Re: [prog. in c] how to make stable paricles

fir pisze:

see that simulation
 https://www.youtube.com/watch?v=pP-uCjLp3RY
 there is a group of lights/dots/balls/points (say atoms) and such fotces
are applied
1) if the ball is into wall apply a force proportional to depth of
being into wall
        point[i].vx -= H* wall_penetration_x ;
       point[i].vy -= H* wall_penetration_y ;
 2) if the ball is to much close to other ball same the force related to depth of penetration is applied (though here its square)
      point[i].vx+= (dx/dist)* (n_pen*B)*(n_pen*B) /point[i].mass ;
    point[i].vy+= (dy/dist)* (n_pen*B)*(n_pen*B) /point[i].mass ;
  i chose it partially conceptually - knowing a=F/m - and partially experimantally..it work wery well imo
 the problem is i want to apply some ball-ball force additional to
this repeling one that would make those ball build in a kinda multi-ball
relativly stable "particles" ... and the question is what this force should be?
 here above on this video the lines you see are drawed when this third force start to work.. (so teh lines arre kinda optical debug showing
how it work) i applied some but the particles that forms are much to dynamical, oscilate llike crazy then sometimas shot one of atoms out and so on
 and i need a stable one
 i experimented with this additional link force but not much
what i experimented was like adding link it is attraction force
that enebles when two points are closer than 1.5*link distance, which is 20 units (pixels) it is this attract force work on distance closer than
30 pixels
 the force is something like
   double link_dist = 20;
   double dx = point[i].x - point[n].x;
   double dy = point[i].y - point[n].y;
   double Mi = point[i].mass;
    double dist = sqrt(dx*dx + dy*dy );
   if(dist>link_dist*1.5) return;
    double L = 1.;
   double p =  (dist - link_dist)*L;
      point[i].vx+= -(dx/dist)*(p*p)*sign(p)/point[i].mass;
     point[i].vy+= -(dy/dist)*(p*p)*sign(p)/point[i].mass;
    if(!m_toggler)
    DrawLine2d(point[i].x, point[i].y, point[n].x, point[n].y,0xffffff );
 but as i said the results are to messy and i need stable little "solid bodies" i mean more "stiff" particles
 some ideas as to what this link force should be?

note as to "stiff" particles (not sure as to this word becouse weak english) they are possible as check some of my old demo
https://www.youtube.com/watch?v=0bw_R8nIIG4
this 9 dot "Rectangle" is made by simple 1-1 link forces but the difference is i set up those initial distances like 100 and sqrt (100) = 141.42 and so on by hand from init function in code
and this is very tiresome to set upo by hand so i think it would be give some force that works liek chemistry it is glues that balls together naturall way not by tedious set up
- but its the proof that it is generaly possible to have stiff particles
(more to say its making like rotating objests with angular momentum without any code of angular momentum just simple lineas bal-ball forces and newtonian a=F/m)