7 Replies Latest reply on Jun 12, 2018 3:49 PM by Mark Summers

    Thermoelectric coolers

      Has anyone played with TEC in SW flow simulation?


      Currently I am doing benchmark testing on electronics cooling that uses customized TEC.


      We can set DC current, Heat transfered at cold side, temperature difference and voltage at their maximum.


      However, not electrical resistivity, and no input on TEC elements.


      And I found that the temperature difference disagrees with empirical data.


      I suspect that it's due to incomplete customization to this black box component.


      Anyone facing the same problem?

        • Re: Thermoelectric coolers
          Amir Amir

          Have you found any more detail on how the TEC simulation works in Flow?

            • Re: Thermoelectric coolers
              Jared Conway

              Amir, what trouble are you having? Have you taken a look at the help in Flow Simulation or page 48 of the technical reference? Those 2 resources are the best for information about the thermoelectric cooler. It is a "what you see is what you get" type element, so if you have different inputs, you need to convert them or develop a workaround. With more information about what you're facing, someone may be able to help.


              Another reference that might be helpful is the SolidWorks Knowledgebase.

                • Re: Thermoelectric coolers
                  Amir Amir

                  Well the issues that I have are two things. First, in the TEC library, when I try to choose a TEC to insert to the study, I don't see the TEC thickness to be a parameter at all. and since SW doesn't show the TEC physically, I never know if all the faces are in contact correctly. Second, is that I can not run the TEC in heating mode.

                    • Re: Thermoelectric coolers
                      Jared Conway

                      Hi Amir, the TEC "smart part" assumes that you have modeled the TEC as a solid cube in your model already. Since you have a part, there is no need to add a thickness because you've directly modeled the overall geometry.


                      Here's what the help has:

                      "In the flyout FeatureManager design tree select a component which you want to consider as the Component to Apply the Cooler. You can also select a component in the graphics area. You can select only one component. To remove the component from this list, select it in the list and press Delete or right-click it and select Delete or select the component again in the graphics area."


                      Don't forget to check the assumptions and recommendations in the technical reference doc. (I don't think the solving engineering problems doc has additional info but it wouldn't hurt to check)

                      For example:

                      "The specified TEC’s "hot" face must be in contact with other solids, i.e. it must not be in contact with any fluids. In a typical setup the "cold" face is in contact with the cooled heat source while the "hot" face has a heatsink attached."


                      You can check that things are touching by using the interference detection tool. Just choose the option "treat coincidence as interference".

                        • Re: Thermoelectric coolers
                          Rick Champagne

                          Hi Jared,


                          When specifying a TEC, what value is to be entered in the I field directly below the hot face box?  Also what if the mfr only gives the max current, voltage, and power at one temperature value?



                          Rick C.

                          • Re: Thermoelectric coolers
                            Mark Summers



                            Thanks for your responses to the early question(s).  I have three additional questions for you and/or the other participants.


                            1) I notice the heat flux through the TEC is not balanced.  The heat flux is higher on the hot side than the cold side.  Also, I checked and no heat is going out the edges of the TEC.  This is the case with my analysis as well as the sample problem you suggested.  It seems like they should be equal.  Does this make sense and/or do you think I could correlate the ratio of the heat in to heat out as the TEC's COP (coefficient of performance)?


                            2) I found a comment elsewhere that transient analysis is not recommended with TECs.  Do you have experience with running transient analysis with TECs?


                            3) I was not able to apply a thermal joint between the hot side of my TEC and my heat sink nor one between the cold side of my TEC and my condensing plate.  I read that it must be solid to solid contact, but I would have thought a thermal joint (thermal grease) would work.  Any ideas on how to accomplish / accommondate this situation?

                              • Re: Thermoelectric coolers
                                Mark Summers

                                Some answers to my own questions, and a new question.  I would appreciate hearing from others on these topics.


                                1) Since the specifications for the TECs list Qc (not Qh), I am going to use the heat flow on the cold side as Qc.  Qh seems to be what it needs to be to make the other parameters balance out.


                                2) Would still be interested in hearing opinions on this.


                                3) Would still be interested in hearing opinions on this.


                                4) It is not clear how the four tables in the TEC "material" definition provide enough information when the unit is operated at less than maximum current.  If the TEC has table values at the maximum current of 6.1A, how does Flow Simulation know what parameters to use when the input is say 1.5A?  My Flow Simulation show much higher performance than expected when I operate it at 1.5A.