relative to rotating reference frame. I checked Flow Simulation, it looks like it is also using the same or similar approach, but difference is in my project, I am using internal flow, it means that fluid domain is automatically detected by Flow Simulation; if I am using external flow, I guess I will have to model fluid region.

                 

                I guess I am just confused at how Flow Simulation handles rotating flow.