It looks like you're new here. If you want to get involved, click one of these buttons!
My print is too small to print the base_gear normal to its flat face. Instead, I can print it at an angle with supports but worry it may fail. How did you print yours? Laying down or standing up or at an angle?
Printing it on an angle might prove to be a bit tricky with an FDM process but with DLP printer it should work out fine. It will be cool if it works but its better if it can be printed flat. One of the methods you can use to print this is splitting. Most 3D printer software allows you to split a model into parts. You could split this model in half down the middle and print each half on its own.
The bolts on the bottom should allow you to bolt each half to the bottom of the shoulder. The belt tension will also help keep the two halfs together.
I will try printing a split half and see how it goes.
That was a great suggestion. I split the part down the middle and printed it. One of the parts came out perfect, while the other warped (image below) and so I am reprinting the other half now and will let you know how it turns out. I've prepared the build plate with glue, painters tape, and have paperclips to hold the raft down just in case. Without a heated build plate, I have to babysit my printer.
The right part looks perfect! Are you printing PLA or ABS? If you are using ABS you can use the ABS pipe cement to glue the two halves together.
I am printing in PLA and was just going to use crazy glue and hope for the best lol
I generally use epoxy when gluing PLA as crazy glue doesn't stick particularly well (depending on your ABS), particularly with such a large surface area.
Although PLA is supposed to print without a heated platform, it does a heck of a lot better if you have one. This gear is only the beginning. When you get to the giant base, and the shoulder, I'm imagining you're going to have no end of frustrations.
If you can at all manage, I'd get a heated bed.
I printed on a heated glass bed using elmers glue, and my prints are glass flat with PLA.
I've been really looking at heated beds. I am still stock with my makerbot platform. It's a little pricey, but I might as well splurge if I'm to be doing large prints like these. I anticipate many frustrations with the larger parts. I'll have to get some epoxy, thanks for the suggestions!
The larger parts are a study in patience. I had a blackout during the printing of the geared base ring. Then I had a child poke their finger into the off button while printing out the wrist gears the first time around (70% through the print). They are full day prints. If there are any imperfections in your printer, they will be amplified.
lithium grease your various linear slides before you start
make sure your bed is as level as is humanly possible, then make it level again
Do a temperature calibration, on each filament. I have filament that is rated from 180 - 225, but generally prints well at 210 in my cold garage right now.
Print fairly slowly. My prints were done at 40mm/s. I bumped it up to 45 for the last couple to compare. The last thing you want is stringing, or bumps, and temperature and speed can affect those tremendously.
Print at a reasonable layer height for the part. .3mm is probably fine for those giant parts (up through the shoulder)
hand tune the placement of support material if at all possible with your slicer. Nothing worse than using the automatic placement, and ending up with more support than print.
And definitely get the heated bed. You'll spend more in wasted plastic and frustration than the heated bed is likely to cost.
And good luck! With a single printer, it's probably a week or more of printing, keeping it humming roughly 24/7.
Wow that's unfortunate that you had some complications while printing. Fortunately, you can maximize your throughput with having six printers. I've been luricating my axis between every print and even during prints if I hear squeaking. I have not had any warping after the first print now that I've enclosed my printer and applied glue stick to my painters tape covered glass build plate.
For the larger shoulder parts, I've found a free scripting 3D CAD program called "OpenSCAD," and a function for it called puzzlecut. This will quickly allow me to cut models into interlocking parts to decrease gaps I'm seeing in flat surface cuts. Surprisingly, the crazy glue had enough surface area to bond to, so for the base_gear part I believe it to be sufficient enough. And I intend to do 0.25-0.30 depending on time and remaining PLA.
That's a good point about the heated bed I hadn't considered. With all the frustrations of warping, it would definitely be worth it. I'll see if I experience any more troubles now that I'm enclosed and glued, and if I do, I will be inclined to purchase one.
I've split the shoulder part adding interlocking pins to fit it together. This allows small build volume printers to be able to print this. Hopefully things line up when printed and I'll post the result. :-)
That's a great option. I wonder if you could go even further, and split into essentially 4 pieces, for slightly different reasons.
The big loopy part that forms the 'shoulder' at the top could be sliced flat. Then I could lay that flat against the print bed. Do that for the two sides.
Then, the tear drop curvy pieces at the base can each be split into two large pieces, essentially identical, so print one for each side.
With the big should pieces printing flat, the major stresses will be along long continuous lines, rather than across layers as they currently are.
This could also solve your small print bed problem, while at the same time giving us an option to print pieces in an orientation that provides a part that better matches the stresses of the application.
And that looks like 3D Builder. Did it fix whatever error it had with that part (the red under square).
That's a great idea! I was trying to keep it under 2 print jobs, but I like to idea of being able to print the shoulders flat. Contingent I find a suitable way to interconnect them. If the shoulder pieces had male pins on them, printing that side flat on the bed would need many supports. If we make both of the shoulders female connectors, that would work well for those pieces. However, the tear drop pieces would then require supports for printing the male pins. Of course they can all be printed flat without interconnecting pins and just epoxied or glued together, but considering the application, something to reduce some of the forces acting on the part would desirable. Pins would make sense for any shearing force acting on the part(s). And maybe there is little shearing force to begin with and these aren't necessary.
My printer's limitation for the shoulder is its z-axis. The part is just outside of my printers limitations. I had planned to lay the parts flat with supports on the outside of the shoulder piece. Unfortunately, this would likely result in a part with a rough surface from removing supports/rafts. Your idea of printing them flat on the inside of the shoulder is a good option.
Oh and it is 3d Builder! It rendered a better visual for snipping tool than OpenSCAD. The red box was an error that was repaired.
If you do flat shoulder pieces, with female pin places on all parts, you can actually use steel pins to join them together. glue and pins should make for a part that's fairly solid, if you go that route.
I decided to split the parts and make one shoulder (call it left) with male pins and one shoulder (call it right) with female pins. Then the two middle pieces were basically M-F connectors between the two. I printed the left shoulder with the pins facing down with supports and a raft. And I printed the right shoulder just flat on its surface. I made the mistake of not gluing the right shoulder so it curled a tiny bit, but I was at work and the process was already 50% through the 17hr print when I had noticed. It wasn't curled to the point of rendering it useless, but it did result in a gap between the M-F connectors and the right shoulder. I had to break the male pins off on that side as no matter how much I dremeled with my conical sand head, it still wasn't going to be flush. So 3 out of the 4 connections are mated with M-F pins (I didn't leave enough tolerance for the female pins so I had to dremel them a bit larger too). Then, I covered them in super glue. Not sure if this will present a structural/strength problem later, but if it does I am prepared to either ask my friend with a larger build volume to print this part for me or reprint myself in a different orientation or with different slice settings to make mating the parts more structurally sound. I am nearly done after 10 days of continuous printing. I am printing the wrist now and am halfway through my 3rd roll of 1kg Hatchbox filament. (I had two failed prints in 10 days estimated to be a total 200g of wasted plastic.)
I am also considering spot welding the pieces with an iron or a cheap 3d printing pen to bond the pieces together.
That looks great! This is a really good part splitting. From the pictures I can tell you that it will assemble just fine. Have you ever tried bonding with one of the 3D print pens? I wonder if that would work.
I can tell you that the shoulder is the hardest parts. Most of the other parts I had success on my first try.
That looks pretty darned good. That's pretty much how I envisioned the split going. I really like that blue color. Haven't seen that before. I might try one myself, although without the male pins as I want to get both shoulders flat on the build plate.
Nice to see that Hatchbox is working for you as well. I printed my first set using mostly Hatchbox, and I really liked the ease of printing, a high quality.
Soon enough we'll be sharing pics of finished and flailing robots!