18 Replies Latest reply on Jul 2, 2014 6:46 AM by Simone Storace

    Problem with cavitation in a simple model

    Simone Storace

      Hi,

      let me introduce myself: my name is Simone and I am a novice in this community and in SW.

      I've just started to use SW-FS and I immediately found a question about cavitation.

      I'll attach 3 pictures and explain:

      img 1) -> Contours of static pressure used to find where cavitation can grow, that is the part in which water pressure, at a constant and fixed temperature, drops below the value of the vaporization pressure . So I redo the calculation with cavitation method activated.

      img 2) 3) -> contours of pressure and mass fraction of vapor in the second calculation. In this case the pressure seems to have a higher value of pressure than before and now the minimum value is greater than vaporization pressure.

      How is possible to have a non-zero value for a mass fraction of vapor with a value of pression greater than vaporization pressure?

       

      Boundary condition:

      - Transferred boundary condition in inlet (this is a submodel)

      - Fixed value in pressure outlet (P=110000 Pa)

      - Adiabatic wall and stationary wall (no rotating zone)

      - Temperature =362K

      - Vaporization Pressure = 70000 Pa

       

       

      Thanks,

      best regards

      Simone

        • Re: Problem with cavitation in a simple model
          Jared Conway

          Have you tried this without transferred bcs?

           

          What are your convergence criteria?

            • Re: Problem with cavitation in a simple model
              Simone Storace

              No, I've not tried yet, but I'll do a test today with a simple model with "standard" bc pressure/pressure.

               

              My convergence critetia is the same suggesed in Flow Simulation Help Topic, 2.5 travel fixed and global goal density.

               

              Yesterday I did another model with a longer outlet pipe; the cavitation zone moved a bit to outlet, but the problem has remained (mass fraction of vapor non-zero with static pressure > P vaporiz).

               

              Thank you

                • Re: Problem with cavitation in a simple model
                  Jared Conway

                  Haven't played with cavitation a lot. Have you confirms your expectations match the cavitation model in the software and that you are within the physical limits it works within? Documented in the tech reference and the solving eng problems document?

                    • Re: Problem with cavitation in a simple model
                      Simone Storace

                      By the model used, my expectation was:

                      a zone with cavitation near last edge and a following pressure recovery.

                      If you take the picture n° 1 in the first post, I believe that the zone with green circle would have cavitation and after only water (the pressure recovery).

                      I've made the second test with tha same geometry (but with cavitation mode activated) to validate the first results, but now I more confuse than before..

                      Obviously I used water pre-define as fluid

                      The range written in tecnical reference is:

                      • 280.15 K < T < 583.15 K
                      • 10^3 Pa < P < 10^7 Pa
                        • Re: Problem with cavitation in a simple model
                          Jared Conway

                          Would suggest uploading your model

                            • Re: Problem with cavitation in a simple model
                              Simone Storace

                              Sure, how can I do that? The .zip file is 83MB and i can't attach it

                              I've the results of the simple test with "standard bc" and the problem is the same. I'll upload this calculation

                              Thanks

                                • Re: Problem with cavitation in a simple model
                                  Jared Conway

                                  pack and go the assembly to a zip

                                  then right top corner of replying, use advanced editor

                                  upload the file there

                                    • Re: Problem with cavitation in a simple model
                                      Simone Storace

                                      I've tried to upload the file, but the maximum dimension to upload is 25 MB..

                                        • Re: Problem with cavitation in a simple model
                                          Jared Conway

                                          just post the files, not the results

                                          will have to rerun the results but that is ok

                                            • Re: Problem with cavitation in a simple model
                                              Simone Storace

                                              Ok,

                                              thank you for the help.

                                              This is a part of my model.

                                                • Re: Problem with cavitation in a simple model
                                                  Jared Conway

                                                  Haven't had a chance to take a look at your model or dig into cavitation. Did you end up finding any examples with a known answer?

                                                   

                                                  Or a reference that would show that your expectation would be reality? It sounds logical, but some references for others to look at might get you a quicker answer while I look for time to look at your model.

                                                   

                                                  I also assume you've done things like vary mesh and double check boundary conditions..etc.

                                                    • Re: Problem with cavitation in a simple model
                                                      Simone Storace

                                                      Hi Jared,

                                                      It seems impossible, but I haven't found anyhing about Solidworks cavitation yet. Few examples and few documents.

                                                       

                                                      Yes I've tried to change and increase the mesh quality, to use transferred boundary condition and "standard" boundary condition, to use a greater range of values, but the results have always the same problem.

                                                       

                                                      If you can suggest other examples or documents to read (different from solidworks tecnical reference) I would be grateful.

                                                      Thank you for the help

                                                        • Re: Problem with cavitation in a simple model
                                                          Jared Conway

                                                          hi simone, my understanding is there is no cavitation example or validation model. so my suggestion is mostly around finding a known physical case and using it to check the software. that way we aren't testing the software against our assumptions, we're testing it against a known case.

                                                            • Re: Problem with cavitation in a simple model
                                                              Jared Conway

                                                              so i pulled down your file and am still running it but cavitation is definitely showing in the model

                                                               

                                                              that being said, i did make some modifications to your model

                                                              1. i got rid of the 2 lids that you had in the middle of your model and just ran the whole thing. one lid created a super small step and the other looked like it was just used for measuring mass flow rate.

                                                              2. i merged all your bodies together. your lids had overlaps in your model which i wouldn't recommend.

                                                              3. i went down to level 3 mesh to get a fast solve time. i also re-enabled the optimize thin wall option. the developers recommend always keeping this enabled.

                                                              4. added goals at inlet and outlet for pressure, temp, velocity, flow rate, mass rate so that i could keep an eye on what was going on

                                                              5. fixed your calc control options. i think 2 things that could be causing you issue, you set max travels to 2.5, leave it on auto and disable it. for global density, you set the convergence criteria to 2.5kg/m^3, this could be a pretty wide goal. so i just went with automatic. i set the interval check to 2.5. i disabled all convergence criteria. the reason for these changes is these are the recommendations in the KB, tech ref and solving eng problems doc.

                                                              6. in the solver window, disabled most of the windows you had open. all of them take cpu. so having them disabled allows the solver to use the CPU to maximum extent.

                                                               

                                                              overall it is running without issue and showing cavitation starting where the model turns 90degrees. i'm at around 2500 iterations and only about 1.5 travels so it hasn't measured for convergence yet.

                                                               

                                                              the other thing i wanted ot mention about tutorials is that what escaped my memory is that there is a validation example on cavitation in the tech ref. and in the solving eng problems document there are several examples that you oculd replicate and compare against.

                                                               

                                                              make these changes and let me know if it changes what you are seeing. note, i am using 2015 beta 1. this could also be a factor but i don't think so. wanted to mention it because that is the reason i can't share the files.

                                                                • Re: Problem with cavitation in a simple model
                                                                  Jared Conway

                                                                  here's the result after 3 travels

                                                                    • Re: Problem with cavitation in a simple model
                                                                      Simone Storace

                                                                      Hi Jared,

                                                                      let me say that you are a hero

                                                                      Could you upload the result of static pressure, so I can compare it with mass fraction of vapor, please?

                                                                      I try to reply to your points:

                                                                      1)I couldn't study the whole model because it crashed after a thousand iteration (also with different mesh). I don't know why. So I used only a part of the model to attach you some result in few time.

                                                                      The first lid was very bad with the small step, you're right, i had not noticed that.

                                                                      The second lid was created only to check mass flow rate

                                                                      2) "i merged all your bodies together" I'm sorry but I didn't understand. Which bodies and why? the main body and the lids?

                                                                      3)I believed that "thin wall option" was useful only for the solid cell (for example solid conduction), was I wrong?

                                                                      5) So you suggest to disable all convergence criteria and check and stop manually the calculation? In Fluent I forced the end of the calculation watching the residual without any convergence criteria. In Flow simulation how can I understand the moment when stopping the run?

                                                                       

                                                                      I've seen the tech. ref you were saying, I'll try to run one of the example next days.

                                                                      What do you mean for "the solving eng doc". Is it a document inside SolidWorks help?

                                                                       

                                                                      Thank for your time

                                                                      Greetings

                                                                      Simone

                                                                        • Re: Problem with cavitation in a simple model
                                                                          Jared Conway

                                                                          i think i lost the file, i can't fnd it anymore. with the changes you should be able to get the pressure plot

                                                                           

                                                                          regarding your questions

                                                                          i merged the lids because it made for easier meshing, and reduced cell count. this will also minimze bad mesh and crashes due to bad mesh or solving errors due to bad mesh.

                                                                           

                                                                          you're correct about thin wall, but your walls are still used as part of the meshing algorithm as far as i understand. no reason to disable it.

                                                                           

                                                                          solving eng problems document is in the install directory > lang > english > docs