Week 14 - Mechanical Design
Assignment
Design a machine that includes mechanism+actuation+automation
build the mechanical parts and operate it manually
Process
Gathering ideas and sketching the mechanism
Being in a group with no mechanical ground, me and my mates expected that the only limit for this assignment was our imagination. The hit with reality was hard, but still really fun and interesting especially because we wanted to craft a completely automatized catapult.
After several sketches and hours of brainstorming we got to a rather complete idea of the machine and its mechanism. A particularly helpful inspiration was a video with a similar war machine operated with an Arduino.
In the division of the tasks, I got to design and test the operativity of a vartical bar used to control the range of movement of the catapult lever.
Designing a prototype
The design my mechanism, the first thing that I did was to study the gears operating some of the best machines made by a Fab Lab. I’m talking about FellesVerkstedet which also manages a well-documented Github repo used to store all of their machine projects.From this source, I retrieved the thorough explaination of their chamfer rail mechanism and chose to adopt it for my system.
In the repo, there was a really handy dxf drawing of the rack and pinion system onto which I started to desing my mechanics. In order to use this file, I opened it on Adobe Illustrator and started stripping it of all the unnecessary parts
Essentially, what I needed was the profiles of the rack and pinion, so to extrude them on a solid shape in a CAD. Using Onshape as my weapon of choice, I needed a dxl file, and so I exported the drawing as such with a scale of 1 cm per unit (otherwise, the importation on Onshape would have resulted scaled).
Afterward, I sketched few elements that would have made a first prototype of the mechanism I envisioned with the dxl drawings applied and extruded onto various elements. Essentially, what I needed was:
a cart onto which lodge a motor and the bar
a pinion moved by the motor and passed through by the bar
a pillar with the rack engraved on it
In my idea, these parts would be the basic mechanic elements of the whole system.
I proceeded to quickly print these parts to see if they would fit and the mechanism worked.
Surprisingly the proportions worked just fine and my mechanism worked accordingly.
This was only a prototype - a proof of concept - and from here I started to develop in more detail the mechanics of my assignment.
Forking for a final design
Exploiting OnShape version control system, I branched the project and moved to set new parameters and parts of the final design of the mechanism.
Most importantly, I sketched a new pillar, made of layers of woods of different sizes and a 3D printed rack, I resized the various components to fit into the larger project and redesigned the shape of the pinion.
After designing these parts I virtually assembled them together and added the STLs of a Nema 17 stepper motor and a 22m circle bearing taken from the engineering resources site GrabCad to complete CAD of the mechanical system.
Crafting and assembling
With parts and sizes properly defined I proceeded to make the necessary pieces. For first thing I wanted to cut a couple of pieces of OSB wood to make the pillars, and since my group mate Eleonora needed to cut out some pieces of osb for the mechanics too, we merged together our files and set up a milling file for the ShopBot CNC machine.
I also needed few layers of plywood, so I made an SVG file on Adobe Illustrator to load to our Trotec Laser Cutter.
With all the pieces ready, I took screw and screwdriver and started to assemble them.
In the meantime, I set up in a 3D printer the file for the rack piece. For this purpose, I needed a robust piece but I wanted not to set up a 100% infill print because of material economy and strict time management.
So, I came up with the idea of setting up more thick walls. In this way, especially for the adjacent rack teeth, I can obtain a lot of support even with minimal infills. Furthermore, I adopeted a quarter cubic infill, which guarantees more resistance in every direction of the print.
Moreover, I printed the part laid on a side. I this way, the printing lines would be aligned with the direction of the rack, hoping (in my opinion) to not have them jamming the pinion or making it lose steps while climbing.
On Cura, I set up the print with the following parameters:
Layer height: 0.1mm
Trace width: 0.4 mm
Wall thickness: 1.2mm
Infill: 20%
Infill geometry: Quarter cubic
The result had a very good feeling of robustness and convinced me to apply to this time-saving gimmick for other pieces of this project (me and my mates were always overlapping and struggling to find moments with the 3D printers available. Saving time was very important for both getting resultd quickly and not annoying other people in desperate need of printing).
The following pictures is of the pillar properly mounted.
The other important piece was carrier that would use to keep all the parts together. The functions of this part is to hold the pinion, the bar and a stepper motor in place and aligned and it would climb all over the pillar, as depicted in the prototype model.
I made the hole where the bar would pass of exact size of a 22mm bearing, so that it could hold and make turn the bar much better.
After a couple of failed tries, I made a hole tester with the laser cutter with holes differing in radiuse of just few mm tenths. These helped me figure out which would be the best hole size to hold the bar inside the pinion that I had to print.
I found out that an optimal size is 8.04 mm - just enough the make the bar pass but not too loos - and set that dimension in my CAD software to design the pinion and printed it.
It its design I also added two screwholes to allow headless screws to fit in and anchor the pinion to the bar
In this way, I hoped to ensure the best traction from the pinion and rack movement.
Mistakes!
Just after printing the parts I needed I started to assemble the mechanism, but soon encountered a very problematic issue!
The carrier piece had an unresolvable flaw I didn’t notice in my designing: the hole for the bearing is misaligned why the pinion, rendering it completely useless.
I later noticed that the issue resulted in me sketching on the wrong face of the carrier 3d model, and after evaluating few solutions I decided tu just print a new, fixed carrier.
Few hours of printing later, I had all the my pieces I needed and now they were fitting just fine.
I also added few extra holes in the carrier for two reasons: 1. quicken the printing process; 2. have more room to work on the assembling process.
The result seemed to be solid and properly designed, but I had to face another important problem: the whole mechanism was frictioning to much! It was cumbersome, slow, heavy and would often get stuck.
Although I hadn’t made tries with a proper motor, it was way too obvious that this clunky system wouldn’t work in any way. I don’t know if the fault was in the material is used, in the sturdiness of the design or simply in the wrong application of mechanic principles, but I had to face the fact that I needed a differente solution.
Brand new mechanics
I had to simplify a lot of things. First of all, I needed a more fluid mechanism, with less parts and less complications. Furthermore, I wanted to save as more wheight as possible because otherwise my system would have affected also my mates’ works.
Reassessing all these things, I figured out that the way to proceed was by using a belt mechanism that relied on a couple of vertical bar onto which two slide bearing would move a horizontal bar. In my idea, this new system needed less parts and was more efficient, especially because it would rely on more apt ready-made components such as bars, bearings and belt. But still I needed to print few custom parts to to realize it.
First I needed something to support the upright bars. This wasn’t too hard as a task, also because I already made s model for a similar purpose for another part of the machine and it only took me the time to print a couple of them to get this done.
More non-trivial was to design the pieces that I would need to stick together the various parts of the mechanism. For the slide bearings, I sketched an holder that could also hold in the place the horizontal bar. The most counter-intuitive part of this item was setting the a sizes for the holes that wouldn’t be too tight (as my previous test for the bar diameter proved), but the addition of screwholes for blocking bars and bearings into place proved to be a really efficient safety measure.
Finally, I envisioned a belt clamp that could be reliable, versatile and easy to adopt, my point being that having to work with belts requires a lot putting into place, testing, removing, stretching, putting into place back again and so on, so having took the care to consider some time to save in this process resulted in a lot of less frustration and a more efficient work.
As it’s possible to see in the design above, when properly fitted, the belt’s ends emake a bend and are stuck inside a narrow space where they held tightly. With this piece, adjusting a belt is just matter of pulling it and pushing it back to the length requires.
Also in this case, I added screwhole for stopping purpose, but actually never had to use them.
Result
Albeit being really simpler, this mechanism worked way better than foresaw and also provided stability and easiness of use.
Another really strong point was that, except for the printed parts, I made it entirely out of makeshift components found around the lab or in other machines, making save time to me and budget to the lab.
Gist & Further development
I hardly thought that mechanics could be such a great experience, even in moments of failure. Every single issue, obstacle and flaw I met was the device for learning something new and applying different topics of knowledge. Adding fun to the experience was the fact that even though mechanic is a subject where some theory is fundamental to get a proper result, working with your own hands is the final test to see if anything can really work or not (and my first, failed attempt is surely a demonstration of that).
The real end of this assignment was in the integration of the other mechanisms frommy mates and their automation, which can be checked in my next assigment and in the group page.
Tools and software used
A vast amount of hardware and workshop tools