HDPE Filament production

I know nobody has mentioned this for a while but While testing my heater I managed to create about 8mm of .75mm filament by manually pushing the feed for about 10 seconds has hard as I could. I then pulled pulled the freed filament back and created about 0.5m of similarly sized filament by stretching with very little force in less than a second. 625 times more productive!

That got me wondering about the problem of making HDPE filament from recycled milk bottles again. Rather than pushing the filament through a small die with enormous pressure should we be stretching it by pulling with a constant speed take up system from a molten reservoir?

---

[www.hydraraptor.blogspot.com]

Hmm, if memory serves that's how they make fiber optics. I wonder how you would regulate the diameter... how precise does it need to be for the MkII? It's got the spring-loaded guide, so I would think tolerances could be rather loose.

Kyle Corbitt Wrote:
-------------------------------------------------------

> It's got the spring-loaded guide, so I
> would think tolerances could be rather loose.

Well, probably not. The mass flow rate is implicitly dependent on the speed of the feed pump. If you get your filament cross-sectional area too far off you are going to make changes to your flow rate that will compromise your print quality.

The spring is there more to force the thread of the pump studding to cut into the filament than to accommodate changes in the cross-sectional area of the filament, although it does that, too.

Edited 1 time(s). Last edit at 07/30/2007 10:16PM by Forrest Higgs.

Presumably if you fed powdered material into a temperature controlled heating vessel at a constant rate and pulled a thread from it a constant rate then you would geta constant diameter filament?

I am not certain but I got the impression that the pulled filament has different mechanical properties to the extruded one. Possible a bit clearer and more flexible, a result of the molecules being aligned longitudinaly perhaps (out of me depth here).

---

[www.hydraraptor.blogspot.com]

Thats basically the way it is. You can either do it by stretching/compressing or chemicaly by changing the level of very linear sections (not suprising called plastiziers) or hightly branched sections (crosslinkers) thats the difference between PVC and CPVC the C stands for crosslinked. Crosslinked materials have more heat and stress resistance and so are tougher materials in general but also far harder to process.

Milk jugs have relativly high levels of crosslinking, low crosslinked versions are used to make the thin plastic bags for grocery stores. That does suggest that adding plastic bags to the milk jugs might be a way of adding plastiziers while increasing the total material recycling. That would also have the effect of lowering the melting point of the material and making it easier to handle.

It comes at a cost of the very linear filiment being lower strength (both tensile and transverse but especially transverse) which might create a stiffness problem. The different may or may not be noticable at the work loads we are talking about.

Of course at the speeds were talking about there isnt any reason why you couldnt reinforce the material.

Eventually I imagine that the filament production will occur concurently with the material usage so that recycling and material adjustment can be done quickly and easily. Or for that matter a hopper could deliberately be filled with a high crosslinked material for the part base, then a mid range material for the bulk of the part and a high plastizier material for sections that have to be able to shift during use.

Mike
The thoughts and ideas expressed in this post do not reflect those of my employer and are intended only as communications between individuals. Any attempts at implement are at your own risk

Edited 1 time(s). Last edit at 09/12/2007 09:13PM by ohiomike.

Could you use a drawplate (like jewelers do to size wire, or scale modelers do to get wooden dowels to the correct size) to get the filament to the correct final size?

Are 2 Litter pop bottles made of HDPE?

I think transparent pop bottles are PET not HDPE. [en.wikipedia.org]

---

[www.hydraraptor.blogspot.com]

Forrest Higgs Wrote:
-------------------------------------------------------
> Kyle Corbitt Wrote:
> --------------------------------------------------
> -----
>
> > It's got the spring-loaded guide, so I
> > would think tolerances could be rather loose.
>
> Well, probably not. The mass flow rate is
> implicitly dependent on the speed of the feed
> pump. If you get your filament cross-sectional
> area too far off you are going to make changes to
> your flow rate that will compromise your print
> quality.
>
> The spring is there more to force the thread of
> the pump studding to cut into the filament than to
> accommodate changes in the cross-sectional area of
> the filament, although it does that, too.

Maybe an extruder head that could handle this kind of variation could be devised. I've been thinking of a way to put a pressure sensor in the head; it would weigh the force being applied to the filiment being fed though the heater/extruder so that adjustments could be made to keep the pressure consistant. Variations in temperature could cause problems, though, because the viscosity of the melted plastic would change.

drvanthorp Wrote:
-------------------------------------------------------
> Variations in temperature
> could cause problems, though, because the
> viscosity of the melted plastic would change.


That shouldn't be too hard of a problem to solve. Software has control of the means of heating and a sensor for feedback; I would think you could keep temperature pretty constant.

Of course, it would be easier to just avoid the whole issue and just be sure to use a constant-diameter filament!

Kyle

I had wondered this all along and have come to these conclusions. It seems that the sturdiness of the motors and the mauling of the filament is what makes it work so far. I shouldn't be surprised that if you have varying thickness of filament you wouldn't notice much of a difference in the output. Output should be the right size and as long as the pressure is sufficient. At a lower limit if your filament became rather thin, then the force it applies might drop below that needed to drive the flow. At an upper limit the filament might become too big for your drive to accommodate it. And within those boundaries I predict production.

I just had a thought.

This wouldn't be practical for long filaments, but for sticks, it might work.
How about rolling out a sheet of the stuff, then using a jig and sharp knife to slice rectangular lines off of the sheet?
The stuff would only be as long as the width of the sheet you were working with, but gets around any difficulties in either extruding or drawing.

Of course, then you need a consistent thickness to slice from. I'm thinking just pour some molten material out on a surface it won't stick to, or can be peeled off of.

I started to say you could pour it into a container of the same material, but one made thin enough to be worth it. Then I realized, you'd want this stuff to cool equally. If you poured it hot enough to pool before it started cooling, it'd probably soften any tray of the same material enough to stick to whatever was below it. If you poured it at just above its melting temp., it'd probably harden at less than level.

Of course, you could also slice the slices themselves off of a larger block, so you could pour your plastic into an ingot, then slice the ingot, then slice the slices.

Also, on the idea of incorporating the properties of two different mixes of the same basic plastic, could you use two smaller filaments, and melt them together at the head?

Perhaps for Mendel, a head can be developed that allows for multiple feeds into a single heating element. This would allow for variations to be made in the plastic, or other material. These variations could include pigments, softening agents, agents to increase rigidity, material to decrease transparency, either towards opacity or toward translucency, possibly even, with some materials, ad-hoc variations in conductivity.
Each "ink" could be a filament with an excess of the choice modifier, mixed with just enough of the base material to get it to extrude correctly. This could be mixed with a base filament to alter its properties. The base filament would probably be significantly thicker than the other filaments.
It might be possible to use a single "drive" motor, and several smaller actuators to clutch in the other filaments. In this way, the head doesn't have to get ridiculous in its dimensions just to support all the gearmotors.

There is the possibility that a fiber that hasn't been used in awhile might conduct enough heat along its length that it is too soft to drive with a screw, however.

It is a mistake to think of molten plastic as a fluid except at VERY high pressures and high temperatures. If you heat it well beyond its melting point and try to dump it on a teflon lined tray, for example, it will just lay there and take maybe half an hour to noticeably slump, never mind spread out into a layer.

I suppose that rolling the molten mass to a 3mm thick sheet with a cylinder would be the way to go. You may need the cylinder to be heated. Maybe a hollow stainless steel cylinder filled with boiling water would keep the whole thing warm for long enough to achieve an uniform sheet thickness.
The cylinder would need to be mounted on some sort of rails to ensure a uniform 3mm sheet thickness. It's a cumbersome operation but IMHO, it's the best solution i have heard for now.

I guess that for the working surface, a steel pane sprayed with teflon or silicon lubricant would be good enough to ensure that the polymer doesn't stick to it.

The easiest way would be to use a cookie sheet with fairly high sides, telfon lined being ideal. The put a known weight of ground plastic in it and shove it into an oven set at about 300F, well above melting but not hot engough to ignite. Then periodically pull out the sheet and use a marble rolling pin to weld the pieces together and smooth them out. At a known surface area the weight of material necessary to reach 3mm thickness should be constant.

Another way would be to build a pouch out of silicon infused parchment paper and then use a clothes iron to flatten it. Provided you could make the pouch study enough to handle the abuse its volume shouldnt change a meaningful amount.


Mike

The thoughts and ideas expressed in this post do not reflect those of my employer and are intended only as communications between individuals. Any attempts at implement are at your own risk

Edited 1 time(s). Last edit at 09/12/2007 09:12PM by ohiomike.

ohiomike Wrote:
-------------------------------------------------------
> The put a known weight of ground plastic in it and
> shove it into an oven set at about 300F, well
> above melting but not hot enough to ignite.

This begs the question of how you are going to grind plastic. If you come up with a way that works well, please let me know. No sarcasm is intended whatsoever. I've just never managed to find or invent a way that works very well.

Well my milling machine does a pretty good job of reducing plastic to dust. Not very fast, but probably faster than RepRap can extrude it.

---

[www.hydraraptor.blogspot.com]

Plastic like in milk jugs?

I haven't machined any HDPE yet but I read somewhere that you can. The closest that I have would be Delrin / acetal. Perpex / acrylic makes the the nicest powder. The mystery metal loaded resin that I made my extruder body from makes a horribly fine dust that blocks up the pores of my vacuum clean bags.

---

[www.hydraraptor.blogspot.com]

Oh you can machine HDPE all right. You have to use sharp cutting tools for both machining and drilling, however. I was mostly wondering if anybody has found a small scale grinder that could be used to recycle plastic.

I just attacked a milk bottle with a 3mm endmill at about 20000 RPM. It ripped into it and produced fine swarf.

[picasaweb.google.co.uk]

---

[www.hydraraptor.blogspot.com]

Yup, I bought a tungsten carbide grinder that will let me do that with my Demel a few days ago. The problem is that in my case, at least, it isn't a practical way of reducing a big box of used milk just to swarf that I could use in a filament drawing or extruding scheme.

for larger operations i found these wood chippers:

[www.lowes.com]

cheepest one comes at 480$ new.....

or at ebay for 140$:

[cgi.ebay.com]

Edited 1 time(s). Last edit at 08/01/2007 02:36PM by Fernando.

Thinking about grinders, ain't better building one?

probably using polymorph "skeleton" or another thing...

See ya! []'s