3 Replies Latest reply on Jul 18, 2013 9:53 AM by Euan Long

    Flow simulation breaks heat laws?

    Euan Long

      I'm attempting to carry out a flow simulation to look at the temperatures within a solid; for purposes of simplicity I have created a cube to test on. The boundary conditions necessary are:

      boundary conditions.png

      The simulation uses air and titanium as the fluid and solid properties.

      On applying these boundary conditions using "Total pressure" and "Static pressure" for the inlet and outlet flows respectively, the following cut plot is attained:

      cut plot.png

      Which looks OK until you realise that the minimum temperature, 955 K, is below any of the input conditions (min. 986 K at flow inlet).

      From the moment of entry in fact, the flow looses heat (which does not make sense as it has a lower temperature than the rest of the block), see below: NB different temp. scale

      cut plot.png

      If anyone has any ideas on what is going on, and why this is an unsuccessful simulation, that would be greatly appreciated.

      Additional comment: For the outlet I would like to set simply a pressure outlet without having to define end temperature which it asks for under thermodynamic properties - I think it is over-constrained as is...?

      Thanks, Euan

        • Re: Flow simulation breaks heat laws?
          Chris Michalski

          The temperature you provide at the outlet is simply to ensure that if there are eddy currents that result in outside gas being pulled into the model that it knows what temperature to use.


          What did you set the initial fluid temperature to?  Did you do this via a fluid domain condition?

          What goals did you use for convergence?  It look as if the model has not actually converged based on the temperature gradients shown in your last image.

            • Re: Flow simulation breaks heat laws?
              Euan Long

              Hi Chris, thanks for explaining the temperature setting on the static pressure, that makes sense to me now.

              The initial temp was 273 K and I changed it to 1200 K to see if this would help but it has no impact on the outcome, which makes sense as it's an initial temperature and will converge to the boundary conditions. This is in a fluid domain yes. All goals were used for convergence, but as I understand there is no necessity for defined goals in the first place. I must admit that I am not entirely sure that I understand what 'convergence' is used for though.

              I have continued working on the simulation and seeked help from a colleague that noticed that the mach number may have had a role to play. Using some theory in the form of the steady flow energy equation (SFEE) it can be shown that:

              T0 / T = 1 + ((gamma-1)/2)*Mach^2

              This equation shows how the temperature can in fact drop at higher Mach numbers. This has explained the odd plots.


              Thanks for your help and suggestions Chris

            • Re: Flow simulation breaks heat laws?
              Rich Bayless

              What materials and conditions have you assigned to the lids?  Sometimes assigning 'insulator' as the lid material is useful.