Re: Patent 5,506,607 (3-D Model Maker)

From: Erkut Negis (enegis@lynx.dac.neu.edu)
Date: Thu May 16 1996 - 05:12:10 EEST


Dear Andrew,

At 08:49 AM 5/13/96, you wrote:

>MR ALBIN A HASTBACKA wrote:
>
>> For those of you who have an interest in the Sanders Model Maker,
>> Patent # 5,506,607 was granted for the invention on April 9, 1996.
>
>We are interested in this and have had a part produced as a sample by the
>UK agents. However, the 3D systems machine, for a similar price seems to
>offer more, being 96 jets as opposed to two:
>
>Is it quicker?
>Is it any better?
...

Actually their target markets are different:
   3D Systems' Actua 2100 is targeted for concept modeling applications in
which throughput and ease of use are most important factors.
   Sanders' MM-6PRO is currently targeted (or most suitable) for fabrication
of small, complex shaped and precision investment casting patterns. Surface
finish and accuracy are most important factors for these applications.

I compared some of the important aspects of the two machines listed below. I
hope these will be helpful for you.

THROUGHPUT:
  Actua 2100 should be much more faster than MM-6PRO. Depending on surface
to volume ratio of the part, throughput can be 20 to 60 times faster than
MM-6PRO. (Actua 2100 is better for producing parts that have high ratio of
volume/surface_area)

However if Sanders commercializes the technology* explained in their newly
issued patent, their throughput can be increased by 10 to 100 times
(depending on part size and geometry). The number can be even higher
depending on number of jets that is used for void filling.

*: A third, dedicated larger orifice jet or array of jets are proposed for
quick raster filling of voids...

WORKING ENVELOPE:
Actua 2100: 10" x 8" x 8"(z)
MM-6PRO : 6" x 6" x 6"
Due to the fact that currently MM-6PRO has a very slow build rate it is not
very practical to build parts larger than 3" x 3" x 2".

MATERIAL STRENGTH:
  I felt that the wax used in Actua 2100 is very weak even for concept
modeling. Similarly Sanders' wax is not strong enough for safe handling
however it is nearly 2 or 3 times stronger than the build material used in
Actua 2100. Both companies should be working to introduce new and stronger
materials... However drop-on-demand injection of stronger, higher melting
point and viscous materials is not an easy task.

SURFACE FINISH:
  Actua 2100 has generally good surface finish except supported regions.
  MM-6PRO has the best surface finish in the Additive Automated Fabricator
market because it constructs models using very thin layers (up to ten times
thinner compared to other fabricators) and support material can be dissolved
easily without damaging supported surfaces.
Instead of raster scanning used in Actua 2100, MM-6PRO uses vector plotting
to construct peripheries of each layer. This also greatly helps in
production of smooth surface finish.

ACCURACY:
  Shrinkage of the build material upon solidification is biggest cause of
warpage and dimensional errors. In MM-6PRO, milling process after
construction of each layer helps in reducing internal stresses. Also leaving
a space between droplets in the first pass and filling between during second
pass reduces shrinkage buildup.
As far as I know Actua 2100 doesn't have any specific deposition procedure
or mechanism to reduce shrinkage buildup.
  Z axis accuracy in MM-6PRO is closely controlled by milling each layer.
Again, no specific procedure is used in Actua 2100.

MINIMUM FEATURE SIZE:
  Actua 2100 has slightly smaller droplet size (0.085mm) than that of
MM-6PRO (0.1mm).
However due to raster building operation of Actua 2100, you can not place a
droplet to anywhere that you want, so you may lost some small features. So
we can say that feature resolutions are similar in x-y plane.

RELIABILITY:
 Both machines have some level of jet clogging problems.
 After constructing each Layer, MM-6PRO checks the jets and if there is a
problem "erases" previous layer and reconstructs it.
In Actua 2100, no specific jet checking and layer repairing procedure has
been employed.

EASE OF USE:
  MM-6PRO requires a chemical solvent (non-toxic) and a separate
on-the-shelf ultrasonic cleaning device for support removal.
  Actua 2100 doesn't requires a special chemical or cleaning device however,
manual support removal from difficult to reach locations (without breaking
the part) might be a tedious task.
I expect that use of Actua 2100 in general, should be more straightforward
and easy in terms of software and hardware.

SIZE AND WEIGHT:
  Actua 2100: W 137x D 76 x H 112 cm and 318kg
 
  MM-6PRO : W 52 x D 44.5 x H 59.7cm and 32.3kg
  However you will also need a separate PC (controller cabinet) and a vacuum
support system unit which weights about 30kg.

=====
Another proposed method in Sanders' recently issued patent (5,506,607) is
interesting:
In the embodiment, the model shaving (milling) system is replaced by a
powder distribution hopper system that is movable in the same manner as the
shaving system for movement in the X coordinate direction. MC (Modeling
Compound) is selectively injected onto the powder, bonding predetermined
regions. The MC powder in this embodiment serves not only to increase the
rate of which models can be produced but also serves as a support medium for
cantilevered or overhanging portions of models.

It sounds very similar to the MIT's 3D Printing technology which was
patented long before and currently commercially available (for ceramic
powder) through Soligen's PartsNow service bureau.(I heard that a new
company licensed MIT's technology for use with polymer materials. It will be
interesting to watch the competition).

I think these methods can be very suitable for very fast concept modeling
with an easy support removal procedure...

Best Regards.

------------------------------------------------------------
Erkut Negis

New Silkroad Int'l, Inc. Phone : (617) 730-9512
75 Saint Alphonsus St. #301 Fax : (617) 734-3128
Boston, MA 02120 E-Mail: enegis@lynx.neu.edu
------------------------------------------------------------



This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:37:20 EEST