Subject:
|
Re: POV-Ray Chain Program
|
Newsgroups:
|
lugnet.cad.dev
|
Date:
|
Mon, 2 Aug 1999 10:06:24 GMT
|
Viewed:
|
972 times
|
| |
|
|
In lugnet.cad.dev, Jonathan Wilson writes:
Hi,
I'll check out the math and look for a more LDraw-like approach. I had already
some problems with the elastic parts, and now indeed need a chain for OMR 8844
(blue small helicopter). My first intention was/is to search for 3D-spline-
editors instead of bezier-curves. Your idea is a good new idea I'll look at
that too. When I've desided which way to go (3D-splines versus 2D polygon-
fitting and rotating (which is the basic idea of the page you mentioned)) I
think I'll start a discussion to get out a functional description of such a
program. (I've programmed a lot the last 5 years and also have some good
programming friends)
Bart Titulaer
>
>
> Adam Howard wrote:
>
> > Hi All!
> >
> > I was surfing the web the other day and discovered a wonderful plug-in for
> > POV-Ray that can create chains. Considering all the talk about hoses and a
> > desire for the chain piece in Ldraw I thought it would be great if this
> > programs ideas could be used to develop a similar program for ldraw. I sent
> > an email to the author asking what math he used and how he went about
> > programming it. He replied with the following email:
> >
> > "Hi - I forwarded this to me at work, because my internet connection was
> > down
> > at home...
> > anyhow...
> >
> > If you downloaded the .inc, and you know POV the math
> > is there for all to see;-)
> >
> > In short (and going from memory):
> >
> > I use the formula for a catenary (look for Andrews
> > treasure trove of mathematics if you can't find it
> > anywhere else, I don't have the URL handy).
> >
> > I curve fit the catenary formula given the two points,
> > mapping one to 0,0 and the other to x,y. (After the
> > chain is built I rotate it and translate it to the
> > proper location.
> >
> > because there are an infinite number of curves that
> > can pass through two points, I use a variable to set
> > the "length" (through the droop) of the curve.
> >
> > I initially calculated the slope to map each link, but this
> > really didn't work. In my final solution I determined
> > the endpoint of the each link by iteratively solving (binomial
> > search) the curve equation and comparing the distance between
> > this point and the current point using an error variable so the
> > loops didn't go too deep. Repeat for the next link.
> >
> > I initially planned to reverse this, so you could specify
> > the length of the chain (number of links) and it would
> > determine the missing value to scale the chain appropriately
> > but doing it iteratively caused POV to break.
> >
> > If the source isn't clear enough, drop me a line back.
> >
> > -Rob"
> >
> > You can find his program at http://www.geocities.com/SoHo/7865/chain.htm
> >
> > I am not a good computer programmer. So I just wanted to throw this idea
> > out and see if anyone else would like to work on a chain/hose/rope plug-in
> > for ldraw.
> >
> > Thanks,
> > Adam
>
> cool idea.
|
|
Message is in Reply To:
3 Messages in This Thread:
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
