Inconsistent results about thermal exchange in channel

    • Mickael Perrin


      I am currently facing a thermal problem and found inconsistency in my results.
      My problem is a basic heat transfer between a solid and fluid. A flux of 10.000 W/m² is generated in the solid. The fluid flows next to the solid and cools it. The geometry is know : both solid and fluid are square channel of 0.5 x 0.5 x 10 mm. The fluid enter as a velocity inlet at 300K and 0.000514 m/s. Except inlet, outlet and interfaces, all the system is insulated, there are nor heat loss nor fluid loss. Fluid is water, solid is aluminum.

      Considering no heat loss, outlet fluide temperature can be calculated :
      Q x S = m x C x DT
      DT = (Q x S) / (m x C)
      with : the heat flux Q [W/m²] ; the surface giving the flux S [m²] ; the mass flow rate m [kg/s] ; the heat capacity C [J/Kg/K] and the temperature difference between inlet and outlet DT [K].

      NA : with m = rho x velocity x section :
      DT = (Q x S) / (rho x velocity x section x C)
      = (10000 x 5e-6 ) / (997 x 0.000514 x 2.5e-7 x 4177) = 93.4 K
      Considering all the flux is absorbed by water, its temperature will rise about 96K and will be 393.4K at outlet.

      A first simulation of only the fluid channel confirm this result. Please, find respectively in the next figures the boundary conditions, a contour of temperature (side view) and the surface heat flux balance.

      Figure 1 : Boundary conditions (simulation fluid channel only)


      Figure 2 : Contour of temperature (side view), (simulation fluid channel only)


      Figure 3 : Surface heat flux balance (simulation fluid channel only)


      Now, the real simulation : instead of inject directly the flux in the water, the flux is introduced in the solid which is in contact with the water. Except this interface, the solid is totally insulated. The channel of solid is the same of fluid meaning the same flux enter the system.
      In theory, the water will rise the exactly same DT because the boundary conditions didnt change, only the position on the flux.

      However, my resultats doesn't show this. Please, find respectively in the next figurtes the boundary conditions, a contour of temperature (side view) and the surface heat flux balance.

      Figure 4 : Fluid boundary conditions (simulation fluid and solid)


      Figure 5 : Solid boundary conditions (simulation fluid and solid)


      Figure 6 : Contour of temperature, side view (simulation fluid and solid)


      Figure 7 : Surface heat flux balance (simulation fluid and solid)

      As you can see, 10.000 W/m² enter the solid but only the half enter the water which rise its temperatrure only 27K, which totally inconsistency. I tried to modify the mesh, the residual criteria, the model used or interface type (coupling wall or not) but nothing correct the discontinuity. Every try, the solution is converge but give inconsistent result. The only parameter modifying the results is the thermal conductivity of the solid. Setting its value like the water's giving finally the expected result. Please find in the next figures the contour of temperature (side view) and the surface heat flux balance when i simulate aluminum with the same heat conduction than water.


      Figure 8 : Contour of temperature, side view (simulation fluid and solid, k solid = k fluid)


      Figure 9 : Surface heat flux balance (simulation fluid and solid, k solid = k fluid)

      Please, can someone explain to me what i do wrong in my simulation ? How can simulate a regular solid and obtain correct results ?

      I also tryied to change the source type for volumetric heat source [m/m3] but nothing changed. 


      Thank you all for your help.




    • C N
      Ansys Employee


      I recommend you to run the simulation for a longer time with solid surface being in contact with fluid case. Theoretically, the entire amount of 10000w/m^2 heat flux has to transfer to the fluid. Please provide the  boundary conditions you used at the interface. I need your meshing details for further understanding your problem .Please refer this link for further understanding on how to model heat transfer problems

      14.2. Modeling Conductive and Convective Heat Transfer (ansys.com)

      If you cannot access it, kindly refer this link Using Help with links (ansys.com)


    • Mickael Perrin


      Tank you for your reply. Here are more information about my case :

      • at the interface, i only specified "coupled wall" option.
      • i run my case as steady flow so there is no way to run a 'longer time'
      • i try to rise residuals and tested many models and many methods to calculate energy equation but nothing worked
      • concerning the mesh, i activated face meshing and used element size on every edge. I tested global size of 50µm, 25µm and 10µm but the results only changed about few K.

      As suggested in the user guide you linked, i tried different value for under-relaxation factor of energy equation but it didnt solve my problem.


      Thanks for your help,

      Sincerely, best wishes for the new year,


Viewing 2 reply threads
  • You must be logged in to reply to this topic.