Subject:
|
Re: Nice Viper
|
Newsgroups:
|
lugnet.cad
|
Date:
|
Mon, 13 Dec 2004 17:57:04 GMT
|
Viewed:
|
899 times
|
| |
|
|
In lugnet.cad, Tim David wrote:
> I was browsing Brickshelf when I saw this
> http://www.brickshelf.com/cgi-bin/gallery.cgi?f=101824
> Not only is it a nice model and nice renders the thing that struck me most was
> how much work must have gone into placing and rotating the parts into their
> non-standard positions
>
> Tim
Hi all,
This is my first post on lugnet, so don't blame me if anything goes wrong :-)
First of all, thank you for the complements on my cadwork, I plan on selling
this model in the future, so I don't publish the actual file yet. Therefore, all
I can do is tell some things about it. Since I'm a bit of a precision freak, I
never just place parts under an angle, and move them into place, I use loads of
reference parts to have other parts turn around and most of the angles are
carefully calculated. If that is too much of a job, then I tune them to 0.1 or
0.01 degrees precise. Now this is not all that hard to do, it's the hoses that
were killing me. In the past, I have used Lsynth to do the ribbed hoses, Lsynth
works fine for most bedible objects but not for hoses, basicly because of two
things: the parts don't have realistic sizes (the ribbed segments are too thin
due to integer Ldaw units they were created in) and Lsynth places way too many
of them per inch. Therefore, I had to addapt the hoseparts (make them half a
unit broader) and manually place them. That was hell. As soon as I have a decent
website I'll write down elaborate tips and tricks on how to do everything from
angled universal joint connections to these hoses, but I think it will not hurt
to illustrate how I did the last part. My starting point is always the first
link and the last link of the bended part of the hose. I write down the matrix
rotation vectors of one of the parts, isolate both in a subfile (exact same
location), derotate both around the written down vectors, with that part as a
pivot. Now that one part has a normal angle, I start with extending the hose on
that side, with just as many segments as the real one. This can already be
tricky given the half and quarter units one has to work with, especially when
the referencepart has unorthodox values. A calculater wouldn't be overkill here.
Now comes the hard part of the story: Take the second upto the last part of the
straight (in that order) and turn the whole thing about some degrees (depending
on how the hose is bended) around axle you want the hose to be bended around.
After this, take the third segment upto the last one and bend them around the
third, then the fourth upto the last one, around the fouth etc. To get a smooth
curve will take some practise, and it can be very fustrating when it doesn't fit
right when you reach the other starting part with your hose. You'll have to do
some more bending operations to get the hose to be smooth then. The next thing
to do is to return all parts around the first reference part with the written
down matrix rotation vectors. Then insert the submodel into your basemodel (on
0,0,0 ofcourse, unangled) and see if you need more addaptions. Usually, hoses
are symetric, thus the exact opposite is demanded on he other side of the model,
the mirrorhose. It's madness to write down all your steps and recreate them on
the other side (for you want them to be the exact same), but there is a easy
trick to not have to do this. Part by part, copy them, mirror their coordinates
(usually just put a minus in front of the x-coordinate) and then use the matrix
rotation vectors to mirror their angles. The easiest way to find out how to
mirror is to place two parts and turn them mirrorred around the same angles, and
see how the matrices behave (which of the nine values stay the same and which
inverse). When you know this, delete the experimental parts, and apply the
matrix mirror factor (-1) to all parts that you mirrored. This will give you a
perfect mirror hose.
As you might have noticed, this is quite a job, and thus takes very long to get
right. I hope better synthesizers will be available soon.
The Viper cad is not finished yet (as you can see) the real model is though, the
only thing I need is some white hoses and the right wheels (I use silverchamp
wheels now but I'm dying for 8448 wheels). Then I'll buy all the parts that it
takes to build a copy, which I'll build cadding along two new files, that will
become the instructions (one with perfect positioning for every step, and one
big file for Lpub to write BOM's).
That's pretty much it for my first post here...
Misha
|
|
Message has 1 Reply:
 | | Re: Nice Viper
|
| Misha, As an AFOL, CAD teacher, Mopar nut, and Le Mans fan, all I can say is HOLY CRAP that's a great MOC! Have you thought about advertising (once you're ready to sell) in Viper magazine? When you can afford a $70,000 sportscar you get your own (...) (20 years ago, 13-Dec-04, to lugnet.cad)
|
Message is in Reply To:
| | Nice Viper
|
| I was browsing Brickshelf when I saw this (URL) only is it a nice model and nice renders the thing that struck me most was how much work must have gone into placing and rotating the parts into their non-standard positions Tim (20 years ago, 11-Dec-04, to lugnet.cad)
|
4 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
