RE: Void Normal Direction

From: Blasch, Larry (LBlasch@OPW-FC.com)
Date: Fri Nov 14 2003 - 14:58:17 EET


Ian,

Yes I've built many closed volume parts in SLA. In fact, years ago, I
developed a technique for creating hollow parts of a specific wall thickness
from "non-hollow" STL files. It produces trapped volume parts consistently.

This technique was especially useful when prototyping rotationally molded
parts which are closed hollow parts.

You bring up a good point with the support generators though.

In the past (I've been using SLA for 13+ years) I had to create a solid and
a hollow version of the object and have the support generator create the
supports for the solid, then substitute the hollow version to build the
part. But the more recent support software lets me remove the interior
supports.

The are several other issues:
1) How do I get the liquid resin out of the part without damage?
2) When building thin wall trapped volume parts, the build cycle "must not"
finish by raising the object up out of the resin or the weight of the
trapped resin may rupture the part walls.

Now when I need to create a trapped volume part I just build it with
drain/vent holes and build the plugs along with the part. Most of the
problems are avoided that way.

Sincerely
Larry Blasch
Design Engineer
CAE Systems Administrator
     
OPW Fueling Components
P.O. Box 405003
Cincinnati, OH 45240-5003 USA
Voice: (513) 870-3356
Fax: (513) 870-3275

-----Original Message-----
From: Ian Gibson [mailto:igibson@hkucc.hku.hk]
Sent: Thursday, November 13, 2003 8:44 PM
To: rp-ml@rapid.lpt.fi
Subject: RE: Void Normal Direction

Larry

I was wondering whether you or anyone else can verify (or has already done
so) this through a simple experiment/test?

I'm not quite so convinced that the problem comes from the CAD. I was
thinking that maybe it could also be a bug in the support generation
software. Or indeed it could be deliberate to avoid weird postcure effects
when the part is removed and placed in the oven.

I don't have an SLA machine else I would try it myself.

IG

At 04:33 AM 11/14/2003, you wrote:
>Jonathan,
>
>Yes, you can build a cube of Swiss cheese with an SLA machine. The totally
>enclosed "Bubbles" within the block would build as a enclosed volume of
>liquid resin.
>
>I believe the problem you are experiencing stems from your modeling
software
>not the RP method.
>
>When you attempt to model a theoretical construct that contains only a
>single twisted surface, you are creating the problem. Single surface
objects
>don't exist in nature unless they are a solid shape like a sphere. Even a
>soap bubble has a thickness, and therefore has an inside and an outside
>surface.
>
>If the twisted surface that you are attempting to create wraps back onto
>itself, most solid modeling software, by default, will assume the trapped
>volume is a solid.
>
>When exporting the data for RP, the CAD modeler will follow the rules of
STL
>file generation and export the STL surface data as a set of triangles, each
>with a single normal facing in only one direction.
>
>See if you can create your surface with a thickness of some small amount.
>The modeler can then export the STL data as a pair of surfaces "back to
>back" and may let you build the object as you envision it.
>
>Sincerely,
>
>Larry Blasch
>Design Engineer
>CAE Systems Administrator
>
>OPW Fueling Components
>P.O. Box 405003
>Cincinnati, OH 45240-5003 USA
>Voice: (513) 870-3356
>Fax: (513) 870-3275
>
>
>-----Original Message-----
>From: Jonathan Chertok [mailto:chertok22@hotmail.com]
>Sent: Thursday, November 13, 2003 11:35 AM
>To: rp-ml@rapid.lpt.fi
>Subject: RE: Void Normal Direction
>
>
>List,
>
>After a little thought on the responses to this question I am wondering the
>following:
>
>Could you build essentially a cube of swiss cheese using SLA RP for
>instance?
>
>I understand that there is an issue with the fact that the voids should
have
>
>no draft type angles in the downward build direction that would trap
>material.
>
>But would you need to put tiny holes through all the voids in the "cheese"
>in order to get it to build?
>
>I'm curious about this,
>
>Jonathan
>
>
>
>Universal Joint: Buildings + Digital Design
>
>Jonathan Chertok. Principal
>Austin, Texas [512] 407 9628
>
>Full Service Design and Construction
>
>_________________________________________________________________
> >From Beethoven to the Rolling Stones, your favorite music is always
playing
>on MSN Radio Plus. No ads, no talk. Trial month FREE!
>http://join.msn.com/?page=offers/premiumradio

Dr. Ian Gibson

Currently on study leave at
National University of Singapore,
Dept. Mechanical Engineering
9 Engineering Drive 1
Singapore 117576
Tel: +65 6874 1917
Mob: +65 9087 3512

"Everything really is stupidly simple, and yet all around is utter
confusion,
don't look around to find the sound that's right beneath your feet"



This archive was generated by hypermail 2.1.7 : Sat Jan 17 2004 - 15:18:21 EET