Don't know... I have no plans for an IK solver. It's actually pretty tricky to write one that performs fast enough to be of use in a real- time environment. We're writing one for Bongo2.0, I can talk to those developers and see if the code is portable.
-- David Rutten Robert McNeel & Associates On Nov 19, 11:33 am, Evert <[EMAIL PROTECTED]> wrote: > Hi Taz and Andy, > > Thanks a lot for your input. > > Taz, I saw that Korean blog about a week ago and I think they did it > by formula and that's where I started figuring out arcs, circles, > relations in formulas relationships. > > Andy, > > I say your animation about a month ago. I really liked it. > > I second your question to David about IK implementation in > grasshopper, perhaps the maths people at Mcneel could have some input > as well. > > Evert Amador > > On Nov 18, 7:23 pm, Rchitekt <[EMAIL PROTECTED]> wrote: > > > This is an interesting topic. It seems like the most logical choice > > for something like this would be to use an inverse kinematic system... > > only this type of system doesn't exist in grasshopper yet. David, > > would this be something worth pursuing in future releases of the plug- > > in? I know the IK solvers are quite complex, and it would probably be > > very difficult to do, if you were just trying to figure out the math > > behind all of the rotations and translations. I did a project a while > > back where I tried to re-create one of Theo Jansens walking mechanisms > > in 3D. I was using 3D Max, but I originally started by trying to > > solve the math behind all of the joints... When that proved to be too > > much for my tiny brain, I decided that IK would work better, and was > > able to generate a pretty convincing 3D version of his sculpture. > > Here's a link to the > > animation:http://www.liftarchitects.com/journal/2008/4/15/animaris-rhinoceros-m... > > It would be really awesome if IK systems were enabled in grasshopper, > > but I don't know how hard that would be to program. > > -Andy > > > On Nov 18, 10:27 am, taz <[EMAIL PROTECTED]> wrote: > > > > Oops... wrong link. > > > >http://tinyurl.com/6jpebx > > > > taz > > > > On Nov 18, 1:22 pm, taz <[EMAIL PROTECTED]> wrote: > > > > > Evert, > > > > > I'm not sure about the procedure that was used (it looks to be mostly > > > > formula based...) but here's an example of a 4-bar linkage from a > > > > Korean blog that you might find interesting. I'll leave answering > > > > your actual questions up to other people. > > > > >http://tinyurl.com/5mbopk > > > > > taz > > > > > On Nov 18, 12:02 pm, Evert <[EMAIL PROTECTED]> wrote: > > > > > > Hi David and All, > > > > > > It's very interesting this topic more so because I've been trying to > > > > > explore this kind of geometries. > > > > > > My question is if there is a way to constraint the length of the bars > > > > > to mimic the movements of a real structure with rotating nodes. > > > > > > I would think you can do it with geometric formulas and or equations > > > > > to move and rotate everything at the same time based on the possible > > > > > trayectories of everything but that seems to complex! > > > > > > Does any body know how catia, parasolid, etc put this into operation? > > > > > > Thanks a lot in advance. > > > > > > Evert Amador > > > > > > On Nov 18, 9:28 am, David Rutten <[EMAIL PROTECTED]> wrote: > > > > > > > The Realflow plugin uses the same technique as Grasshopper at the > > > > > > moment. I draw my own gunk in a display conduit and save out the > > > > > > viewport images. > > > > > > If you're rendering with the mesh option, then I do create 'real' > > > > > > geometry. Typically this is a bad idea, since continually adding/ > > > > > > adjusting/replacing/deleting objects will flood the undo buffer. > > > > > > > -- > > > > > > David Rutten > > > > > > Robert McNeel & Associates > > > > > > > On Nov 17, 8:08 pm, visose <[EMAIL PROTECTED]> wrote: > > > > > > > > You mean exactly this animation or any particular animation? > > > > > > > To create an animation: right click in a slider component and > > > > > > > select > > > > > > > "animate". This will create a series of images by capturing the > > > > > > > perspective viewport. The first image will have the slider's > > > > > > > lowest > > > > > > > value, the last image will the highest and the ones in between the > > > > > > > corresponding in-between values. > > > > > > > After you've created the series of images (frames), use any video > > > > > > > editing software (i used 'virtualdub', its free) to create a video > > > > > > > from them. > > > > > > > Main limitations are that you can only use one slider per > > > > > > > animation > > > > > > > and you can't render the animation. I guess it wouldn't be too > > > > > > > complicated to have a render animation feature: It would have to > > > > > > > be > > > > > > > able to bake geometry, render, delete geometry, bake again, render > > > > > > > again, delete again, etc. > > > > > > > > Question for David: Doesn't the real flow plug in do something > > > > > > > like > > > > > > > this to render an animation? > > > > > > > > On Nov 17, 5:17 pm, Danny Dong <[EMAIL PROTECTED]> wrote: > > > > > > > > > Wow! > > > > > > > > > that is impressive, > > > > > > > > > but I have no clue how you animated it once you created the > > > > > > > > geometry > > > > > > > > in GH. > > > > > > > > > it would be super helpful if you could give a tutorial on how > > > > > > > > you > > > > > > > > created that animation > > > > > > > > > D > > > > > > > > > On Nov 15, 9:19 pm, visose <[EMAIL PROTECTED]> wrote: > > > > > > > > > > Well, I tried to do what I described above. Don't compare it > > > > > > > > > to the > > > > > > > > > video posted above, please :). This is much much crappier. The > > > > > > > > > rotation doesn't actually fit the movement. It shrinks > > > > > > > > > depending on > > > > > > > > > the slope of the terrain. No scripting > > > > > > > > > involved.http://grasshopper3d.googlegroups.com/web/anim_dodecahedron.avi > > > > > > > > > > On Nov 16, 12:43 am, visose <[EMAIL PROTECTED]> wrote: > > > > > > > > > > > That video is really cool. I don't think grasshopper is the > > > > > > > > > > ideal > > > > > > > > > > platform to achieve something like that. To make it roll > > > > > > > > > > over a flat > > > > > > > > > > surface would be easy, but to make it roll over a terrain > > > > > > > > > > you'll need > > > > > > > > > > some sort of collision detection..... I'm thinking having a > > > > > > > > > > linked > > > > > > > > > > move and rotation component (so it rotates while it moves), > > > > > > > > > > then it > > > > > > > > > > projects the closest point of the object to the terrain and > > > > > > > > > > gets its > > > > > > > > > > normal, and this normal is linked to the vector in the move > > > > > > > > > > component > > > > > > > > > > so it's always perpendicular to the terrain normal.... To > > > > > > > > > > make > > > > > > > > > > something half decent you'll probably need some scripting > > > > > > > > > > components. > > > > > > > > > > Of course to make it look real you'll need some sort of > > > > > > > > > > physics > > > > > > > > > > engine. > > > > > > > > > > To deform it as it rolls, I can't think of a way to get > > > > > > > > > > even close to > > > > > > > > > > what the video produces, but something really simple i > > > > > > > > > > guess would be > > > > > > > > > > to place a scale component that squishes the vertices (only > > > > > > > > > > vertices, > > > > > > > > > > not the whole geometry so the tubes are not squished, only > > > > > > > > > > shortened) > > > > > > > > > > in the z axis or perpendicular to the surface. > > > > > > > > > > Nevertheless, I'm sure there are more qualified people here > > > > > > > > > > that could > > > > > > > > > > help you. I'd be interested too to see what kind of complex > > > > > > > > > > motion can > > > > > > > > > > be achieve in grasshopper. > > > > > > > > > > > The dodecahedron > > > > > > > > > > definition:http://grasshopper3d.googlegroups.com/web/dodecahedron.ghx > > > > > > > > > > I added a couple of components that remove duplicated lines. > > > > > > > > > > > On Nov 15, 11:58 pm, Danny Dong <[EMAIL PROTECTED]> wrote: > > > > > > > > > > > > Amazing stuff! > > > > > > > > > > > > Both from Topmod3d and the models Visose made. > > > > > > > > > > > > Do you have the GHX file for those shapes? I would really > > > > > > > > > > > like to see > > > > > > > > > > > how you made them... > > > > > > > > > > > > hmm, for the manipulating individual points part, maybe I > > > > > > > > > > > can explain: > > > > > > > > > > > I want to to create a structure that will be able to > > > > > > > > > > > deform, be able > > > > > > > > > > > to walk perhaps. > > > > > > > > > > > > If I start out with a standard shape like a dodecahedron > > > > > > > > > > > that has > > > > > > > > > > > mathematically based patterns could i deform them based > > > > > > > > > > > on a series > > > > > > > > > > > limiting parameters like topography so that the shape > > > > > > > > > > > would deform as > > > > > > > > > > > it "rolled" over it or as wind blew on it from one side? > > > > > > > > > > > > I do want to create an animation eventually, to see how > > > > > > > > > > > a structure > > > > > > > > > > > reacts.... Could I say increase the triangulation on the > > > > > > > > > > > west side > > > > > > > > > > > only when needed and return it back to the original after > > > > > > > > > > > its not > > > > > > > > > > > needed. > > > > > > > > > > > > Sorry if this is unclear, But I have an idea in my head > > > > > > > > > > > and am not > > > > > > > > > > > sure which program(s) to execute them in. > > > > > > > > > > > maybe this youtube video will help > > > > > > > > > > > clarifyhttp://www.youtube.com/watch?v=txZMLS7YD6Q > > > > > > > > > > > > On Nov 15, 2:54 pm, visose <[EMAIL PROTECTED]> wrote: > > > > > > > > > > > > > and here's a dodecahedron using almost same method but > > > > > > > > > > > > getting the > > > > > > > > > > > > vertices using pointXYZ components instead of > > > > > > > > > > > > rectangles:http://grasshopper3d.googlegroups.com/web/dodecahedron.jpg > > > > > > > > > > > > > On Nov 15, 8:20 pm, visose <[EMAIL PROTECTED]> wrote: > > > > > > > > > > > > > > I don't know much about this stuff, but i did an > > > > > > > > > > > > > icosahedron using the > > > > > > > > > > > > > "cartesian coordinates" i found in its wikipedia > > > > > > > > > > > > > page. I doubt this is > > > > > > > > > > > > > a good method for doing different types of > > > > > > > > > > > > > Archimedean > > > > > > > > > > > > > solids:http://grasshopper3d.googlegroups.com/web/icosahedron.jpg > > > > > > > > > > > > > Btw, why do you want to manipulate individual points > > > > > > > > > > > > > in grasshopper? > > > > > > > > > > > > > Once you start manipulating indivual points, the > > > > > > > > > > > > > relationships that > > > > > > > > > > > > > build the structure are lost, so you might as well > > > > > > > > > > > > > bake it and do it > > > > > > > > > > > > > in rhino. > > > > > > > > > > > > > > On Nov 15, 2:01 am, Danny Dong <[EMAIL PROTECTED]> > > > > > > > > > > > > > wrote: > > > > > > > > > > > > > > > HI, > > > > > > > > > > > > > > > I've used > > ... > > read more »
