Extruder Heater Materials
PTFE Breaks down at 230C into toxic substances, so replacement materials are required in order to safely process plastics with high melting points such as PLA and HDPE. Besides, PTFE rod is hard to come by for most folk.
Several high-heat areas are present in the nozzle and its insulator, with different requirements:
- The heater barrel - Low thermal resistance needed.
- The heater cladding - High thermal resistance, some "give".
- Heater insulator - High thermal resistance, rigid.
Promising materials are:
- BBQ Paint, or other "high heat" paint.
- Plaster of Paris
- Fibreglass insulation
- Perlite/vermiculite (possibly pumice)
- Cement or concrete
The first proposed approach is to use ceramic-filled BBQ paint to insulate the heater barrel from the nichrome heating coil. BBQ paint is good for up to 600C+ but does outgass on its first heating.
To protect and enclose the heater element, I proposed a fibreglass wrap saturated in plaster of paris, much in the same way a plaster cast is made. The fibreglass can be arranged to make sure the thermistor and all protruding wires do not rip through the plaster jacket if tugged. It should be resilient enough to cope with thermal expansion and contaction of the heating element.
To attach the barrel to the extrusion mechanism, a rigid block of insulating concrete is used. This need be no more than 50mm in diameter and 60mm high. Any narrower in diameter, and it will be hard to get enough concrete around the threaded rod that will attach it to the extruder mechanism.
Concrete itself is quite stable at temperatures up to 300C (540F), and can take twice this for brief periods without significant damage. However, it may shed particles at high heat, which is not good news for nozzles.
A concrete is made from cement mix with added perlite and vermiculite. Due to the small size of the insulator, a fairly damp mix may be needed to ensure the concrete cures before it dries out. This, of course, weakens the concrete so a happy medium needs to be found. Absolute minimum use of water and precautions against evaporation should be tried.
A 3mm knitting needle through the centre of the block will maintain the central cavity, and hold the heater barrel in allignment. The needle will be removed before the concrete fully hardens and the hole tidied manually. Hardener or ceramic paint may be necessary to prevent wear at the edges.
One end of the block will have threaded rods embedded in it to facilitate attachment to the extrusion mechanism; washers or nuts may be embedded to improve adhesion. The other end will hold the heater barrel by means of a cluster of nuts and washers. This cluster of nuts and washers will be embedded, allowing the barrel to be unscrewed for maintenance or replacement. Some smoothing of the finish and washing will be necessary for the concrete component, to keep debris to a minimum.
The removable nozzle will ensure that the barrel and central cavity can be properly cleaned of debris before extrusion commences. Chance of successful alignment on first pass is slim, but there is some lattitude for error. A jig will simplify the alignment process.
The design has to hold a 200C+ degree nozzle on one end, and maintain a temperature of <60C on the other. This will allow Polymorph/polycaprolactone extrusion. Other plastics require a higher nozzle temperature, but in turn can take more heat at the other end.
While at full temperature, the 4 threaded rods inserted in the block need to withstand a force of at least 10N each, and the nozzle itself 20N.
Test effect of using black BBQ paint on concrete portion to increase heat radiation and protection of concrete exterior.
Fairly obviously, we don't want to use any materials that degrade at <300C.
For small quantities of premixed cement, ask your DIY store for a cement "patch pack", used for small repairs.
Check out several brands of black "high heat" paints. Prices vary by a factor of 2. Automotive stores also sell this paint for spraying engines and exhaust systems.
White cement varies from ordinary grey cement only in terms of colour and additional cost.
Pelican Wire (and presumably other suppliers) do 0.2mm (32-33AWG) nichrome coated with silicone and fibreglass insulation that is good up to 400C+. Nice if you can get it as the heater barrel does not then need the BBQ spray for electrical insulation: http://www.ec-securehost.com/PelicanWireCo.Inc./Nichrome_80.html
A brief forray in to the world of concrete puring produced the first test article. This showed that an insulating concrete mix would stand up to mechanical handling and blowtorch-quality heat.
A 52mm diameter plastic cylinder hacked off a plastic flowerpot was used as the former, 65mm high. A tin lid with a 6mm hole in the middle was secured to the bottom with gaffer tape. It then occurred to me that a disposable cup of the appropriate dimensions would have sufficed. Oh well.
A nut, washer, nut, washer and final nut were put on one end of the 6mm brass heater barrel. Thin glavanised steel wire was twisted about the top two nuts and splayed out in loops to give the concrete something to gain purchase on. This was then inserted into the former, and the free end of the heater barrel was passed through the hole. A final nut under the lid secured it in place and made the join leak-proof.
A zig-zagged cylindrical form of wire was then added on top, to strengthen the concrete. All wires terminated at least 10mm below the edge of the former and in loops where possible.
A 3mm knitting needle with a sipke on both ends was then inserted into the former, and placed firmly into the hole at the top of the heater barrel.
A variety of long M3 bolts and threaded rods were attached to the bottom of the Mk2 extrusion mechanism, so that they would engade with the concrete. This included 2 x 65mm lengths of M3 with the last 15mm bent into a J-shape. Two 45 degree bends are required, closely spaced, or you'll break it.
5 Heaped desert spoons of cement mix were mixed with 3 heaped spoons of vermiculite/pearlite and a small quantity of water. Underdo the water if anything - the plasticisers in ready-mix need remarkably little water.
This was packed into the form within 5mm of the top, and vibration applied with the pad of a small orbital sander. This causes bubbles to rise to the surface by turning the concrete into quicksand. It also causes concrete to dribble out of every gap in the bottom of the form.
The extrusion mechanisim with protruding bolts was then inserted into the concrete, sliding down the knitting needle. The concrete was vibrated again, topped up a bit, vibrated once more and left to cure. A large dollop of leftover concrete was kept to act as a guide to the set of the concrete in the form.
Results of First Concrete Form
After 3 1/2hrs, the stuff still isn't set. This batch is much slower for no obvious reason. After 12 hrs the needle was removed. After 48hrs the mould was cut away with a Dremel tool, revealing some very soild concrete. I'm letting this dry after some rudimentary shaping to remove the leaked bits of concrete. Images of the form are shown below.
- The first concrete insulator just out of its mould (10-Feb-2006):
Note that the concrete is quite dense, though some bubbles are visible. The whiter patches are mostly vermiculite/perlite.
After 3 days, the concrete is finally starting to go a lighter grey. The "temporary" wadding used to make the FDM'd extruder component stay on the bolts embedded in the concrete turns out to have been extraordinarily effective. All I did was stick a bit of insulated wire down the side of the bolt, and it seems to have had a rawl-plug-like effect. We might need an FDM rawlplug design :).
Unfortunately an end from the reinforcing wire has sprung out into the central cavity. It is easily deflected and does not render the item useless, but better care needs to be taken with the placement of wires.
Hauling a wad of wire wool through the 3mm central cavity has also removed any stray debris. I am unable to get sand to come off with a Polymorph rod, though that will no doubt change when the heat is applied the first few times. Best run it empty for a bit, I think.
After 4 days, the concrete continues to lighten in colour, and also in weight as the moisture comes out of it. No additional shedding is occurring.
The M6 brass heater barrel unscrewed easily from the 3 nuts in the concrete insulator, and was dried in the family oven at 250C after receiving 5 generous coats of BBQ paint. It will shared the oven with a roast for 4hrs (not at 250C) and it is believed that this process cured the paint.
8 ohms worth of 0.2mm nichrome wire was then wrapped around the bolt - it only being a 50mm bolt, that's all that would fit. THe ends were taped down with masking tape, and a mixture of fibreglass wall insulation (3 cc) and Plaster of Paris (2 tsp) was applied. Fingers don't work - use spoon handles.
While the plaster jacket dried, a 6mm hole was drilled in the termistor's tag (tagged ones were going for 10c a piece...) to allow it to be threaded onto the heater barrel. After 1/2hr the plaster was shaped with a rasp to remove unwanted protrusions.
The design is servicable, but should it jam and plastic enter the concrete section it gums up and is very hard to remove. The high temperature capabilities of the concrete can be exploited here. First remove as much as possible with some form of ramrod - I used a bamboo skewer - then heat a metal rod (classic Kiwi No. 8 wire is good) to dull red and ream out the unwanted plastic with it.
The specimen produced here is probably a bit too long. Approximately 20mm could be shaved off, reducing the risk of jamming. It is possible that a silicone baking paper lining might reduce the gumming up issues and this will be tested later.
The bolts that connect the concrete to the poly-holder should be held on with 2 nuts each. This will make it possible to remove the concrete from the poly-holder; a flaw with the model described above is that I can't get the darned thing off!
-- Main.VikOlliver - 04 Feb 2006