Crazy Extruder Nozzle Idea

The issue with most machinable ceramics is they are designed to be good thermal conductors. There are exceptions - but they are even harder to find!

Maybe we need a few old tiles off the Space Shuttle?

Si

Macor thermal conductivity is not too different from Glass, slightly better (Being about as good as a house brick !)

Macor being 1.46 W/m°C

Glass being 1.05 W/m°C

Aluminium is 250 W/m°C
Brass 109 W/m°C

Concrete Brick 1.4 W/m°C

PTFE - 0.25 W/m°C

But as it can withstand 800 deg C you could just use crank up the heater power... but then what do you use to insulate that ? It's getting hard to see where I'm going with this, sorry.

Machinability isn't my main concern. I really like the idea of glass because it can be drawn into a fine, precision needle-tip without having to use micro drills.

Continuing my attempts to put together an off-the-shelf extruder nozzle, the following are proposed designs and have not been experimented with yet.

All prices and part numbers are from the McMaster-Carr catalog.

Extruder A:
(1) 6710A72 Luer Lock Dispensing Tip
(2) 51465K155 Luer Lock Threaded Tube Fitting
(3) 5182K793 Stainless Steel Female compression fitting
(4) 89895K218 Stainless Steel seamless tubing [enough tubing for four or five extruders]

Est. price for one set of parts:
$2.61
$6.92
$10.44
+ $9.11
= $29.08 for an extruder

Substituting the stainless steel compression fitting (3) for brass 5272K313 brings the cost down to $23.49, but it's not certain if this would work as well.

To make twelve extruders, the cost of each is about

$2.14
$6.92
$10.44
+$2.28
= $21.78

Substituting (3) for brass again brings the cost down to $16.19


You can substitute parts (1) and (2) for screw-connection needle 7590A45.
This reduces the cost by $2.56, but the screw-tip needles are 2" long and would need to be significantly shortened somehow. I'm not sure what tools could be used to do this.


A bowden-cable extruder could be even simpler, and there are many options.

Extruder B:
(1) 6710A72 Luer Lock Dispensing Tip
(2) 51465K115 Barbed coupler for 1/4" ID tubing (Not sure about the ID of the coupler; perhaps 5194K35 is better)
(3) 5321K14 Worm drive hose clamp (pack of 10)

Est. price for one set of parts (not including the bowden cable):
$2.61
$6.60
$6.48
= $15.69

In larger quantities, the cost of each set of parts drops to $9.39.


The idea here is to widen the last few centimetres of the bowden cable to accept the barbed coupler. Then the bowden cable fits directly to the barbed coupler, and the plastic filament fits through both. The heating coil would need to be put around or inside the coupler.

The downside is that this requires a larger diameter Bowden cable, which could be expensive, plus the widening step. An alternative to the barbed connector is another compression fitting. In this case, we'd want a setup like this:

Extruder C:
(1) 6710A72 Luer Lock Dispensing Tip
(2) 51465K155 Luer Lock Threaded Tube Fitting
(3) 5272K313 Compression Tube Fitting (or 5272K281 for 1/4" tube)

Est. price for one set of parts (not including the bowden cable):
$2.61
$6.92
$4.85
= $14.38

Here, the Bowden cable fits directly into the compression fitting, which screws to the luer lock fitting, which mates with the needle. Simple to assemble, hassle-free bowden cable design.


For the Bowden cable, tubing such as these Teflon varieties should work:

52355K12 ($1.84 per foot)
51805K32 ($2.06 per foot)
5239K11 ($1.09 per foot)


For a multi-head system, perhaps one of these colour-coded, semi-clear PTFE tubes would be appropriate:
53935K221
53935K222
53935K223
53935K224
53935K225
($1.25/foot)

... i used hollow glass-microspheres in waterglass as air hardening paste with good thermal isolation capabilities.

The mechanical properties -when hardened- are similar to styrofoam, but it can withstand 600 centigrades, until the glass spheres start to melt ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

There's also Mica and Micalex. But like I said, I think using machinable glass mostly plays against the advantages using glass...

A more elaborate Bowden cable extruder could use one of these to join the PTFE tube to a stainless steel tube. This would also let it screw into a mounting plate (up to 0.5"/12.7 cm thick). For a thinner plate, that also provides room to file a hole and install a solenoid valve to prevent oozing.

5272K504 Through-wall coupling $8.90

Which would make a complete Bowden-cable extruder from these parts:

Extruder D:

(1) 6710A72 Luer Lock Dispensing Tip
(2) 51465K155 Luer Lock Threaded Tube Fitting
(3) 5272K313 Compression Tube Fitting
(4) 89895K218 Stainless Steel seamless tubing
(5) 5272K504 Through-wall coupling
(6) 53935K225 PTFE colour-striped tube (let's say, 4 feet)

Cost for one:
$2.61
$6.92
$4.85
$9.11
$8.90
+$5.00
= $37.39

Or $30 each, when obtained in larger quantities.

The whole extruder could be disassembled and re-assembled with ease, because all of the components fit together with standard compression fittings. And the tip can also be easily exchanged for other diameters.

I'm starting to think this could be quite an ideal setup as an off-the-shelf extruder. All it needs is testing.

On bowden cables.

I have had cause to make up brake lines for my land-rovers and can not help but think of borrowing some of the techniques for extruder construction.

If the extruder barrel were to be drilled out to the diameter of the bowden ptfe tube. and a small internal chamfer applied to the nozzle end this would make it possible to flare the end of the PTFE tube and clamp the tube between the barrel and the nozzle/heater.

The flared end of the teflon tube would make a good heat proof washer and securely anchor the tube.

The seal should be excellent and the ptfe lining goes all the way to the hot bit helping to reduce jamming.

If the extruder barrel were made out of PEEK and the nozle end were threaded you would also have an excellent thermal break.

Replacing the tube is fairly easy this way.

I guess it would be nice to find a break cable sleeve that had the correct inside diameter to take the ptfe tube and reduce the stretch that contributes to ooze.

Thoughts for what they are worth.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

Do you have a drawing of the idea? I'm finding it hard to picture. But it sounds interesting!

I have a diagram and if you want it your gonna have to pm me an email address.

I have wasted enough time arsing about with firefox's reluctance to upload files for today.

Life is too short and there is much to do.

I have 3 machines here all running Ubuntu and using Firefox. Of late any attempt to upload files using Firefox gets me a server not responding error. But windows PC's and IE are working just fine.

Cheers

aka47

---

Necessity hopefully becomes the absentee parent of successfully invented children.

Here's AKA47's illustration. I'm going to try something similar, though not exactly this - although it's a good idea, I want to first see if I can get away with using just off-the-shelf parts. Then maybe I'll try it next.

So I've put in an order to McMaster-Carr to try testing this, a luer-lock extruder, and also some bowden cable ideas. It was an expensive order, but it's for the sake of science. I'm excited to see how they work out.

Here's what I got:

Glass tube extruder:
5949K61 - x1 - vibration resistant metric compression fitting, 6mm - should hold glass firmly but with a soft touch
5305K301- x1 - Metric reducing compression fitting, 6mm x 1/4" - potentially a good adapter between a 1/4" PTFE bowden cable and a 6mm glass extruder nozzle. A bit pricey though.

Luer lock extruder:
6710A72 - x3 - 0.4 mm stainless steel luer-lock nozzle
51465K155 - x1 - Brass luer-lock adapter for 1/8" NPT thread
5272K281 - x1 - Brass 1/8" NPT adapter to compression fitting for 1/4" bowden cable

Bowden cable tests:
5033K31 - 3 ft - 1/4" PTFE bowden cable tube
8503K245 - 1 ft - 1/4" PEEK rod (I'll have to drill out the interior to 3/16").
89895K225 - 1 ft - 1/4" stainless steel tube
53935K225 - 2 ft - 3/16" PTFE narrow bowden cable tube tube

Here I'll try to see if a cheap and flexible 3/16" PTFE tube, with the end placed into either a stainless steel or PEEK 1/4" exterior for mounting to the print head, can replace the stiffer 1/4" bowden cable. My hope is that the flexible tube will still support a load over a short distance, without putting as much drag on the print head. The last 5 to 10 cm or so will be bonded to the stiff insulating tube, and squeezed into a 1/4" compression fitting.

Looks like a great idea. Is there any way I can help? What was the combined total after shipping?

I won't find out the shipping total until it arrives, but pre-shipping total was $74.76 US. The most expensive part was the 6mm-1/4" reducing coupling at $16.55, because it's made out of type 316 stainless which is always quite expensive. When I test the glass nozzles, I'll check if the 1/4" side of the coupling will work, since 1/4" is just 6.35 mm, and I don't want to compress the glass too tightly anyway. Then I could get away with a 1/4"-1/4" coupling, which are available in much more variety and much more cheaply, for example, $3.34 for 51745K201, or $2.10 for 5220K23.

Come to think of it, I could also try wrapping a few turns of kapton tape around the glass tube to get it up to 6.35 mm, and then it'd fit for sure... with a cushion to boot. That might have been a much better idea. ;-)

Edited 2 time(s). Last edit at 08/05/2010 07:14PM by jbayless.

An aside

The idea of using a flared PTFE/Teflon tube as a liner in a peek barrel (thermal break) is also doable on a Wade style extruder. Not just a Bowden Cable rig.

Just cut the tube short enough to take the feed from the pinch wheels.

Got some more ideas on pinch wheels yet, but these can wait till later.

Cheers

aka47

---

Necessity hopefully becomes the absentee parent of successfully invented children.

Nice graphic, aka47. But I am concerned that the teflon bowden cable will still be too soft, and after flaring it to fit the PEET, the back pressure will push it out. JB, I am also looking into long thin needle extruders to handle building a large part in stages. When you fill the gap between a completed section that may be an inch or two thick to the layer by layer build of the new section, you will need and extruder that can either work right next to a 'cliff' where the existing part ends and the first layers of the new on begins, or a long enough needle reach down past the existing piece. Most extruder nozzles only work with a cliff of a single layer, not up to 100 layers.


This can be a needle like extruder.


Or it could be tilted to one side and chamfered to fit.

I have the first few sample glass nozzles without any embedded wiring. I think I need to try and rig up a test extruder using my original PTFE to brass welding tip nozzle, and an all glass with taped on wiring nozzle, and see which requires more force to get roughly 1 mm³ per second (I think that is the current design rate?). The simple glass nozzles, without embedded wiring, are turning out to be quite easy to make. I think I can mass produce them in quantity for something like $5 each delivered. And they should be good up to about 1000C.

Why would you need to build next to a cliff - wouldn't it be better to print those layers before the cliff is printed?

Also, in aka47's design, the bowden tube is squeezed between the brass nozzle and the PEEK block, so I think it's unlikely it could slip out. I don't have too much experience with PTFE though, so I might be wrong.

Ah, I missed the brass in the graphic. I looked again and it makes more sense now.

The reason for the cliff is that I am considering making the major structural parts of my Gada prize entry on the machine itself. Of necessity, they will be much bigger than the build table. So my plan was to make the first half of the length, then move it almost completely out of the build volume, clamp it down, and build the second half starting at the edge of the first half. That is why I expect to have a large difference in layers. The first half will be completed, maybe a 100 layers or more thick, cold, solid. The new stuff will start off just one layer thick, hot, hopefully gooey enough to permanently bond to the original, cold part, and then build up from there.

My apologies for the oversized graphics. I will shrink them later when I have some time.

Mike

Hm... Is it not possible to use fasteners or adhesives? Or print snap-clips to hold the parts together? I think it'd be hard to move a part on the print head and still have it aligned precisely enough for the machine to know where it is.

BTW, here's a crazy extruder idea: A big "paintbrush" extruder head for doing rapid infill of large parts.

I was thinking of having three different nozzle sizes, either as 3 separate extruder heads, or better yet like the three lenses on a microscope that rotate into position. One perhaps slightly smaller than the current 0.5mm for fine detail work, a 1.0mm for building up edges, then a 2.0 that should shoot plastic like made to fill in 'between the lines'.

As for your first question, ABS glue or good fasteners might work also, but snap together would not be strong and rigid enough for the main structural members. Think in terms of a Mendel with the threaded and smooth rod replaces with thicker pieces of RP plastic.

Mike

I like the 'microscope style' rotate idea, but you still have the problem of sealing it shut after rotation.

Does anyone have a removable Hot-end design? some sort of screw-fit?

I always wondered why you can't just build a framework in ABS and then have another extruder (2-3mm nozzle) that just squirts out extruded 'hot glue' let it flood the inside of the part, settle and cool a bit then cap it off with a top layer of ABS? It should be much faster and quite strong.
I guess that's also asking for a warp-problem, or maybe not? is the hot core working with us? It would be neat to try it manually with a glue-gun, anyone want to have a go and let us know if it's a disaster or not.

Sorry it's getting a bit off-topic.

I can't wait to be experimenting, hopefully I will get my motors and electronics next week. And another thing on my list to try when I get my own kit working.

A problem I see is controlling the "flooding" of the interior. It's easy for a human, but not necessarily for a machine, unless some thought is put into it.

For example, if you pour infill/hot glue material into a part and it doesn't have a smooth, level top, then when you try to print the top layer of ABS, it won't have a level surface to print on, and the print head might crash into a lump. If the material is very low-viscosity it could work, but not many materials are.

For a very fat nozzle (say 2 or 3 mm), there's the issue of the filament being a cylinder. Ideally the extruded infill material is a rectangular prism, so that it's space-filling. But for a round nozzle, the extruded material will be several cylinders packed together. When the nozzle is very fine, this is not a problem, but for a wide nozzle the height deviation becomes very large, and the packing efficiency quite low.

If you try to solve the problem by moving the nozzle close to the print surface, so that the extruded material gets smears out, then there's a situation where more material is being pushed out at the center than at the edges, which I believe would put more drag on the material and reduce print speeds and quality.

That's why I think using a narrow but broad extruder tip (ie, taking a 3mm diameter steel tube and squeezing the end in a vise around a thin rectangle) might be a better approach for rapid infill. It would print at the same layer thickness as usual, but with much broader strokes.

The downside is that it could only move in the direction perpendicular to the nozzle, and it would leave gaps. The fine nozzle would have to fill in the gaps where the corners are. The parts couldn't be cross-hatched, which might make them prone to failure in certain directions. There's four solutions that I see:
- Stacking the wide infill layers like bricks, instead of cross-hatching them, might give an acceptable level of strength.
- Having two infill nozzles perpendicular to each other would easily solve the problem, but it would require two dedicated extruders for this.
- Having a motor that can rotate the nozzle would be an ideal, but complex, solution.
- Alternating between broad-stroke infill layers in one direction and thin infill layers using the fine-detail nozzle in the perpendicular direction would work very well, but then the prints would not be much faster than in the first case.

I would eventually like to try setting up an infill nozzle that extrudes a mixture of maybe 40% hot glue and 60% fine sand, instead of plastic.

The bowden cable discussion here strayed pretty close to some of the work I did on my mini mendel before I decided to scrap the cable and modify my Wade's to fit the x carriage, so I thought I'd share my findings.

I worked for some time on using the standard barrel insulators to hold my 1/4" OD PTFE tube. I had an idea to thread the tube (dies don't work, but turning an M6 nut onto the tube does) and screw it directly into the insulator. This worked well, and for a while I was using PTFE insulators on both ends to hold the bowden cable. One attached to my Wades in the standard way, the other with the entire length drilled and tapped to an M6x1.0mm thread. This setup failed and ruined the modified PTFE insulator when I accidentally ran the extruder motor while the hot end was't living up to its name, but it was the BRASS HEATER BARREL, not the PTFE tube that stripped the insulators threads and popped out. This means that screwing a 1/4" OD PTFE tube into M6x1.0mm threads is quite strong.

After ruining the PTFE insulator, I tried replacing it with a PEEK insulator, but because I didn't have a lathe/drill press to do the drilling, I ended up drilling the hole at an angle, causing the filament to catch on a lip when trying to transition from the PTFE tube to the brass heater barrel. So this was a failure, but possibly only a failure of my machining capabilities.

A useful technique for getting barrel and insulator alignment correct is to drill tap and fit the barrel into the insulator.

Then drill the feed hole using the barrel as a drill guide.

If you then want to try the flared end on the bowden cable use the feed hole in the insulator as the pilot to drill it out large enough for the bowden tube. (remove the barrel first though.

incidentaly because the cutting face of the drill bit is angled this may do as the seating for the flared end of the tube.

Questions on welding tip nozzles. has anyone used these with any success.

0.6 mm seems to be a big hole for what we are doing.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

Can't speak for MIG tips (although just yesterday an engineering PhD friend of mine suggested using them too). But the extruder nozzle currently used by the SpoolHead team was probably closer to 0.9 mm. It still produces acceptable prints, but much less fine than the parts we got from Wade (who uses an 0.5 mm nozzle I believe). We really desperately just needed to get the printer working for our school project, so that served our needs. But now I'm keen to replace it, which is one of the reasons I'm in this thread! =)

is wades nozzle a welder tip or something else ??

---

Necessity hopefully becomes the absentee parent of successfully invented children.

I'm pretty sure it's something else. Probably a finely drilled brass dome nut. But Wade would know better than me. =)

I have an extruder made with a 0.6 welding nozzle: [hydraraptor.blogspot.com]

I had to drill out the hole to almost the end because otherwise it has a long thin hole at 0.6mm, which causes too much resistance.

You could make the hole smaller by filling it with HMP solder and drilling that with a fine drill. I used that technique here: [hydraraptor.blogspot.com]

---

[www.hydraraptor.blogspot.com]

I'm planning on using MIG tips for my first go...but even though they have .6 stamped on them.. that is not the hole size...they are a sliding fit for .6mm welding wire... thus are quite a bit bigger than .6mm (mine are bigger then .7mm)

I Tried squeezing the end (before cutting it off and drilling out the stub) in a vice, but didn't manage to decrease the size of the hole by much... got it down to .6mm I think, after rounding it out by squeezing around the tip.

I have an easy out here... I own a TIG welder, and can just melt the end over. But I'll try making a cone shaped mandrel and decreasing the hole by driving the end in, or rolling it somehow in the middle to stretch it before cutting it off to length. They are cheap and easy to get, so experimentation is encouraged. They have 6mm thread on the top. So work well with a resistor block. (just tap the 6mm THRU)

Al...

---

[araspitfire.blogspot.com]

hmmm shame about the welding tips. I am currently suffering the urge to have a mess with inductive heating and was looking to take advantage of a small off the shelf nozzle. Rather than make one.

the nozzle could be almost completely sited within the peek insulator then. Leving just enough for clearence and to get a grip when tightening it down on the flared end of the ptfe tube.

i guess i am looking at applying nopheads principle of minimum hot zone with shortest transition zone.

a quick guesstimate sugests 12 to 24 of 0.5mm high temp magnetgwire (copper) should be enough.

the nozzle itself is actualy a shorting turn secondary.

it could be novel and greatly reduce the extruder size.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

Here is my 2cents. BTW I don't have a RepRap yet and I am still learning and understanding the problems.

There are already Brass Tips and brass ferrules made from 6mm Hex available by some members. If the inner end of the brass ferrule (the insulator end) was flarred (as it is) and the outer surface was reduced to a round ferrule to allow the PTFE tube (I must check but I presume this is also Teflon) to slide over the reduced brass ferrule, say maybe 1/4" in length, might this work. We still have the insulation issue to deal with, but would this solve the nozzle problems you are all seeing???

As a test for the above, it could just be inserted into a wooden dowl. Do you see any problems with this idea???

Hmmm shame that.

I guess if I need to mess about drilling etc I may as well turn one out of an M6 bolt or machine screw.

Cluso99, sounds as doable as just flaring brake pipe style so equally good.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

dear sir

i have a problem in extrusion brass products me manufacturing the brass auto stem valaves for scooters, trucks, and motorcycles now i have a question that i made all products from brass extruded rods only the customer said dat we got ur product wid leakage problem i dnt know how it be come in brass extruded rods is dat true ???? leakage problem common in extrusion or i need ur some guidence for this we are just doing turning and cutting from that road ?? why needed leak testing ? and how it will b come in our rods ? any have u idea or suggestion please send it to me soon


thanx dhairya