From: Alex Do (do@oxy.edu)
Date: Wed Sep 21 2005 - 08:07:38 EEST
Thank you for the detailed response.
I am considering the merits of direct SL tooling as another tool in
evaluating part-and-process. I don't have much experience with mold
analysis software so I'm not sure if this process is complementary or
superseded by such software. I also see that from RA Harris' work that SL
tooling is not valid for materials with a propensity for
crystallinity/lamellae formation because of differences in heat transfer
between SL and metal molds. But for highly amorphous polymers it appears to
be a valid low-cost prototyping process to examine DFM issues which arise
from part geometry and or gate location.
I am building a hand-mold set which will cost hardly over $1,000 for basic
testing with ABS plastic (my parts are fairly small). I'll be able to
modify the design and swap out cavity or core inserts at a relatively low
cost ($400). I'll let everyone know how it goes.
~alex
-----Original Message-----
From: owner-rp-ml@rapid.lpt.fi [mailto:owner-rp-ml@rapid.lpt.fi] On Behalf
Of E.Mueller@gom.com
Sent: Thursday, September 15, 2005 1:34 PM
To: do@oxy.edu
Cc: owner-rp-ml@rapid.lpt.fi; rp-ml@rapid.lpt.fi
Subject: Re: [rp-ml] Question for Moldflow (and similar) users about famous
molding case
Hello Alex,
here my late 5 cents.
Simulation is well established for critical tooling work. e.g. for metal
stamping and plastic injection production processes. The simulation is based
on CAD data, typical material properties and process properties. An
experienced user is also required to get good results.
As the material and the process properties are complex and vary with the
batch, time, temperature, you name it, rework of the tooling in pilot
production is still very common.
Scanning can define the reworked tooling geometry and the deviation of the
actual product from CAD data.
If a simulation software can "import" the captured data and "compare" these
data with the CAD data and the "first guess" tooling data, the simulation
software should be able to come up with an "educated" guess how to rework
the tool to get good results. In addition the feedback information from the
scanner can be used to "adjust" the material and process parameters for
succeeding simulations.
Most simulation software providers are still busy developing their
mathematical models and nice user interfaces. Simple tests are made to
extract material and process parameters to enhance the model calculations.
Some providers would not like to be offended by the deviation between their
simulation result and the real product form.
Power users will have to push the simulation software developers to make use
of the available feedback data to get more precise simulation results and
more target oriented rework. But even in the future, using all available
resources, pilot production and first article inspection will still be
necessary.
With best regards (from the scanner side)
Ernst Mueller www.gom.com
GOM International AG e.mueller@gom.com
Bremgarterstr. 89b Tel. ++41 56 631 04 04
8967 Widen Fax:++41 56 631 04 07
CH - Switzerland
"Alex Do"
<do@oxy.edu>
Sent by: To
owner-rp-ml@rapid <rp-ml@rapid.lpt.fi>
.lpt.fi cc
Subject
12.09.2005 07:40 [rp-ml] Question for Moldflow (and
similar) users about famous molding
case
Please respond to
<do@oxy.edu>
There is a famous case about the Oral-B CrossAction toothbrush that was
published in Mechanical Engineering 2001 ... basically after all the fancy
design when they went to mold the part, it unexpectedly warped 0.2" after
cooling. To solve the problem they re-designed the part with a counter warp
so that when it cooled it took its intended shape.
I've been unable to test out mold analysis software (can't get a license),
but I wonder would it have solved or quantitatively predicted that problem?
It seems like they still would have needed to see how much it warped by to
make an educated guess about how to redesign the part.
Thanks for the feedback ... just gathering data for a paper on SL injection
molding I'm writing.
~alex
Alexander Do
Research Assistant
Ford Prototyping Studio
University of California, Berkeley http://kingkong.me.berkeley.edu 2117
Etcheverry Hall Berkeley, CA 94720-1750
(510) 643 9486
This archive was generated by hypermail 2.1.7 : Mon Jan 02 2006 - 08:09:21 EET