Last week, Bob Morton used this forum to comment on a presentation that I made at the Modern Mold and Tooling Conference in Chicago, IL on October 6, 1999. While I have to admit that my daughter has on more than one occasion referred to me as a "grumpy old man", I must take exception to the other comments made by Mr. Morton. I think Mr. Morton has misinterpreted the point of the presentation entirely.
For those of you who did not hear the presentation, I gave a talk entitled "How Accurate is Pattern Based Tooling?" that was based on work I first presented at the SME/RPA conference in Dearborn, MI in April1998. In that paper I developed an expression that predicts tooling accuracy as a function of pattern build accuracy, pattern finishing accuracy, tooling material shrink, and plastic shrink. I then used experimentally determined values of pattern and finishing accuracy and published values of shrink to draw conclusions about the accuracy capabilities of current pattern based rapid tooling processes. I would be happy to send a copy of the presentation and the paper it was based on to anyone who requests it. The presentation I gave last week was essentially the same as one I gave last June at the ANTEC conference in NY.
My work was an extension of work done by Dr. Paul Jacobs that he first presented at the Gwaltney Manufacturing Symposium at Georgia Tech in October 1997. Dr. Jacobs has presented updates to his work several times since, including the SME/RPA conference in 1998. I suspect that well over 2000 people have heard one or both of our presentations. This is the first time, however, that I have heard it referred to as a "sick vendetta".
Mr. Morton questioned the accuracy and validity of my methods, apparently without ever having read the paper. In addition, he paraphrased my conclusions and stated that they are "very flawed and biased". For the sake of accuracy, the conclusions in my presentation were, word for word:
1. Rapid prototyping processes in general are not accurate enough to create patterns within generally accepted tolerances for production tooling
- Dimensional errors in build process are outside tolerance limits
- Finishing adds significant error
2. Combination of dimensional errors in the pattern with those from the tooling build process generally result in tolerances unacceptable for prototype tooling applications
For all of Mr. Morton's lengthy protestations about my conclusions, statements he makes about the way he runs his business indicate that he clearly agrees with them. He states that "we typically leave stock in areas where Z depth is critical" and "We highly recommend that the tooling masters be finished only by master craftsman". The fact is that if rapid prototyping processes were accurate enough to create patterns within generally accepted tolerances for production tooling, it would not be necessary to leave finishing stock or only used skilled craftsmen to finish the patterns. Mr. Morton admits that they routinely correct dimensional errors in the pattern by tweaking dimensions after the fact.
Mr. Morton also states that "any moldmaker worth his stuff will always know where the critical dimensions are and make sure that there is stock there for final tweaking and finishing." He also states that 20% of the blocks they create are scrapped. If his process were capable of maintaining generally accepted tolerances for prototype tooling, there wouldn't be a need to tweak dimensions, nor would he have to scrap every fifth tool he builds. I have no doubt that the tools that Mr. Morton delivers are within the tolerances specified by his customer. However, from his comments, it is clear that his tools are within tolerance because Fusion Engineering corrects dimensional errors, after the fact, on both the pattern and on the finished tool.
The point of my conclusions is that generally accepted tolerances for prototype molding cannot be achieved with the current state of RP equipment and rapid tooling systems. Mr. Morton's comments clearly demonstrate that. The simple fact is, without the intervention of skilled toolmakers, as in the case of Fusion Engineering, rapid tooling does not live up to most of the accuracy claims made in the marketplace.
Mr. Morton further states in his comments that "The number of critical dimensions have a direct bearing on the final price. We have built 3D Keltool molds for automotive certified production molds. The cost was not very much less than a conventional mold because we had to certify each dimension. On the other hand, we have built other molds using 3D Keltool Process which the customer was willing to accept whatever we were able to provide as long as the part functions. These molds were able to be delivered much more quickly and inexpensively, approximately half the cost of any other method available today."
Mr. Morton again confirms my conclusion and goes a step further. My interpretation of these statements are: If you want the tolerances of a machined mold, it will cost as much as a machined mold. If you are willing to forego accuracy, then we can give you the lower cost and shorter delivery time that we promise.
Does my presentation do "a great injustice to our struggling industry" as Mr. Morton claims? I can hardly believe that work to identify and quantify sources of dimensional error in rapid tooling can be considered an injustice. On the contrary, I think that shedding light on the mechanisms that drive tooling accuracy enable further progress to be made.
In fact there is evidence that such progress has already begun based on Dr. Jacobs work. In presentations at the same conference last week, representatives of both DTM and Extrude Hone made reference to their efforts to reduce the mean shrink rate of their materials to reduce dimensional errors due to variations in shrink rate. DTM in particular has made impressive progress in this area, reducing the mean shrink rate of Rapid Steel from 2.5% to 0.2%. Even Keltool, the process so dear to Mr. Morton, reduced the mean shrink rate from 0.8%, where it has been for several years, to 0.6% in the two years since Dr. Jacobs first presented his information. I have to believe that the work done by Dr. Jacobs to identify a major cause of dimensional error in rapid tooling is at least in part responsible for inspiring the concentrated effort by manufacturers of rapid tooling systems in this area. I think it is ironic that recent improvements in the process Mr. Morton defends so vigorously may be in part inspired by work that he describes as "nonsense".
I believe the greater injustice to the industry has been done by those vendors who overstate the accuracy of the tools they sell. I still hear salesmen claim that their process can deliver a tool within 0.005" in a matter of days at half the cost of a machined tool. A long history of questionable claims and customers disappointed in the tools they received has reduced the credibility of the industry to dangerously low levels. If the industry is to improve customer confidence in its products, a better understanding of the true capabilities of rapid tooling, by both vendors and customers, will be necessary to avoid selling rapid tools into applications where they are inadequate. It appears that Mr. Morton would prefer to pretend that rapid tooling has no limitations.
I do not, as Mr. Morton would have you believe, have a "vendetta" against the rapid tooling business. Even with its current limitations, I believe that, for some types of parts, it offers real value to the customer compared to conventional machined tooling. In fact, I think Fusion Engineering has probably done a better job than anyone at identifying and exploiting those market niches. It is unfortunate that Mr. Morton has chosen to respond to my presentation with personal attacks instead of using the information to further the state of the industry.
Tom Mueller
President
Express Pattern
1574 Barclay Blvd.
Buffalo Grove, IL 60089
(847) 215-0001
(847) 215-0002 fax
tmueller@expresspattern.com
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 - 22:53:00 EEST