Conference paper
Design to Process Capabilities: challenges for the use of Process Capability Databases (PCDBs) in development
In approaches such as Robust Design, Tolerance Management, Design for Six Sigma (DfSS), etc. there is little disagreement that a better understanding of inevitable production variation is conducive to the success of development projects [Eifler et al. (2013), Arvidsson and Gremyr (2008), Karmakar and Maiti (2012), Breyfogle (2003)].
At the same time, information on the achievable manufacturing accuracy or the supplier’s performance is usually inaccurate and largely qualitative in early development stages. Design decisions as well as the choice of manufacturing processes, therefore, often rely on experiential approaches or expert judgment.
There are numerous examples that this subjective assessment of potential variation and a mostly informal communication between design and production engineers can result in non-satisfying product or manufacturing solutions. Whereas overestimated production capabilities may lead to low yields and a cost/time overrun, conservatively underestimated capabilities affect quality through the reduced design space, or through increased play, rattle/noise, size or weight.
A possibility to overcome the subjective assessment of variation in development projects is a Process Capability Data Base (PCDB) offering valuable insight into the actual or expected performance of production processes (Tata and Thornton, 1999). But although the potential benefits as well as initial challenges for the use of PCDBs have been addressed in earlier research, e. g. by Delaney and Phelan (2008), Kern (2003) or Tata and Thornton (1999), a widespread adoption in industry cannot be observed.
As already stated by Okholm et al. (2014), there are many open questions and a methodical support how to generate, provide and use generalized production variation data still seems to be missing. To foster the use of corresponding databases and to stimulate further research, this paper gives an overview about the scope of potential applications for a PCDB in product development.
Furthermore, the expected manufacturing accuracy of specific product characteristics/features is discussed. For the generalization of measurement data, a DOE (Design of Experiments)-based approach is proposed to identify influencing factors related to the production accuracy of each geometric feature, using the example of metal shear forming processes.
Language: | English |
---|---|
Year: | 2014 |
Proceedings: | 25th Symposium Design for X |
Types: | Conference paper |
ORCIDs: | Eifler, Tobias and Howard, Thomas J. |