Iterations on a corner. Printed with 0.3mm layers at 150mm/s and 20% infill.

3D Printing
, , , ,
1

Iterations on a corner. Printed with 0.3mm layers at 150mm/s and 20% infill. Surprisingly(!) solid for such a light/sloppy print.

Note the ultimate aim of my exercise is fast printing with a fairly large (~500mm cube) volume. That means a rapidly accelerating print head. That means the frame needs to be extremely rigid.

To my eye, a metal corner cube on the end of a 500mm lever is over-matched. The usual small triangle braces or L-brackets for the inside corners are also both over-matched.

Using extrusions as part of a linear bearing (in Z at least), the inside-corner braces cut into the print volume.

As the extrusions are aluminum, I want the major braces to be aluminum. Metals with different rates of thermal expansion are a source of error I want to avoid.

If I had a CNC router (a project for another month), I might cut side-braces out of sheet aluminum. For this iteration, strips of aluminum with a hole in each end are something I can readily fabricate. Tedious to assemble and align, but possible and perhaps sufficient.

My aim is to have a frame of rigid triangles. With such a frame, plastic corners might be sufficient.

This is seat-of-the-pants engineering. Also my Amazon deliveries included a laser pointer. I expect to fabricate a clamp for the pointer, and collect deflection measurements on the frame (with the laser projecting ~30 feet to an opposite wall). Also received a linear slide, linear bushings, and an assortment of pulleys, gears, and bearings to test and measure. Maybe not just seat-of-the-pants. :slight_smile:

Printed plastic corners can serve quite a lot of purpose, and simplify the overall design - if they work. The corners could also serve as… Feet to elevate the frame off the desk. Stand-offs for a printer enclosure. Mounts for stepper motors and worm screws. Supports for belt/pulleys. Could potentially reduce the BOM and simplify fabrication.

At least the above is a bunch of theory. On to the next experiment…

65e0fe5c1cb237526653d86a7b59591a.jpeg
65e0fe5c1cb237526653d86a7b59591a.jpeg4048×3036 2.62 MB

2

Solid block corners out of high-shore-index polymers have been around for a while. No reason to think yours would perform worse :slight_smile:

Although… my 8x4 CNC router uses cast alu right-angles and it’s a whole heck of a lot more rigid than a 3D printer.

As to acceleration, I was concerned with such forces when designing a printer with a carbon fibre/polymer print carriage, but using smoothie or tinyG seems to result in sinusoidal curves and much less jarring/clattering.

Please post the results of your laser testing - really interested in the opportunities for cast/printed polymer with aluminium.

3

Nice work, keep updating and maybe sharing the parts, I would like to try them when I finish my hypercube build.

4

If it’s an enclosure then the corners only need to serve as guides to keep the ends in place as the side panels are installed. Without panels, you’ll need diagonal bracing of some form, and that’s what gussets do on rectangular frames.

5

With metal corner brackets, I recommend the length of the bracket be about 10% of the length of the beam. So for 500mm extrusions I would aim for about 50mm of metal L-bracket length on the extrusion. It’s a rough rule of thumb.

Plastic corner pieces are fine IF you add triangular bracing or bolt on an enclosure panel.

6

@Mike_Thornbury In the current experiment, the corners are printed out of PLA with 20% infill. You think I might have the match of a dense, solid, high grade polymer? :slight_smile:

In truth, fast light prints save me time, and allow more iterations. Also I do not expect the current iteration to be good enough - but will measure to see.

Pretty sure your CNC machine handles substantial loads. Success in this case is equivalent rigidity - but with much lighter loads.

7

@Jeff_DeMaagd Yes, exactly!

If the enclosure is solidly mounted (there is some argument for damped mounting, to reduce noise). If the side(s) are currently attached (as you might want to remove a side to monitor a print - especially on an experimental printer).

If I had a CNC router, I could cut plexiglass for at least the four vertical sides (with large diameter mounting holes to avoid stress cracks), and perhaps be done with diagonals.

But … building a CNC router is a later project. :slight_smile:

At the present, my main concern is the rigidity of the frame around the XY movement of the printhead. Printing another absurd experiment, at the moment.

8

@Ryan_Carlyle … and you are pretty much on the focus of my current experiment.

Keep in mind I have 500mm sticks of 2020 extrusion. I have the frame assembled (with light plastic corner cubes), and it is as flexible as expected. Metal corner brackets with only 10% of the side dimension are at mechanical disadvantage. For smaller frames and slower print speeds, I suspect your rule is pretty good.

But I want to push both size and speed.

So … assuming my cheap i3 clone does not break (and it is making some slightly alarming noises), I will have more empirical answers. :slight_smile:

9

You don’t think 10% bracket engagement will be stiff enough? More is better, of course, but the standard 40-60mm x 3mm aluminum corner plates Adafruit sells for 2020 are extremely stout. I usually use those plus Misumi right-angle brackets and end up with printer frames strong enough to batter down doors, plus kinematically guaranteed to be square/true within the fabrication tolerances of Misumi.

10

@Preston_Bannister maybe not the match, but ill bet its got a high shore rating :slight_smile: Theres always casting and moulding…