| Thread Previous • Date Previous • Date Next • Thread Next |
Václav Šmilauer a écrit :
I don't understand the problem. F_s and ΔX_s are in the global coordinate system, not in any contact plane.(16) is supposed to update shear force F_s (vector): F_s=F_s+K_s*ΔX_s but this is not correct, since you add ΔX_s (which is in the _current_ contact plane) to F_s (which is shear force from the previous step, i.e. in the _previous_ contact plane)!!Yes, they are in global coordinates, but, at the same time, current F_s must be always perpendicular to current n (in global coords); otherwise, we wouldn't call is F_s, right? Now, F_s from previous step is perpendicular to _previous_ n, while ΔX_s is perpendicular to _current_ n. Summing these 2 together doesn't make sense, since the sum not perpendicular to anything (neither current, nor previous n).
Well, in the code you rotate F_s before summing dF_s : no problem.In a paper, you want to split the problems usually, and you comment on the contact law disregarding the rest. But ok, the equations for rotation should be explained somewhere.
The incremental approach is very common in the DEM litterature, but usually not explained in details in papers (perhaps because it is so common?). I'll send you a detailed paper in a separate mail. It is using the same "approximate" rotation equations than yade, with cross-product(oldN,newN), and then it defines the shear increment like in (16).
Bruno
v _______________________________________________ Mailing list: https://launchpad.net/~yade-users Post to : yade-users@xxxxxxxxxxxxxxxxxxx Unsubscribe : https://launchpad.net/~yade-users More help : https://help.launchpad.net/ListHelp
-- _______________ Bruno Chareyre Associate Professor Grenoble INP Lab. 3SR BP 53 - 38041, Grenoble cedex 9 - France Tél : 33 4 56 52 86 21 Fax : 33 4 76 82 70 43 ________________
| Thread Previous • Date Previous • Date Next • Thread Next |
