expert or not, you are exactly right about the difficulties that can be
encountered in building a production tool. however, let me clarify what i
the "rapid" and the "inexpensive" are what is at issue here. ideally,
between rapid tooling and production tooling, the gap for both of these
variables will be as large as possible. when you have to do lots of things
besides build an rp pattern, set it up in a casting box, and pour a tool
around it, the gap closes. the method that al proposes is certainly an
intelligent way of getting an accurate mold, but its' application will be
limited to certain niches where the cost of the tool and the criticality of
its features and application are already substantial. say, a large
enclosure part with a dozen screw boss locations and a shadow line all the
way around the joint, and where spending twice as much money and time as a
simpler rt approach would require is justified by the scope of the project.
don't get me wrong. i'm a huge fan of rapid tooling, when used correctly.
however, anytime you add more human factors and shrinkage factors and
economic factors, etc. etc., just because you can, then you are asking for
trouble. that was my only point.
Sent: Friday, April 16, 1999 2:02 PM
Subject: RE: Improved Accuracy Tools
I'm only a college professor, so I don't consider myself an expert at mold
making, but my understanding is that designing and building an injection
mold is time consuming, expensive, and fraught with opportunities for
This is true no matter what design methodology you use. The wonderful
advantage of rapid tooling is that it provides a rapid and inexpensive way
to give your design a try without investing in production tooling. Al's
technique just organizes the iteration process to optimize the final mold
If you can make it right and within tolerance the first time every time,
you don't need this process (and you can probably work anywhere you want
and name your salary!)
>from a product development standpoint, i have to say that this all seems
>theoretically sound, but also time consuming, expensive, and fraught with
>opportunities for error.
>i suppose that if you could do it all and still realize worthwhile time
>cost savings over traditional methods, then it would be viable.
>product development engineer
>Sent: Wednesday, April 14, 1999 1:50 PM
>To: MR ALBIN A HASTBACKA
>Subject: Re: Improved Accuracy Tools
>Why not go one step further?
>The mold was designed to make a plastic part, right? The mold design
>involved several assumptions about shrinkage, etc. (and these assumptions
>are only nominally correct.)
>Why not prototype,
>Build the tool,
>Measure the parts,
>Correct the CAD file,
>We did some work with this concept for soft foam tooling in a vaccuum
>>-- [ From: Al Hastbacka * EMC.Ver #2.5.3 ] --
>>When making injection molding tools, one of Sanders customers has found
>>a way to improve on the very high accuracy of the Sanders Model Maker II
>>(best accuracy of the commercially available US RP systems).
>>This is how it was done..
>>1. Fabricate the tool pattern on the ModelMaker II
>>2.Send it to NDM for fabrication of the tool
>>3. Measure the resulting tool
>>4. Resize the CAD file to compensate for the measured errors
>>5. Fabricate the second tool pattern on the ModelMaker II keeping all of
>>the configuration files as they were in step 1.
>>6. Send the second tool pattern to NDM for fabricating the tool in
>>7. Measure the second version of the tool
>>The measurements of the second tool were found to track within 1/2 to
>>one mil of the original CAD file.
>>For more information about the rp-ml, see http://ltk.hut.fi/rp-ml/
>Charles L. Thomas
>Department of Mechanical Engineering
>University of Utah
>50 South Central Campus Drive
>Salt Lake City, Utah 84112
>FAX (801) 585-9826
>For more information about the rp-ml, see http://ltk.hut.fi/rp-ml/
Charles L. Thomas
Department of Mechanical Engineering
University of Utah
50 South Central Campus Drive
Salt Lake City, Utah 84112
FAX (801) 585-9826
For more information about the rp-ml, see http://ltk.hut.fi/rp-ml/
This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:51:30 EEST