7 Replies Latest reply on Apr 8, 2014 11:53 AM by Jared Conway

    Mixing tank with rotating stirrer

    Xenofon Kalogeropoulos

      Dear All,


      I would like to simulate a mixing system. My setup is the following (also see attached picture):


      • Tank: Normally the top of tank is open to the atmosphere.
      • Stirrer: The stirrer is a long circular rod with 2 propellers fixed on it.
      • Fluid Medium: Water for the moment


      Should I set up the study as internal analysis with a lid on the water-level of my tank and define this as environmental pressure opening?

      For the rotating stirrer I have created a separate part that completely encloses the stirrer and defined local rotating region @ 200 RPM.


      I have tried with this setup but unfortunately the solver terminates abnormally after reporting that there is negative pressure (1.8e+81 or some sort of crazy number like that).

      What do you think the problem is?


      Also, would it be a good idea to define the walls of the tank as "Wall Boundaries"? Are there any goals I should set to help the solver converge? if so, what would an appropriate goal be for a general feeling of how good/bad the water is being mixed?


      Any ideas or comments would be more than welcome!

      Thanks in advance guys.




        • Re: Mixing tank with rotating stirrer
          Jared Conway

          this one is going to be a challenging analysis, if you havent' done so already, go through all the tutorials and read through the documentation on rotating regions.


          from a setup perspective, a lid at the top is required, you can't do 2 phase flow in simulation (ie, it won't recognize the boundary of air and fluid in the pot)


          likely your solver terminated abnormally problem is due to geometry, breaking the rules for rotation or bad geometry. see above comments and also check out this article: http://www.hawkridgesys.com/blog/flow-simulation-4-starting-flow-simulation-analysis/


          no need to do anything to the walls

          for goals i don't really know what you mean by "how good the water is mixed". what is your measurement for success? how would you do it in the real world?

            • Re: Mixing tank with rotating stirrer
              Xenofon Kalogeropoulos

              Thanks Jared! Your comments are very welcome.


              If you have the time can you please ellaborate a little on why the setup will be a challenging analysis?


              I guess we would have to leave goals for the moment, since the project hasn't got a definite goal as part of its scope yet. My primary focus is to get the setup correct and working.


              I have checked the geometry a couple of times with the Check Geometry tool. It seems fine. I changed the rotating region geometry to just a plain cylinder (before it was basically a cylinderical contour following the impellers at 1mm offset). Now the solver works nice. I also removed the Envrionmental Pressure BC from the lid, as I want to simulate it as a closed volume (I didn't want water to go through the lid).


              I have attached a velocity plot for reference. Let me know what you think.I am a bit concearned about the region in red below the second impeller. the transition seems a bit "sharp".


                • Re: Mixing tank with rotating stirrer
                  Jared Conway

                  its looking better but i don't think you're following the recommendations for a rotating region


                  you may want to try splitting each blade into its own region to see if it works better.


                  the reason i say this is going to be hard is that it seems like the problem is basically like a perpetual problem which may or may not have a real steady state solution. the other thing is that rotating regions are made to have fluid flow though them axiallally and axisymetrically and you're going to have a lot of mixing.


                  i'd do some reading up on rotating regions and speak with your resller to see if this is a good application for rotating regions. if someone brought this type of project to me as a consulting or trianing project, I would want to check with the developers first.

                    • Re: Mixing tank with rotating stirrer
                      Xenofon Kalogeropoulos



                      I have now changed a couple of things, and there seems to be an improvement. I have now got 3 different rotating regions (impeller1, impeller2, shaft between them). I have changed the analysis to transient for simulating the first 60 seconds.


                      Today I ran into another problem. Up to now I had been simulating with normal water @ 273K. Now I want to simulate using Refrigerant HFA134a, which does not exist in SW but I have gone ahead and defined it as a custom fluid from the company's datasheets.


                      I am repeating the study, with the exact same setup just different fluid and 218K temperature. The solver gives me this constant warning "Negative Pressure, minimum pressure=4.2e+8 Pa; dV/V=19.8544" , see attached screenshot below.


                      Does anyone know what is troubling the solver with the fluid changeover? anything I am overlooking?


                      Also when trying to simulate with the HFA134a fluid, the solver takes many times longer to solve....and I mean ages longer? is that typical? or again I am missing background theory?


                      Any comments/thoughts are welcome!

                      Thanks all!

                        • Re: Mixing tank with rotating stirrer
                          Jared Conway

                          Rotating region on the shaft would not be recommended.


                          What result do you get for steady state?


                          Do you have cavitation enabled?


                          What are the main differences between the fluids that you must check the other fluid? What are the expected result changes?



                          Have you followed all the recommendations for rotating regions? Does your application fit the boundaries of rotation?

                            • Re: Mixing tank with rotating stirrer
                              Xenofon Kalogeropoulos

                              Hi Jared,


                              I resolved the problem with the negative pressure. There was a mistake in the definition in the viscisity of my custom fluid. The solver runs just fine now which is good.


                              I just gave a go at running the analysis at steady state. I have copied 2 plots below.



                              These were performed with 2 rotating regions on the impellers. The connecting shaft is set at stator.


                              I had a closer look at the technical reference as well as the Solving Engineering Problems document. One thing I found out is:

                              "If you consider gravitational effects in your analysis, the rotation axis must be parallel to the gravity vector."  Now my model contradicts this statement because the stirrer is inclined from the vertical axis that gravity acts upon.



                              In the Solving Engineering Problems document I have been trying to follow the following:



                              Thanks for helping out