11 Replies Latest reply on Jun 3, 2013 11:12 AM by Jared Conway

    heating of water in flow simulation

    Mohamad Nafis

      i'm doing a simulation on water heating where heat is applied at the bottom of the tank..the thing is, the water inside the tank increase only a few degrees although i have applied 1000K and 2000W/m2K..can anybody help me?

          • Re: heating of water in flow simulation
            Jared Conway

            it would be helpful to post the file or use the report function and upload the report with the details of the setup if the model is propreitary and you can't setup a sample model.


            from the screenshot, it looks like you setup an internal analysis which is steady state and that your default wall is likely adiabatic and it is being heated from the bottom with the 2000W/m^2K. and I think you're saying that the initial condition is that the water was at 1000K?


            that setup doesn't really make sense so you might want to describe what you're trying to simulate as well.

              • Re: heating of water in flow simulation
                Mohamad Nafis

                1000K is the heat applied to the bottom surface of the tank..i select the outer wall in boundary condition and put 1000K for external fluid temperature and 2000W/m^2K for heat transfer coefficient..i just assume the parameters..


                what i'm trying to simulate :  a tank filled with water is heated from the bottom surface..like boiling water in a pot..


                can you please describe the right steps to do this simulation? flow simulation is new to me and i can't find any similar simulation on the internet..

                  • Re: heating of water in flow simulation
                    David Paulson



                    Because of your heat source of 1000K, I don't think that you will get accurate results from Flow.  1000 K. is very much beyond the boiling point of water.  The heat exchanbge at the bottom of your tank will gas/liquid and gas/gas.  Flow does not have the ability to determine the relative heat exchange surface that is gas/liquid vs. gas/gas.  This is beyond even the limitation of computing two phase flow that is beyond the ability of Flow.  This also requires the ability to compute the nucleate boiling surface, which is a function of the geometry of the heat exchange surface. 


                    Any software that could purportedly solve this issue would also have to be tested with actual field calibration.

                    • Re: heating of water in flow simulation
                      Jared Conway

                      Hi Mohamad, David's got some valid points here. This is a bit outside the realm of what Flow Simulation can do. You can do some of it, but you're going to have to make assumptions on much of it and neglect some of the physical behavior that actually happens.


                      What exactly do you want to learn in this simulation? That might help us point you in the correct direction.


                      That being said, your choice of boundary conditions doesn't make much sense  to me. What you've told flow simulation is that your pot is in a 1000K environment with 2000W/m^K heat transfer coefficient. This would be like being inside a very very very hot oven where the air is being blown over at a very high speed. http://www.engineeringtoolbox.com/overall-heat-transfer-coefficients-d_284.html


                      What exactly is the environment that it is in?


                      And 1000K heat at the bottom. That might be ok. What are you trying to simulate here?


                      The closest we've done in our consulting work is a camping heater. To simulate the boiling water we created a component that was at 100deg C and neglected the actual phase change and the properties changes in the water. What we were able to read is the amount of energy required to get the system to equilibrium which we could compare to known values. Through that, we were able to compare design to design. In this simulation we also had to create some workarounds for combustion but overall the method came pretty close to experiment.