Subject:
|
LDraw-Mode vs. BezGen... problems?
|
Newsgroups:
|
lugnet.cad.dev
|
Date:
|
Sun, 9 Jan 2000 20:51:42 GMT
|
Viewed:
|
417 times
|
| |
|
|
after playing with BezGen recently, I've managed to get a complete hold on
using it and implementing hoses in my models...
but up until now i've only used the prefab-length hoses
(part #73590B http://lugnet.com/cad/ldraw/parts/ref/images/73590b.gif)
and this is created in BezGen (or the old DOS-based Bezier proggy) by using
a series of sphere primatives (8-8sphe.DAT), placed so close together by the
bezier routine that they overlap and create the illusion of the ribbed hose
shown above.
today i decided to start work using pneumatic hoses in a model and then
drafting it up in MLCad... but my tests using BezGen have shown a problem... a
problem that appears to have been addressed successfully in Fredric Glockner's
LDraw-Mode (http://www.math.uio.no/~fredrigl/technic/ldraw-mode/bezier-curve/)
when using segments that are symmetrical (in *all* 3D axes, such as a sphere)
there is no problem and you get output such as the examples Terry Keller showed
us a while ago
(http://www.ldraw.org/memorial/archive/bezier.html)
but if you change the segment object (8-8sphe.DAT) to something asymmetrical
such as Fredric's hose segment (hose-seg.DAT) you realize the limitation with
BezGen (and probably Richard Goldfinder's Bezier.exe, since BezGen credits him
for the bezier formulary)
in most 3D modeling software you have the ability to "loft" a shape along a
path... which is essentially what the bezier programs i'm speaking of do... the
path is a bezier curve that the program calculates and the shape is the .DAT
that is repeated along the path to create the hose (usually 8-8sphe.DAT)...
when you "loft" shapes, you usually have the option to "bank" the shape as it
travels along the path. this means that as the path curves and turns, the shape
is rotated in 3D space to keep 2 of its axes perpendicular to the path, thus
leaving the final shape with a uniform cross section down the entire length of
the loft...
LDraw-mode appears to implement this "banking" in its calculations, which is
why Fredric's hose looks correct in his examples.. but trying to use his hose
cross section in BezGen doesn't give the same results (at least not for me)
because the cross section is not rotated along the path, so as you curve your
hose, it "collapses" and becomes flat... sort of like writing with a caligraphy
pen, the lines are thicker or thinner depending on the direction of the stroke
in the letters...
i'd prefer to not load up emacs on my system just to create hoses... and i'm
very curious about where Fredric got the information on handling bezier
formulas, because his seems the best...
any feedback or other info guys??
J
|
|
Message has 2 Replies:
 | | Re: LDraw-Mode vs. BezGen... problems?
|
| (...) If you have MS Excel, John Van's hoser spreadsheet works wonderfully. It rotates sections as needed, and displays the length of the hose in one of the cells (I don't remember which--John?) By default, it uses the "official" hose sections, but (...) (25 years ago, 9-Jan-00, to lugnet.cad.dev)
| | | Re: LDraw-Mode vs. BezGen... problems?
|
| (...) I'm happy to hear that you found my program interesting, and I realize that LDraw-mode will not be a mainstream application by far (not yet anyway), as most people here don't seem to use Emacs. My program probably does fairly much the same as (...) (25 years ago, 10-Jan-00, to lugnet.cad.dev)
|
13 Messages in This Thread:
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
