Talk:SpoolHead

From RepRapWiki
Jump to: navigation, search

I propose that in a multiple spool/pinch wheel design-> in order to facilitate "multi-head" printing: the "new" wire could be adhered to the "current" wire by means of an adhesive such as superglue and a crimping clamp. The "current" wire would then feed both wires past the adhesive crimping clamp, grab the "old" wire, and cut it at a point past the grabber, but prior to the bond.

The "new" wire can be then towed to the pinch rollers by the "current wire."

In practice: it is most likely more practical to have multiple "spool heads" which contain pinch rollers and the spool of wire as a complete independent assembly, separate from the stepper driver, and which can be changed out by the means of a "tool changer."

--Lewisptaylor


>> There are three proposed methods for this as well: Using superglue, using UV-curing resin, or using heat.

A fourth option may exist as a viable alternative - use mechanical means to secure the wire.

Print a layer or two with a linear recess with a small complete or segmented overhang, then press the wire into the recess, snapping it past the overhangs. The recess should be as close or slightly larger than the wire diameter, and the overhangs just slightly less... in cross sectional ASCII art:

  O   <- mechanical pressure forcing the wire past the overhangs.
__ __  
 (_)  <- surface created with a small recess allowing the wire to snap into it.

--BeagleFury

  • Interesting method! This could work well for larger diameter wires than what we're considering at the moment (<0.5mm), but I don't think our resolution or accuracy are good enough for doing that on the small scales we're starting out with.

We are planning to use similar mechanical methods to help position the wire though; for example, we'll print guides on the inner radius of corners and such, so that the wire's tension doesn't pull it out of place.

By the way, can I ask who wrote the above idea? :)

--Jbayless 19:31, 28 January 2010 (UTC)

I also mentioned in the forum the idea of 'zigzagging' thru pre-printed small overhangs. new ASCII art image:

  <  <-.
  |     \
  >    <-+-- tiny overhangs to prevent the wire from moving laterally
  | 
  <
  |  <- wire

I'm not sure how you would secure the initial bit of wire.. perhaps wrapping a small amount around a post that would be snipped by the operator when attaching external wiring. The idea would be that you would advance the wire enough, then switch to printing plastic on top of that wire to firmly secure it. With two print heads, one wire head, and one plastic extruder, and the ability to retake up slack given from the wire tool, you could possibly calculate points that would allow you to more firmly secure wiring by wrapping it around a temporary post, extruding extra plastic on top of the just layed wire, then returning to the wire wrap tool which unwraps the temporary end back off the post.

--BeagleFury

Hi Guys, As I talked about on the forums, you could extrude the metal thread from an extruder orifice to 'lubricate it' with thermoplastic and at the same time keep it a the right temperature in order to fuse with the material below it. The fact that there will be thermoplasts around the wire will really help bonding with the rest. Stopping the regular extrusion would allow you to keep a part of the wire bare (for galvanic connections). It could be functional for conductors, but keep in mind that structurally the parts could also greatly benefit from composites (like anchored mortar). The extruder tip might degrade from the wire, so I'm not sure whether it would be useful as a normal extrusion head too. Otherwise it would be promising to combine the two.

Also, for some wire material, fusing it through induction heating could be an option. And for the nichrome, a current could be applied to heat it up.

--ErikDeBruijn

... i saw in past some methods of feeding wire-wraps along paths and edges, where on a thermoplast sheet were many prefabbed short columns, the wire was feeded between or around a column and then a heated stamp pressed on the column, so the tip melted and fixed the wire beneath the collar ...

--VDX

That's very interesting! We're already planning to pre-print short columns/ guide curves along the inside of turns, to improve accuracy. Melting them onto the wire would be very helpful I bet.

--Jbayless 17:53, 1 February 2010 (UTC)

... maybe another idea: - fab or mould a sheet with a grid-array of holes ... then insert only there some plastic-pins, where you want to feed wires ... or fab some clips/pin-modules with the right grid-dimensions for adding all sorts of electronic or mechanical components by simply clicking in the holes.

Then you'll have some versatile module like a bread-board for electronics or mechanical construction ...

--VDX

How about heating the wire so it melts the plastic just a little, similar to how each plastic layer bonds to the previous? Then you could embed it in the surface, and hopefully the plastic will grab it firmly enough to stop it curling up as the head moves away.

--Triffid_Hunter

... i think this is not so good as the wires would be of different lengths => diferent currents for the same temperature - and the temperature profile along the wire wouldn't be uniform too - as the local heating per current is highly depending of contact to the plastic surface, what's nearly impossible to guarantee over the complete length.

Maybe better try with strictly local heating per hot stamps, induction, light (if the wire is dark and the plastic white) or laser ...

--VDX

Yes, I think you're both right, and this is precisely what we're planning to do. VDX is correct that it would be nearly impossible to heat the wire using its own resistance, due to varying lengths/materials/uncontrolledness of the situation (not sure if that's what Triffid hunter). But we are planning to heat the plastic surface locally, and the wire too, using a separate nichrome heating element. Then we should be able to bond the wire to the plastic either by laying it on top, or perhaps with some additional pushing/stamping force from place to place. We will experiment to see how little stamping we can get away with, to avoid messing up the printed part.

--Jbayless 00:00, 6 February 2010 (UTC)


... i'm brainstorming with an aditional toolhead feeding really thin plastic-wires (0.1mm fishing line or thinner), touching with the end of the wire the surface and with a dode-laser (see [Development:Laser Cutter]) melting/fusing the tip down to the surface.

Retracting the fiber while a piece of the tip is molten and in contact with the surface should snap the wire from the molten blob, so you can place single droplets of plastic.

With moving and rotating the tool-head and feeding the wire with the moving-speed you'll draw a contour-line similar to FDM but with the thickness of the molten wire (e.g. 0.05mm or even smaller?)

For better absorbing the laser-energy you should use dark coloured or pigmented palstic ... with a CO2-laser (10600nm Wavelength) instead of a diode-laser (typically 808, 905 or 975nm) and very low powers you can melt all light or transparent plastics too ...

If someone is in need of even smaller droplets, i've developed a method of mechanically placing droplets of high viscous materials (fluids, pastes or molten materials) in diameters from some ten micrometers down to submicron range ;-)

--VDX

Interesting technique. I hadn't thought of using a laser to heat the plastic, but that might be useful even for the current SpoolHead approach. Laser heating would be a lot more precise than our current plan, using radiative and convected heat to melt the surface. We'll have to consider doing that as a fallback plan.

One concern about your plan for ultra-thin wires is that they tend to be very frail and kink easily, so it would be hard to have a needle that can feed them, and it might be hard to control their position precisely.

--Jbayless 23:17, 26 February 2010 (UTC)


... i had to handle gold-wires with 25 micrometers and platinum-wires with 10 and 1 micrometers (1 micron = Wollaston-wire), so i have some experience with really thin and fragile wires ;-)

And when remember this, i'm rethinking my proposal in respect to melting/fusing metal-wires with a stronger laser - i have [email protected] with smallest possible spot of 5 microns diameter or can bundle some of the diode-lasers with 5Watts per fiber to a spot of 0.3 to 0.5mm diameter and powers as sum of the diodes-count (around 30Watts with 7 diodes, or more with higher counts)

--VDX


Wow, VDX, it sounds like you've had a lot of experience in high precision industries. Where did you work? I would love to learn from your experience. When you were handling these fine wires, how did you feed them?

I've done wirebonding work, handling micron-scale aluminum wires using tweezers under a microscope. The main frustration was their tendency to kink. It was easy to advance them under tension, by pulling the wire through the needle once it's already been threaded. But if the wire snapped in the wrong place, I would have to thread it manually (tweezers again) through the needle, which meant pushing on the wire. Needless to say this was frustrating work, which took a lot of experience before I could get it right. But I'm pretty young and I haven't had a lot of other industrial experience, so I don't know what the best solutions are.

Unfortunately the only solution I can think of for the SpoolHead is to drive it from the rear; that is, to push on the wire (at least until it adheres to the build surface). I wonder if a machine will be able to handle the wire as gently as a human hand.

When you were working with these wires, was it a machine that was handling them? If so, did it ever need to "push" on the wire? If so, how was it achieved?

--Jbayless 22:23, 27 February 2010 (UTC)


... the 25microns-goldwire was feeded through a bonder-needle with an inchworm-actuator, but stuck sometimes in the capilary, so i had to clean the feeding hole and reinsert the wire two to three times in an hour bonding time.

The 10microns-wire was stiffer than the 25microns-goldwire, so was more stable in the (smaller) feeder, but i had to manipulate it manually for soldering, glueing and brazing, so handled it mostly with a tweezer.

The 1micron Wollaston wire couldn't be handled with a tweezer or bonder (simply to thin and brittle), so i soldered a piece of the 50microns-Wollaston-wire to two tips, etched the silver in the middle region away, so the 1micron-platinum core came free over some milimetres and parted the two tips with the remained silver-pieces, so the 1micron-wire was stretched until it was straight ... then i moved it with a manual 5D-microstage until it was in position and glued and soldered it at the goldpads.

I have fishing-wire with 100 microns and steel-wires with 25, 35 and 50 microns and they are much stiffer than the 25microns-goldwire, so should be easy to handle and feed through a hollow needle with an inchworm-drive or maybe with a motor and two disks - a hard one on the motor and an elastic one as counterpart ... so you have enough grip for feeding the wire and not enough pressure for deforming it ...

--VDX

Yes, I think microwelding is the way to go. Lasers would be great because we could probably use the same laser to do other useful things too. But strong lasers are also dangerous for amateurs. I've also heard of something called "microticking", where a capacitive discharge can be used to weld two wires together even when they're highly conductive like copper. Do you know anything about this technique?

--Jbayless 19:51, 4 March 2010 (UTC)

... yes, i made some developments for welding thin Kovar-wires to PCB-pads by discharging high value capacitors (20000uF/75VDC, charged with 12 Volts) - but stopped when we bought a comercial micro-pulse-welder. I discharged simply with a big mechanical switch through a wolfram-pin pressing the wire to the pad. For automated discharging you need a fast switching high-power SSR or similar ...

--VDX