5 Replies Latest reply on Oct 7, 2014 3:15 PM by Robert Gould

    Heat transfer simulation setup

    Robert Gould

      Ok well, if someone could quickly run down how to set this up it would be appreciated. I have a graphics card I want to cool, and I am using a block sitting on a processor. The plastic block has a copper plate on the bottom which sits on a larger aluminum plate. the copper plate slightly overlaps the aluminum p[late and cools it, there is a second aluminum rectangular  block where the fluid enters the loop in the front through a barb. OK so I set up the flow simulation as an external sim (good?), i set the inlet and outlet as environmental pressure, and using an impeller for my rotating region in the plastic block. I am assuming the rotating region will suck the fluid in and throw it out basically. I set up the main fluid (fluid subdomain) as the internal water and set it at 35 degrees. the other fluid is air surrounding the assembly which I also set up at 35 ambient. I also added surface heat generation rates to 4 surfaces on the bottom, 300watts for the center, 50 watts, 5 watts, and 8 watts on the outer pads. I basically just selected them and added them as surface heat sources (will this work?) One thing that confuses me is that there is temperatures you can set for everything, inlet temp, initial conditions temp, environmental pressure temp....do all these have to be set? and how does this affect my results? Also, my numbers seem low. I set all those temperatures i mentioned at 35 degrees, and in the screenshots i hit a max temp of 48 degrees on the outside of the processor around the outside of the copper. I was picturing higher temperatures. Is this even set up correctly or is there something I am missing like goals or like maybe my heat sources are not generating, I set this up as steady state or default sim not a time dependent one. Inital mesh is default. heat conduction is on, radiation is on, gravity is on, rotation is on. when i see flow trajectories the fluid is rotating and exiting as expected. any suggestions on this//???? I may not be able to pack and go this because of possible NDA issues, so hopefull you run through a good set up to get accurate results without it...

        • Re: Heat transfer simulation setup
          Jared Conway

          external analysis > yes, this is ok if you want to look at the cooling inside and outside

          rotating region is an impeller > the rotating region should be a simple cylinder

          setup the main fluid as water > this means that your whole environment is water, aka, your graphics card is underwater, is that what you want? you will need to set a fluid subdomain on the water part but those will need to be closed off.

          surface heat sources > that should be fine. volume heat sources are also fine.

          initial temps > not important in steady state

          inlet/environmental pressure temp > yes, these are important as they are the temperatures of the fluid coming in and out and could affect the results

          temperatures seem low > compared to what? intuition may be one thing but it should be a known solution you are comparing to. if only one value isn't as expected and the others aren't, you may have a setup issue or it just isn't working as well as you'd like

           

          my suggestion would either to be to walk before you run. start with just a couple of compnents and then slowly add complexity. this is very similar to several tutorials. if you're not picking up where they end off, feel free to contact me at jared@hawkridgesys.com if you're interested in working through this with someone. or post more info about what your analysis goal is here and i'll do my best to help through the forum.

            • Re: Heat transfer simulation setup
              Robert Gould

              I actually did set up a fluid subdomain and selected the inside surface of the loop as the water, so I am assuming air would be the secondary fluid? So then its ok to set up the inlet as a boundary layer environmental pressure and temp at 35 degree C, and the outlet also as a boundary layer  environmental pressure at 35 degrees C also? I would have put the inlet volume flow, but what I want is the rotating region to create the movement not an inlet flow boundary condition, if that makes sense?

                • Re: Heat transfer simulation setup
                  Jared Conway

                  I actually did set up a fluid subdomain and selected the inside surface of the loop as the water, so I am assuming air would be the secondary fluid?

                   

                  >>your wizard settings apply to all fluid regions unless overridden by a fluid subdomain. so the way you describe what you did, you overrode a water global domain with water subdomain.

                   

                  So then its ok to set up the inlet as a boundary layer environmental pressure and temp at 35 degree C, and the outlet also as a boundary layer  environmental pressure at 35 degrees C also?

                   

                  I would have put the inlet volume flow, but what I want is the rotating region to create the movement not an inlet flow boundary condition, if that makes sense?

                   

                  >>this should work but you should check the swx kb and help. for some reason i remember rotating regions liking to have at least one flow rate. but for your purpose, environmental/environmental with a rotating region in between should work