2 Replies Latest reply on Dec 2, 2015 9:10 AM by Amit Katz

    Contact Resistance Clarification

    Tim Bevins

      I am doing a flow analysis study with a fan pushing air across an oil can with some number of watts being dissipated inside and a fin array on the outside to be used as a heat exchanger. The  oil can is mounted to an aluminum base, which is mounted to some structure at the customers site; for lack of better information, I assumed it was a flat plate. The can needs to operate at 90°C, and the junction/film temperatures of the components inside the can must stay below 125°C, which is proving very difficult. My question is regarding contact resistances between different surfaces. When selecting surfaces for contact resistance values, do you have to select BOTH surfaces, or just one of them? I.e., if I know there is going to be a thermal interface material of some thermal resistance, if I were to select both the surface the TIM is applied to, and the mating surface, am I "adding another layer" of material? I realize often times, these values are so small that they are negligible, but even in this setting, every single degree is important. I can provide more information if need be.

       

      Thanks,

       

      Tim

        • Re: Contact Resistance Clarification
          Amit Katz

          I would assume you need only apply it to one face (in this case, the part of the oil can resting on the plate.) But I don't know for sure, maybe one of the VAR guys can chime in on this.

           

          Why not run a simple experiment, simulate heat conduction between two cubic prisms and apply your contact resistance to one face only. You can sketch a line through the centers and use the XY plot to see what your temperature looks like.

          • Re: Contact Resistance Clarification
            Amit Katz

            OK so I got curious and ran this simulation experiment myself. Two cubic prisms, 100MM on a side each. 1 W of heat transfering through, with a 1 K*m^2/W resistance in the middle. The heat flux through the resistance is 100 W/m^2, so the temperature difference should be 100K, and indeed it is:

             

            Capture.PNG

            Capture.PNG

             

            So in short: use just 1 face for your contact resistance.