Topic: Polymer Extrusion - A Brief Tutorial.
Okay guys, I'm going to try to aggregate the lessons of myself and others here.
If you haven't yet, take a look at the thread that started it all.. There is a wealth of information in that thread, but I am going to attempt to summarize most of it here in place of that 58 page thread.
Basic Concepts
The Filastruder is very comparable to a single screw industrial extruder. One of the main differences is the screw type - industrial extruders have a tapered center section to increase shear and compression. The Filastruder has a constant crossectional area, primarily due to cost. That means that the Filastruder runs at a significantly lower pressure (~150psi/10bar). Otherwise, they are similar in operation.
Raw Material
What goes in, comes out. This includes dust, dirt, and other debris. You will want to keep things clean. With a 0.6mm nozzle and no on-printer filament cleaning, I have gone through over 15kg of Filastruder filament without a clog. With a 0.35mm nozzle and no on-printer filament cleaning, I experienced a clog every 2kg or so. This may have been from dust settling on the filament as it was fed in, though. With the addition of a melt filter (should be in the store by September 1st), you should not experience any contamination in your filament. It is still important to be mindful of debris, so you do not have to clean the melt filter too often. More on that later.
The Filastruder will feed both pellet and powdered resins. Powdered resins may have a tendency to stick/clump in the hopper and not feed well if not agitated. Pellets feed great as long as the maximum pellet dimension is less than 6mm in length.
Polymer Selection
The Filastruder will process just about anything you put in it that meets the above (pellet size, mostly). Two aspects of the polymer significantly affect output rate:
1.) MFI/MFR (melt flow index/melt flow rate) This is a measurement of how easily a polymer flows through an orifice. If you take the Melt Flow Rate as tested at 230°C/3.8 kgf (ASTM D1238) and multiply by the following multipliers, you will get an estimated output rate in inches per minute with 1.75mm filament size:
v.1.0, original motor multiplier 0.9
v.1.1+, original motor multiplier 0.7
Post kickstarter motor: 1.8
The 230C/3.8kgf MFI of MG94 is 11.7. So if you have a v1.0 with original motor, you'd expect around 10.5in/min. If you upgraded the motor, you'd expect 19in/min. This would be with an appropriate nozzle temp (185C in our case), not the ASTM D1238 test condition temperature.
2.) The other is the coefficient of friction between the polymer and the heated barrel. You want this to be high, and you want the COF between the polymer and screw root to be low. Think of it as rotating a bolt while holding the nut. The more friction between the nut and the bolt, the greater the torque required (increased motor load). If there was no friction between your fingers and the nut, the nut would not advance on the motor. Unfortunately this isn't really a specification on a datasheet. It does mean the barrel needs to have some internal roughness, otherwise the polymer to barrel COF will drop and melt pressure will drop with it, resulting in reduced output. Warning: math ahead!
The relevant variables here are P(z) (pressure profile), f_b (barrel to polymer dynamic COF) and f_s (screw to polymer dynamic COF). As you can see, increasing f_b increases pressure, decreasing f_s increases pressure. [1] In fact, some screws are fluoropolymer (PTFE or similar) coated. Too bad that's way out of our little Filastruder's price range.
ABS
The recommended ABS resin to use is Sabic's MG94. It is what is included with your Filastruder, and it is available through Open Source Printing. I have processed half a dozen varieties of ABS, and they all work well with little hassle. Basically, they are the easy button. MG94 has the highest MFI of any ABS resin I've found so far, which is why I selected it as the official resin to include with the Filastruder. Expect tolerances of +/-0.05mm on 1.75mm filament for any ABS you throw at it.
PLA
PLA is a tougher nut to crack, owing it its affinity for moisture and its melting properties. In short, it must be very dry to work well in ANY extruder, not just the Filastruder. With proper drying, tolerances can approach those experienced with ABS. It also helps to have a winder, like IanJohnson's Filawinder. When switching from ABS to PLA, it is important to either run to empty and then remove the nozzle and pull out any remaining ABS, or to run PLA at ABS temperatures for a few hours to purge all remaining ABS.
Nylon
To my knowledge, the only nylon that has been run through a Filastruder is this. It is the same nylon Shapeways uses in their strong and flexible processes. If you can keep it feeding into the hopper well, the Filastruder will pressurize and extrude it just fine. Sometimes the powder clumps in the hopper if it isn't dry enough, but you can poke it with a piece of filament periodically to keep it flowing. Tolerances are on par with ABS (+/[email protected])
LDPE
One of our users, R.J.A.Allen, has extruded LDPE: http://www.soliforum.com/post/28185/#p28185
TPE
I have extruded Dynaflex G2755 successfully using the Filastruder. I actually got very fast output rates, about double what I see with ABS. It is not possible to accurately measure the diameter with calipers because the filament is so flexible - I could easily tie it in a knot and untie it. However, the tolerances looked about like I see with ABS, visually. Thermoplastic elastomer is like string. Unfortunately I have no idea how you would print with it, since it is so flexible.
Effects of Extrusion Temperature
Increasing temperature yields higher output rate, smaller diameter. Eventually, there is a point that a surface roughness on the polymer can be felt. My hypothesis is that this is due to internal stresses and unstable flow. It may also be due to moisture content. If the filament is not smooth, reduce nozzle temperature. Increasing temperature also reduces drag on the motor, letting it work less (even though output rate is increased). Another side effect of increasing temperature is the possibility of tangles at the nozzle (we call this ramen). With the filament guide set properly and nothing for the filament to hit on its way down, ramen should not be an issue at any temperature.
Polymer Changeover
Changing from a polymer with a lower melt temperature to one with a higher melt temperature is easy - just run until the hopper is empty, load the new material, and wait for the old material to be flushed out. Going the other direction is not so easy. You will need to remove the nozzle to clean out all remnants of the first polymer. Be sure to wear gloves while doing this, as the hotend will still be hot!
Masterbatch
Masterbatch is a solid or liquid (we use solid) additive to a resin to add a particular property. In our case, it is to add color. Hence, colorant and masterbatch are used interchangably. Here's some natural ABS with blue colorant [2]:
When, which run through the Filastruder, produces this [2]:
It really is that simple. For most colorants, you want a ratio between 32:1 and 48:1. Simply measure this just like you would if baking a cake - grab a tablespoon and measuring cup, and get to work. There are 48 tablespoons in a cup, so one tablespoon of colorant per cup of natural resin will get you in the ballpark. For more information, check out this thread:
http://www.soliforum.com/topic/2123/col … sterbatch/
Recycling
There's a few hurdles in recycling printed plastic. They are as follows:
Shredding/cutting the prints into pellets. The Filastruder needs chunks of plastic that are no larger than 5mm on any side.
Contaminants. Any dirt or dust you get on your print, ends up in your filament. Same for whatever method you use to cut/shred your prints - any metal debris you get in the shredded plastic ends up in filament.
Heat history. The more you heat cycle a polymer, the weaker it gets as each heat cycle breaks the polymer chains. You will want to mix in virgin material to maintain strength, though this is not a huge issue typically [3]
As far as contaminants, this should be a non-issue if you wash and dry the plastic, and use the melt filter we provide on the site. The other two issues are left up to the user. There are some hobbyist shredders beginning to be made, but we have no experience with them.
Furthermore, I can only recommend recycling plastic that you are 100% sure you know the properties of. Some plastics can offgas toxic fumes near their melt temperature - PVC, for example.
Do not attempt to extrude plastics you are not familiar with.
Do not attempt to extrude plastics you are not familiar with.
Do not attempt to extrude plastics you are not familiar with.
Die design/melt filtering
The output side of an industrial extruder looks like this:
The breaker plate looks like this:
The screen pack is on one side of that, and the die is on the other side.
I've gone through 5 revisions of a combined screen pack/breaker plate/extrusion die, and have come up with the following:
It uses a 250 micron wire mesh filter, and has 75sqmm of effective filtration area. It is CNC machined out of brass - this in and of itself is an upgrade from the standard nozzles which are made on a drill press. I have tested this final revision for over 100 hours of extrusion, with no measurable decline in output rate. Because the local velocity of the polymer at the screen pack is so low, output rate is not diminished until the filter is nearly full of contaminants.
This melt filter is completely reverse compatible with all versions of the Filastruder - simply heat up the Filastruder to operating temperatures, unplug it, remove the thermocouple, loosen the nozzle 1 turn each minute while the Filastruder cools, install the new nozzle, reinstall the thermocouple, and heat the Filastruder back up. You may need to tighten the nozzle another half turn after heating. Of course, wear gloves and safety glasses during this!
References:
[1] Polymer Extrusion By Chris Rauwendaal
[2] Images provided by DePartedPrinter
[3] http://www.journalamme.org/papers_vol37_2/37221.pdf
