Hi, I think this is entirely possible: you define your own engine that will control the criterion for stage transition (e.g. stabilization of the box compression) and if it is satisfied, it can do whatever it wants (including dispatcher additions/removals from rootBody) and then go ahead to the next stage. I think Bruno had something similar with his triaxial test?
And for singleaction, someone else knows better probably? Regards, Vaclav > Hi all, > > I wonder if it is possible to simulate multistages in YADE. Let's say I want > to simulate the following steps: > > 1. A box with 7 walls is created (4 at sides, 1 on top and 2 at the bottom, > one for the hole we will remove after) > > 2. A box with stresses applied on a wall on top (similar to odometer test), > and then, after the simulation stable (with the value of all the translations > of the bodies equal zero), we remove the walls on top (erase the ID of the > top) and remove the wall from the bottom. > > 4. When a new stage is calculated, other engines can be applied to calculate > the motion of the bodies (like contact laws, etc..) (no idea how to deal with > it) > > However, i dont know if we can process the above stuffs with the > fileGenerator derived class since we must redefine the dispatcher engines. > > Btw, i'm not sure about the task of a method singleaction(), what is it used > for?:) > > Not sure if I exposed my question clearly, > > Thank you so much > > Kien Dang > > > Thank you so much > > > _______________________________________________ > Yade-users mailing list > [email protected] > https://lists.berlios.de/mailman/listinfo/yade-users > _______________________________________________ Yade-users mailing list [email protected] https://lists.berlios.de/mailman/listinfo/yade-users
