Re: SLS Somos problem

From: Chuck Alexander (
Date: Thu May 14 1998 - 19:13:05 EEST

At 09:42 AM 5/14/98 -0400, you wrote:
>We have been having difficulty running the DuPont Somos 201 in our
>Sinterstations. The problem seems to appear on mostly tubular parts in
>which the edge leading into the roller pulls away from the part bed.
>Eventually the part lifts and the build needs to be aborted. DTM has
>told us to increase the layer thickness to .008", lower the bed temps,
>lower the laser power and slow down the roller but it hasn't resolved
>our problems. .................

>Drew Brown

Sounds like what you're running into is what I call the "purse-string
effect". Imagine a bag with a draw-string to close the open end. As you pull
the string, the bag closes. Also, as you pull the string, the open end of
the bag puckers. This happens because the open end of the bag has a fixed
diameter (tubular bag) which the string is pulling to a smaller diameter.

Same thing is happening with your part. The bag is your tubular part. The
string is the Somos 201 shrinking. The fixed diameter of your STL file is
being pulled smaller by shrinkage.

If shrinkage is uniform, no problem, but shinkage is not uniform. The O.D.
of your tubular cross-section shrinks more than the I.D. Also, the top of
the layer shrinks more than the bottom. This is why the layers of your part
curl up (toward the laser) and into the center of the tube.

On an SLA we address this problem by spacing laser scan lines to have
minimal contact with each other on a first pass, allowing the layer to
shrink partially. Then a second scan, again with minimal contact. Any left
over uncured material is taken care of in post-cure. This is the theory
behind the WEAVE drawstyles developed by Joe Allison and others at 3D
Systems. Hopefully the Sinterstation will allow you to modify scanning
parameters similarly.

The bottom line is - shrink is the enemy -. It is also why with any material
being developed for RP equipment, it is imperative that the shrinkage of the
material be included in the material specifications.

I trust something in the previous rambling helps,

Chuck Alexander

Solid Concepts Inc.
Rapid Prototyping Software & Services
ph. 805/257-9300
fax 805/257-9311

For more information about the rp-ml, see

This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:45:37 EEST