Welcome to mirror list, hosted at ThFree Co, Russian Federation.

git.blender.org/blender.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorBrecht Van Lommel <brechtvanlommel@pandora.be>2008-11-13 00:16:53 +0300
committerBrecht Van Lommel <brechtvanlommel@pandora.be>2008-11-13 00:16:53 +0300
commitbdfe7d89e2f1292644577972c716931b4ce3c6c3 (patch)
treed00eb50b749cb001e2b08272c91791e66740b05d /source/blender/blenkernel/intern/collision.c
parent78a1c27c4a6abe0ed31ca93ad21910f3df04da56 (diff)
parent7e4db234cee71ead34ee81a12e27da4bd548eb4b (diff)
Merge of trunk into blender 2.5:
svn merge https://svn.blender.org/svnroot/bf-blender/trunk/blender -r12987:17416 Issues: * GHOST/X11 had conflicting changes. Some code was added in 2.5, which was later added in trunk also, but reverted partially, specifically revision 16683. I have left out this reversion in the 2.5 branch since I think it is needed there. http://projects.blender.org/plugins/scmsvn/viewcvs.php?view=rev&root=bf-blender&revision=16683 * Scons had various conflicting changes, I decided to go with trunk version for everything except priorities and some library renaming. * In creator.c, there were various fixes and fixes for fixes related to the -w -W and -p options. In 2.5 -w and -W is not coded yet, and -p is done differently. Since this is changed so much, and I don't think those fixes would be needed in 2.5, I've left them out. * Also in creator.c: there was code for a python bugfix where the screen was not initialized when running with -P. The code that initializes the screen there I had to disable, that can't work in 2.5 anymore but left it commented as a reminder. Further I had to disable some new function calls. using src/ and python/, as was done already in this branch, disabled function calls: * bpath.c: error reporting * BME_conversions.c: editmesh conversion functions. * SHD_dynamic: disabled almost completely, there is no python/. * KX_PythonInit.cpp and Ketsji/ build files: Mathutils is not there, disabled. * text.c: clipboard copy call. * object.c: OB_SUPPORT_MATERIAL. * DerivedMesh.c and subsurf_ccg, stipple_quarttone. Still to be done: * Go over files and functions that were moved to a different location but could still use changes that were done in trunk.
Diffstat (limited to 'source/blender/blenkernel/intern/collision.c')
-rw-r--r--source/blender/blenkernel/intern/collision.c1677
1 files changed, 1677 insertions, 0 deletions
diff --git a/source/blender/blenkernel/intern/collision.c b/source/blender/blenkernel/intern/collision.c
new file mode 100644
index 00000000000..9db3dda94eb
--- /dev/null
+++ b/source/blender/blenkernel/intern/collision.c
@@ -0,0 +1,1677 @@
+/* collision.c
+*
+*
+* ***** BEGIN GPL LICENSE BLOCK *****
+*
+* This program is free software; you can redistribute it and/or
+* modify it under the terms of the GNU General Public License
+* as published by the Free Software Foundation; either version 2
+* of the License, or (at your option) any later version.
+*
+* This program is distributed in the hope that it will be useful,
+* but WITHOUT ANY WARRANTY; without even the implied warranty of
+* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+* GNU General Public License for more details.
+*
+* You should have received a copy of the GNU General Public License
+* along with this program; if not, write to the Free Software Foundation,
+* Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+*
+* The Original Code is Copyright (C) Blender Foundation
+* All rights reserved.
+*
+* The Original Code is: all of this file.
+*
+* Contributor(s): none yet.
+*
+* ***** END GPL LICENSE BLOCK *****
+*/
+
+#include "MEM_guardedalloc.h"
+
+#include "BKE_cloth.h"
+
+#include "DNA_cloth_types.h"
+#include "DNA_group_types.h"
+#include "DNA_mesh_types.h"
+#include "DNA_object_types.h"
+#include "DNA_object_force.h"
+#include "DNA_scene_types.h"
+
+#include "BKE_DerivedMesh.h"
+#include "BKE_global.h"
+#include "BKE_mesh.h"
+#include "BKE_object.h"
+#include "BKE_modifier.h"
+#include "BKE_utildefines.h"
+#include "BKE_DerivedMesh.h"
+
+#include "Bullet-C-Api.h"
+
+#include "BLI_kdopbvh.h"
+#include "BKE_collision.h"
+
+
+/***********************************
+Collision modifier code start
+***********************************/
+
+/* step is limited from 0 (frame start position) to 1 (frame end position) */
+void collision_move_object ( CollisionModifierData *collmd, float step, float prevstep )
+{
+ float tv[3] = {0,0,0};
+ unsigned int i = 0;
+
+ for ( i = 0; i < collmd->numverts; i++ )
+ {
+ VECSUB ( tv, collmd->xnew[i].co, collmd->x[i].co );
+ VECADDS ( collmd->current_x[i].co, collmd->x[i].co, tv, prevstep );
+ VECADDS ( collmd->current_xnew[i].co, collmd->x[i].co, tv, step );
+ VECSUB ( collmd->current_v[i].co, collmd->current_xnew[i].co, collmd->current_x[i].co );
+ }
+ bvhtree_update_from_mvert ( collmd->bvhtree, collmd->mfaces, collmd->numfaces, collmd->current_x, collmd->current_xnew, collmd->numverts, 1 );
+}
+
+BVHTree *bvhtree_build_from_mvert ( MFace *mfaces, unsigned int numfaces, MVert *x, unsigned int numverts, float epsilon )
+{
+ BVHTree *tree;
+ float co[12];
+ int i;
+ MFace *tface = mfaces;
+
+ tree = BLI_bvhtree_new ( numfaces*2, epsilon, 4, 26 );
+
+ // fill tree
+ for ( i = 0; i < numfaces; i++, tface++ )
+ {
+ VECCOPY ( &co[0*3], x[tface->v1].co );
+ VECCOPY ( &co[1*3], x[tface->v2].co );
+ VECCOPY ( &co[2*3], x[tface->v3].co );
+ if ( tface->v4 )
+ VECCOPY ( &co[3*3], x[tface->v4].co );
+
+ BLI_bvhtree_insert ( tree, i, co, ( mfaces->v4 ? 4 : 3 ) );
+ }
+
+ // balance tree
+ BLI_bvhtree_balance ( tree );
+
+ return tree;
+}
+
+void bvhtree_update_from_mvert ( BVHTree * bvhtree, MFace *faces, int numfaces, MVert *x, MVert *xnew, int numverts, int moving )
+{
+ int i;
+ MFace *mfaces = faces;
+ float co[12], co_moving[12];
+ int ret = 0;
+
+ if ( !bvhtree )
+ return;
+
+ if ( x )
+ {
+ for ( i = 0; i < numfaces; i++, mfaces++ )
+ {
+ VECCOPY ( &co[0*3], x[mfaces->v1].co );
+ VECCOPY ( &co[1*3], x[mfaces->v2].co );
+ VECCOPY ( &co[2*3], x[mfaces->v3].co );
+ if ( mfaces->v4 )
+ VECCOPY ( &co[3*3], x[mfaces->v4].co );
+
+ // copy new locations into array
+ if ( moving && xnew )
+ {
+ // update moving positions
+ VECCOPY ( &co_moving[0*3], xnew[mfaces->v1].co );
+ VECCOPY ( &co_moving[1*3], xnew[mfaces->v2].co );
+ VECCOPY ( &co_moving[2*3], xnew[mfaces->v3].co );
+ if ( mfaces->v4 )
+ VECCOPY ( &co_moving[3*3], xnew[mfaces->v4].co );
+
+ ret = BLI_bvhtree_update_node ( bvhtree, i, co, co_moving, ( mfaces->v4 ? 4 : 3 ) );
+ }
+ else
+ {
+ ret = BLI_bvhtree_update_node ( bvhtree, i, co, NULL, ( mfaces->v4 ? 4 : 3 ) );
+ }
+
+ // check if tree is already full
+ if ( !ret )
+ break;
+ }
+
+ BLI_bvhtree_update_tree ( bvhtree );
+ }
+}
+
+/***********************************
+Collision modifier code end
+***********************************/
+
+/**
+* gsl_poly_solve_cubic -
+*
+* copied from SOLVE_CUBIC.C --> GSL
+*/
+
+#define mySWAP(a,b) do { double tmp = b ; b = a ; a = tmp ; } while(0)
+
+int
+gsl_poly_solve_cubic (double a, double b, double c,
+ double *x0, double *x1, double *x2)
+{
+ double q = (a * a - 3 * b);
+ double r = (2 * a * a * a - 9 * a * b + 27 * c);
+
+ double Q = q / 9;
+ double R = r / 54;
+
+ double Q3 = Q * Q * Q;
+ double R2 = R * R;
+
+ double CR2 = 729 * r * r;
+ double CQ3 = 2916 * q * q * q;
+
+ if (R == 0 && Q == 0)
+ {
+ *x0 = - a / 3 ;
+ *x1 = - a / 3 ;
+ *x2 = - a / 3 ;
+ return 3 ;
+ }
+ else if (CR2 == CQ3)
+ {
+ /* this test is actually R2 == Q3, written in a form suitable
+ for exact computation with integers */
+
+ /* Due to finite precision some double roots may be missed, and
+ considered to be a pair of complex roots z = x +/- epsilon i
+ close to the real axis. */
+
+ double sqrtQ = sqrt (Q);
+
+ if (R > 0)
+ {
+ *x0 = -2 * sqrtQ - a / 3;
+ *x1 = sqrtQ - a / 3;
+ *x2 = sqrtQ - a / 3;
+ }
+ else
+ {
+ *x0 = - sqrtQ - a / 3;
+ *x1 = - sqrtQ - a / 3;
+ *x2 = 2 * sqrtQ - a / 3;
+ }
+ return 3 ;
+ }
+ else if (CR2 < CQ3) /* equivalent to R2 < Q3 */
+ {
+ double sqrtQ = sqrt (Q);
+ double sqrtQ3 = sqrtQ * sqrtQ * sqrtQ;
+ double theta = acos (R / sqrtQ3);
+ double norm = -2 * sqrtQ;
+ *x0 = norm * cos (theta / 3) - a / 3;
+ *x1 = norm * cos ((theta + 2.0 * M_PI) / 3) - a / 3;
+ *x2 = norm * cos ((theta - 2.0 * M_PI) / 3) - a / 3;
+
+ /* Sort *x0, *x1, *x2 into increasing order */
+
+ if (*x0 > *x1)
+ mySWAP(*x0, *x1) ;
+
+ if (*x1 > *x2)
+ {
+ mySWAP(*x1, *x2) ;
+
+ if (*x0 > *x1)
+ mySWAP(*x0, *x1) ;
+ }
+
+ return 3;
+ }
+ else
+ {
+ double sgnR = (R >= 0 ? 1 : -1);
+ double A = -sgnR * pow (fabs (R) + sqrt (R2 - Q3), 1.0/3.0);
+ double B = Q / A ;
+ *x0 = A + B - a / 3;
+ return 1;
+ }
+}
+
+
+
+/**
+* gsl_poly_solve_quadratic
+*
+* copied from GSL
+*/
+int
+gsl_poly_solve_quadratic (double a, double b, double c,
+ double *x0, double *x1)
+{
+ double disc = b * b - 4 * a * c;
+
+ if (a == 0) /* Handle linear case */
+ {
+ if (b == 0)
+ {
+ return 0;
+ }
+ else
+ {
+ *x0 = -c / b;
+ return 1;
+ };
+ }
+
+ if (disc > 0)
+ {
+ if (b == 0)
+ {
+ double r = fabs (0.5 * sqrt (disc) / a);
+ *x0 = -r;
+ *x1 = r;
+ }
+ else
+ {
+ double sgnb = (b > 0 ? 1 : -1);
+ double temp = -0.5 * (b + sgnb * sqrt (disc));
+ double r1 = temp / a ;
+ double r2 = c / temp ;
+
+ if (r1 < r2)
+ {
+ *x0 = r1 ;
+ *x1 = r2 ;
+ }
+ else
+ {
+ *x0 = r2 ;
+ *x1 = r1 ;
+ }
+ }
+ return 2;
+ }
+ else if (disc == 0)
+ {
+ *x0 = -0.5 * b / a ;
+ *x1 = -0.5 * b / a ;
+ return 2 ;
+ }
+ else
+ {
+ return 0;
+ }
+}
+
+
+
+
+/*
+* See Bridson et al. "Robust Treatment of Collision, Contact and Friction for Cloth Animation"
+* page 4, left column
+*/
+int cloth_get_collision_time ( double a[3], double b[3], double c[3], double d[3], double e[3], double f[3], double solution[3] )
+{
+ int num_sols = 0;
+
+ // x^0 - checked
+ double g = a[0] * c[1] * e[2] - a[0] * c[2] * e[1] +
+ a[1] * c[2] * e[0] - a[1] * c[0] * e[2] +
+ a[2] * c[0] * e[1] - a[2] * c[1] * e[0];
+
+ // x^1
+ double h = -b[2] * c[1] * e[0] + b[1] * c[2] * e[0] - a[2] * d[1] * e[0] +
+ a[1] * d[2] * e[0] + b[2] * c[0] * e[1] - b[0] * c[2] * e[1] +
+ a[2] * d[0] * e[1] - a[0] * d[2] * e[1] - b[1] * c[0] * e[2] +
+ b[0] * c[1] * e[2] - a[1] * d[0] * e[2] + a[0] * d[1] * e[2] -
+ a[2] * c[1] * f[0] + a[1] * c[2] * f[0] + a[2] * c[0] * f[1] -
+ a[0] * c[2] * f[1] - a[1] * c[0] * f[2] + a[0] * c[1] * f[2];
+
+ // x^2
+ double i = -b[2] * d[1] * e[0] + b[1] * d[2] * e[0] +
+ b[2] * d[0] * e[1] - b[0] * d[2] * e[1] -
+ b[1] * d[0] * e[2] + b[0] * d[1] * e[2] -
+ b[2] * c[1] * f[0] + b[1] * c[2] * f[0] -
+ a[2] * d[1] * f[0] + a[1] * d[2] * f[0] +
+ b[2] * c[0] * f[1] - b[0] * c[2] * f[1] +
+ a[2] * d[0] * f[1] - a[0] * d[2] * f[1] -
+ b[1] * c[0] * f[2] + b[0] * c[1] * f[2] -
+ a[1] * d[0] * f[2] + a[0] * d[1] * f[2];
+
+ // x^3 - checked
+ double j = -b[2] * d[1] * f[0] + b[1] * d[2] * f[0] +
+ b[2] * d[0] * f[1] - b[0] * d[2] * f[1] -
+ b[1] * d[0] * f[2] + b[0] * d[1] * f[2];
+
+ /*
+ printf("r1: %lf\n", a[0] * c[1] * e[2] - a[0] * c[2] * e[1]);
+ printf("r2: %lf\n", a[1] * c[2] * e[0] - a[1] * c[0] * e[2]);
+ printf("r3: %lf\n", a[2] * c[0] * e[1] - a[2] * c[1] * e[0]);
+
+ printf("x1 x: %f, y: %f, z: %f\n", a[0], a[1], a[2]);
+ printf("x2 x: %f, y: %f, z: %f\n", c[0], c[1], c[2]);
+ printf("x3 x: %f, y: %f, z: %f\n", e[0], e[1], e[2]);
+
+ printf("v1 x: %f, y: %f, z: %f\n", b[0], b[1], b[2]);
+ printf("v2 x: %f, y: %f, z: %f\n", d[0], d[1], d[2]);
+ printf("v3 x: %f, y: %f, z: %f\n", f[0], f[1], f[2]);
+
+ printf("t^3: %lf, t^2: %lf, t^1: %lf, t^0: %lf\n", j, i, h, g);
+
+*/
+ // Solve cubic equation to determine times t1, t2, t3, when the collision will occur.
+ if ( ABS ( j ) > DBL_EPSILON )
+ {
+ i /= j;
+ h /= j;
+ g /= j;
+ num_sols = gsl_poly_solve_cubic ( i, h, g, &solution[0], &solution[1], &solution[2] );
+ }
+ else
+ {
+ num_sols = gsl_poly_solve_quadratic ( i, h, g, &solution[0], &solution[1] );
+ solution[2] = -1.0;
+ }
+
+ // printf("num_sols: %d, sol1: %lf, sol2: %lf, sol3: %lf\n", num_sols, solution[0], solution[1], solution[2]);
+
+ // Discard negative solutions
+ if ( ( num_sols >= 1 ) && ( solution[0] < DBL_EPSILON ) )
+ {
+ --num_sols;
+ solution[0] = solution[num_sols];
+ }
+ if ( ( num_sols >= 2 ) && ( solution[1] < DBL_EPSILON ) )
+ {
+ --num_sols;
+ solution[1] = solution[num_sols];
+ }
+ if ( ( num_sols == 3 ) && ( solution[2] < DBL_EPSILON ) )
+ {
+ --num_sols;
+ }
+
+ // Sort
+ if ( num_sols == 2 )
+ {
+ if ( solution[0] > solution[1] )
+ {
+ double tmp = solution[0];
+ solution[0] = solution[1];
+ solution[1] = tmp;
+ }
+ }
+ else if ( num_sols == 3 )
+ {
+
+ // Bubblesort
+ if ( solution[0] > solution[1] )
+ {
+ double tmp = solution[0]; solution[0] = solution[1]; solution[1] = tmp;
+ }
+ if ( solution[1] > solution[2] )
+ {
+ double tmp = solution[1]; solution[1] = solution[2]; solution[2] = tmp;
+ }
+ if ( solution[0] > solution[1] )
+ {
+ double tmp = solution[0]; solution[0] = solution[1]; solution[1] = tmp;
+ }
+ }
+
+ return num_sols;
+}
+
+
+// w3 is not perfect
+void collision_compute_barycentric ( float pv[3], float p1[3], float p2[3], float p3[3], float *w1, float *w2, float *w3 )
+{
+ double tempV1[3], tempV2[3], tempV4[3];
+ double a,b,c,d,e,f;
+
+ VECSUB ( tempV1, p1, p3 );
+ VECSUB ( tempV2, p2, p3 );
+ VECSUB ( tempV4, pv, p3 );
+
+ a = INPR ( tempV1, tempV1 );
+ b = INPR ( tempV1, tempV2 );
+ c = INPR ( tempV2, tempV2 );
+ e = INPR ( tempV1, tempV4 );
+ f = INPR ( tempV2, tempV4 );
+
+ d = ( a * c - b * b );
+
+ if ( ABS ( d ) < ALMOST_ZERO )
+ {
+ *w1 = *w2 = *w3 = 1.0 / 3.0;
+ return;
+ }
+
+ w1[0] = ( float ) ( ( e * c - b * f ) / d );
+
+ if ( w1[0] < 0 )
+ w1[0] = 0;
+
+ w2[0] = ( float ) ( ( f - b * ( double ) w1[0] ) / c );
+
+ if ( w2[0] < 0 )
+ w2[0] = 0;
+
+ w3[0] = 1.0f - w1[0] - w2[0];
+}
+
+DO_INLINE void collision_interpolateOnTriangle ( float to[3], float v1[3], float v2[3], float v3[3], double w1, double w2, double w3 )
+{
+ to[0] = to[1] = to[2] = 0;
+ VECADDMUL ( to, v1, w1 );
+ VECADDMUL ( to, v2, w2 );
+ VECADDMUL ( to, v3, w3 );
+}
+
+
+int cloth_collision_response_static ( ClothModifierData *clmd, CollisionModifierData *collmd, CollPair *collpair, CollPair *collision_end )
+{
+ int result = 0;
+ Cloth *cloth1;
+ float w1, w2, w3, u1, u2, u3;
+ float v1[3], v2[3], relativeVelocity[3];
+ float magrelVel;
+ float epsilon2 = BLI_bvhtree_getepsilon ( collmd->bvhtree );
+
+ cloth1 = clmd->clothObject;
+
+ for ( ; collpair != collision_end; collpair++ )
+ {
+ // only handle static collisions here
+ if ( collpair->flag & COLLISION_IN_FUTURE )
+ continue;
+
+ // compute barycentric coordinates for both collision points
+ collision_compute_barycentric ( collpair->pa,
+ cloth1->verts[collpair->ap1].txold,
+ cloth1->verts[collpair->ap2].txold,
+ cloth1->verts[collpair->ap3].txold,
+ &w1, &w2, &w3 );
+
+ // was: txold
+ collision_compute_barycentric ( collpair->pb,
+ collmd->current_x[collpair->bp1].co,
+ collmd->current_x[collpair->bp2].co,
+ collmd->current_x[collpair->bp3].co,
+ &u1, &u2, &u3 );
+
+ // Calculate relative "velocity".
+ collision_interpolateOnTriangle ( v1, cloth1->verts[collpair->ap1].tv, cloth1->verts[collpair->ap2].tv, cloth1->verts[collpair->ap3].tv, w1, w2, w3 );
+
+ collision_interpolateOnTriangle ( v2, collmd->current_v[collpair->bp1].co, collmd->current_v[collpair->bp2].co, collmd->current_v[collpair->bp3].co, u1, u2, u3 );
+
+ VECSUB ( relativeVelocity, v2, v1 );
+
+ // Calculate the normal component of the relative velocity (actually only the magnitude - the direction is stored in 'normal').
+ magrelVel = INPR ( relativeVelocity, collpair->normal );
+
+ // printf("magrelVel: %f\n", magrelVel);
+
+ // Calculate masses of points.
+ // TODO
+
+ // If v_n_mag < 0 the edges are approaching each other.
+ if ( magrelVel > ALMOST_ZERO )
+ {
+ // Calculate Impulse magnitude to stop all motion in normal direction.
+ float magtangent = 0, repulse = 0, d = 0;
+ double impulse = 0.0;
+ float vrel_t_pre[3];
+ float temp[3];
+
+ // calculate tangential velocity
+ VECCOPY ( temp, collpair->normal );
+ VecMulf ( temp, magrelVel );
+ VECSUB ( vrel_t_pre, relativeVelocity, temp );
+
+ // Decrease in magnitude of relative tangential velocity due to coulomb friction
+ // in original formula "magrelVel" should be the "change of relative velocity in normal direction"
+ magtangent = MIN2 ( clmd->coll_parms->friction * 0.01 * magrelVel,sqrt ( INPR ( vrel_t_pre,vrel_t_pre ) ) );
+
+ // Apply friction impulse.
+ if ( magtangent > ALMOST_ZERO )
+ {
+ Normalize ( vrel_t_pre );
+
+ impulse = magtangent / ( 1.0 + w1*w1 + w2*w2 + w3*w3 ); // 2.0 *
+ VECADDMUL ( cloth1->verts[collpair->ap1].impulse, vrel_t_pre, w1 * impulse );
+ VECADDMUL ( cloth1->verts[collpair->ap2].impulse, vrel_t_pre, w2 * impulse );
+ VECADDMUL ( cloth1->verts[collpair->ap3].impulse, vrel_t_pre, w3 * impulse );
+ }
+
+ // Apply velocity stopping impulse
+ // I_c = m * v_N / 2.0
+ // no 2.0 * magrelVel normally, but looks nicer DG
+ impulse = magrelVel / ( 1.0 + w1*w1 + w2*w2 + w3*w3 );
+
+ VECADDMUL ( cloth1->verts[collpair->ap1].impulse, collpair->normal, w1 * impulse );
+ cloth1->verts[collpair->ap1].impulse_count++;
+
+ VECADDMUL ( cloth1->verts[collpair->ap2].impulse, collpair->normal, w2 * impulse );
+ cloth1->verts[collpair->ap2].impulse_count++;
+
+ VECADDMUL ( cloth1->verts[collpair->ap3].impulse, collpair->normal, w3 * impulse );
+ cloth1->verts[collpair->ap3].impulse_count++;
+
+ // Apply repulse impulse if distance too short
+ // I_r = -min(dt*kd, m(0,1d/dt - v_n))
+ d = clmd->coll_parms->epsilon*8.0/9.0 + epsilon2*8.0/9.0 - collpair->distance;
+ if ( ( magrelVel < 0.1*d*clmd->sim_parms->stepsPerFrame ) && ( d > ALMOST_ZERO ) )
+ {
+ repulse = MIN2 ( d*1.0/clmd->sim_parms->stepsPerFrame, 0.1*d*clmd->sim_parms->stepsPerFrame - magrelVel );
+
+ // stay on the safe side and clamp repulse
+ if ( impulse > ALMOST_ZERO )
+ repulse = MIN2 ( repulse, 5.0*impulse );
+ repulse = MAX2 ( impulse, repulse );
+
+ impulse = repulse / ( 1.0 + w1*w1 + w2*w2 + w3*w3 ); // original 2.0 / 0.25
+ VECADDMUL ( cloth1->verts[collpair->ap1].impulse, collpair->normal, impulse );
+ VECADDMUL ( cloth1->verts[collpair->ap2].impulse, collpair->normal, impulse );
+ VECADDMUL ( cloth1->verts[collpair->ap3].impulse, collpair->normal, impulse );
+ }
+
+ result = 1;
+ }
+ }
+ return result;
+}
+
+//Determines collisions on overlap, collisions are writen to collpair[i] and collision+number_collision_found is returned
+CollPair* cloth_collision ( ModifierData *md1, ModifierData *md2, BVHTreeOverlap *overlap, CollPair *collpair )
+{
+ ClothModifierData *clmd = ( ClothModifierData * ) md1;
+ CollisionModifierData *collmd = ( CollisionModifierData * ) md2;
+ MFace *face1=NULL, *face2 = NULL;
+ ClothVertex *verts1 = clmd->clothObject->verts;
+ double distance = 0;
+ float epsilon1 = clmd->coll_parms->epsilon;
+ float epsilon2 = BLI_bvhtree_getepsilon ( collmd->bvhtree );
+ int i;
+
+ face1 = & ( clmd->clothObject->mfaces[overlap->indexA] );
+ face2 = & ( collmd->mfaces[overlap->indexB] );
+
+ // check all 4 possible collisions
+ for ( i = 0; i < 4; i++ )
+ {
+ if ( i == 0 )
+ {
+ // fill faceA
+ collpair->ap1 = face1->v1;
+ collpair->ap2 = face1->v2;
+ collpair->ap3 = face1->v3;
+
+ // fill faceB
+ collpair->bp1 = face2->v1;
+ collpair->bp2 = face2->v2;
+ collpair->bp3 = face2->v3;
+ }
+ else if ( i == 1 )
+ {
+ if ( face1->v4 )
+ {
+ // fill faceA
+ collpair->ap1 = face1->v1;
+ collpair->ap2 = face1->v4;
+ collpair->ap3 = face1->v3;
+
+ // fill faceB
+ collpair->bp1 = face2->v1;
+ collpair->bp2 = face2->v2;
+ collpair->bp3 = face2->v3;
+ }
+ else
+ i++;
+ }
+ if ( i == 2 )
+ {
+ if ( face2->v4 )
+ {
+ // fill faceA
+ collpair->ap1 = face1->v1;
+ collpair->ap2 = face1->v2;
+ collpair->ap3 = face1->v3;
+
+ // fill faceB
+ collpair->bp1 = face2->v1;
+ collpair->bp2 = face2->v4;
+ collpair->bp3 = face2->v3;
+ }
+ else
+ break;
+ }
+ else if ( i == 3 )
+ {
+ if ( face1->v4 && face2->v4 )
+ {
+ // fill faceA
+ collpair->ap1 = face1->v1;
+ collpair->ap2 = face1->v4;
+ collpair->ap3 = face1->v3;
+
+ // fill faceB
+ collpair->bp1 = face2->v1;
+ collpair->bp2 = face2->v4;
+ collpair->bp3 = face2->v3;
+ }
+ else
+ break;
+ }
+
+#ifdef WITH_BULLET
+ // calc distance + normal
+ distance = plNearestPoints (
+ verts1[collpair->ap1].txold, verts1[collpair->ap2].txold, verts1[collpair->ap3].txold, collmd->current_x[collpair->bp1].co, collmd->current_x[collpair->bp2].co, collmd->current_x[collpair->bp3].co, collpair->pa,collpair->pb,collpair->vector );
+#else
+ // just be sure that we don't add anything
+ distance = 2.0 * ( epsilon1 + epsilon2 + ALMOST_ZERO );
+#endif
+
+ if ( distance <= ( epsilon1 + epsilon2 + ALMOST_ZERO ) )
+ {
+ VECCOPY ( collpair->normal, collpair->vector );
+ Normalize ( collpair->normal );
+
+ collpair->distance = distance;
+ collpair->flag = 0;
+ collpair++;
+ }/*
+ else
+ {
+ float w1, w2, w3, u1, u2, u3;
+ float v1[3], v2[3], relativeVelocity[3];
+
+ // calc relative velocity
+
+ // compute barycentric coordinates for both collision points
+ collision_compute_barycentric ( collpair->pa,
+ verts1[collpair->ap1].txold,
+ verts1[collpair->ap2].txold,
+ verts1[collpair->ap3].txold,
+ &w1, &w2, &w3 );
+
+ // was: txold
+ collision_compute_barycentric ( collpair->pb,
+ collmd->current_x[collpair->bp1].co,
+ collmd->current_x[collpair->bp2].co,
+ collmd->current_x[collpair->bp3].co,
+ &u1, &u2, &u3 );
+
+ // Calculate relative "velocity".
+ collision_interpolateOnTriangle ( v1, verts1[collpair->ap1].tv, verts1[collpair->ap2].tv, verts1[collpair->ap3].tv, w1, w2, w3 );
+
+ collision_interpolateOnTriangle ( v2, collmd->current_v[collpair->bp1].co, collmd->current_v[collpair->bp2].co, collmd->current_v[collpair->bp3].co, u1, u2, u3 );
+
+ VECSUB ( relativeVelocity, v2, v1 );
+
+ if(sqrt(INPR(relativeVelocity, relativeVelocity)) >= distance)
+ {
+ // check for collision in the future
+ collpair->flag |= COLLISION_IN_FUTURE;
+ collpair++;
+ }
+ }*/
+ }
+ return collpair;
+}
+
+int cloth_collision_response_moving( ClothModifierData *clmd, CollisionModifierData *collmd, CollPair *collpair, CollPair *collision_end )
+{
+ int result = 0;
+ Cloth *cloth1;
+ float w1, w2, w3, u1, u2, u3;
+ float v1[3], v2[3], relativeVelocity[3];
+ float magrelVel;
+
+ cloth1 = clmd->clothObject;
+
+ for ( ; collpair != collision_end; collpair++ )
+ {
+ // compute barycentric coordinates for both collision points
+ collision_compute_barycentric ( collpair->pa,
+ cloth1->verts[collpair->ap1].txold,
+ cloth1->verts[collpair->ap2].txold,
+ cloth1->verts[collpair->ap3].txold,
+ &w1, &w2, &w3 );
+
+ // was: txold
+ collision_compute_barycentric ( collpair->pb,
+ collmd->current_x[collpair->bp1].co,
+ collmd->current_x[collpair->bp2].co,
+ collmd->current_x[collpair->bp3].co,
+ &u1, &u2, &u3 );
+
+ // Calculate relative "velocity".
+ collision_interpolateOnTriangle ( v1, cloth1->verts[collpair->ap1].tv, cloth1->verts[collpair->ap2].tv, cloth1->verts[collpair->ap3].tv, w1, w2, w3 );
+
+ collision_interpolateOnTriangle ( v2, collmd->current_v[collpair->bp1].co, collmd->current_v[collpair->bp2].co, collmd->current_v[collpair->bp3].co, u1, u2, u3 );
+
+ VECSUB ( relativeVelocity, v2, v1 );
+
+ // Calculate the normal component of the relative velocity (actually only the magnitude - the direction is stored in 'normal').
+ magrelVel = INPR ( relativeVelocity, collpair->normal );
+
+ // printf("magrelVel: %f\n", magrelVel);
+
+ // Calculate masses of points.
+ // TODO
+
+ // If v_n_mag < 0 the edges are approaching each other.
+ if ( magrelVel > ALMOST_ZERO )
+ {
+ // Calculate Impulse magnitude to stop all motion in normal direction.
+ float magtangent = 0;
+ double impulse = 0.0;
+ float vrel_t_pre[3];
+ float temp[3];
+
+ // calculate tangential velocity
+ VECCOPY ( temp, collpair->normal );
+ VecMulf ( temp, magrelVel );
+ VECSUB ( vrel_t_pre, relativeVelocity, temp );
+
+ // Decrease in magnitude of relative tangential velocity due to coulomb friction
+ // in original formula "magrelVel" should be the "change of relative velocity in normal direction"
+ magtangent = MIN2 ( clmd->coll_parms->friction * 0.01 * magrelVel,sqrt ( INPR ( vrel_t_pre,vrel_t_pre ) ) );
+
+ // Apply friction impulse.
+ if ( magtangent > ALMOST_ZERO )
+ {
+ Normalize ( vrel_t_pre );
+
+ impulse = 2.0 * magtangent / ( 1.0 + w1*w1 + w2*w2 + w3*w3 );
+ VECADDMUL ( cloth1->verts[collpair->ap1].impulse, vrel_t_pre, w1 * impulse );
+ VECADDMUL ( cloth1->verts[collpair->ap2].impulse, vrel_t_pre, w2 * impulse );
+ VECADDMUL ( cloth1->verts[collpair->ap3].impulse, vrel_t_pre, w3 * impulse );
+ }
+
+ // Apply velocity stopping impulse
+ // I_c = m * v_N / 2.0
+ // no 2.0 * magrelVel normally, but looks nicer DG
+ impulse = magrelVel / ( 1.0 + w1*w1 + w2*w2 + w3*w3 );
+
+ VECADDMUL ( cloth1->verts[collpair->ap1].impulse, collpair->normal, w1 * impulse );
+ cloth1->verts[collpair->ap1].impulse_count++;
+
+ VECADDMUL ( cloth1->verts[collpair->ap2].impulse, collpair->normal, w2 * impulse );
+ cloth1->verts[collpair->ap2].impulse_count++;
+
+ VECADDMUL ( cloth1->verts[collpair->ap3].impulse, collpair->normal, w3 * impulse );
+ cloth1->verts[collpair->ap3].impulse_count++;
+
+ // Apply repulse impulse if distance too short
+ // I_r = -min(dt*kd, m(0,1d/dt - v_n))
+ /*
+ d = clmd->coll_parms->epsilon*8.0/9.0 + epsilon2*8.0/9.0 - collpair->distance;
+ if ( ( magrelVel < 0.1*d*clmd->sim_parms->stepsPerFrame ) && ( d > ALMOST_ZERO ) )
+ {
+ repulse = MIN2 ( d*1.0/clmd->sim_parms->stepsPerFrame, 0.1*d*clmd->sim_parms->stepsPerFrame - magrelVel );
+
+ // stay on the safe side and clamp repulse
+ if ( impulse > ALMOST_ZERO )
+ repulse = MIN2 ( repulse, 5.0*impulse );
+ repulse = MAX2 ( impulse, repulse );
+
+ impulse = repulse / ( 1.0 + w1*w1 + w2*w2 + w3*w3 ); // original 2.0 / 0.25
+ VECADDMUL ( cloth1->verts[collpair->ap1].impulse, collpair->normal, impulse );
+ VECADDMUL ( cloth1->verts[collpair->ap2].impulse, collpair->normal, impulse );
+ VECADDMUL ( cloth1->verts[collpair->ap3].impulse, collpair->normal, impulse );
+ }
+ */
+ result = 1;
+ }
+ }
+ return result;
+}
+
+static float projectPointOntoLine(float *p, float *a, float *b)
+{
+ float ba[3], pa[3];
+ VECSUB(ba, b, a);
+ VECSUB(pa, p, a);
+ return INPR(pa, ba) / INPR(ba, ba);
+}
+
+static void calculateEENormal(float *np1, float *np2, float *np3, float *np4,float *out_normal)
+{
+ float line1[3], line2[3];
+ float length;
+
+ VECSUB(line1, np2, np1);
+ VECSUB(line2, np3, np1);
+
+ // printf("l1: %f, l1: %f, l2: %f, l2: %f\n", line1[0], line1[1], line2[0], line2[1]);
+
+ Crossf(out_normal, line1, line2);
+
+
+
+ length = Normalize(out_normal);
+ if (length <= FLT_EPSILON)
+ { // lines are collinear
+ VECSUB(out_normal, np2, np1);
+ Normalize(out_normal);
+ }
+}
+
+static void findClosestPointsEE(float *x1, float *x2, float *x3, float *x4, float *w1, float *w2)
+{
+ float temp[3], temp2[3];
+
+ double a, b, c, e, f;
+
+ VECSUB(temp, x2, x1);
+ a = INPR(temp, temp);
+
+ VECSUB(temp2, x4, x3);
+ b = -INPR(temp, temp2);
+
+ c = INPR(temp2, temp2);
+
+ VECSUB(temp2, x3, x1);
+ e = INPR(temp, temp2);
+
+ VECSUB(temp, x4, x3);
+ f = -INPR(temp, temp2);
+
+ *w1 = (e * c - b * f) / (a * c - b * b);
+ *w2 = (f - b * *w1) / c;
+
+}
+
+// calculates the distance of 2 edges
+float edgedge_distance(float np11[3], float np12[3], float np21[3], float np22[3], float *out_a1, float *out_a2, float *out_normal)
+{
+ float line1[3], line2[3], cross[3];
+ float length;
+ float temp[3], temp2[3];
+ float dist_a1, dist_a2;
+
+ VECSUB(line1, np12, np11);
+ VECSUB(line2, np22, np21);
+
+ Crossf(cross, line1, line2);
+ length = INPR(cross, cross);
+
+ if (length < FLT_EPSILON)
+ {
+ *out_a2 = projectPointOntoLine(np11, np21, np22);
+ if ((*out_a2 >= -FLT_EPSILON) && (*out_a2 <= 1.0 + FLT_EPSILON))
+ {
+ *out_a1 = 0;
+ calculateEENormal(np11, np12, np21, np22, out_normal);
+ VECSUB(temp, np22, np21);
+ VecMulf(temp, *out_a2);
+ VECADD(temp2, temp, np21);
+ VECADD(temp2, temp2, np11);
+ return INPR(temp2, temp2);
+ }
+
+ CLAMP(*out_a2, 0.0, 1.0);
+ if (*out_a2 > .5)
+ { // == 1.0
+ *out_a1 = projectPointOntoLine(np22, np11, np12);
+ if ((*out_a1 >= -FLT_EPSILON) && (*out_a1 <= 1.0 + FLT_EPSILON))
+ {
+ calculateEENormal(np11, np12, np21, np22, out_normal);
+
+ // return (np22 - (np11 + (np12 - np11) * out_a1)).lengthSquared();
+ VECSUB(temp, np12, np11);
+ VecMulf(temp, *out_a1);
+ VECADD(temp2, temp, np11);
+ VECSUB(temp2, np22, temp2);
+ return INPR(temp2, temp2);
+ }
+ }
+ else
+ { // == 0.0
+ *out_a1 = projectPointOntoLine(np21, np11, np12);
+ if ((*out_a1 >= -FLT_EPSILON) && (*out_a1 <= 1.0 + FLT_EPSILON))
+ {
+ calculateEENormal(np11, np11, np21, np22, out_normal);
+
+ // return (np21 - (np11 + (np12 - np11) * out_a1)).lengthSquared();
+ VECSUB(temp, np12, np11);
+ VecMulf(temp, *out_a1);
+ VECADD(temp2, temp, np11);
+ VECSUB(temp2, np21, temp2);
+ return INPR(temp2, temp2);
+ }
+ }
+
+ CLAMP(*out_a1, 0.0, 1.0);
+ calculateEENormal(np11, np12, np21, np22, out_normal);
+ if(*out_a1 > .5)
+ {
+ if(*out_a2 > .5)
+ {
+ VECSUB(temp, np12, np22);
+ }
+ else
+ {
+ VECSUB(temp, np12, np21);
+ }
+ }
+ else
+ {
+ if(*out_a2 > .5)
+ {
+ VECSUB(temp, np11, np22);
+ }
+ else
+ {
+ VECSUB(temp, np11, np21);
+ }
+ }
+
+ return INPR(temp, temp);
+ }
+ else
+ {
+
+ // If the lines aren't parallel (but coplanar) they have to intersect
+
+ findClosestPointsEE(np11, np12, np21, np22, out_a1, out_a2);
+
+ // If both points are on the finite edges, we're done.
+ if (*out_a1 >= 0.0 && *out_a1 <= 1.0 && *out_a2 >= 0.0 && *out_a2 <= 1.0)
+ {
+ float p1[3], p2[3];
+
+ // p1= np11 + (np12 - np11) * out_a1;
+ VECSUB(temp, np12, np11);
+ VecMulf(temp, *out_a1);
+ VECADD(p1, np11, temp);
+
+ // p2 = np21 + (np22 - np21) * out_a2;
+ VECSUB(temp, np22, np21);
+ VecMulf(temp, *out_a2);
+ VECADD(p2, np21, temp);
+
+ calculateEENormal(np11, np12, np21, np22, out_normal);
+ VECSUB(temp, p1, p2);
+ return INPR(temp, temp);
+ }
+
+
+ /*
+ * Clamp both points to the finite edges.
+ * The one that moves most during clamping is one part of the solution.
+ */
+ dist_a1 = *out_a1;
+ CLAMP(dist_a1, 0.0, 1.0);
+ dist_a2 = *out_a2;
+ CLAMP(dist_a2, 0.0, 1.0);
+
+ // Now project the "most clamped" point on the other line.
+ if (dist_a1 > dist_a2)
+ {
+ /* keep out_a1 */
+ float p1[3];
+
+ // p1 = np11 + (np12 - np11) * out_a1;
+ VECSUB(temp, np12, np11);
+ VecMulf(temp, *out_a1);
+ VECADD(p1, np11, temp);
+
+ *out_a2 = projectPointOntoLine(p1, np21, np22);
+ CLAMP(*out_a2, 0.0, 1.0);
+
+ calculateEENormal(np11, np12, np21, np22, out_normal);
+
+ // return (p1 - (np21 + (np22 - np21) * out_a2)).lengthSquared();
+ VECSUB(temp, np22, np21);
+ VecMulf(temp, *out_a2);
+ VECADD(temp, temp, np21);
+ VECSUB(temp, p1, temp);
+ return INPR(temp, temp);
+ }
+ else
+ {
+ /* keep out_a2 */
+ float p2[3];
+
+ // p2 = np21 + (np22 - np21) * out_a2;
+ VECSUB(temp, np22, np21);
+ VecMulf(temp, *out_a2);
+ VECADD(p2, np21, temp);
+
+ *out_a1 = projectPointOntoLine(p2, np11, np12);
+ CLAMP(*out_a1, 0.0, 1.0);
+
+ calculateEENormal(np11, np12, np21, np22, out_normal);
+
+ // return ((np11 + (np12 - np11) * out_a1) - p2).lengthSquared();
+ VECSUB(temp, np12, np11);
+ VecMulf(temp, *out_a1);
+ VECADD(temp, temp, np11);
+ VECSUB(temp, temp, p2);
+ return INPR(temp, temp);
+ }
+ }
+
+ printf("Error in edgedge_distance: end of function\n");
+ return 0;
+}
+
+int cloth_collision_moving_edges ( ClothModifierData *clmd, CollisionModifierData *collmd, CollPair *collpair )
+{
+ EdgeCollPair edgecollpair;
+ Cloth *cloth1=NULL;
+ ClothVertex *verts1=NULL;
+ unsigned int i = 0, k = 0;
+ int numsolutions = 0;
+ double x1[3], v1[3], x2[3], v2[3], x3[3], v3[3];
+ double solution[3], solution2[3];
+ MVert *verts2 = collmd->current_x; // old x
+ MVert *velocity2 = collmd->current_v; // velocity
+ float distance = 0;
+ float triA[3][3], triB[3][3];
+ int result = 0;
+
+ cloth1 = clmd->clothObject;
+ verts1 = cloth1->verts;
+
+ for(i = 0; i < 9; i++)
+ {
+ // 9 edge - edge possibilities
+
+ if(i == 0) // cloth edge: 1-2; coll edge: 1-2
+ {
+ edgecollpair.p11 = collpair->ap1;
+ edgecollpair.p12 = collpair->ap2;
+
+ edgecollpair.p21 = collpair->bp1;
+ edgecollpair.p22 = collpair->bp2;
+ }
+ else if(i == 1) // cloth edge: 1-2; coll edge: 2-3
+ {
+ edgecollpair.p11 = collpair->ap1;
+ edgecollpair.p12 = collpair->ap2;
+
+ edgecollpair.p21 = collpair->bp2;
+ edgecollpair.p22 = collpair->bp3;
+ }
+ else if(i == 2) // cloth edge: 1-2; coll edge: 1-3
+ {
+ edgecollpair.p11 = collpair->ap1;
+ edgecollpair.p12 = collpair->ap2;
+
+ edgecollpair.p21 = collpair->bp1;
+ edgecollpair.p22 = collpair->bp3;
+ }
+ else if(i == 3) // cloth edge: 2-3; coll edge: 1-2
+ {
+ edgecollpair.p11 = collpair->ap2;
+ edgecollpair.p12 = collpair->ap3;
+
+ edgecollpair.p21 = collpair->bp1;
+ edgecollpair.p22 = collpair->bp2;
+ }
+ else if(i == 4) // cloth edge: 2-3; coll edge: 2-3
+ {
+ edgecollpair.p11 = collpair->ap2;
+ edgecollpair.p12 = collpair->ap3;
+
+ edgecollpair.p21 = collpair->bp2;
+ edgecollpair.p22 = collpair->bp3;
+ }
+ else if(i == 5) // cloth edge: 2-3; coll edge: 1-3
+ {
+ edgecollpair.p11 = collpair->ap2;
+ edgecollpair.p12 = collpair->ap3;
+
+ edgecollpair.p21 = collpair->bp1;
+ edgecollpair.p22 = collpair->bp3;
+ }
+ else if(i ==6) // cloth edge: 1-3; coll edge: 1-2
+ {
+ edgecollpair.p11 = collpair->ap1;
+ edgecollpair.p12 = collpair->ap3;
+
+ edgecollpair.p21 = collpair->bp1;
+ edgecollpair.p22 = collpair->bp2;
+ }
+ else if(i ==7) // cloth edge: 1-3; coll edge: 2-3
+ {
+ edgecollpair.p11 = collpair->ap1;
+ edgecollpair.p12 = collpair->ap3;
+
+ edgecollpair.p21 = collpair->bp2;
+ edgecollpair.p22 = collpair->bp3;
+ }
+ else if(i == 8) // cloth edge: 1-3; coll edge: 1-3
+ {
+ edgecollpair.p11 = collpair->ap1;
+ edgecollpair.p12 = collpair->ap3;
+
+ edgecollpair.p21 = collpair->bp1;
+ edgecollpair.p22 = collpair->bp3;
+ }
+ /*
+ if((edgecollpair.p11 == 3) && (edgecollpair.p12 == 16))
+ printf("Ahier!\n");
+ if((edgecollpair.p11 == 16) && (edgecollpair.p12 == 3))
+ printf("Ahier!\n");
+ */
+
+ // if ( !cloth_are_edges_adjacent ( clmd, collmd, &edgecollpair ) )
+ {
+ // always put coll points in p21/p22
+ VECSUB ( x1, verts1[edgecollpair.p12].txold, verts1[edgecollpair.p11].txold );
+ VECSUB ( v1, verts1[edgecollpair.p12].tv, verts1[edgecollpair.p11].tv );
+
+ VECSUB ( x2, verts2[edgecollpair.p21].co, verts1[edgecollpair.p11].txold );
+ VECSUB ( v2, velocity2[edgecollpair.p21].co, verts1[edgecollpair.p11].tv );
+
+ VECSUB ( x3, verts2[edgecollpair.p22].co, verts1[edgecollpair.p11].txold );
+ VECSUB ( v3, velocity2[edgecollpair.p22].co, verts1[edgecollpair.p11].tv );
+
+ numsolutions = cloth_get_collision_time ( x1, v1, x2, v2, x3, v3, solution );
+
+ if((edgecollpair.p11 == 3 && edgecollpair.p12==16)|| (edgecollpair.p11==16 && edgecollpair.p12==3))
+ {
+ if(edgecollpair.p21==6 || edgecollpair.p22 == 6)
+ {
+ printf("dist: %f, sol[k]: %lf, sol2[k]: %lf\n", distance, solution[k], solution2[k]);
+ printf("a1: %f, a2: %f, b1: %f, b2: %f\n", x1[0], x2[0], x3[0], v1[0]);
+ printf("b21: %d, b22: %d\n", edgecollpair.p21, edgecollpair.p22);
+ }
+ }
+
+ for ( k = 0; k < numsolutions; k++ )
+ {
+ // printf("sol %d: %lf\n", k, solution[k]);
+ if ( ( solution[k] >= ALMOST_ZERO ) && ( solution[k] <= 1.0 ) && ( solution[k] > ALMOST_ZERO))
+ {
+ float a,b;
+ float out_normal[3];
+ float distance;
+ float impulse = 0;
+ float I_mag;
+
+ // move verts
+ VECADDS(triA[0], verts1[edgecollpair.p11].txold, verts1[edgecollpair.p11].tv, solution[k]);
+ VECADDS(triA[1], verts1[edgecollpair.p12].txold, verts1[edgecollpair.p12].tv, solution[k]);
+
+ VECADDS(triB[0], collmd->current_x[edgecollpair.p21].co, collmd->current_v[edgecollpair.p21].co, solution[k]);
+ VECADDS(triB[1], collmd->current_x[edgecollpair.p22].co, collmd->current_v[edgecollpair.p22].co, solution[k]);
+
+ // TODO: check for collisions
+ distance = edgedge_distance(triA[0], triA[1], triB[0], triB[1], &a, &b, out_normal);
+
+ if ((distance <= clmd->coll_parms->epsilon + BLI_bvhtree_getepsilon ( collmd->bvhtree ) + ALMOST_ZERO) && (INPR(out_normal, out_normal) > 0))
+ {
+ float vrel_1_to_2[3], temp[3], temp2[3], out_normalVelocity;
+ float desiredVn;
+
+ VECCOPY(vrel_1_to_2, verts1[edgecollpair.p11].tv);
+ VecMulf(vrel_1_to_2, 1.0 - a);
+ VECCOPY(temp, verts1[edgecollpair.p12].tv);
+ VecMulf(temp, a);
+
+ VECADD(vrel_1_to_2, vrel_1_to_2, temp);
+
+ VECCOPY(temp, verts1[edgecollpair.p21].tv);
+ VecMulf(temp, 1.0 - b);
+ VECCOPY(temp2, verts1[edgecollpair.p22].tv);
+ VecMulf(temp2, b);
+ VECADD(temp, temp, temp2);
+
+ VECSUB(vrel_1_to_2, vrel_1_to_2, temp);
+
+ out_normalVelocity = INPR(vrel_1_to_2, out_normal);
+/*
+ // this correction results in wrong normals sometimes?
+ if(out_normalVelocity < 0.0)
+ {
+ out_normalVelocity*= -1.0;
+ VecMulf(out_normal, -1.0);
+ }
+*/
+ /* Inelastic repulsion impulse. */
+
+ // Calculate which normal velocity we need.
+ desiredVn = (out_normalVelocity * (float)solution[k] - (.1 * (clmd->coll_parms->epsilon + BLI_bvhtree_getepsilon ( collmd->bvhtree )) - sqrt(distance)) - ALMOST_ZERO);
+
+ // Now calculate what impulse we need to reach that velocity.
+ I_mag = (out_normalVelocity - desiredVn) / 2.0; // / (1/m1 + 1/m2);
+
+ // Finally apply that impulse.
+ impulse = (2.0 * -I_mag) / (a*a + (1.0-a)*(1.0-a) + b*b + (1.0-b)*(1.0-b));
+
+ VECADDMUL ( verts1[edgecollpair.p11].impulse, out_normal, (1.0-a) * impulse );
+ verts1[edgecollpair.p11].impulse_count++;
+
+ VECADDMUL ( verts1[edgecollpair.p12].impulse, out_normal, a * impulse );
+ verts1[edgecollpair.p12].impulse_count++;
+
+ // return true;
+ result = 1;
+ break;
+ }
+ else
+ {
+ // missing from collision.hpp
+ }
+ // mintime = MIN2(mintime, (float)solution[k]);
+
+ break;
+ }
+ }
+ }
+ }
+ return result;
+}
+
+int cloth_collision_moving ( ClothModifierData *clmd, CollisionModifierData *collmd, CollPair *collpair, CollPair *collision_end )
+{
+ Cloth *cloth1;
+ cloth1 = clmd->clothObject;
+
+ for ( ; collpair != collision_end; collpair++ )
+ {
+ // only handle moving collisions here
+ if (!( collpair->flag & COLLISION_IN_FUTURE ))
+ continue;
+
+ cloth_collision_moving_edges ( clmd, collmd, collpair);
+ // cloth_collision_moving_tris ( clmd, collmd, collpair);
+ }
+
+ return 1;
+}
+
+
+// return all collision objects in scene
+// collision object will exclude self
+CollisionModifierData **get_collisionobjects(Object *self, int *numcollobj)
+{
+ Base *base=NULL;
+ CollisionModifierData **objs = NULL;
+ Object *coll_ob = NULL;
+ CollisionModifierData *collmd = NULL;
+ int numobj = 0, maxobj = 100;
+
+ objs = MEM_callocN(sizeof(CollisionModifierData *)*maxobj, "CollisionObjectsArray");
+ // check all collision objects
+ for ( base = G.scene->base.first; base; base = base->next )
+ {
+ /*Only proceed for mesh object in same layer */
+ if(!(base->object->type==OB_MESH && (base->lay & self->lay)))
+ continue;
+
+ coll_ob = base->object;
+
+ if(coll_ob == self)
+ continue;
+
+ if(coll_ob->pd && coll_ob->pd->deflect)
+ {
+ collmd = ( CollisionModifierData * ) modifiers_findByType ( coll_ob, eModifierType_Collision );
+ }
+ else
+ collmd = NULL;
+
+ if ( collmd )
+ {
+ if(numobj >= maxobj)
+ {
+ // realloc
+ int oldmax = maxobj;
+ CollisionModifierData **tmp;
+ maxobj *= 2;
+ tmp = MEM_callocN(sizeof(CollisionModifierData *)*maxobj, "CollisionObjectsArray");
+ memcpy(tmp, objs, sizeof(CollisionModifierData *)*oldmax);
+ MEM_freeN(objs);
+ objs = tmp;
+
+ }
+
+ objs[numobj] = collmd;
+ numobj++;
+ }
+ else
+ {
+ if ( coll_ob->dup_group )
+ {
+ GroupObject *go;
+ Group *group = coll_ob->dup_group;
+
+ for ( go= group->gobject.first; go; go= go->next )
+ {
+ coll_ob = go->ob;
+ collmd = NULL;
+
+ if(coll_ob == self)
+ continue;
+
+ if(coll_ob->pd && coll_ob->pd->deflect)
+ {
+ collmd = ( CollisionModifierData * ) modifiers_findByType ( coll_ob, eModifierType_Collision );
+ }
+ else
+ collmd = NULL;
+
+ if ( !collmd )
+ continue;
+
+ if( !collmd->bvhtree)
+ continue;
+
+ if(numobj >= maxobj)
+ {
+ // realloc
+ int oldmax = maxobj;
+ CollisionModifierData **tmp;
+ maxobj *= 2;
+ tmp = MEM_callocN(sizeof(CollisionModifierData *)*maxobj, "CollisionObjectsArray");
+ memcpy(tmp, objs, sizeof(CollisionModifierData *)*oldmax);
+ MEM_freeN(objs);
+ objs = tmp;
+ }
+
+ objs[numobj] = collmd;
+ numobj++;
+ }
+ }
+ }
+ }
+ *numcollobj = numobj;
+ return objs;
+}
+
+void cloth_bvh_objcollisions_nearcheck ( ClothModifierData * clmd, CollisionModifierData *collmd, CollPair **collisions, CollPair **collisions_index, int numresult, BVHTreeOverlap *overlap)
+{
+ int i;
+
+ *collisions = ( CollPair* ) MEM_mallocN ( sizeof ( CollPair ) * numresult * 4, "collision array" ); //*4 since cloth_collision_static can return more than 1 collision
+ *collisions_index = *collisions;
+
+ for ( i = 0; i < numresult; i++ )
+ {
+ *collisions_index = cloth_collision ( ( ModifierData * ) clmd, ( ModifierData * ) collmd, overlap+i, *collisions_index );
+ }
+}
+
+int cloth_bvh_objcollisions_resolve ( ClothModifierData * clmd, CollisionModifierData *collmd, CollPair *collisions, CollPair *collisions_index)
+{
+ Cloth *cloth = clmd->clothObject;
+ int i=0, j = 0, numfaces = 0, numverts = 0;
+ ClothVertex *verts = NULL;
+ int ret = 0;
+ int result = 0;
+ float tnull[3] = {0,0,0};
+
+ numfaces = clmd->clothObject->numfaces;
+ numverts = clmd->clothObject->numverts;
+
+ verts = cloth->verts;
+
+ // process all collisions (calculate impulses, TODO: also repulses if distance too short)
+ result = 1;
+ for ( j = 0; j < 5; j++ ) // 5 is just a value that ensures convergence
+ {
+ result = 0;
+
+ if ( collmd->bvhtree )
+ {
+ result += cloth_collision_response_static ( clmd, collmd, collisions, collisions_index );
+
+ // apply impulses in parallel
+ if ( result )
+ {
+ for ( i = 0; i < numverts; i++ )
+ {
+ // calculate "velocities" (just xnew = xold + v; no dt in v)
+ if ( verts[i].impulse_count )
+ {
+ VECADDMUL ( verts[i].tv, verts[i].impulse, 1.0f / verts[i].impulse_count );
+ VECCOPY ( verts[i].impulse, tnull );
+ verts[i].impulse_count = 0;
+
+ ret++;
+ }
+ }
+ }
+ }
+ }
+ return ret;
+}
+
+// cloth - object collisions
+int cloth_bvh_objcollision ( Object *ob, ClothModifierData * clmd, float step, float dt )
+{
+ Cloth *cloth=NULL;
+ BVHTree *cloth_bvh=NULL;
+ int i=0, numfaces = 0, numverts = 0, k, l, j;
+ int rounds = 0; // result counts applied collisions; ic is for debug output;
+ ClothVertex *verts = NULL;
+ int ret = 0, ret2 = 0;
+ CollisionModifierData **collobjs = NULL;
+ int numcollobj = 0;
+
+ if ( ( clmd->sim_parms->flags & CLOTH_SIMSETTINGS_FLAG_COLLOBJ ) || ! ( ( ( Cloth * ) clmd->clothObject )->bvhtree ) )
+ {
+ return 0;
+ }
+
+ cloth = clmd->clothObject;
+ verts = cloth->verts;
+ cloth_bvh = ( BVHTree * ) cloth->bvhtree;
+ numfaces = clmd->clothObject->numfaces;
+ numverts = clmd->clothObject->numverts;
+
+ ////////////////////////////////////////////////////////////
+ // static collisions
+ ////////////////////////////////////////////////////////////
+
+ // update cloth bvh
+ bvhtree_update_from_cloth ( clmd, 1 ); // 0 means STATIC, 1 means MOVING (see later in this function)
+ bvhselftree_update_from_cloth ( clmd, 0 ); // 0 means STATIC, 1 means MOVING (see later in this function)
+
+ collobjs = get_collisionobjects(ob, &numcollobj);
+
+ if(!collobjs)
+ return 0;
+
+ do
+ {
+ CollPair **collisions, **collisions_index;
+
+ ret2 = 0;
+
+ collisions = MEM_callocN(sizeof(CollPair *) *numcollobj , "CollPair");
+ collisions_index = MEM_callocN(sizeof(CollPair *) *numcollobj , "CollPair");
+
+ // check all collision objects
+ for(i = 0; i < numcollobj; i++)
+ {
+ CollisionModifierData *collmd = collobjs[i];
+ BVHTreeOverlap *overlap = NULL;
+ int result = 0;
+
+ if(!collmd->bvhtree)
+ continue;
+
+ /* move object to position (step) in time */
+ collision_move_object ( collmd, step + dt, step );
+
+ /* search for overlapping collision pairs */
+ overlap = BLI_bvhtree_overlap ( cloth_bvh, collmd->bvhtree, &result );
+
+ // go to next object if no overlap is there
+ if(!result || !overlap)
+ {
+ if ( overlap )
+ MEM_freeN ( overlap );
+ continue;
+ }
+
+ /* check if collisions really happen (costly near check) */
+ cloth_bvh_objcollisions_nearcheck ( clmd, collmd, &collisions[i], &collisions_index[i], result, overlap);
+
+ // resolve nearby collisions
+ ret += cloth_bvh_objcollisions_resolve ( clmd, collmd, collisions[i], collisions_index[i]);
+ ret2 += ret;
+
+ if ( overlap )
+ MEM_freeN ( overlap );
+ }
+ rounds++;
+
+ for(i = 0; i < numcollobj; i++)
+ {
+ if ( collisions[i] ) MEM_freeN ( collisions[i] );
+ }
+
+ MEM_freeN(collisions);
+ MEM_freeN(collisions_index);
+
+ ////////////////////////////////////////////////////////////
+ // update positions
+ // this is needed for bvh_calc_DOP_hull_moving() [kdop.c]
+ ////////////////////////////////////////////////////////////
+
+ // verts come from clmd
+ for ( i = 0; i < numverts; i++ )
+ {
+ if ( clmd->sim_parms->flags & CLOTH_SIMSETTINGS_FLAG_GOAL )
+ {
+ if ( verts [i].flags & CLOTH_VERT_FLAG_PINNED )
+ {
+ continue;
+ }
+ }
+
+ VECADD ( verts[i].tx, verts[i].txold, verts[i].tv );
+ }
+ ////////////////////////////////////////////////////////////
+
+
+ ////////////////////////////////////////////////////////////
+ // Test on *simple* selfcollisions
+ ////////////////////////////////////////////////////////////
+ if ( clmd->coll_parms->flags & CLOTH_COLLSETTINGS_FLAG_SELF )
+ {
+ for(l = 0; l < clmd->coll_parms->self_loop_count; l++)
+ {
+ // TODO: add coll quality rounds again
+ BVHTreeOverlap *overlap = NULL;
+ int result = 0;
+
+ // collisions = 1;
+ verts = cloth->verts; // needed for openMP
+
+ numfaces = clmd->clothObject->numfaces;
+ numverts = clmd->clothObject->numverts;
+
+ verts = cloth->verts;
+
+ if ( cloth->bvhselftree )
+ {
+ // search for overlapping collision pairs
+ overlap = BLI_bvhtree_overlap ( cloth->bvhselftree, cloth->bvhselftree, &result );
+
+ // #pragma omp parallel for private(k, i, j) schedule(static)
+ for ( k = 0; k < result; k++ )
+ {
+ float temp[3];
+ float length = 0;
+ float mindistance;
+
+ i = overlap[k].indexA;
+ j = overlap[k].indexB;
+
+ mindistance = clmd->coll_parms->selfepsilon* ( cloth->verts[i].avg_spring_len + cloth->verts[j].avg_spring_len );
+
+ if ( clmd->sim_parms->flags & CLOTH_SIMSETTINGS_FLAG_GOAL )
+ {
+ if ( ( cloth->verts [i].flags & CLOTH_VERT_FLAG_PINNED )
+ && ( cloth->verts [j].flags & CLOTH_VERT_FLAG_PINNED ) )
+ {
+ continue;
+ }
+ }
+
+ VECSUB ( temp, verts[i].tx, verts[j].tx );
+
+ if ( ( ABS ( temp[0] ) > mindistance ) || ( ABS ( temp[1] ) > mindistance ) || ( ABS ( temp[2] ) > mindistance ) ) continue;
+
+ // check for adjacent points (i must be smaller j)
+ if ( BLI_edgehash_haskey ( cloth->edgehash, MIN2(i, j), MAX2(i, j) ) )
+ {
+ continue;
+ }
+
+ length = Normalize ( temp );
+
+ if ( length < mindistance )
+ {
+ float correction = mindistance - length;
+
+ if ( cloth->verts [i].flags & CLOTH_VERT_FLAG_PINNED )
+ {
+ VecMulf ( temp, -correction );
+ VECADD ( verts[j].tx, verts[j].tx, temp );
+ }
+ else if ( cloth->verts [j].flags & CLOTH_VERT_FLAG_PINNED )
+ {
+ VecMulf ( temp, correction );
+ VECADD ( verts[i].tx, verts[i].tx, temp );
+ }
+ else
+ {
+ VecMulf ( temp, -correction*0.5 );
+ VECADD ( verts[j].tx, verts[j].tx, temp );
+
+ VECSUB ( verts[i].tx, verts[i].tx, temp );
+ }
+ ret = 1;
+ ret2 += ret;
+ }
+ else
+ {
+ // check for approximated time collisions
+ }
+ }
+
+ if ( overlap )
+ MEM_freeN ( overlap );
+
+ }
+ }
+ ////////////////////////////////////////////////////////////
+
+ ////////////////////////////////////////////////////////////
+ // SELFCOLLISIONS: update velocities
+ ////////////////////////////////////////////////////////////
+ if ( ret2 )
+ {
+ for ( i = 0; i < cloth->numverts; i++ )
+ {
+ if ( ! ( verts [i].flags & CLOTH_VERT_FLAG_PINNED ) )
+ {
+ VECSUB ( verts[i].tv, verts[i].tx, verts[i].txold );
+ }
+ }
+ }
+ ////////////////////////////////////////////////////////////
+ }
+ }
+ while ( ret2 && ( clmd->coll_parms->loop_count>rounds ) );
+
+ if(collobjs)
+ MEM_freeN(collobjs);
+
+ return MIN2 ( ret, 1 );
+}