Re: SLS Nylon Composite

From: David K. Leigh (
Date: Wed Feb 14 1996 - 11:10:18 EET

> Date: Wed, 14 Feb 1996 16:57:21 +0800
> From: (Ian Gibson)
> Subject: SLS Nylon Composite

> 1. Do you encounter problems with curling, caking, bonus-z, etc. when using NC?
Every one of those problems will be experienced unless you get your
powder consistent and the settings right. (also assuming you don't
have problems with your IR Sensor)

> 2. Which problems do you encounter the most?
For me, caking and bonus-z. I tend to overcompensate at first
because I hate curling. You cannot easily fix that.
> 3. Are you able to diagnose problems
> before build?
Not really. Unless you call warm-up before the build.
> during build?
Yes. Density of part can be verified by looking at the "clarity" or
lack of it during the melt. Curling can be detected because the part
will either raise out of the bed or the edges don't appear to have as
much powder. Caking is more difficult. In fine nylon, the material
will actually melt between parts, causing a bridge due to heat
> prior to the next build?
Examine the cake. Solid Rock - Lower Temp. Drastically. Tough break
-out - Lower the part bed a couple of degrees. Delamination of part,
increase laser. Very rough texture (mottled), the powder is in BAD
shape. Too much growth - decrease laser, may be bad powder.
> 4. If you are able to diagnose problems before build, is this because you
> use the same settings for each build or because you can identify possible
> changes in the default settings from the part geometry?

> 5. Do you adjust the part orientation to minimise the possible problems? If
> so, what features do you try and avoid?
Yes. Chalk it up to experience. Build on angles, minimize large
area cross sections, center the part. Overhangs can be a problem.
Small features need to be oriented in X-Y plane for strength.
> 6. If you are able to diagnose problems during build, are there visual clues
> that lead you to make changes?
Many! See notes above.
> 7. Do you find that you have to reduce part feed temperatures at periodical
> intervals during the build? If so, what scheme do you use (e.g. 1 degree per
> inch)?

No. Heat the feeds until they are streaking. Then, keep them there.

> 8. Do you find you have had to rebuild a part because of incorrect settings?
> Can you give an example? What do you commonly have to change for the next build?
Yes, but not too often. Shrink can be a problem - fix is to run the
scale parts and calculate shink every build. Part bed temp. can also
be a culprit.
> 9. Have you experienced any problems resulting from recycling material too
> many times or from powder contamination?

Yes. Yes. Yes. You need to keep a consistent mixture of virgin to
used. For example: 50% virgin. Always run with 50% virgin. If you
don't, the physical properties and settings will vary from build to
build, making your life miserable. It is also suggested that you use
a very fine mesh screen to sift with, eliminating alot of the bad

> 10. What percentage mix new/old powder do you use for NC?
DTM suggests 50%. But you will start to build up a stockpile of used
material. I wouldn't suggest going much less than that if you do not
have a good sifter. If you are able to sift with a 70 micron or
similar screen, you can go to as low as 10-20% virgin.

> 11. Are there any other problems that you have encountered that may be
> relevant to this questionnaire?
> 12. Is there any other information you feel may be useful (e.g. normal part
> type, application, etc.)
I'd suggest using the User's Group for this. That's what most of the
meeting is about.

David K. Leigh
Harvest Technologies
Rapid Prototyping Center
Phone (817) 742-1822

This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:37:12 EEST