6 Replies Latest reply on Jul 22, 2018 9:51 PM by Bill Dempsey

    Simulation Flow with open DC Blowers

    Tung L.

      I'm fairly new to SolidWorks Simulation Flow.

       

      I'm having trouble trying to put together a simulation flow through my heat sink using DC blower fans. I can't properly set up lids to do a boundary property with the inflow of the DC flowers and then properly set the out flow.

       

      Can anyone suggest or guide me with the right path with this simulation?

       

      Attached is the parts

        • Re: Simulation Flow with open DC Blowers
          Bill Dempsey

          I'm wondering if you still need help with this?  Can you post the Solidworks file instead of STEP?  The fan type should be set to internal and then select the inlet face and the outlet face on the fan body you have.  

          If you post up a SWX file I could help you set it up but I can't with just a STEP file

            • Re: Simulation Flow with open DC Blowers
              Tung L.

              I was able to get the simulation working as I went the "internal" simulation route. I used only the outlet of the fan and used "external inlet fan" to simulate the fans.

               

              My only issue is that I'm trying to properly set up the Meshing without causing the simulation to run over a long period of time. Also, my "environment pressure" area keeps giving the error "Face is not laying on the boundary region between solid and fluid region".

               

              I'm trying to push the boundary layer for my "environment pressure" as it keeps giving "vortex forming across boundary layer" warnings, but then I have a component that has some faces that doesn't need any results for.

               

              Please give me some time to post up the SWX file as there are some material on here that can't be shown in public.

                • Re: Simulation Flow with open DC Blowers
                  Bill Dempsey

                  For this kind of simulation (electronics in a box) I prefer to do external simulation but use an "internal" fan -- that way your fan represents "real-world" how it pull/push air.  The fan will starve, for example, if you choke the input flow and you will see that in the sim.  You also get rid of the "face not laying on boundary region" since you're doing an external flow analysis. That in turn will remove the vortex at boundary layer problem.

                   

                  I assume you have entered the fan data from the manufacturer in the model?  Or have a very very close model?

                   

                  Once you do an internal fan then the only reason you might be getting excessive meshing is how you have the global mesh set up.  You only need to mesh the faces of the internal fan and keep flow passages sized correctly through the fluid space.  This might mean adding local meshes or using control planes.  There are good examples of both methods in the Flow documentation.  A good starting point for mesh is global level 3 and then define the minimum gap.  Run the mesher only and examine the full mesh.  How many cells did you generate?  What areas can be reduced?  What areas need to be increased?  It takes a few tries usually to get the meshing groomed correctly before you hit "run" on the solver.

                    • Re: Simulation Flow with open DC Blowers
                      Tung L.

                      Hi Bill,

                       

                      I initially did an external simulation for my simulation but then I noticed that my exit air from my heat sink is actually flowing right back into the intake of my fan, which isn't supposed to be happening. I have moved the environmental pressure boundary further out from the exit of fins in hopes to solve this problem.

                       

                      I used a tool to get the fan curve from the fan's data sheet so I do believe that my initial parameters of my fans are correct. I took your advice with the global mesh and fan gaps, the mesh seems to look a little more accurate in the vicinity that I needed to analyze. I was also able to get in contact with an application engineer to go over my initial parameters as I felt some of them were wrongly set up.

                       

                      Would you have any experience on how to properly set up a thermal simulation involving IC chips to housing to heat sink? I noticed that my thermal gradient is showing a huge in accuracy in its results.