Re: Concept Modeling

From: Erkut Negis (
Date: Tue Feb 06 1996 - 22:05:47 EET

On, Sun, 4 Feb. 1996 "Rolf Hubert" (Sanders Prototype, Inc.) wrote,

>... Are we to assume that engineers can no longer
>see their creation on their sofisticated workstations with their sofisticated

 You can even use 3D/EYE's Trispectives-Pro or Caligari's TrueSpace2 to
create photorealistic images with specified surface color, texture,
roughness etc. So if your purpose is visualization only, it is better to use
rendering or virtual reality software.
 However sometimes you may need some QUICK and rough physical models to feel
ergonomics or some primitive levels of functional testing. In that sense MJM
could be an answer because they are currently using more than one nozzle
(96) to build a part, so it might have some speed advantage over other
systems. I think that, a desktop Additive Automated Fabricator, priced less
than $10,000, that is capable of fabricating (let's say) a computer mouse in
less than 30 minutes could be feasible to buy.
 But I think there is nothing to do with BPM's machine. It is a very SLOW
and VERY ROUGH concept modeler... They have no chance to implement multiple
jets because in that case they sould put a 4-axis control mechanism for each

>The article states that Sanders has created a niche market, jewelry. If
>Sanders catered only to the jewelry industry, why is our machine a 6 inch
>cubed. Has anyone seen rings or braclets or whatever that was that large? If
>you are interested in building something that doesn't require sanding, try the
>Sanders machine. One of our many non jewelry customers sent us a STL file
>created by Delcam of a shower head. We produced it at .0015 inches or 0.037
>mm. The model was in 2 parts. They said "this was the first time this model
>was produced and actually fit together." The model required no post
>processing and was extremely smooth.

I'll appreciate if you post more information about your success stories? I'm
especcially interested in rapid tooling applications via investment casting,
Keltool process or Spray metal tooling...

>And of course, let's not forget Terry Wohler's statement that "raster is

If you have thousands of (reliable) jets with a spacing of less than 0.01mm
and if this assembly costs less than say $5,000, one can say that raster is
better than vector.
However, currently vector scan is much more accurate and practical.
Actually, as far as I know, Model Maker uses best combination of vector and
raster scan;
-> Vector scan for perimeters where surface quality and precision is important
-> Raster scan for filling inside of each crossection rapidly, where you can
even use grids to speed up the process...

 I want to add some comments on advantages of Sanders' 3DP system based on
classification and comparison of support structure implementation methods
used in commercialized (in USA) Additive Automated Fabricators;

 1- Support regions should be filled with a material that have DIFFERENT
phsical and/or chemical properties than build material. Only in this way you
have the chance of easily removing supports without damaging the part
(especcially for complex parts). There are five systems in this category:
DTM's SLS, Soligen's DSPC, Cubital's SGC, Stratasys' FDM and Sanders'
3D-Plotting. So, eliminate 3D Systems' STL, Helisys's LOM, 3D System's MJM
and BPM's Personal Modeler.

 2- To prevent unwanted deformations and/or dislocations during fabrication,
support regions must be filled with a SOLID material; in that way you can
produce accurate and precise parts. There are only three systems in this
category: Cubital's SGC, Stratasys' FDM and Sanders' 3DP. So, eliminate
DTM's SLS and Soligen's DSPC,.

 3- Support material must be easily (automatically) removable using CHEMICAL
and/or heat means. There are only two systems in this category: Cubital's
SGC and Sanders' 3DP. Currently Stratasys' FDM requires manual support
removal that is impossible to remove from diffucult to reach areas... So,
eliminate Stratasys' FDM

 4- Supports must be applied only when it is required; SELECTIVE support
generation saves material and makes easier to remove it. However in SGC
process every single space (other than build regions) is filled with support
wax. Only Sanders' 3DP system generates supports selectively. So I eliminate
Cubital's SGC.

 As a result I believe that Sanders' 3DP uses best combination of
technologies for support generation and removal; It is the only system that
SELECTIVELY constructs supports using a DIFFERENT SOLID material which is
easily removable by CHEMICAL and/or heat means.

 P.S. If you classify commercial Additive Automated Fabricators according to
the methods used to control z-axis accuracy there are only two systems which
uses milling; Cubital's SGC and Sanders' 3DP. However Sanders' Model Maker
is much more cheaper and smaller than Cubital's SGC process.


Erkut Negis
Engineering manager

New Silkroad Int'l, Inc. Phone : (617) 730-9512
75 Saint Alphonsus St. #301 Fax : (617) 734-3128
Boston, MA E-Mail:

This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:37:10 EEST