6 Replies Latest reply on Jul 29, 2014 11:40 PM by Jared Conway

    Wide temperature range

    Robert Ellett

      Has anyone seen a flow study where the average temperature would be something that is expected, but the maximum temperature is completely unrealistic. Recent studies I've run have had average temperatures at 80C, or less in most cases, but as high as 2700C. No, that is not an extra zero. There isn't any possible way a 13W volume heat source can heat up a metal can (in this case the air around it) to 2700C, so what is wrong with my setup? 

      Here are my settings for the internal study

      Heat conduction in solids, checked

      Exclude cavities without flow conditions, unchecked

      Gravity, checked

      Fluids (gases), Air

      Flow type, laminar and turbulent

      Default solid, Insulator

      Default outer wall, Heat transfer coefficient, 5.5 W/m^2/K

      Temperature of external fluid, 24.05C

      Initial conditions, pressure 14.6959473 lbf/in^2, pressure potential checked, temperature, 24.05C

      Initial Mesh, Default

      All solid bodies had material assigned. Mild steel was assigned to the metal can, electronics (PCB material assigned) are inside the metal can, inside a ventilated ABS plastic housing.

      Those are the settings. If anyone has any ideas please let me know.

      FYI - file is too big to attach

      Thank You


        • Re: Wide temperature range
          Chris Michalski

          Robert -

          a picture at least would be helpful, preferably a screen capture to see your Flow feature tree.


          1st thought - your "insulator" as default... did you apply materials inside of Flow for the other parts?  Or simply in their model files?  If it wasn't inside of Flow then it will use the Flow default (at least in 2011 which I use).  In this case "insulator" is a perfect insulator and will not allow any heat transfer through the solids.

          • Re: Wide temperature range
            Jared Conway

            yes, something like a singularity in fea where you have bad geometry or the mesh is really really poor or the setup isn't realistic

            • Re: Wide temperature range
              Chris Ma



              A handy tool that might help you is to activate a cut plot of temperature, and go to the dropdown menu Flow Simulation > Results > Display > Global Min/Max. This will put a blue dot on your model where the global minimum temperature is, and a red dot for where the global max is. This may help narrow down the area of concern.


              Also, I would agree with the other posters and try the following:


              1. Double check that all the parts have a solid material assigned to them. A quick tip is to go into the flow simulation tree, shift select all of the solid materials you have defined. Hold the down the ctrl key, right-click in the graphics area (somewhere not part of the model), and select hide. This will hide every part that has a material assigned to it. Whatever is still showing will be defined as your default material.


              2. Use the "Interference Detection" tool and have the "Treat Coincidence as Interference" check box turned on. This will highlight all touching faces. I've had cases where I thought heat conduction was happening, but in reality there was a small gap that I didn't take into account.


              3. Check the mesh. Create a cut plot in the area that that global max temperature exists, and show the mesh in this area (no contour plots active). In the options section of the cut plot property manager, make sure to UNCHECK the "use CAD geometry" option. This will show you if the mesher is truly capturing the geometry correctly, at the mesh level you are at.