Re: [CAD_CAM_EDM_DRO] Making a rapid prototyper

On Wednesday 02 August 2006 18:44, Alan Marconett wrote:

> HI Sebastien, FRED,
>
> Thanks for the comments; my comments below:
> > -----Original Message-----
> > From: CAD_CAM_EDM_DRO@yahoogroups.com
> > [mailto:CAD_CAM_EDM_DRO@yahoogroups.com] On Behalf Of Sebastien Bailard
> > Sent: Tuesday, August 01, 2006 10:03 PM
> > To: CAD_CAM_EDM_DRO@yahoogroups.com
> > Subject: Re: [CAD_CAM_EDM_DRO] Making a rapid prototyper
> >
> > On Tuesday 01 August 2006 12:28, Alan Marconett wrote:
> > > HI Sebastien,
> > >
> > > I hadn't seen the 2nd generation RP head, just the first. Sounds like
> > > a step forward! I'll give it some study. (I wish you'd publish
> > > PDF's.)
> >
> > I assume you want an engineering drawing of the extruder head, in pdf
> > format?
>
> Actually .DXF would be better, but publishing the accompanying article in
> PDF would be great.

The files describing the parts of the nozzle are in a gzipped tar file here:
http://reprap.org/Downloads/extruder-v2.tgz
- a mention of it is buried here,"12. Design and software downloads" :
http://reprapdoc.voodoo.co.nz/bin/view/Main/Version2OfThePolymorphExtrusionHe
ad

I'm afraid for the parts like the Polymorph Rod Guider, we designed them in
ArtOfIllusion, save them as .stl and print them with Adrian Bowyer's
Stratasys-brand RP machine. ArtOfIllusion does not do DXF.

If you are lucky, your cad program may be able read .stl files.

There is no accompanying article, it's all on the website and in that .tgz
file. Let me know what still needs to be documented. I realize PDF's are
useful for the workshop, but if I say "you know, it'd be really nice to have
http://reprapdoc.voodoo.co.nz/bin/view/Main/Version2OfThePolymorphExtrusionHe
ad as a PDF, they'd give me that 'you just volunteered' look again." Right
now, it's near the bottom of my project stack.

> > I'll see about doing one up; the CAD parts are being done in
> > ArtOfIllusion (AoI), a free rendering program. AoI is a good program, but
> > it can't output technical drawings. The renderings on
> > http://reprapdoc.voodoo.co.nz/bin/view/Main/Version2OfThePolymorphExtrusi
> >o nHead
> > are from AoI. The plastic parts in the photographs were printed by
> > Adrian Bowyer's Stratsys rapid prototyper - the stratsys used .stl files
> > from AoI. (Adrian's the project lead.)
>
> STL files for those who plan to directly RP the plastic parts (as published
> in version1) should be well received. I'm half tempted to have YOUR vers2
> parts made just for fun!

Let me know if you do. What would you use as a CNC machine, a sherline?

> > Since I'm trying to use my mill to machine an extruder head, I'll put up
> > the
> > g-code and related files that other folk with CNC mills will be able to
> > use.
> > Neil Gershenfeld, of MIT's Fab Lab, wrote this bit of code for converting
> > .stl
> > files into .g code - we may find it useful for this and for other
> > projects:
> >
> > http://web.media.mit.edu/~neilg/fab/dist/cam.py
>
> I downloaded it. I don't have Python, but it looks a LOT like C. At least
> the algorithms are there for those wanting to understand the process (me).
>
> FRED,

It needs a bit more documentation, but at least the code is there. If you
have any questions about it, go here:
http://fab.cba.mit.edu/forum/phpBB2/viewforum.php?f=20
The fab lab's don't have a wiki yet, which is a mistake on their part. The
nice thing about wikis is that the documentation grows organically.

> It occurs to me the StlWork might do some STL to Gcode conversions like RP
> requires? Something to do with water line?
>
> I was considering generating Gcode initially for a simplified object by
> drawing the "layers" in Vector CAD/CAM. RP would seem to be the INVERSE of
> milling a mold cavity. But you'd have to stay on the same Z level before
> moving on to the next one. Hmmm, real intellectual stuff...
>
> > I'd also like to put up .dxf files of the thing, which will happen soon
> > after
> > I buy and start using a CAD program - either varicad or something else
> > that
> > will work under linux/linux+wine. Any suggestions for linux-friendly CAD
> > programs?
>
> Not familiar with CAD on Linux. I'm using Vector CAD/CAM. Make sure the
> DXF files you export are about vers. 12 or so. Newest ones can't be
> imported by all CAD programs.

Will do. I'm trying BRL_CAD. The user interface is daunting.

> > > RepRap is using lower temp materials then I was considering. We
> >
> > recently
> >
> > > purchased a ProtoPulsion Dimension SST RP, and it uses ABS "wire". I'm
> > > told it melts at about 165C, a little warmer! And, the inside of the
> > > enclosure for the RP is heated. Lovely machine!
> > >
> > > How hard is the Poly material you're using when it's cooled? The ABS
> > > is fairly hard. And the .07" ABS wire (1.77mm) comes on spools already
> > > (no need to make your own PolyMorph rods! I'm told I can have the
> > > tail-ends of the spools; they leave quite a bit on them in the
> > > magazines.
> >
> > We've looked at using ABS, but the lower temperature thermoplastics may
> > be better for our purpose, since we can recycle them with a double boiler
> > and a ladle. Anything that melts over 100C is much harder to process.
> > These are the stuff that we're thinking about using:
> >
> > http://reprapdoc.voodoo.co.nz/bin/view/Main/ThermoPlastic
> >
> > We're probably going to stick with spools of 3mm diameter filament of
> > some thermoplastic, caprolactone or something else. Forrest Higgs has
> > done some good work on extruding 3mm filament, written up here:
> > http://reprap.blogspot.com/2006/07/new-wrinkle-on-capa-filament.html
> >
> > Often, feedstock for rp machines can be expensive - over at a university
> > I know, parts from their printer cost them about $6/cubic inch! For
> > comparison, solvay sent me a 20kg bag of polycaprolactone (CAPA 6800)
> > pellets for about $5/kilo.
>
> That's a good point to recommend low-temp thermoplastics. A magazine of
> ABS is about $250. I don't know the length or weight of the ABS supplied.

You'll find that restuffing your magazines with ABS sourced from a local
plastic shop violates your service contract.

ABS in the wild is pretty cheap, I think. We're using low-temp
thermoplastics because you can recycle the parts easily.

> Having more then one material of interest suggests that the extruder head
> should be made to accommodate various diameters and temps of material.
> This might be done by simply replacing the "Polymorph Rod Guide", although
> perhaps the High-temperature Nozzle would have to be adjusted as well.

Probably. It's got an OD of 6mm, so you could probably drill it out to a ID
of 4-5 mm, but that's it.

Also, right now it's counterproductive to look at lots of variations on
nozzle design. (Big improvements yes, many versions no.) After the 1.0
release, we can diversify.

> > CAPA 6800 has an elastic modulus of 440, a yield stress of 16 Mpa, a
> > Shore A hardness of 94, and a Shore B hardness of 50, according to:
> > www.solvaycaprolactones.com/.../attachments/a.Props.___Processing_of_CAPA
> >_ Thermoplastics_ADS.pdf
> > I'm just parroting here, I don't know beans about material science, and
> > couldn't tell you how it compares to delrin or ABS. Also, most numerical
> > data is going to be for a solid ingot or rod of plastic - when I have a
> > rod made from filaments, I'll try to make relevant measurements.
>
> I don't have data yet for the ABS material I'm considering. I'll see about
> getting a sample of CAPA.

www.solvaycaprolactones.com has a good sampling policy, and there's a form on
their website here:
http://www.solvaycaprolactones.com/sample/0,5683,3486-_EN,00.html
(Don't even try to get it from DOW. Extremely uncooperative.)

For more info about sampling, check out this discussion in the emc-users
mailing group:
http://sourceforge.net/mailarchive/forum.php?thread_id=30159447&forum_id=3314
0 I'd mentioned the reprap project and asked if anyone wanted to help design
a reprappable wire-edm head, and we got to talking about CAPA.

Let me know if Solvay gives you any difficulty in getting a CAPA sample.

I should have asked here about EDM as well. Anyone interested?

> > Also, we'd like to source our thermoplastics or biopolymers from plant
> > starch
> > or something similar, as opposed to from petroleum. However, we'll look
> > into making our own plastic after the 1.0 release.
>
> Sounds a bit sticky!

It's not my field - I don't know if this is a three-month project for a grad
student, or something that would take n years for a dedicated team. We also
haven't decided if we're going to look into it now, or after the 1.0 release.

> > > I've designed and have been building a "Baby PCB Router", which at 12"
> > > x 12" x 12" high overall might be an appropriate size to take an
> > > extrusion head. It's currently designed with a 3.25" x 5" tooling plate
> > > on the Z axis, and I have a bracket for DeWalt DW670 trim router
> > > mounted on it.
> >
> > It sounds perfect. The extruder head is 20-25 cm tall, and you may need
> > an adapter plate to hold it. If a mill can move a dremel around, it can
> > move an
> > extruder around. (It should be possible to build an adapter plate
> > sockets into a collet on a bridgeport and holds our fiddly little
> > extruder head, so that people with big machines could do this too.)
>
> I think I'd just design my own extruder using metal and Delran, and it
> would be a simple job to bolt up to my Z axis tooling plate. Or a Sherline
> Head spacer block as well! The Taig could no doubt be easily accommodated
> as well (if I had Taig info).

The taig's got a bunch of slots in the block that holds the spindle. It's
utterly perfect for this kind of thing. I've got a crude prototype of a
laser scanning head on mine.
http://www.dubsen.org/twiki/bin/view/Main/RepRapLaserScanningHead
I don't have a picture of the device bolted onto my machine, because the
device contains my digital camera.

> > > Although I'm currently implementing USB with the PIC 18F4550 uP, my
> >
> > Step4
> >
> > > controller program runs under DOS. I don't yet have a convenient
> > > driver for USB under DOS. And I am a LONG way from getting Step4
> > > ported over to Windoz (if ever). I am investigating USB-CNC, a windoz
> > > controller
> >
> > program.
> >
> > > The extruder control program could run as a windoz program as well!
> > > All we'd need would be some simple handshake between the two processes
> > > (I think, help me here).
> >
> > If it is easy, we'll have a command line switch so that the reprap
> > controller
> > software also works with parallel port + emc controlled machine tools.
> > Our
> > controller is java based - it will work under linux, osx, windows.
>
> Why not make the extruder controller appear as a spindle? Send it "speed"
> commands? (S2000). M3/M4/M5.

We're not going to use gcode for this, mostly because we want our device to
work over USB with a java program as a driver, running it on Win/OSX/Linux.
If emc worked on all 3 platforms, we'd have gone with emc and g-code, maybe.
USB (or maybe ethernet) is necessary because modern computers don't have
parallel ports.

> > > I think "self replication" is a long way off (Star Trek, NG), I'm
> > > simply interested in being able to explore a new fabrication method.
> >
> > I agree with you. I have no idea how to fabricate a microcontroller.
> > Possible a microfluidic plate with conductive/semiconductive plastic in
> > the
> > channels, for a hand-waving solution. Once we get the 1.0 release out
> > the door, we'll worry about esoteric stuff like that. More usefully and
> > realistically, the gang has looked at printing circuit boards out of
> > conductive materials and thermoplastic. Right now, it's silly to work on
> > that.
> >
> > > Have you considered a two-material extrusion head? The SST also has a
> >
> > dual
> >
> > > extrude head; the second material is a water-soluble structural support
> > > material.
> >
> > Yeah, we'll have some kind of a support material going for the 1.0 or 1.1
> > release, along with a higher diameter orifice for thermoplastic to make a
> > bulk feed head. Every new material we add to the feedstock bus makes it
> > heavier and more complicated (especially if it means another motor). I
> > wrote
> > up some thoughts on materials in an evening or so:
> > http://reprapdoc.voodoo.co.nz/bin/view/Main/MaterialsScience
> > I'd enjoy hearing what folks would suggest,
> > [ ...especially in a form I can copy and paste into the wiki ;) ]
>
> Copy away!

Certainly. Any bit in particular?

> > > A sintering head would really be making progress! I know nothing about
> > > the process, but expect it needs higher heat and pressure that a
> > > hobbyist can easily handle. Please prove me wrong!
> >
> > You can sinter clay (like your tea cup or a rp fabricated object from
> > extruded
> > kaolin slurry), precious metal clay, or aluminum slurry in a kiln or
> > furnace.
> > Extruding the material requires a syringe in a squeezing mechanism or a
> > peristaltic pump clipped onto a feedline ending in a syringe tip, so that
> > is very easy.
>
> Funny you should mention peristaltic pumps. We make commercial laundry
> controller equipment at work, and make equipment with various sized
> peristaltic pumps. I don't know how they'd work with clay or a slurry
> 'tho.

Adrian's got one he RP'ed
here:http://reprap.blogspot.com/2006/07/peristaltic-pump-go-after-all.html
I want an RP!

> > If you want to make carbide lathe tool bits inserts, you'll need a
> > pressurized
> > mold in a furnace. That is hard. Luckily, most objects won't need to be
> > pressurized when they are sintered.
> >
> > There's some work out there on putting down a layer of material,
> > sintering it
> > with a laser, putting down another layer, etc. We'll worry about it
> > later.
>
> Yes, later. No room for a furnace.

I'm buying my jewelry's instructor's enameling furnace. It should work. I
need something to burn out jewelry flasks full of invested wax or CAPA,
before doing silver casting.

> > > Can't we make the extruder out of METAL? Would that make it smaller
> > > and more compact? I'd also consider a stepper motor for the drive. As
> > > I mentioned, I'm really not targeting self-replication. I'd make the
> > > best use of materials available to me.
> >
> > Yes. If you have a mill, it makes more sense to make the extruder out of
> > metal. For the reprap project, we need to be able to make the extruder
> > out
> > of thermoplastic and a drilled out brass bolt made with a lathe. (We
> > only need a crude lathe based on a hand drill or rotary tool:
> > http://staff.bath.ac.uk/ensab/replicator/Downloads/Afghan_lathe/Afghan_La
> >t he.html
> > I'd suggest using a real lathe if you have one.)
> > This way, we can make copies of the reprap using a reprap and a rotary
> > tool.
> > If we cannot, it's not truly self-replicating, and we haven't met our
> > project
> > requirements. (The assumption is that the user does not have a mill or
> > ECM/EDM head, but does have a reprap or repstrap and a spool of
> > thermoplastic.)
> >
> > If you do design a smaller extruder out of metal, the reprap group would
> > be very interested to hear about it and would gladly host the files and
> > write-up; drop me an email or log onto to objects.reprap.org and upload
> > the info.
>
> I can do that!

Hmmm... How would you do it? (Remember you need to keep the metal away from
the thermoplastic parts extruder head holder.
(Regarding the stepper motor for the drive, probably not. Stepper motors get
expensive, and we'd like to keep costs down. See:
http://reprap.blogspot.com/2006/08/rethinking-everything.html

> > Eventually, not today but by 3 years from now we hope, we'll be making
> > loaner
> > extrusion heads/repraps, and sending them out to folk who want to make
> > their
> > own machine, and would be willing to make their own loaner extrusion
> > heads/repraps.
> >
> > It's far too early to start a sign-up sheet. Sort of like counting your
> > chickens before inventing the first proto-chicken. I'm hoping to have a
> > stack of loaner extruder heads ready within 2-6 months. We'll see.
>
> Ha! Delivery time is now down to 2 months! I'm watchin' my mail! ;>)

We will see. I think you'll have one done before I do, if you are serious
about making one.

> > Anyone have plans for a "flat-pack" cnc machine they'd care to release
> > under the gpl?
>
> What's a flat-pack cnc machine? There are some wooden hobby routers to be
> made. I'm using 3/4" x 3" and 1.25" x 1.24" aluminum bar stock!

The flat-pack RepStrap (FPRpSp) machine is an idea of Vik Olliver's - if he
were to design it, we'd publish it as a file anyone can take to a CNC Router
shop / signmaker's shop and cut out the structural elements for a RepStrap,
just waiting for steppers, drill rod, threaded rod, and an extruder head.
http://reprap.blogspot.com/2006/07/sign-for-reprap.html

> I know, get a Duck!

The Duck has dodged a bullet - the FPRpSp is unlikely to be strong/vibration
resistant enough to move a rotary tool around. Hmm... Ballendo's got a good
price on steppers; the steppers on my taig are starting to look pretty
weedy...

I kind of favor the "buy a benchtop mill, CNC convert it" pathway to making a
RepStrap and then a RepRap. Vik's idea is cheaper, and therefore better.

> Cheers,
>
> Alan KM6VV
> Central coast, CA, USA
> SherlineCNC list

Regards,

Sebastien Bailard

-------------------------------------------------------