Blender Git Loki

Kaikki Blender Git kommitit.

Page: 5116 / 8462

January 20, 2015, 08:29 (GMT)
Cleanup, removed unused old code.
January 20, 2015, 08:29 (GMT)
Optimized matrix filling using the Eigen triplets method.

Otherwise the construction of matrices becomes very slow for larger
vertex counts because adding a new element is O(n), making it O(n^2) in
total.
January 20, 2015, 08:29 (GMT)
Added back external effector forces for the cloth solver.
January 20, 2015, 08:29 (GMT)
Reenabled goal springs and removed the arbitrary structural scaling
factor.
January 20, 2015, 08:29 (GMT)
Fixed bending forces for cloth springs.
January 20, 2015, 08:29 (GMT)
Correctly implemented and verified gravity, drag, structural springs.
January 20, 2015, 08:29 (GMT)
Minor fix in debug print, so solver prints can be diff'ed.
January 20, 2015, 08:29 (GMT)
A couple of defines to disable various influences on the cloth sim for
debugging.
January 20, 2015, 08:29 (GMT)
Debug printing code for large sparse vectors and matrices, to compare
solver input/output of the old and new methods.
January 20, 2015, 08:29 (GMT)
Added a new internal header for implicit solver defines.

Should become its own intern wrapper library eventually.
January 20, 2015, 08:29 (GMT)
Reenabled air drag force for cloth sim.
January 20, 2015, 08:29 (GMT)
Arbitrary scaling factor for the structural spring forces (stretching).

This is not nice at all, but for some reason (possibly time scale) the
old force values are much too high and cause the solver to become
unstable. These will be revisited later anyway, so for now such scaling
should be fine.
January 20, 2015, 08:29 (GMT)
Switched to the modified CG method that supports constraints, and
added back structural stretch springs.
January 20, 2015, 08:29 (GMT)
Implemented a modifier conjugate gradient algorithm to support stable
constraining.

The algorithm is described in the paper "Large Steps in Cloth Simulation"
(Baraff/Witkin 1998). The same method was (incorrectly) implemented in
the old cloth solver.

It is based on restricting the degrees of freedom (ndof) of vertices
using a block matrix and a vector of target velocity deltas.
See chapter 5 of the paper for details.
January 20, 2015, 08:29 (GMT)
Reimplemented Goal springs for the Eigen CG solver method.

Note that goal springs currently are really bad ... They have a factor
on hairs that "fades" goal influence from the root to the tip. The last
point on the hair is completely free, which makes the goal springs
pretty much useless on their own without supporting bend stiffness.
Can only assume this was added to compensate unphysical behavior of
goal springs when using uniform weight, but it's a poor replacement for
true localized bending forces ...
January 20, 2015, 08:29 (GMT)
To make gravity a true force vector, multiply by the mass.
January 20, 2015, 08:29 (GMT)
Alternative new solver for cloth using the Eigen CG solver instead of
a custom built solver.

The old cloth solver is broken unfortunately. Eigen is a designated
linear algebra library and very likely their implementation is a lot
better (can't compare until it's implemented though).

Only basic gravity is active atm, spring forces, external force fields,
damping and volumetric friction have to be added back by converting
the data into the Eigen format.
January 20, 2015, 08:29 (GMT)
Don't try to enforce a velocity change for pinned vertices; their
velocity is defined externally by hair root motion.
January 20, 2015, 08:29 (GMT)
Use squareroot of the current epsilon error target for CG, since we
square this value anyway.
January 20, 2015, 08:29 (GMT)
Disabled collision culling on the inside of the collider faces for now,
this seems to remove too many contact points somehow ...
Tehnyt: Miika HämäläinenViimeksi päivitetty: 07.11.2014 14:18MiikaH:n Sivut a.k.a. MiikaHweb | 2003-2021