RE: SLA *&$#@& Trapped volumes

From: Stahlhut, Todd A. (stahlht@po2.nawc-ad-indy.navy.mil)
Date: Wed Jan 17 1996 - 13:38:00 EET


Todd Grimm wrote:

> My interest is specifically in building traps in 5180. But, for the
benefit
> of the all users, responses with respect to any resin would be welcome.

> We have tried many solutions:
> - Build no traps
> - Holes and plugs
> - Multiple sweeps
> - Lengthen Z wait

Todd,

I too share your love for trapped volumes on the SLA. I am using the epoxy
resins
rather than the 5081 resin - but strangely enough the epoxies are about as
bad
even though they are thinner,

For those that don't know what a trapped volume is on a SLA - imagine
building a
coffee cup upside down. This is not a trapped volume and causes no problems

except that you have a bunch of supports up on the inside holding the bottom
up
(which is inverted at the top of the build).

Ideally, you would turn the cup over and build the bottom at the 'bottom'.
 This
makes supporting much easier since all you need is some from the platform to

the cup bottom. However .... The cup is now a trapped volume and holds
resin.
The resin inside the cup is trapped in the part geometry and cannot flow as
the
layers change. It also forms a meniscus at the top. This interferes with
leveling
since resin inside the trapped volume tends to flow under the wiper blade
and up
behind it - and not leave the trapped volume. If left uncorrected - the
resin
buildup on the edge cures and continues to get higher than it should be.
Eventually, the blade catches this raised edge and at best makes a poor part
 -
and at worst rips the platform off the supports destroys the build (and
possibly
breaks your system)

Of all the choices listed above - we try just not building them the most.
 This
usually means mores supports - but getting a good part is worth it. If this
won't
 work - we try a combination of all of the rest of the choices: Model a
hole in the
bottom of the trapped volume - and a plug to fill it, sweep multiple times,
and
lengthen the wait time between layers. We try to add at least a .5 hole, go
3
sweeps and 45 -60 sec z wait. Most of the time this will work but results in
a
longer build time.

We had a large 10 x 10 x 5 chassis fail yesterday which had a 1 " diameter
hole
in the bottom. It started hitting about 2.5 inches up - and failed about .5
inch
further. Today, we are cutting the box in half on the cad system, flipping
either
side 90 degrees - and building - then gluing them back together. We often
put
little alignment tabs on the part to help in the gluing back together. This
often saves time, effort, and aggravation.

Regards, Todd Stahlhut



This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:37:06 EEST