19 Replies Latest reply on Mar 27, 2013 1:22 PM by Jared Conway

    FloXpress Issue "Meshing in progress..."

    Dan Stevenson

      Hello all,

       

      I'm a student working on my final project and I could really use some help. I'm an Electrical Engineering student and I'm reuqired to do a bit of SolidWorks for our project. So here is the issue:

       

      I have a fuselage for an airplane and I need to run a floXpress Simulation on it to show the pressures inside and the airflow around the body. My group has calculated both the input and output pressures for the simulation at our required height and speeds. I've run through the Solidworks FloXpress tutorial provided by their help section so I know what I'm doing as far as the steps it takes to run the simulation.

       

      I've tried numerous times and I cannot get the computer to run the simulation correctly. I made a tube around my fuselage with caps (not merged with the tube), selected one face as input, one face as output and inserted my pressure values for each. Pressure values are 90000 Pa for input and 89738.75 Pa for the output. Alls fine up to the last page. I click the green triangle "go" button and the progress bar pops up. Good, right? Except it will not pass the "Meshing in progress..." phase no matter the length of time I give it.

       

      We've run it on a number of computers as well. All computers are running Windoes 7 64 bit with 4GB of RAM and they all have Intel processors. We've run it on core 2 duos up to the i7s to no avail. My advisor believes this may be an issue with the graphics cards our school runs (I believe they are all NVIdia Quadro 580).

       

      I'm hoping that by posting here somebody will have some insight as to whether we should be looking at the graphics cards, OS, Ram, etc. or if we should be looking at the simulation itself. Do the pressure values seem correct? Is the fuselage maybe too complicated for the meshing?

       

      Anyways, any help is greatly appreciated. I'll talk to my advisor about uploading our fuselage here but I doubt it will be an option for privacy purposes...

       

      Thanks!

       

      Dan

       

       

      EDIT:

       

      It was able to be run on another machine. Has anybody run into issues with this on Windows 64-bit? That's the only thing different on the computer other than permissions.

        • Re: FloXpress Issue "Meshing in progress..."
          Chris Michalski

          The model is likely increadibly thin compared to the area.  This will take a large number of elements.  If you look at the processor usage I would assume that when you give up during the "meshing in progress" that it is still working. 

           

          Did you model your fuselage as a solid or a hollow body?  If it is hollow try making it a solid if you are only interested in the external flow.

            • Re: FloXpress Issue "Meshing in progress..."
              Dan Stevenson

              The tube around the fuselage has an area roughly 3x that of the fuselage. We've run it for 7 hours on an i5. I'm running it now on a core 2 duo and the processor is indeed still running high (around 70%). Would you recommend a computer cluster or a supercomputer to run this then?  We have the resources I believe just a matter of reserving them. The odd thing is however that even with the same values and a straight tube with nothing inside it, the simulation won't pass the meshing phase.

               

              We're interested in the pressures and such inside the fuselage as well so we need to keep it open.

                • Re: FloXpress Issue "Meshing in progress..."
                  Chris Michalski

                  Clustering is unlikely to help unless the meshing stage is greatly different from the convergence stage. 

                  Part of the problem may be limitations of FloXpress (I only use full Flow Simulation so I'm unsure how much you can customize meshes and such.

                   

                  But if you have a plain tube that takes a long time you likely have other issues.  Is that tube thin wall also?  it may be wasting time and cells by defining the thickness of your boundary.  Can you make it a much thicker tube and see if the empty tube meshes acceptably?

                    • Re: FloXpress Issue "Meshing in progress..."
                      Dan Stevenson

                      I am thinking there are larger issues here than the computer hardware as well. We're speaking with IT as well but they aren't generally very helpful.

                       

                      I made a tube with a 6"ID and a 46" OD (Extremely thick walls...) The tube is 12" long. It has end caps on the 6" ID of 1" thickness from the face of the tube outwards.

                       

                      Step 1: Fluid Volume is 6.012"

                      Step 2: Air

                      Step 3: Pressure, selected inlet at 90000PA and 281.2K

                      Step 4: Pressure, selected outlet at 89398.75PA

                      Step 5: Solve - > Progress bar "Meshing in progress..."

                       

                      I just started but so far nothing, I would think it wouldn't take long for such a small area. I'll let it run for a bit and update if anything changes but nothing so far.

                       

                      For our pressure values we used 90000 because that is the environmental pressure at 1000m above sea level (where we will be flying) and 89398.75 from bernoulli's equation with our velocity at ~25m/s, just to provide as much info as I can

                • Re: FloXpress Issue "Meshing in progress..."
                  Jared Conway

                  Where are you at on this?

                   

                  I checked the model you uploaded, there was no study included.

                  I checked a simple setup on it, 101325 to 50000 and it runs without issue. Can you try that on your end? Also can you tell me what you have set for the smallest fluid passage. If that value is too low, meshing is going to take a long time. I used the default which was about 300mm.

                  if you're still having trouble. do other models work? make sure they are located on your desktop.

                  Note, is there a reason for the large outer diameter?

                   

                  Bottom line, for the type of model it sounds like you're going to be using, you're not going to have enough control over the mesh with floXpress, you need Flow Simulation. You'll need adaptive meshing and controls to do it. Most likely the FloXpress study is upset because of the limitations with automated meshing in that tool. If your school has SolidWorks, you should have access to the full Flow tool.

                   

                  A couple other notes:

                  -Whether you use FloXpress or Flow Simulation, there is no clustering. It will use SSDs and the RAM you give it and solve time improvements start to level off at around 6 CPUs.

                  -my suggestion for setup would be velocity in, pressure out. you really need the full flow simulation to do this right but this is as close as you're going to get. you're going to have the effect from the walls affecting your model. if all your calcs are right, the pressure at the inlet should be what you calculated.