16 Replies Latest reply on May 25, 2012 8:43 AM by Dan Hofstetter

    Transient, Recirculating Flow - Possible?

    Devin Spratt

      Hello!

       

      Without getting into too many details, I'm attempting to model a two-fluid heat exchanger, with the fluids as hot air and cold liquid.  In the system, the liquid would start at 25°C, go through the heat exchanger (heating up), run through a series of other components and return back to the inlet of the heat exchanger.  However, when it comes back around, it won't be at 25°C anymore, it will be warmer.  For the sake of reason, let's say the liquid comes back at the temperature at the exit of the heat exchanger.

       

      I haven't come across any way to link the temperature of the entrance boundary condition to that of the exit (can I create a variable from the bulk average leaving temperature?), leading me to believe that in the current configuration it isn't possible.

       

      My only other alternative would be to connect the inlet and outlet with a pipe.  If this is done, how would a boundary condition be set such that the flow would in fact, flow?

       

      Any other work-arounds?

        • Re: Transient, Recirculating Flow - Possible?
          David Harrold

          this is an intriguing question. for what it's worth: I have no idea how to link an inlet boundary condition to the result at an exit. I haven't found a way. that would be cool if you could.

           

          another workaround could be to model the pipe connecting the outlet to the inlet, then add a heat source to that pipe to approximate the other components in the system. If you did it this way, you'd want to set the inlet flow conditions (temp, flow, etc) as you already are, and set the final outlet condition to a pressure opening. Atmospheric, or whatever is appropriate for your system. That should solve the recirculating flow.

          • Re: Transient, Recirculating Flow - Possible?
            Bill McEachern

            You might want to put it at a domain boundary - both BC's -  and try the cyclic option. I haven't tried it but it might work.

            • Re: Transient, Recirculating Flow - Possible?
              Dan Hofstetter

              I am looking for the same type of information.  I am running an internal simulation for a rectangular manure tank with a circular fan inlet on the top of the tank at one end, and a rectangular slot outlet on the top at the other end.  Hydrogen Sulfide gas is being generated by the manure on the bottom of the tank, and there is an initial concentration of H2S in the tank (142.9 ppm).  Field studies have shown that in still air (no wind), 30% of the exhausted mixture of air and H2S gets recirculated back to the fan inlet.  I am doing a transient study to show how long it takes to drop the level of H2S in the tank below 10 ppm, and need to consider this recirculation.

               

              I am told most other CFD packages allow the outlet mixture to be linked back to the inlet fairly easily - how can I do this in Flow Simulation?

                • Re: Transient, Recirculating Flow - Possible?
                  Bill McEachern

                  use an internal fan with one surface on the  outlet and the other on the inlet. this will take the out flow and put it on the inlet. If this is what you want it is easy peasy. Scaling it back by 30%  is not something I knowhow to do directly. You could build some ducts and another interna fan to build a dilution curcuit - shouldn't be too hard. You could also build a box that allows the flow path to recirculate you will get what you get but at least you can measure it. Whomever gave you the 30% number is generalizing unless for the exact situation at hand they have measured it. If that is the case build the whole geometry that enables the recirculation with whatever external conditions are present if you want to get it right. Put in sensors to measure what ever is being reciculated. Not to difficult either. Transient will take awhile. I would run steady state studies till you tune it for the deesired recirc. for whatever its worth.

                    • Re: Transient, Recirculating Flow - Possible?
                      Dan Hofstetter

                      Yes, that 30% number was based on measured levels for a research facility.  I am trying to set up a model with the same geometry to determine how closely SW Flow can predict the actual results.  Phoenics was used previously and was within 10% of the actual measurements.

                       

                      Is there a way to place a sensor at the outlet and use a variable and equation to set the inlet concentration equal to 0.30 x outlet concentration?  i.e. if 10 ppm is exiting, then ~3 ppm would be entering the fan.  This would need to be in an equation so that the level could change over time.

                       

                      Thanks

                      • Re: Transient, Recirculating Flow - Possible?
                        Dan Hofstetter

                        Well, I tried to use the "build a box" approach, and got something to work, but I am having trouble with the concentrations.  I was hoping someone here could offer some more advice.  I set up the tank inside the box, and set a fluid subdomain within the tank walls with an initial concentration of H2S of 142.9 ppm by volume (air = 1 - 0.0001429).  I then built an outer box that was many times larger, and set the initial conditions for the simulation so that the initial concentration was air 1, H2S 0.  Using an internal fan as suggested, I set the outlet face just inside the lid of the tank and the inlet face just above the lid, so that fresh air would be drawn from the larger box into the tank, and air gets exhausted from a narrow slot on the opposite wall of the tank into the larger box.  I can see a plume of H2S being exhausted into the larger box, which is great, but instead of the concentration level within the tank dropping over time to zero, it actually increases above the initial fluid subdomain value of 142.9 ppm.  What it seems to be doing is setting the concentration for the entire computational domain to the 142.9 ppm value, and recirculating some of that into the tank, adding to the initial 142.9 ppm value.  I can't figure out how to make this work the way it should - any ideas?  I can post (if forum allows) or email the model if it would be helpful.

                         

                        Thanks,

                         

                        Dan

                          • Re: Transient, Recirculating Flow - Possible?
                            Ron Reiners

                            All,

                             

                            I am a little late to this discussion. Way behind on e-mail review. If you are still working on this project I have some information that might help.

                             

                            In my typical model I monitor the temperature of a specific component and use that temperature to vary the input wattage of a heat source in order to hold a steady temperature. I do this by defining some "Equation Goals". My 1st equation goal measures the current wattage of the heat source. My 2nd equation goal is a math formula that adjusts the wattage value based on the current temperature. The result of the 2nd equation goal is then used as an input value for the heat source dependency table. Every iteration of Flow uses a new wattage value based on these goals.

                             

                            I think this information might work for both the heat exchanger and the manure tank.

                             

                            Let me know if you need more information.

                             

                            Ron

                              • Re: Transient, Recirculating Flow - Possible?
                                Dan Hofstetter

                                Ron,

                                 

                                That sounds like what I was looking for - can you post an example of the equations?  For the time being, I run the simulation for one time step, then look at the average volume fraction of H2S on the surface of the outlet lid, multiply that value by 0.24, and change the inlet volume fraction of H2S to match for the next time step.  Since I'm only running a total simulation time of 80 seconds and a time step of 10 seconds, this isn't too painful for the moment.  I wouldn't try it for a time step of 1 second or less though.

                                 

                                I have a surface goal for the outlet lid for the volume fraction of H2S, so I created the following equation goal:

                                 

                                {Inlet Volume Flow 1:Volume Fraction Hydrogen Sulfide:< Dependency >}=0.24*{SG Av Volume Fraction of Hydrogen Sulfide 1}

                                 

                                Currently the volume fraction of H2S in the inlet air is based on an F(time) table.  Should I change this to formula?  If so, how do I link the formula in question 1 to the equation goal above?  I tried to paste the text for the eq goal but it gave me an error.  I would probably need to do this for the volume fraction of air also, basically setting it equal to 1,000,000 - vol fraction H2S (in ppm).

                                 

                                Thanks!

                                 

                                - Dan

                                  • Re: Transient, Recirculating Flow - Possible?
                                    Ron Reiners

                                    I think you are on the right track. The equation the you show might be a little too much. I would think it would just be 0.24*{SG Av Volume Fraction of Hydrogen Sulfide 1}. Then whatever the name of this equation is, would be used as the input value. 

                                     

                                    I am not sure if you need a dependency table or not. I use the dependency table to set end limits so the value cannot go below 0 or above a predefined limit.

                                     

                                    I have attached part of my setup file for your reference. I know at time this can be a little confusing which is why I made my own setup instructions.

                                     

                                    You might have to go through some simple test files and setups to refine your configurations of the formula's.

                                     

                                    Let me know if any if this helps or if you need more information. I am not an expert, just someone that did a lot of trial and error on my specific requirements, but I try to help where I can.

                                     

                                    Ron