Re: benchmarking

From: Don Wilde (Don@PartsNow.com)
Date: Thu Jan 30 1997 - 20:31:15 EET


> Geometry (just what is most important to users)
> Method of measurement ( hand tools, CMM, laser or other )
> Process parameters and method (for results verification)
> Data to be presented and where (papers, internet)
> Discussion limits (No flaming those who results fell off charts.)
> speed, and cost of models. Which are the most expensive
  processes, which the least?!
Add cost and time to functional metal part
Add cost and time to production tooling

        Our company (Soligen) is built on the premise that real users need real
metal parts, not just visualisation prototypes (although there are lots
of users for those, too, don't get me wrong). We can't compete with some
of the technologies on surface finish and fine detail, but for parts
which end up as metal production items, I don't believe any other system
can touch us, especially if you only get one try!
        If I remember right, Chrysler used SLS to make a proto for the Viper
V-10 manifold a while back. That would be an ideal comparison for us,
although a lot of the others would be left out because of the volume of
the part. Why don't we look at some part with some neat hollow cores in
it which is about 9"x9"x9" or so, with the caveat that the Sanders guys
can build it 1/4 scale. I think that would work for everybody.
        What will happen, of course, is that we will all "win" and promote our
particular selling point. We (Soligen) won't talk about detail
resolution results, but we'll gloat about build time (43 jets, raster
pattern, slice-on-the-fly) and time to production tooling.
        Finally, I know of a site that already did such a comparison about a
year ago which included SLA, SLS, and Soligen, but I don't want to
mention names until I get their approval to do so. It was very unbiased,
and everybody had their strong points and everybody had their weak
points, like I said.

-- 
    oooOOO O O O o * * *  *   *   *
   o     ___
   V_=_=_DW ===--- Don Wilde [ don@PartsNow.com ]
  /oo0000oo-oo--oo-ooo---ooo-ooo---ooo-ooo---ooo



This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:39:18 EEST