17 Replies Latest reply on May 6, 2013 6:45 PM by James Lin

    Unresolved fan

    James Lin

      I have setup an internal fan in a enclosed box. I picked a predefined fan from the database. When I ran the simulation, it said "Unresolved fan" in the warning info. Could anyone help on this issue or explain what this mean and how can I prevent this to happen. The set up has no boundary condition because it is a sealed box with gravity turned on in the initial setup.

       

      thanks,

        • Re: Unresolved fan
          Jared Conway

          is the fan inside the computational domain?

           

          are you using an internal analysis?

           

          how is your fan setup?

           

          have you tried using a really simple model to check that your setup is valid? (box with a fan inside of it) if that works, it is most likely computational domain/model related (like leaks).

            • Re: Unresolved fan
              James Lin

              Thanks for you reply

               

              Yes the fan is inside the computational domain

               

              I am using external analysis which the box is sealed and submerged in water.

               

              I put a circular disk inside the box to represent the fan.

               

              The model is very simple, I wonder if the box volume is too small for the fan to work inside.

               

              CONCEPT HT ISSUE1.JPG

                • Re: Unresolved fan
                  Jared Conway

                  are you using an internal fan?

                   

                  can you get it to run as an internal analysis? does solidworks recognize the computational domain properly when you make that switch?

                   

                  for the external analysis, what is the size of the computational domain that you've chosen relative to the tube? you might be right that they are incompatible. unresolved is usually because of mesh or the problem setup. (couple of good kb articles about that)

                    • Re: Unresolved fan
                      James Lin

                      Yes I am using internal fan

                       

                      I just set it up for internal only which will simplify the domain for gas only (previously having two domain, water and air as sub). With just one domain (air only inside the enclosed volume), the fan still unresolved.

                       

                      I will try to simplify the model even more, leaving only the fan inside the tube.

                       

                      Thanks alot for giving me a hand here.

                      • Re: Unresolved fan
                        James Lin

                        It works now because I make the tube size much bigger. It looks like the fan won't work if the tube size is too small.

                          • Re: Unresolved fan
                            Jared Conway

                            check out dropbox or yousendit. that will allow you to post the file on the web. put the link up here for us to download it.

                            • Re: Unresolved fan
                              Jared Conway

                              i'm not sure what version/service pack you're using but 2013 sp2.0 runs that problem without issue. (without the tube wall thickness change)

                               

                              it does produce some fan mass flow rate errors but they are nothing to be concerned about.

                               

                              (i ran the one that says with all 8)

                                • Re: Unresolved fan
                                  James Lin

                                  yes, you are right that i am still using 2011 version which is not as good as the 2013 version. I would think in 2011 version the fan will not work in a enclosed volume that is too small for its space geometry to develop. I mean the fan has a physical size plus a physical addition zone in its built in parameter.

                                   

                                  i make the tube 10 times bigger and the fan will work inside the tube.

                                   

                                  thanks alot for the help and input

                                    • Re: Unresolved fan
                                      Jared Conway

                                      james, you could make the tube the same diameter as the outside diameter as your fan and it should still work. i have a feeling it has something to do with mesh size choice.

                                       

                                      is the model you attachedh ere the 10x size version or the one without the size change?

                                        • Re: Unresolved fan
                                          James Lin

                                          The model I posted yesterday was the small tube size without 10x size change. I have tried to do mesh control on the fan volume to get the element size "reasonable" (may be i did not get it right).

                                           

                                          On the other hand, if we make the fan OD same as the ID of the tube, that may mean we won't allow the air circulates because there is not a return path from the outlet face to the inlet face of the fan. If Flowwork is smart enough, it would just give us a presure diference with no volume flow rate (for simplicity). Because it is a internal fan, I may think it has different mass flow rate balance issue; this may be the cause of the unresolve fan warning (this is my wild guess).

                                            • Re: Unresolved fan
                                              Jared Conway

                                              james, we checked here at Hawk Ridge with your model, in 2011 SP5.0 that model runs without issue (no warnings).

                                               

                                              my example was simply as an example of what i've seen the software do before such that your setup shouldn't be an issue. (it should run regardless of how close the wall is to the fan)

                                                • Re: Unresolved fan
                                                  James Lin

                                                  Thanks for helping me to check with Hawk Ridge, i am using 2011 SP5.0 also. Because the workstation is running  alot tasks now, i have not had a chance to rerun the model and cut some screen shots. I will try to find a gap this week to rerun the model.

                                                   

                                                   

                                                    • Re: Unresolved fan
                                                      James Lin

                                                      Hi Jared,

                                                       

                                                      Thanks for the help last time. This time I got another problem. May you take a look at this very simple model? I wonder why this time the solver terminated abnormlly.

                                                       

                                                      Thanks,

                                                        • Re: Unresolved fan
                                                          Jared Conway

                                                          hi james, please start a new thread for this issue. so that we don't fill this one with troubleshooting for issues that aren't related.

                                                           

                                                          in this case i created a brand new study with the basic setup and it worked. meaning that it is a setting that you have or boundary condition.

                                                           

                                                          i suppressed everything you had in the tree. same problem.

                                                           

                                                          switched your selections for fluid and solid. same problem.

                                                           

                                                          when i dug into your initial conditions it looks like you set turbulence length to 0m. i entered a value and it worked. generally it isn't recommended to change that value unless you have data for it.