Re: Applications

From: Ian Gibson (University of Hong Kong)
Date: Friday, March 17, 1995

From: Ian Gibson (University of Hong Kong)
To: RP-ML
Date: Friday, March 17, 1995
Subject: Re: Applications
>this lull I would like to probe for some insights from this group. 

I agree there does seem to be somewhat of a lull in proceedings, where is 
everyone?

>        - Concept Models
>                - look, feel, show to client
>        - Design Verification
>                - form and fit

I have no problem with the above although I do believe that there should be some mention of the words parallelism and iteration, I said iteration (bad 
joke). Part of the difficulty in explaining the technology to those 
uninformed and sceptical industrialists that remain out there revolves 
around the philosophical changes to company structure affected by RP. What 
I mean is that you are providing a tool that does not just replace another 
tool, it can allow designers and developers the flexibility to evaluate a 
variety of designs beyond the concept stage in much more depth than 
previously possible for anyone other than the big budget designers.

>        - Testable Parts
>                - form, fit and some function

Here I am starting to have problems. What is a testable part? Just how 
close are the test results to the real thing? Perhaps we should split this 
category into 2 separate areas, one that shows an accurate representation 
of test data, the other that shows a correlative representation of test 
data. In other words the ABS from FDM machines can be used to represent an 
accurate ABS part in terms of thermal behaviour but would not give accurate 
drop test results due to interfacial separation of plastic beads. However a 
positive drop test result would mean that the true ABS part is going to be 
more than satisfactory.

>        - Patterns for "Soft" Tooling
>                - RTV, spray metal, etc.
>        - Patterns for Metal Casting
>                - investment, sand, flask, etc.
>        - Hard Tooling for Manufacturing

Here we must also be honest. This is RP as part of a manufacturing process. 
In other words, in conjunction with other machine based processes. Doug Van 
Putte at Kodak described using Quickcast parts which were then 
milled, surface ground and wire EDMed. I dont dispute that RP is a valid 
part of this chain, but it does give me problems in terms of distinguishing 
the above two (or perhaps three since RTV mouldings, etc. can be used 
to master wax patterns) categories. What are these patterns going to be 
used for? Can we really make parts that dont require further machining? Id 
like to hear about it.

>Are there better ways to segregate these? What is missing?

Missing? What about parts that would be extremely difficult or impossible 
to build using other approaches? What about the mechanisms that dont 
require assembly, or the direct translation of CAT scans? They kind of fit 
into the above categories but may also represent subcategories.

The reason I mention this is twofold. People have been ranting on about 
conservative attitudes and then directed their attention at unimportant 
issues like names. My view is that the conservatives amongst us are 
ignoring the fact that microfabrication (or nanotechnology, or LIGA, or 
????, you think we have difficulties with names?) techniques have been 
around for some time. Okay some say that RP is time saving, but surely its 
real benefit is that it is geometrically independent? The same can be said 
about microfabrication. Eventually someone will come up with a MF system 
with a direct CAD input that allows you to make items that could never have 
been dreamed of before.

Secondly, when are we going to stop thinking in layers? We design 
parametric CAD systems that allow us to think in terms of features and 
characteristics and then we constrain the output to build the parts in 
layers. Surely there must be something fundamentaly wrong in the 
mechanics of this approach? Shouldnt we be looking at building 
parts in an omnidirectional way? Crystals grow from a central seed. People 
can steer the growth of crystals to build simple structures like turbine 
blades. When are we going to see an RP system that builds in the same way? 

I feel better for that.


Previous message | Next message
Back to 1995 index