STONE SOUP

From: Alain BERNARD (Alain.Bernard@irccyn.ec-nantes.fr)
Date: Fri Dec 14 2001 - 12:45:05 EET


Dear All,

Is there a normative procedure that allow to compare technologies one with
the others ? Not at all. Many os us have proposed parts and process
procedures in order to be able to qualify layer-manufacturing machines but
the results remain very difficult to generalize.

My experience is that what is also difficult is to fit the user
requirements (some elements related to product development objectives at
different stages of this development) with process capabilities.
I think there are two levels of decision :
- the first one is to decide between two layered-manufacturing machines or
some "one-step" technologies
- the second level is to see if there is a combination of technologies
(that should be software-based transformations or physical transformations)
in order to decide about a complete process.

For the fisrt one, one PhD student is doing is reaserch on an initiative
based both on the analysis of the geometry and on the functionnal of the
part. The idea is to decompose the part in different pieces and manufacture
each piece with a given geometry and assemble the all pieces at the end.
This is what many companies with the combination of both conventionnal/RP
techniques for some tools (molds or dies).

We have tried and are corrently trying to improve an approach for the
second level of decision (work published with Mr Deglin in different
conferences). The approach is based on the constitution of a case-studies
database. Data are modelled related a conceptual model that we propose.
Decision making is based on case-based reasoning and also on ascendant
generation of process. We also have tested graph theory.

Of course we agree that with the same means two companies should abtain
different results, mainly due to process parameters and men experience.

This problem remains an interesting problem. But I agree with Richard about
the comparison between the technologies. Our reasoning is based on general
parameters (kind of material, speed, quality, etc...) but this is very
difficult to evaluate precisely because many of them are validated in each
company.

Best Regards,

Prof. A. BERNARD
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Prof. Alain BERNARD
IRCCyN - Ecole Centrale Nantes
1, rue de la NoŽ BP 92101
44321 - NANTES Cedex 03
tel : + 33 (0) 2 40 37 69 53
fax : + 33 (0) 2 40 37 69 30
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

For more information about the rp-ml, see http://rapid.lpt.fi/rp-ml/



This archive was generated by hypermail 2.1.2 : Fri Jan 04 2002 - 09:58:18 EET