My granule hopper concept.

Having nothing to do for a few days really does make the brain come up with some good ideas.

Here is a link to the hand drawn concept.
[cjganttphotography.com]

First I will explain the parts. Each number corresponds with its respective arrow on the image. I will try and get a technical drawing up after some discussion.

1. Archimedes Screw

2. Hopper

3. Screw shaft

4. Top skate bearing holder

5. Second skate bearing holder

6. First Nichrome wire wrap

7. Seconf Nichrome wire wrap

8. Extruder tip

9. Just pointing out the support for the second skate bearing holder

The whole concept is based on the way an Archimedes screw (known from here out as AS) [en.wikipedia.org] moves materials. It can do two things.

A. Lift materials up.
B. Push Materials down.
Here is an animation on how they work. [upload.wikimedia.org]

In our case we will focus on the AS pushing materials down.

The hopper would be filled with plastic granules and would be forced down the tube by the AS. When the granules "fall off" the last part of the AC then the back side of that last flute would intern push those granules down into the granules creating some pressure. Heat zone 1 (arrow 6) would melt the plastic to fluidity. Heat zone 2 (arrow 7) would keep the plastic at the tip liquid and allow it to flow better out of the tip. This would require 2 controller boards but they could both run off of the same input channel from the main board.

The top of the conical hopper would have a cross brace with a bracket mounted (or printed) in the center to hold a skate bearing or other bearing to reduce the friction caused by simple metal on metal bearings. Close to the top of the AC there will be another cross brace with a bearing holder in the center. This will provide further support for the AS drive rod.

I propose that from the top of heat zone 2 all the way to the tip be 1 removable tip. That way we can have several tips machined to expand our options on extrusion thickness.

The only draw back to this design is that the print head would have to be stationary and the X, Y, and Z axes be movable.

Materials:

The hopper could either be A. Printed or made from some simple aluminum flashing.

The AC could be made from combining a drive rod with circles cut out of thin gauge aluminum. You would then punch a hole in the center of the circle the same OD as the Drive rod. You would then cut a straight line from the side of the circle to the center. Bend one side of the cut up and the other down then slide the whole thing onto the bottom of the rod and braze it in place. Repeat until you have enough to fill the extrusion tube with about 3-5 flutes of the AC and still have 2-3 flutes sticking above the extrusion tube.

The rod supports could be made of aluminum angle or square profiles.

The tip and extrusion tube could be machined out of brass or copper.

So what do you guys think? In theory could it work?

Your sketch is pretty sketchy, but I can see one thing you may have missed. You are trying to force granules down into the heated zone, right? Doing that is going to create a thrust in the opposite direction where your gearmotor is. How are you going to absorb that thrust. If you let it go into your gearmotor, it will destroy it sooner or later.

Mind, this is a common problem with even the cheaper sort of commercial plastics extrusion machines. They use standard gearmotors and don't last too long before the reduction gear boxes have to be replaced.

Some sort of safety clutch?
like those: [www.maedler.de]

But what I'm wondering is, it looks quite similar to the one that has been tested a while ago.
There was a problem with heat conduction up the screw clogging the entrance and stopping the flow.
the link with the result
[reprap.org]

So I think we need the screw to be made of ...hmmm duroplast?
I have no idea!

'sid

Edited 1 time(s). Last edit at 12/12/2008 11:03AM by sid.

sid Wrote:
-------------------------------------------------------
>
> So I think we need the screw to be made of ...hmmm
> duroplast?
>
Stainless steel. It's a terrible heat conductor.

---

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

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

I've actually thought of this after flipping through my 111 edition copy of McMaster-Carr (which only mentions "minature cord" in only ONE type of plastic, and is NOT FDA approved, bleah,) and I actually do agree that early heat conduction is a major issue for feed hoppers. Not that it's a bad idea--far from it--but it's hard to crank out multiple runs when it clogs up in between cooling periods.

My suggestion would be to think about adding some distance between the auger screw end and the first heating element--less chance of killing the auger that way, especially if you used a design that stops feeding in such an emergency (such as that blind hole shaft/auger design.) But you'd still run the risk of clogs, IMO. Perhaps a "clear run" with a removeable hopper/auger disconnected to empty out the extruder head?

How do you set the flow rate with an auger?

I could be wrong, but is there a way to calculate plastic flow rate directly from auger rotation speed? I thought there's a lot of interplay between plastic oozing out, air dispersing, and the friction effect between the molten plastic and the walls of the screw chamber; ie, if there wasn't any wall friction, you get no plastic flowing at all, just a helix of rotating melted plastic.

The Archimedes screw depends on the auger being only half full, and the friction between the fluid being pumped and the walls/screw being low compared to the weight of the fluid. If the Archimedes screw is completely filled with water, it becomes a series of siphons.

Commercial injection moulding processes just use tons of pressure and too much plastic, and don't worry much about the exact flow rate; they just let extra plastic flow out the other side of the mold.

Getting good results out of my Darwin required getting predictable plastic flow rates; that's one of the major advantages of the original extruder; as long as there's no slippage, things are good and you get exactly 0.8 mm of feedstock for every turn of the drive screw. At the moment, I'm actually getting a bit of slippage on my worn down extruder, and it means that the density of my printed parts fluctuates all over the place, making the built parts pretty useless.

Wade

Edited 1 time(s). Last edit at 12/16/2008 03:22PM by Wade Bortz.

Wade is exactly right here.

Commercial injection molding machines use variable diameter augers. On some of them they have grooves cut into the ID of the feed barrel. From the book "Plastics: Product Design and Process Engineering" by Harold Belofsky

"In a grooved feed barrel, the first two to five turns in a the feed zone have uniformly spaced shallow grooves machined into the ID of the barrel. This increases coefficient of friction between the solid bed and the barrel, which increases driving force for solids conveying. Grooves may be parallel to the barrel axis or helically inclined. They taper uniformly so groove depth is zero when the barrel reaches the transition zone."

However, these aren't strictly necessary usually and they do come with limitations...

"There are some important limitations with grooved feed. Becasue of more efficient solids conveying, pressures of 70-140 MPa or greater can develop; the extruder barrel must accommodate these higher pressures."

There are other things but I'm getting tired of typing! ;-)

Demented

How about pressurising the whole thing? The auger would then only have to deliver granules to the molten pool of plastic which would be forced out of the nozzle by air pressure. Maybe only feed granules when the pool of plastic reaches a lower limit so there is no conduction path from the molten pool to the granule store.

The obvious problem with this is weight and possibly pressure - I'm not sure how much would be required.

Actually, thinking about this over the past couple of days, you could just use the hopper auger to pressurize the molten plastic, and if you've got a pressure transducer that can handle molten plastic, you can just vary the speed of your auger to keep a constant pressure.

You've still got to map out the pressure and temperature vs extrusion rates per nozzle, but it's doable. Pretty similar to what Adrian's cooking up with his granule extruder, actually. Still, it's a rate measurement vs distance measurement, so you're effectively guessing at the integration constant C; small errors tend to build up over time. A nozzle valve simplifies things a bit though, at least you can nail down your zero flow rate.

The trick with all this stuff is to build it and see if it works. I'm up to my eyeballs in great ideas, but I can't get my reprap to extrude a new extruder at the moment without something tangling or breaking. Hopefully it's a temporary situation.

Wade

There is some good information in this thread. I plan on mocking up an auger and a barrel in the near future and testing my theory. It will take some time to finish as I will have to source out a machinist to turn the heat zone 1, the barrel and the tip. As soon as I get that done it should be as simple as making the auger, forming the hopper out of thick flashing, and designing and building the auger drive rod mounting brackets. Of course my mcwire build and darwin build will take the front burner position, but good things come to those who wait right?

The books I have mention something like that, Wade. They go over the equations governing the system but, as is often the case, we can write down the differential equations much more easily than we can ever solve them. They said that, in practice, a machine is often just built and relevant data is extracted from actual test runs to build up a performance table of how things work under given conditions.

Pain but better than doing the math!

Demented

I was thinking about this some more last night. The auger would have to be 2 inches or so from the first heat zone. With that known we could machine the 2 heat zones (hz 1 barrel diameter and hz2 necked down for the nozzle) out of copper. Then we could machine the barrel out of a poor thermal conductor like stainless steel or even carbon fiber. We could use a heat resistant paste (welders use this to control the heat warping on thin sheet metal) on the threads where HZ1 would thread onto the auger barrel. This would cut the heat transfer down to a minimum. All of this is just thoughts from my head. I have a business degree and an arts minor so none of this is engineering based. Its just simply based off of my experiences making stuff my whole life.

I didn't realize we were talking about machining stuff. If you are going that route, I have plenty of drawings of production extruders down to a very small size. No need to do extra work.

However, the point is making an extruder that you don't need a machine shop to produce. The average joe has to be able to make it, buy it, or print it himself. Otherwise it's just like so much other technical ability...limited to the elite.

Demented

I just do not see the extruder nozzle (the actual hole) being accurately centered in the tip with out proper machine set ups. For the average joe a traditional extruder like the ones we are using now would work just fine. A granule hopper is not something most people would need. It would be beneficial to those who set their machine printing a spool of parts and want to walk away for several hours to a day and just let it print and not worry about running out of plastic.