Re: [DSII] Genre - and details about casting/moulding

>oh, ah, erm, i was using 'fuse' in the sense my physics teacher taught
me,
>meaning 'melt' (as in 'latent heat of fusion'). honest.

of course... :)

>> the
>> cutting tools are driven by 3D CAD data.  It takes a lot of time to
get the
>> CAD data correct, and to then develop mould plans from that data.
>
>is stereolithography used for this? that would seem a natural (if
>expensive) way to make masters. of course, that doesn't help in making
a
>steel mould. oh well. how about electro-discharge machining?
metalcast.com
>says:
>
>"Do you think that stereolithography is the only 'buck rogers' rapid
>prototyping method out there? Think again! EDM *disintegrates metal
with
>electricity*. Thousands of minature lightning bol tscarve the metal
into a
>wide range of three dimensional surface contours."
>
>of course, this is bound to be pricey. pricey, but cool.

Stereolithography is used for making patterns (masters) in the blow
moulding industry (blow moulding is the process that is used to make
plastic soda bottles, shampoo bottles, etc. and a huge variety of
consumer
and industrial stuff - mostly hollow items like that).	Blow moulding is
a
lower pressure process, and uses aluminum tooling.  They use the
stereolith
model as a pattern to make a wax copy of the part, then use a lost-wax
process with sand to cast a rough aluminum mould.  This is then machine
finished.  This isn't used much, however, because the resulting mould is
not as high-quality as a machined mould.  Good for quick stuff and for
prototyping, though.  I worked as a designer for an automotive parts
manufacturer, and we had a whole stereolighography shop set up to do
that
kind of thing.	Fascinating process.

Where you see stereolithography used in the injection moulding industry
is
for, as your EDM quote mentioned, "rapid prototyping".	The resins used
in
that process are unbelievably expensive (I think a gallon can goes for
about $20,000), and are not very strong (quite brittle, actually).  My
first decent summer job while I was in design school was working at a
stereolithography rapid prototyping company.  We used RTV rubber
(industrial stuff - much tougher than what is commonly available to
hobbyists) to make casts of the prototype part, and then used various
urethane-type casting resins to create plastic copies which exhibited
most
of the strength and durability characteristics of an injection moulded
part.  Some of the rubber moulds we did were quite complex - the last
one I
did before I left the company was a five-part mould - all kinds of
inserts
in the middle, etc.  Neat stuff.

Piece of trivia for the list - Steve Jackson Games was investigating
making
cast metal Ogre models from Stereolithography masters, and the company I
worked for did some of the masters - so there were a bunch of little
Ogres
around the place.  That project didn't go anywhere, to the best of my
knowledge, though.

EDM (Electrical Discharge Machining) is a wild thing to watch (I had a
tour
at an EDM tool shop).  They use that process to manufacture production
moulds - though if you use softer metal (ie not steel) you can make
prototype moulds for injection moulding relatively quickly.  Very wet,
though - they use water as a cooling and conducting medium, so there is
water all over the place when the machine is working.  

>
>> For
>> models like GW tanks - picture the Eldar Falcon Grav Tank which is
all
>> curvy - they do not create 3D CAD data - they produce a master model
about
>> 3 times normal size and then use that as a guide for doing the mould
>> cutting.  This is all very time consuming and costly.
>
>i've heard people refer to the use of pantographs in this context, like
>those drawing reducers you get as kids toys.

Yes.  that's exactly what they use - one end of the pantograph has a
cutting tool, the other a stylus they run over the master.  I wanted to
mention it, but forgot what it was called.  Thanks.

>> You get a high reject rate
>> with small figures because they require small moulds.  Small moulds
have
>> tight air passeges (ie the arm or leg of a foot trooper in 25mm), and
there
>> is an extremely high likelyhood that you will trap air in the mould,
>> causing a rejected part.
>
>so air trapping is the problem? hmm. that's just crying out for a
>technological solution. maybe if you used open-cell foam moulds, so you
>could suck the air out, but which the resin wouldn't leak through. if
the
>holes were small enough, surface tension would keep it in. plus, the
>smaller the holes, the less effect they have on the moulded surface. is
>there a polymer chemist / materials engineer in the house?

The real answer to this is to not have air in the mould or disolved in
the
resin to start with.  When I worked in the rapid prototyping company, we
used a vacuum chamber to degas all our RTV and casting resin before
pouring
either.  Cut down on trapped air bubbles by over 90%.  If you want to do
it
really well, though, you do the pouring IN a vacuum chamber which has
been
degassed.  That way there is NO air in the mould (well, not enough to
make
a difference, anyway) as you pour and therefore no trapped air bubbles.
You need a big vacuum chamber, though, with remote manipulators or
something to protect your hands if you stick them in to do the work
with.
That would basically eliminate the air problem.  It wouldn't solve the
"resin is brittle and arms break off minis easily" problem though.
Lead/pewter is the best, 'cause it will just bend if you drop it...

>
>> With spin casting, air is forced out of the mould
>> by centrifugal (centripidal?? I always forget which is which) force
as the
>> mould spins and the molten metal is forced into the mould cavities.
>
>why can't you spin-cast resin?

Because it takes at least several hours to set.  You'd lose the
efficiency
of spin casting if you had to keep the thing going for five hours just
to
get one set of parts.  Also, the resin is liquid at first, and the
moulds
are 2 piece.  If you spun it for that long, you would expel the liquid
resin out the sides of the mold long before it had a chance to set up.

Having said that, I have some casting resin at home that I use to make
the
odd bit here and there...  It sets up in about 45 seconds, and you can
demould in about 2 or 3 minutes.  Great stuff, though you have to work
FAST.  The resulting parts are sandable, drillable, cut-able, etc, and
are
not particularly brittle.  You have to be careful to look for trapped
air
(I don't have a vacuum chamber at home...), but otherwise it is great
stuff
(well, it is a nasty gooey chemical that shouldn't touch your skin, so I
wear latex gloves, but...).  For those who are inclined, it is available
in
North America as "Alumilite" Casting Resin - and it is rather pricy ($40
for a 1 litre kit) - though for the stuff I make (small bits, mostly) it
lasts a while.	This isn't something you could do large vehicles with, I
think, 'cause of the cost - but then again, I don't know what the
company
would sell it for in large quantities.	You can slow the set-up time
down,
by sticking the 2 parts into a refrigerator for a couple of hours.

Most of these casting resins react exothermally, and if you heat them,
they'll react faster.  When we had rush jobs at the RP shop, we would
stick
our moulds full of resin in a hot-air oven and cook them - they'd set up
and go hard in 1/4 the time - the results were often not quite as good,
though.  More chance for trapped air.

Anyway - once again I hope somebody finds this interesting/useful.  I've
done quite a bit of work with making RTV moulds and casting on my own,
and
if anybody has any questions, feel free to drop me a line off-list
(unless
you think "advanced" modelling has a place here too - I guess it
does?!).

Adrian