Recycled Deposition Modelling

Hello Community!

I can use your quality feedback on the following project!!!
PROJECT PAGE
check my project about the devopment of a granule extruder... ultimately to be able to rint with shredded plastic waste chips.


As you might have noticed their are several different projects focussed on the development of such an extruder, I'm currently in the stage that the project will be lifted to a more professional level. (examining proper literature (polymer extrusion by Chris Rauwendaal), collaborating with professionals (DSM, Anatech), precision machined prototypes (at Zuyd University, Jan van Eyck Academy))

short project history:
-Today
-Research and development of first extrusion stage. primitive cold test done with wallpaper paste, promising, new prototype will be manufactured and tested
-Commercial electronic controlled pressure transducers are to expencive (+-$300 (RS online)) self developing one controlled by arduino and standard servo and miniture 3/2 valve.(not yet finished)
-Concept choice made: 2 stage extrusion printer head (not seperately like the filabot project). first by an inexpensive screw/auger> second with the use of accurately controlled air pressure. Two stage, because of expected start stop complication when only using an auger (based on research)
-research done: see project page
-first objective, there is no printer head developed that can print with granules
-initial project goal: to develop cheap and locally manufacturable Big reprap which would be able to recycle plastics

Please post your feedback below, keep it on topic please..

Thanks in advance!
Ward Hupperets

Hi Ward,

how did you seal the bearings?

---

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

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

good question:

not yet sealed, First test was just a try out.

when I will perform the second (cold) test with wallpaper paste I will use an O-ring around the blank shaft of the screw. (I will grind a little round grove in the shaft so it will be axialy fixed)

in the new prototype I will need to have a proper seal, do you have a proposal?

... no, not really - the problematic sealing is the main cause for the common hopper designs, where the bearings are outside the heated and pressurised areas ...

---

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

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

Wouldn't it be simpler just to put the drive above the auger so that it doesn't need any sealing?

I have been thinking about this kind of thing and wondered if having a slightly narrowing clearance around the auger (so in a very slightly conical hole) would help with feed and getting the material to pack-up close together as it melts and the air is excluded.

Very interested to see how it goes, good luck.

George

In the engineering of injection molding, a very narrow channel is good enough as a fluid seal.

since polymer melt is a high viscose fluid it will not flow trough these narrow channels, beceause to much pressure would be required.
between the two mold halves these channels are manufactured to release the air in the molt cavity when fluid is forced in. this works good if properly designed. (0.02 mm range)

If I would manufacture little ribs at the inside of the channel which are just a little smaller than the screw chaft, and force the screw in, the little ribs wear in and you get a proper sealing... (at least for the first prototype it will do the job wouldn't it?)
I designed it in a prototype I printed, but when manufacturing it in alumium it did'nt came out that way. (see picture below)


what do you think of this?

Edited 1 time(s). Last edit at 04/04/2012 07:58AM by Ward Hupperets.

How do you propose to work around or compensate for the massive variations in plastic quality you will encounter by using recycled plastic? What about the degraded physical/chemical properties that regrind has compared to virgin plastic? Have you looked at how these issues will impact print quality, and have you done a cost-benefit analysis to see if the increase in cost, complexity, weight, and plastic quality control issues are worth the benefit of being able to use recycled plastic?

I'm not saying don't pursue this idea. I've seen this sort of concept proposed in the past by people who were ignorant of the pitfalls associated with using recycled plastic, so I'm trying to do 2 things: see if you understand what you're getting into, and if you don't steer you in the right direction of things to look at.

Technically workable is not the same thing is technically feasible.

Pretty neat idea, Ive often thought about it. I just got started with building my afirst printer so something like this is a ways down the road. I will be watching to see how this turns out for you

---

Ryan
Quality Engineer & Hobbyist
thingsandtrains.blogspot.com

@crispy

At first, I will only focus on the use of vergin materials, so I dont have trouble of degraded plastic etc like you say. I'm aware of these problems.

The goal of recycling material is not a realistic achievable goal beceause of the problems your talking about. Initially this was the goal of the project, but I left this goal in the second week, since it requires a properly working granule extruder in the first place, which can be used to test with for other non vergin materials.

The title of the project is still RDM though, something that cant be changed, and reflects the original goal of the project (9 weeks ago), title can't be changed though.


Although I respect your opinion about the matter, now the only task left is to engineer and manufacture prototypes of a working granule extruder. based on work en experience already done by others in the community and commercial world.

Being able to recycle belongs to the future until than.

Edited 1 time(s). Last edit at 04/04/2012 09:37AM by Ward Hupperets.

@gsport
I guess the problem with drive above auger, is that you need a structural support, which ads to the total weight of the printer head (since the forces on the auger are pointed upwords)

"reversed" extrusion direction, solves makes the construction way more simple, but the problem of sealing pops up.


CAN WE TACKLE THE PROBLEM OF SEALING??? ANYONE?

Edited 1 time(s). Last edit at 04/04/2012 09:51AM by Ward Hupperets.

... as you apply heat, pressure and tumbling to the molten plastic, the sealing is a really difficult thing.

You need extreme high precision for the auger-shaft and the throughput - your 0.02mm is only possible with bearings on both ends or a longer shaft and aditional fixture under the drive.

Any inhomogenity withing the squeesed plastic will force the auger to bend some ten microns, what will wear off the sealing with time, so the slit will widen and leak.

I think, you can't seal the throughput over longer time ...

---

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

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

VDX Wrote:
-------------------------------------------------------
> ... as you apply heat, pressure and tumbling to
> the molten plastic, the sealing is a really
> difficult thing.
>
> You need extreme high precision for the
> auger-shaft and the throughput - your 0.02mm is
> only possible with bearings on both ends or a
> longer shaft and aditional fixture under the
> drive.
>
> Any inhomogenity withing the squeesed plastic will
> force the auger to bend some ten microns, what
> will wear off the sealing with time, so the slit
> will widen and leak.
>
> I think, you can't seal the throughput over longer
> time ...

Ill contact sealing specialists to ask for suggestions

I like this idea, and have been needing to get round to trying this myself, but havent had the time recently.

Just a quick thought, but cant a simple homemade injection-moulding type extruder be used with a die? I mean it may take a few passes through a set of these to get a decent enough uniform cross section, but this could potentially be done overnight in large batches.

And who else saw the video that guy did a while back on recycling hdpe milk bottles into strips for his hydroponics? That wouldnt be hard to reproduce. Setting up a couple of them with dies of decreasing diameter from say, 10mm down to the generally used 3mm or less. Gimme some time and i may just try this out.

Ollie

Ward Hupperets Wrote:
-------------------------------------------------------
> -Concept choice made: 2 stage extrusion printer
> head (not seperately like the filabot project).
> first by an inexpensive screw/auger> second with
> the use of accurately controlled air pressure. Two
> stage, because of expected start stop complication
> when only using an auger (based on research)

So, how is this second stage going to work?

If the actual extrusion is going to happen by forcing the plastic out of a chamber by air pressure, it's going to need a complicated valve setup. I'm counting at least two air valves and two valves capable of handling molten plastic.

Because the auger isn't a progressive cavity pump, you'll have to seal it off from the extrusion stage with a valve while printing or the air pressure will force the plastic back up along it. It will have to be cyclic, and during the fill cycle the printing nozzle orifice will have to be sealed off with another valve or it will spill plastic while the auger is refilling the second stage.

I would just run the auger against the orifice valve instead. If that isn't accurate enough, at least you will have developed two important components, a working auger and the orifice valve.

Sure putting the motor at the top will require a little more structure, but structure is easy and "cheap" compared to sealing and you can shift the hot end to be directly below the auger instead of it needing to be off to one side, this means that your heating will be simpler (I assume you will be heating the auger cavity as well as the hot end? They could even be one part maybe?)

Whatever sealing you use the worry will be that plastic will get into the structure above it and gum up the mechanism and deflect the seal.

If you are worried about thrust loads on the auger, then you should probably use a thrust bearing on there as well as the deep groove you seem to have. Personally I dont think it necessary but would always advise using deep groove bearings in pairs to provide proper location.

Also I would suggest making the auger axis slightly off vertical so that it can be retro fitted to existing printers and stick out the front or back without hitting the top member (on say a Prusa) and restricting build volume.

@gsport
do you think the sealing problem could be solved by introducing a counter twisted helix groove inside the blank side of the screw, below the hole? In that way, material that passes the hole will be transported upwords again.

I chose for reversed extrusion direction beceause a thermal barrier at feeder and hopper end will be easier to achieve, when using an overhead extruder this will be more complex to achieve, beceause the thermal barrier must also carry the load of the screw thrust etc... and you will get a less compact design

I will proceed with this setup anyway. I think it has good potential, for several reasons, we can Skype about this perhaps?

Well, it looks like you have already made your concept choice, but I think a seperate filament maker would be much more useful.

@ttsalo

your proposal is good, as you might have seen in the last prototype drawing I made on my project page, it only exists of a screw housing and a screw. with a little hole of 1 mm diameter.

exactly for the reasons you talked about.
maybe with the goal of a two stage extruder I want to achieve to much at the same time.

Ill be producing the last prototype design next tuesday, but I might make some changes based on literature I'm now examining at TU Delft, called polymer extrusion by chris rauwendaal
and also changes based on the choice of sealing

@bobc

we allready have the filabot project for that...
plus, being able to pressurize the polymer melt in a different way than just pushing on the melt with cold filament, creates new possibilities, like using higher molecule weight materials (the higher the better for FDM) and faster extrusion thus faster printing speed.

Edited 1 time(s). Last edit at 04/06/2012 05:20PM by Ward Hupperets.

The reverse thread might well keep the material out of the base but it still seems like a lot of complication to me.

What do you mean by thermal barrier at the top?

Also I assume that all the material needs to be fairly melted by the base of the auger? So you have to heat the entire channel from the auger to the nozzle?

To be honest I think I am missing something in your concept. There seem to be all these extra holes etc and you want to use pressurised air to help feed material but where does this air go?

Also with the Auger so far from the tip how will reversing work? ("the extruded output must be fast controllable, so complex geometry with a lot of start stop movements can be built.") and the longer the channel between the auger (if that is what is controlling the flow of material) and the tip the worse this will be surely?

George

@gsport

allright guys... I realise it is realy hard to get a good picture of the project development so far. Ill explain it short:

first stage:
- Polymer granule is fed into a cold hopper at the top in atmosfere conditions. (continiously fed by a hose or something else)
- The screw in the barrel with the tip pointed upwords and the hex head at the bottom rotates, and so transports the granules down the barrel. (this barrel section needs to be cold)
- the granules come in the melting zone, where the barrel is hot and the core diameter of the screw is increasing. (the standard hex woodthread screw is post machined to get this geometry)
- the melt is gradualy forced downwords and pressure rises due to the constantly added material above.
- The melt pressure is high (5bar) and is forced through a narrow channel which ends up in a buffering chamber this is where stage two takes place

second stage:
- polymer melt is collected in a small chamber, where trapped bubbles can exit the melt. In the buffering chamber the melt has a fluid level, on wich gas pressure is being put on. The gas pressure is variable, liniarly with the demanded extrusion speed for printing.
- the air pressure in the chamber is controlled by a servo controlled valve wich has three hoses connected, a pressure input (continious pressure from compressor), an output (pressure release to atmosfere), and the chamber connection.
- this way the pressure in the buffering chamber can be controlled fast and easily by throttling the input pressure, releasing pressure, or keeping the pressure constant.
- in the second stage the pressure in the chamber is monitored by a pressure sensor and compared to the desired pressure for the demanded extrusion speed, wich is incorporated in a control loop.

Il'l upload a drawing of this shortly

The CURRENT FOCUS OF THE PROJECT IS THE FIRST STAGE, (I've have put effort in the second stage, but the first stage is more valuable when completed, on top of that, it might be that the output of the first stage is sufficient for direct printing)

This is pretty much what I thought.

So the entire base of the first stage, and the channel to the second stage and the small chamber of the second stage and the hot end of course, must be heated... that seems a lot.

Then you have to finely balance the pressure applied by the auger and the air pressure in the second stage chamber so that the air in the second stage chamber doesn't push the melt back to the auger and the pressure from the material coming down the auger doesn't push melt back upstream towards the compressor etc? Seems to me that just putting a "lid" on the chamber in the second stage witht he auger controlling the pressure would achieve a similar result. But in either scenario I dont really see how you "vent" the air from bubbles without causing problems.

I think that the biggest problem with any granule system is going to be getting the air out and I would simply copy what is done commercially if possible.

George

@gspot

you totaly got the picture of the problems that will be encountered when using this concept.

the question is to what extend the venting problems take place after first extrusion, it might be a minor problem or a mayor problem, it will be something that will clear out when first tests are performed. Venting of the melt in the buffer chamber might be achieved beceause you release pressure all the time at the moment you dont want to extrude (small perimeters, start stop movements etc..) it will be atmospheric pressure though, no vacuüm.

Industrial extruders don't always have a venting machanism installed, only when operating with high moisture concentrations in the polymer, or other liquids or gases that emerge when melting the material.
When they do, they use a vacuum pump at aprox 50% of the auger length after a quick core diameter reduction (pressure drop) so the polymer wont flow into the venting hole.
check venting in industrial extruders
I've examined industrial screw layout in detail today in the University library, and read that trapped air pockets can be prevented by gradualy building up pressure. (by having a long conical (4-5xDiameter) section of the core of the screw)

I think Ill manufacture several different screw lay outs, also one with two compression stages, which I can test at the moment that venting is needed, but creating a vacuum will add up to the complexity, so lets hope venting is not needed in the first extrusion stage!

an important thing I learned from the literature today, is the need for axial oriented grooves at the feeder/transport section in the barrel, and a smooth surface finish on the screw itself. in this way the granules are really forced to move in axial direction instead of sticking and rotating together with the screw.

Interesting stuff.

Still makes me think that there is no reason for the second stage and that the motor should be at the top with the hot end directly below. Everything then becomes very simple.

George

Maybe you already found that (not took the time to check all your weblinks on the project page ^^')

but there is another project (working) of a filament extruder : MiniRecycleBot-v2

also on the "let's learn from industrial thing" :

[www.techkits.com] (or other hobby injection machines)

[www.sellbic.com]

---

about // liberapay // flickr // wiki // thingiverse - github

Wow

nice projects you found there,
I didn't knew that more projects have been hiding in the chaos of open source development.

I did some long and boring research work, but these projects (which are very vuluable) I didnt find so far..

thanks for the links! Ill incorparate them in my project page so people will find thair way in the future more easily

Hey!
I found your project on the Facebookpage of FabLab-Leuven.
When you want to work on the plastics-recycling, you may want to try contacting the VKC (Vlaams KunststoffenCentrum).
They try to collect information about plastics from the industry to help research and get correct, verified information into the (Flemish) world.
Good luck!
An interested engineering student (KUL)
Maarten Aerts