NOES! The back button ate my really long post...
And it was so perfect...
Gah!
I'm typing this one in Notepad.... and I'm saving constantly just in case the power goes out or something like that...
[Right then. Brace yourself for much ado about... what were we talking about, again?]
So anyway, I have some basic designs for the bootstrap extruder.
Main Construction Material
The main construction material is variable, and is meant to be able to be replaced by RP... but it is MUCH simpler than something that would require RP (that is, it's much simpler than the v1.1 extruder, even if it is a bit less sophisticated). The basic idea was that I could take an acrylic glass plate 1/2 inch thick and manually cut out all of the pieces. It could also be cut out of wood, PTFE, steel, etc etc. I would think the three materials best suited for use as the main material would be a RepRappable plastic, handcut/lasercut acrylic glass, or wood. As I said before, I'm liking the idea of the handcut acrylic glass, but the use of wood would make this a really cool bootstrap available with even less advanced technology (it's easier to cut a tree down and make wooden logs than to create acrylic glass plates, or so I assume).
Anyway.
Heated Barrel
The heated barrel is the same as that used in the v1.1 extruder. The relevant attributes of the barrel are: it ends in a 16mm diameter PTFE tube with a 3mm hole in the center.
Barrel Clamp/Attachment
The clamp is a two piece shape that can be thought of as a 4cm x 4cm x 1cm block with a 16mm hole (drilled) in the center, and then cut in half, forming two 4cm x 2cm x 1cm blocks with half-cylinder holes in them (and probably even created in such a way).
Something like this (I'm looking at it in Lucida Console font in notepad; you could just copy the text-images and paste them into the reply box to get an idea of what they should look like [looks like you'll actually have to go into the raw source to get the correct number of spaces... anyone know how to maintain spacing in this forum?][Edit- fixed with code tags, thanks Ru.]):
. __ __
. | || |
. | /\ |
. | \/ |
. |__||__|
as a cross section. The diamond looking thing should actually be a circle. I tried using () but it made it look like a 4x5 rectangle instead of a 4x4 square.
Anyway, that's the shape. On the 4x1 side of the blocks, two 3mm holes should be drilled for screws. The centers of the holes should be 5mm from the 4cm-long sides and 9 mm from the 1cm-long sides. That essentially means that each screw will be 3mm above the edge of the PTFE tube; the tube extends to 16/2=8mm above the center of the blocks. Adding 3mm spacing between the tube and the screw hole, we get 8+3=11mm from the center. 20-11=9mm away from the side. [Note that I'm not entirely clear on the dimensions yet, I think they would depend on material strengths etc. and may require some expert input and/or testing on my end.]
The cross section would look like this if you had an x-ray to see the holes:
. __ __ _
. |__||__| |.|
. | /\ | .. or like this from the 4x1 side | |
. |__\/__| |.|
. |__||__| |_|
Not that you needed a visual representation... it's just fun to make them.
Anyway, so you drill the 3mm holes, insert the screws (which should be around 5cm long), and then loosely attach a nut on each of them, such that the nut stays on the screw and stops the two blocks from coming apart, but still allows the blocks to move a bit along the screws (note that the holes should be large enough such that the screws can rotate in the holes, or more importantly move laterally along the holes).
Insert the PTFE tube into the big 16mm hole, tighten the nuts, and voila, you have a working clamp.
[I think I overcomplicated the explanation on that one, sorry.]
[Edit - I just realized the official clamp design doesn't clamp like this... it isn't really a clamp (or rather the clamp isn't used in the long run), more of a fastening device that has screws that actually go straight through the PTFE, holding it in place with nuts and bolts... well, that's depressing, because I was kinda proud of my design... anyway, it can be fixed by simply moving the screws from being just outside of the PTFE tube area, to being just inside of the PTFE tube area... essentially moving the two screws closer to each other (each moves about 6mm, so they are 12mm closer)... some simple changes in construction would be necessary.]
Filament/Screw Holder
Now on to the filament/screw holder. I think they can be created as one piece. It is a bit harder to make, but it is simpler in design and thus smaller and cheaper. The "harder" part means that you'll have to spend more time on it though (and perhaps you might break a piece or two while trying to make it; I still think it would be cheaper, even considering that).
The basic design is a 1cm x 1cm x ?cm rectangular prism. The ?cm means I'm not entirely sure on the length dimension (as stated before, dimensions are up in the air). Anyway, a 3.5mm filament hole and a 3mm screw hole are drilled such that they ever so slightly overlap. The positioning of these holes is up in the air as well; I can't decide on the best locations. Here are a few that I've considered:
1) The two holes placed horizontally in the square cross section, sorta like this:
. ____
. | |
. | OO |
. |____|
The problem with this design is that the piece may break when drilling the holes or at some other time, because the distance between the hole sides and the sides of the 1cm x 1cm square is small: 10-(3.5+3)=3.5/2=1.25mm on each side. Scary.
2) Next I tried putting them in the diagonal, which is 14.14mm in length instead of the 10mm length I placed them in above (pythag theorem/distance formula). It's a bit harder to show in fixed width text, unfortunately. But basically, the distance between the hole edges and the corners would be: (14.14-(3.5+3))/2=3.82mm. Of course, the distance between the hole edges and the square sides is a bit less than this, but it is still better than the 1.25mm in 1). I could calculate it, but it would take a while and frankly it's not worth it (at least right now).
3) Aesthetically (and when building) it might be better to have the filament in the center of the holder:
. ____
. | o|
. | () |
. |____|
The filament hole center would be 7.07mm away from the corner. The edge of the hole would be 7.07mm-(3.5/2)=5.32mm from the corner. The screw hole would be 5.32-3=2.32mm from the edge. I'm guessing this would be structurally unstable, like 1), but I'm not sure. In any case, the screw can be changed to a 2mm screw, changing the distance to 5.32-2=3.32mm.
4) I like symmetry. How about two screws on opposite sides of the filament?
. ____
. | o|
. | () |
. |o___|
I'm not sure if it would work. But if it would, it would give twice the surface area, spreading the pressure across the filament.
5) If 4) could work, maybe this would work too:
. ____
. |o o|
. | () |
. |o__o|
Four screws, each positioned between a corner and the filament.
4) and 5) might require some synchronization of the screws (I really don't know), and might be totally impractical. In any case, the general idea should work (perhaps with some variations).
6) Oh, one more idea: the problem with 1) was that the wall-hole distance was small and might be structurally unstable. We can increase the distance by increasing the holder's dimensions. I'd like to only do this along one dimension, because this will keep the other one at 10mm, which is easily an acrylic glass thickness.
. _______
. | |
. | oO |
. |_______|
This can also be applied to get the filament hole in the center of the holder.
[There are quite a few more variations, such as a cylinder being used instead of a rectangular prism, but for various reasons I prefer the ones on this shortlist.]
The v1.1 extruder has some half-bearing lands in the screw holder... it shouldn't be too hard to implement these similarly. I'm thinking put them at the ends of the holder, where the hole for the screw is slightly widened on one side (the side opposite the filament hole), allowing for placement of some lubricated thingy. (High-tech lab jargon.)
Now that I am reviewing the v1.1 extruder, it seems that the reason for separating the two pieces was to allow for the use of springs to press the screw against the filament. Well, my design can probably do this too. I think maybe if I make the lubricated thingies long enough, they can contact enough of the screw on both sides. I can drill a hole in the side, that meets the screw's hole where the lubricated thingy would be. Inserting the lubricated thingy into the screw hole, and then a spring into this new hole, the spring can press against the lubricated thingy, which can then shove the screw against the filament. (This would be necessary on both sides.) There are quite a few possible variations.
Filament-/Screw-Holder Holder-Block
Now, to attach it to the "clamp" and thus make it flush to the PTFE tube:
More blocks! A block, say 4cm x 4cm x 0.5cm, with a "square hole" cut out of its center, 1cm x 1cm (the size and shape of the screw & filament holder) is used. This can be made in a few ways - directly cutting the square out of the 4x4x0.5 block is the preferred method, but this may be a bit difficult with common tools. Alternately, 4 or 2 pieces can be used:
For the 4 piece method (probably the easiest method):
1) Cut the 4x4x0.5 into a 4x1.5x0.5 and a 4x2.5x0.5.
2) Cut the 4x2.5x0.5 into a 4x1.5x0.5 and a 4x1x0.5. You should now have two 4x1.5x0.5 blocks and a 4x1x0.5 block.
3) Cut the 4x1x0.5 block into a 1.5x1x0.5 and a 2.5x1x0.5.
4) Cut the 2.5x1x0.5 into a 1.5x1x0.5 and a 1x1x0.5. Discard/store the 1x1x0.5. You should now have two 4x1.5x0.5 blocks and two 1.5x1x0.5 blocks.
5) Positioning the blocks in their original positions and then screwing them together (without passing a screw through the center square "hole") yields the proper shape.
[Note that this is for a 10mm x 10mm filament/screw holder with the filament in the center.]
This block is not intended as a clamp, but rather as a positioning guide.
The clamp for the PTFE tube has 4 holes drilled near each corner for screws. The screws need to be really long (as long as the clamp and the filament/screw holder combined, and then some more). 4 holes are drilled in the new block that coincide with the holes in the clamp. 4 screws are pushed through the holes in the clamp, with the heads on the PTFE tube's side. The new block is slid onto these screws, and made flush with the clamp. Nuts are tightened (a loooong way down) to hold the two together, with a LOT of screw left sticking out. The filament/screw holder should be able to be slid into the square hole in the new block, with the filament hole lining up with that of the PTFE tube.
Biplate Geartrain, Motor mounting, and Filament-Screw Divergence
Now for that two-plate geartrain contraption I mentioned in my last post.
The basic idea is that you take two parallel plates with some space between them (of the material - plastic/wood/whatever), place them perpendicular to the screw, and put a motor and geartrain between the plates, attaching to the screw and rotating it.
The first problem with this design is that if you allow the screw and the filament to maintain their relative positioning by keeping them straight, then when you try to attach a gear or something to the screw, you find that it smacks against the filament. This could either cause jamming or prevent placement of the filament/gear (or some other problem). I think this is why the motor drive/gear assembly in the v1.1 extruder is tilted, and uses a bendable steel wire to transfer the motion.
There are three options for solving this that I see. First, use a gear smaller than the screw diameter. This is rather difficult, and might present problems when the screw moves around due to the springs in the holder (the ones used to allow for slight variations in filament size). The second option is to have the filament bend away from the screw. Basically, one hole is drilled on the lower plate, which coincides with the hole in the holder and which is touching the hole for the screw, and then one hole is drilled on the upper plate, which is slightly apart from the hole for the screw. Essentially, the hole is displaced horizontally forcing the filament to go through the two-plate contraption at an angle. The third option is to have the screw bend away from the filament. This is what is done by the v1.1 extruder. The way the v1.1 does it is not, however, the only way to bend a rotating screw and transfer motion! Personally, I thought it would be simpler to use a universal joint or something similar (http://en.wikipedia.org/wiki/Universal_joint), but then I remembered that you still have screw movement due to the springs... so I figure I'll probably stick with the steel wire idea for now (unless anyone has some better suggestions; I don't really like the idea of having to solder a steel wire to the screw; it's not very low-tech).
Then again, if it isn't a problem to bend the filament as described above, then I'd like to do that (I'm not sure if it breaks the filament/messes it up somehow or if some of the harder plastics can't even be bent like that).
The Mk II extruder seems to bend the filament, so I'm hoping it's feasible.
If it is, then the design will basically be like this:
Filament gets bent as described by the section option. Screw goes straight up through the top plate, and is held in the top plate by a bearing and some nuts and washers (I think if you layer it nut-washer-bearing in plate-washer-nut it should work and still rotate, right?). A gear is attached on the screw, right below the top plate (so that the gear doesn't touch the filament). This gear meets a gear next to it, which is the end of the gear train. The gear train is essentially a bunch of gears on rods (possibly just bolts allowed to rotate between the plates) which decreases rotational speed and increases torque (a series of small gear to big gear meetings, from the motor). The motor is placed vertically between the plates, with a small gear on its rod, which contacts the first big gear in the train (might need to drill tiny holes for the leads, depending on the motor).
Four holes are drilled (this is probably one of the first things to be done) in each plate, coinciding with the 4 long screws from the clamp, and nuts are placed on each side of each plate, holding them in place. It may be necessary (depending on plate size) to put additional bolts through the plate to stabilize the structure.
Cartbot Attachment
I'm not entirely clear on the connection of the extruder to the cartbots, but it should be a trivial matter to make an attachment.
Silly Ending Section Title
So that's about it, unless I'm forgetting something. Electronics etc. can be placed between the plates, if necessary (or perhaps above/below, I think this consideration is more aesthetics though).
Seems simple enough to me. Comments?
Oh, and please don't flame me for being silly/naive. It's only my third day of RepRap knowledge.
Edited 1 time(s). Last edit at 07/20/2008 12:38PM by Joshua Merchant.