Re: reconstruction of 3D grid with connectivity

*To*: mathgroup at smc.vnet.net*Subject*: [mg91418] Re: reconstruction of 3D grid with connectivity*From*: Narasimham <mathma18 at hotmail.com>*Date*: Thu, 21 Aug 2008 04:14:26 -0400 (EDT)*References*: <g8gj4j$etq$1@smc.vnet.net> <g8grao$ifa$1@smc.vnet.net>

On Aug 20, 3:23 pm, Jens-Peer Kuska <ku... at informatik.uni-leipzig.de> wrote: > Hi, > > a) your assumption "when u and v are each incremented by given ustep > and vstep" because ParametricPlot3D[] make a irregular triangle mesh > you can see the mesh with > > plt=ParametricPlot3D[{Cos[phi]*Sin[th], Sin[phi]*Sin[th], Cos[th]}, > {th, 0, Pi}, {phi, 0, 2 Pi}, Mesh -> All] > > b) the connectivity can be shown with > > GraphPlot[ > Union[Flatten[(Rule @@@ Partition[#, 2, 1, {-1}]) & /@ > Cases[plt, _Polygon, Infinity][[1, 1]] /. (a_ -> b_) /; > a > b :> (b -> a)] > ]] > > Regards > Jens > > Narasimham wrote: > > For surface ParametricPlot3D[{x = f(u,v), y = g(u,v), z = h(u,v)}, > > {u,umin,umax,ustep},{v,vmin,vmax,vstep}] > > > how to obtain the connectivity matrix (when u and v are each > > incremented by given ustep and vstep), using Delaunay or Voronoi > > triangulations? In this case there would be curved or skewed > > quadrilaterals instead of triangles that discretizes the surface.When > > connectivity matrix and coordinate matrix are given with each point > > ID reference number, the surface should be reconstructed, i.e., > > plotted, and/or Shown without again giving out the above command. > > > Thanks in advance, > > > Narasimham If p1,p2 etc are labels, ( p1= {x1,y1,z1} ; p2 = (x2,y2,z2} ), connectivity p1 -> p2 etc is required. Narasimham