Announcement

Collapse
No announcement yet.

Beer with flip

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Beer with flip

    Im trying to simulate the example beer using the flip solver. Having problem with the foam crashing all the time. Some settings work ok but when I change just slightly it crashes.

    Any tips? or is it better to leave this til the official release ?

  • #2
    can you prepare a setup where it crashes? we already had some complains for foam crashes, but have no scene yet.
    ______________________________________________
    VRScans developer

    Comment


    • #3
      I´ve done nothing extra to the scene, you have it already. Just change to flip and press sim. It crashes most og the time when the foam kicks in at frame 11. I have max 2012 and latest nightly

      Comment


      • #4
        ok, weill try it. actually right now i'm experimenting with the beer (some foam issues are still remaining) but have no crashes. i suppose this is because the lower resolution, the beer sample will be decreased in resolution, because in the old core the high resolution was needed not to have better liquid, but to handle the glass geometry. the flip can handle thin glasses, and no need to simulate with such high resolution.
        ______________________________________________
        VRScans developer

        Comment


        • #5
          Ok. How many cells do you recommend for a realistic beer sim ? Althoug this is particles...

          Comment


          • #6
            170k is enough, however the foam still is not behaving as it should.
            ______________________________________________
            VRScans developer

            Comment


            • #7
              ok. I now use 170k cells. Foam Vel.threshold not working. It crashes most of the times. And rate less than 0,02 crashes also. Still produces too much foam..

              Comment


              • #8
                well, can you send me the log after a crash, and can you test the scene in the attachment? the log is placed in c:\phoenixfd
                the attached scene produces about 95k foam particles until frame 120, the time taken is about 3 min, so you can perform the test very quickly.
                Attached Files
                ______________________________________________
                VRScans developer

                Comment


                • #9
                  Where is this logfile? Is it the "phoenix_maxver14" file ? I can send it late tomorrow. Going away now..

                  Comment


                  • #10
                    yes, it's the "phoenix_maxver14" file
                    ______________________________________________
                    VRScans developer

                    Comment


                    • #11
                      Your scene is simulating without crash My own scene that is crashing turn out to be ok if I turn off and on the liquid. Then after awhile it crashes again. It´s hard to isolate the problem...

                      But I will send you the logfile as soon as it crashes..

                      Comment


                      • #12
                        Another thing is that the foam is not centered with that setup. It´s growing high and above the glas on one side.

                        Comment


                        • #13
                          yes, this problem is under investigation right now. i would recommend to hold any beers sims for a week until the problem is resolved.
                          ______________________________________________
                          VRScans developer

                          Comment


                          • #14
                            test the tomorrows nightly and tell us if it continues the crash
                            about the not centered foam cap - there was a bug producing "bricks" of foam, i.e. compressed foam that behaves like solid body, it was fixed, but you still can end up with foam in the one end of the glass, because the foam is born in the stream, it rises too fast and forms the asymmetric cap. the effect is increased because p2p interaction of the flip is stronger than the one of the grid core (it's not related to the flip, just an improvement). perhaps lower p2p would produce better result, and it's sure that the slow rising foam helps too, it has time to be scrambled in the glass. in few words, it's question of setup, we are experimenting to find the best settings, the sample scenes will be reworked and perhaps new tutorials will be released. for now i can say that the
                            ______________________________________________
                            VRScans developer

                            Comment


                            • #15
                              Ok great. I will have a go again

                              Comment

                              Working...
                              X