Monday, March 23, 2009

 

Wire nozzle

Since Zach’s idea of using a stepper shaft to drive the extruder, the design’s got a lot simpler. We’re now at a point where we’ve only got one non-standard mechanical component in the whole design which needs custom, precision machining: the nozzle.

I had an idea which might get round that: is the Nichrome wire (the heating element) flexible enough to form the nozzle itself? By wrapping the wire around a former, then setting it in an insulating material the wire creates the working profile for the nozzle:



Here JB Weld acts as the insulator. This approach requires no lathe as the former can be made out of ABS filament. The former can be removed by simply turning on the power (ABS melts at 105°C, JB weld decomposes at 200°C) and pulling it out when molten.

Potential advantages for this design:

• No machining needed.
• Direct heating from the element to the polymer (no thermal inertia in the nozzle design).
• Compact.
• One less part.
• Opportunity to alter the draft angles to optimise fluid flow.

The completed nozzle was set in a PEEK cylinder with more JB weld.

Results:



Problems:

• Extrudate was quite fat (~1mm) due to orifice size. This was the first one off though, so will experiment with a former which incorporates a needle to get a tighter wrap at the tip.
• JB Weld decomposes at 200°C which doesn’t leave much of a margin. Will experiment with fire cement instead.
• Wire coming from top of helix is close to the orifice. Will experiment with double helix element so that the wire enters and exits at the 3mm end of the nozzle.
• Potential for non-insulted Nichrome to short circuit if helix is not accurately wound. Will experiment with insulated Nichrome (though fluffy glass insulator is not ideal as a fluid surface and is likely to wear quickly). Note: with a bit of programming, the Cartesian bot might be able to help with the winding.

Comments:
Nice work!

Can you shoot some detail pictures of the tip?
I'm wondering how far in the tip the wire is rolled up.

There goes needing heavy machinery!
 
This sounds really interesting. The nozzle is the once piece of the extruder that I haven't been looking forward to at all.

Good luck!
 
This is cool.
Have you tried using clay as insulator?
And btw. is there a need to make it a spiral?
I have also thought about using the nichrome on the inside of the barrel and thought that simply making it go up and down on the inside might work as well.
The danger of shorts would be near zero and mechanical stress would also be very low.
 
The interesting thing about this, if you think about it, is you could theoretically use extruded plastic as the former - pull the plastic out of a hot nozzle, let it solidify, then wrap the nichrome around the end, and ceramic-ify it in place. As a bonus, once it sets, it's already preloaded to extrude with the plastic!
 
What an elegant design! Isn't JB Weld filled with iron?

You might consider whitewash (aka CaOH aka slaked lime) and/or concrete sealant (aka Na2SiO3 aka water glass) as an insulator on your nichrome. Either would form a permanent bond with fire cement. A thin enough layer of each will leave you with a tough film of CaSiO3, which is one of the stronger components of Portland cement.

One other thing to consider: I'm not sure how I would end a double helix. Have you thought about having the wire run to a copper or brass plate with a hole drilled in it? This would give you a precisely-machined aperture for the polymer, and a way to collect the current that flows through the wire. Perhaps even an old coin would work. It would add a little thermal mass to the system, but might be worth it.
 
On second thought, forget using an old coin. I would just hammer out a scrap of wire or pipe left over from house construction.

Also, it may be worthwhile to cast the whole tube from fire cement, rather than using PEEK.
 
How about also casting the temperature sensor in the nozzle
 
I thought I recognised that face, Ed made an appearance on tonights episode of The Gadget Show (UK). Not a RepRap related appearance but Ed and his colleague Paul from Bath University developed the electronics for a "remote controlled shopping trolley". You can see the clip on YouTube:

http://www.youtube.com/watch?v=EYk2OYLXFI0
 
Great idea!

How about embedding power resistors (along with the temperature sensor) in fire cement instead of using nichrome wire? Even less fiddly work and less risks for shortcuts. The idea of making the orifice from a metal plate with a drilled hole also sounds good.

Maybe it all could be embedded inside a stainless steel pipe section if extra structural strength or some containment for the fire cement is needed.
 
God I love the internet. Thanks guys for all your great suggestions, I'll list them up and get back to you with results...
 
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?

Subscribe to
Posts [Atom]