Cast JBWELD auger bit

So I've been trying to build upon Adrian's findings for a pellet feed extruder head, found here:

[www.reprap.org]

One of his main problems was unwanted heat transfer up the drive screw, for which he recommended casting the screw out of JBWELD. I've been experimenting with doing this. I tried what seemed to me the easiest and most widely available solution, hardware builders silicone, in several variations. The addition of glycerine and a bit of acrylic paint, as well as water to wash out the additives which lead to a slow cure, had been described on a sculpture forum (http://www.sculpture.net/community/showthread.php?t=661) as leading to relatively high-fidelity cheap molds. I found these methods to be severely lacking in comparison to some other mold-making materials. First I tried alginate, which mixes with water and is commonly used in life-casting. This was difficult, as it sets quite quickly, and also hard to get the consistency right. It also sets rather delicate, much too delicate I would find, to be of any use. It was also the cheapest, at about $5 a mold, for the size wood auger bit I used (3/4"x8"). The next compound I tried was alumilite, a two-part putty, which does not degrade, sets up in about 20 minutes, and does not have to be poured, but rather shaped around the part. This did the trick. It costs $10 for the amount needed to cast a screw (I got a $20 container of it, about 1/2lb to each part). It dries to a semi-hard rubbery substance, which can be elastically deformed to a significant degree. This is important. A rigid, halving mold will not work on a screw. I had to cut a seam, and spread the mold along the seam to get the screw out. It snapped right back when I had removed the screw, returning to its proper shape. I then filled the mold with JBWELD (one package worth), and let it set up. The picture below shows the results.

In general, I was quite impressed. There were some small air bubbles, but overall the fidelity was quite good. The mold is also intact and can be used for further casting. However, JBWELD is hardly a structural epoxy. The screw can be bent with only moderate effort. Torsionally, for the amount of torque that the solarbotics motor provides, the screw is very close to deforming significantly, and we'd probably be looking for a beefier motor yet. I wouldn't count on it for robustness. The next one I make is going to have loose 1" chopped fiberglass strands as a filler. I think that will be better, but I think the ultimate solution may be a properly formed composite matrix with JBWELD as the resin and fiberglass as the matrix. This of course would be quite difficult to cast.

Any suggestions or comments would be welcome.



Edited 1 time(s). Last edit at 04/13/2008 12:14PM by Ithacacian.

Baking the JB-Weld at 200C for half an hour seems to make it a lot harder.

---

[www.hydraraptor.blogspot.com]

Nice job!

Nophead,

That was good advice. It made it a lot harder and less flexible. I would say it's eminently workable now. Still bends a little when I flex it, but much less, and torsionally, looks like there's no deformation at all at the torque the solarbotics motor puts out.

I still cast another one with some fiberglass strand mixed loosely in. When mixing it in the pot, it tended to lump together much more and formed a big ball which I had to cut with a razor blade to get individual bits to push into the mold. I think nonuniformity and air bubbles may be more problematic with this method, but I think when set and fired at 200 for a half hour, it could probably be strong enough for just about anything. I will let you know how it turns out.

Also, the JBWELD screw weighs only 30g, whereas the tool steel screw weighed 130g. It seems like this is light enough that it could be incorporated into an extruder head that doesn't weigh a ton.

You need to keep the fibreglass filaments very short, 5-10 mm max. Otherwise, they will indeed make nasty lumps and your mix won't get the tensile strength you need. I learned that with fibreglass concrete years and years ago.

Cast with a bit of fiberglass worked quite well. It had less bubbles than the first one, and it appears to be stronger. For reference, I've attached a picture of the screw bearing a 700g roll of duct tape canitlevered, with no noticeable deformation (certainly not plastic anyway). I think this is probably robust enough to handle pretty much any plastic the JBWELD can handle, with its max working temperature about 300C.

I'm building Adrian's design with his suggested modifications. It should be done in a week or so (at least, that's when it's due in 24-441). One of our improvements was to add interchangeable nozzle tips. One for rough draft mode (1mm) and one for precision (.25mm). It would be quite easy I should think to include a nozzle to extrude at 3mm, making the filament. That way you have both an extruder head and a filament-making machine. It remains to be seen if that little solarbotics motor can handle things. If not, I would recommend using a highly geared DC motor. I've used the kind found at banebots.com on several other projects. They're just insanely high-torque, moderately high-power, very small Chinese low-precision, cheap, dc gearmotors. We drew 120W and 10A stall torque out of a little one weighing 100g, for 360oz-in of torque at 200rpm. Really quite insane. I'm using a number of them in a robotics project. Wouldn't expect them to last forever, but they're a great source. If nothing else, for establishing the kind of torque necessary to get decent extrusion rates.

If you really want torque and low rpm get one of the 12v gearmotors that they use to raise and lower windows in a car. They're big and heavy but they don't draw a bunch of power and they REALLY have torque! They also don't cost that much if you shop.

Forrest Higgs Wrote:
-------------------------------------------------------
> They also don't cost that much if
> you shop.

i.e., pluck one out of your neighbor's car!

---

Jay

Bloody hell! !#$#@@#$#@@

[cgi.ebay.com]

Edited 1 time(s). Last edit at 04/14/2008 09:35PM by Forrest Higgs.

Forrest--

Only a joke. But what you can do is go to a jumkyard and ask them if you can rip a motor out. I've gotten a hundred dollars worth of stuff there for 5 bucks before.

---

Jay

Great work! Now that you've cast an auger, we know that we can cast almost any shape, even with fiber reinforcement.

There's a bunch of high temperature cements that might be strong enough for the auger. They all go to at least 1,200 C. Nophead used Cerastil H-115:
[hydraraptor.blogspot.com]

and it worked great, but unfortunately the minimum quantity that Nophead could buy was 1Kg, which cost over

I didn't notice anyone else mention this, but what you actually need to be looking at is the thermal conductivity, not just it's weight, strength, and max temps.
I emailed Adrian about this a while back and he sent me the link for the site you listed at the top. After that I started doing a decent amount of research on the subject.
Here is a list of the conductivity rating on a bunch of materials. [en.wikipedia.org] and
[www.engineeringtoolbox.com]
Look at the first column, the lower that number the better it would work for this project. Actually, that list looks a little messed up, but here is what I was looking at.
I made up a list of materials with low thermal conductivity, took out all the materials that would be hazardous or too expensive (gold), took out the materials that would melt or break (styrafoam, cork) and was left with a few decent ones such as:

PTFE (teflon) great stuff, low conductivity, not sure how to get it or mold it
Fiberglass was mentioned here and has a really low conductivity
Fireclay brick (clay is easy to cast and can be fired at any local college if there is a nice teacher or somthing)

The copper that Adrian used has a thermal rating of 401K while the fiberglass and clay have a rating of less than 2k, big difference.
I'm probably going to make a clay and fiberglass one.
Anyway, I just thought I'd give my opinion and thought someone might like those links.

Dylan