Sometimes, I make life hard for myself…
2D DRAWINGS
I had not even built anything yet and this was already getting out of hand. Dr. Wettergreen asked me what my idea for the project was, and I responded something along the lines of “some kind of piston like motion? like in an engine?” He approved it, and also said “Probably only do one piston, otherwise things might get too complex” Boy was he right.
I was designing this thing in Fusion 360 and had the basic idea and shapes laid out, and I thought: “Wouldn’t it be cool if it looked like an aircraft engine? A radial engine?” And then: “Ah! wouldn’t it be nice if it had a few valve mockups actuated by a cam like in the real things?” And to finish: “Ooh ooh! And lets make the top shiny aluminum!” Of course, things in Fusion always look easy. You design this, then that, add a little bit of such, and by the end you have a very nice looking design, which you can animate, and everything looks great. You never stop to think that Fusion is a tolerance-free environment. Everything is exactly the size you tell it to be.
It’s even easier to get carried away when you plot all your nice designs in a very pro-looking sheet, with assembly diagrams and fancy exploded views. You go to bed thinking, “wow this project looks real neat! Can’t wait to start building it!”
WHO NEEDS LOW FIDELITY ANYWAYS
When starting to decide on materials for my low-fidelity prototype, I almost immediately realised two things. First, cardboard sucks. It bends, chaffs, squishes, deteriorates, and I’m pretty sure if you look at it hard enough it self-destructs. Second, I already had the entire thing designed in Fusion. I knew that the parts would fit closely enough, and that things like the piston travel distance, crankshaft rotation and such, would work. It made me feel like at this point, any time wasted working with materials other than the final high fidelity ones, was going to be time wasted. Overall, it seemed unncecessary to produce a low-fidelity prototype, and whatever proof-of-concept knowledge I could gain from it I already had from the Fusion motion study. So I decided to start immmediately with the high fidelity prototype.
The first day, things were going great! As long as I could sweet talk the lab techs into lending me more wood, I could keep cutting parts. When I got to the waterjet, it was still great! My only trouble is that one of the parts was too big for the waterjet, so I had to chop it up into thirds, and cut each third independently. I spent the entire day in the OEDK, and by the end, I had the first prototype of one of the pistons, and a few other important parts like the crankshaft core, and the top aluminum cover.
MACHINES BRAKING DOWN? CAN’T LET THAT STOP ME
The second day, which I also spent in its entirety inside the OEDK, was a lot more… eventful. And by eventful, I mean that both the laser cutter and the waterjet bailed on me, hard. Four times total, I walked into the laser cutter room, to see an “Out of order” sign on it. Four times I had to troubleshoot it, and fix it so I could keep cutting with it. I now know more about the new laser cutter than I’d like to.
Speaking of machines I got intimately familiar with, lets talk about the waterjet. Its a very nice machine. Lovely. I was happily sitting there, watching the cut, and all of a sudden water starts bursting out of the bottom of the abrasive container. I did not know it immediateley, but the nozzle had clogged. When that happens, the high pressure water has nowhere to go, except the abrasive feed tube. So in that second, the jet of pressurized water shot all the way back up the abrasive tube, into the abrasive container, and showered me in water and sand. I rush of to turn the machine off, and after a lot of confusedly looking at it, I realise what had happened. Undeterred, I open the browser on the computer, access the webpage of the manufacturer of the waterjet, and look in the maintenance section until I find the “What to do when your nozzle clogs?” page.
I read through the whole thing, and even watch the added video, and then proceed to take apart the whole abrasive feed assembly. I disconnect the tube, remove the container, throw out the now useless wet abrasive, clean all the parts with dry compressed air from the hose in the machine shop, reassemble everything, and do a run without abrasive in the container, to make sure that the water was flowing out of the nozzle again. I see a nice jet of water come out of it, so I refill the container with new, dry abrasive, and do a test cut. Sure enough, the feed is clear, no water coming out of anywhere, and the nozzle is doing very nice cuts on the metal plate.
Finally, tho, I had all of the parts done! Here is an awesome picture of all the parts laid out on the table!
HOME ASSEMBLY TIME – SCREW THE IMPERIAL SYSTEM
By the time I was able to finish cutting the rest of the parts, it was already 10pm, so I had to take all these parts home, to assemble it there. Anyways, this is the fun part. Like a lego, I put it all together, very happy to see physically what I had been toying with in Fusion for a while. All of the parts fitted niceley, save for a few tolerances that had to be adjusted on the piston bulkheads and the camshaft brackets. However, when I got to the parts that have rods going through them, (The piston arms and the crankshaft core) I run into a bit of an issue. The name of that issue is the imperial system.
I allready knew what wooden rods I was going to be working with! I measured them (maybe too quickly) with a ruler before even designing the thing! I knew I had two diameters to work with, 20mm and 10mm! But it all turned out to be a lie. The 20mm rod was actually 3/4 inch, which translates to 19.05mm, and the 10mm rod was 3/8 inch which is 9.525mm. Ugh. This made all of my rods and axles a little bit too thin, which made all the fits very loose. Obviously none of the moving parts worked now. The thing was dead on top of my bed.
MODEL 2
I spent the rest of the night redesigning half of the parts to fit the updated rod sizes and the new tolerances. The next day, I went to the OEDK, and cut them all. On top of that, I decided I could cut the 10mm axles with the laser cutter too, which lets me make them the precise length needed very easily, as well as giving them the burned dark brown tone on the ends, which matches the piston heads.
This time, everything fit snuggly, and after a few taps of the mallet, I had some very nice, sturdy, swively pistons, and some solid axle connections on the main shaft.
I test the rotation on the whole system… and it drags like crazy. It has so much friction in it I can’t even move it. Any ideas?
WAR AGAINST FRICTION
After assembling the Mk2 engine, I realised that if I did not do something about the friction this thing was not going to move. I had originally somewhat planned for this already, by making the casing plates on the crankshaft and the piston heads out of polished aluminum, which slides much better on the wood. However, it seems it was not enough. I brainstormed for a bit and got two key ideas that would get me out of this mess.
The first idea, was that while wood-on-wood contact, especially the burnt edges, had a lot of friction, wood-on-acrylic did not! This helped a ton in the valve actuation mechanism. The cam was getting stuck there all the time, and switching the cam actuators to acrylic made the whole movement a lot smoother.
The second idea, was to insert strategically placed washers in between the disks of wood that had a lot of contact. The crankshaft slid fine on the body, but the two disks that connect it to the main shaft were grinding too much against the outer casing. Now, the OEDK does not have washers this big, so I custom designed and cut two washers out of the thinnest steel I could find (304 Stainless, 0.033 thick). I inserted these washers in between the wooden disks and the casing, and this helped a lot in reducing the overall friction of the system.
With these changes the engine rotated a lot more smoothly. It could still probably use some lubrication, and some tighter tolerances, but it rotates smoothly enough to be easily turned by hand.
THE END PRODUCT
It was done. Completely assembled, rotating beautifully, and rocking that bare wood on bare metal look. The finishing touch? A pair of fancy vinyls on the underside.
(Although if you ask me, It really needs a planetary gerabox on the bottom for added speed, and a thick propeller on the top! – Maybe I’ll add those at some point!)
Here are some shots of the finished model:
Full view:
Underside detail: (You can appreciate the elastic bands that retract the valves)
I’VE WON, BUT AT WHAT COST? LIKE… EXACTLY HOW MUCH?
How much has this thing cost me? Well, if we count the blood, sweat, tears, lost sleep, lost meals… we would never get anywhere, so we will just count the money.
Material cost:
The final product uses a grand total of 7 different materials:
Birch plywood: (1/4 inch thick)
Price = 4.17$/sqft
Used = ~7.5sqft
Cost = 31.28$
Aluminum sheet: (Since I didn’t know what alloy it was, assume it is Alloy 6061, the most common – 1/4 inch thick)
Price = 22.50$/sqft
Used = ~2.5sqft
Cost = 56.25$
Steel sheet: (304 Stainless – 0.033 inch thick)
Price = 17.49$/sqft
Used = ~8sqin = 0.056sqft
Cost = 0.98$
Acrilyc sheet: (Clear – 1/4 inch thick)
Price = 14.61$/sqft
Used = ~12sqin = 0.083sqft
Cost = 1.21$
Machine screws: (Steel M6 screws and nuts – 50mm and 12mm lengths)
Price = 0.18$/50mm screw + nut – 0.10$/12mm screw + nut
Used = 18 50mm screws +18 nuts – 12 12mm screws + 12 nuts
Cost = 4.44$
Vinyl (Black)
Price = 1.20/sqft
Used = ~6sqin = 0.042sqft
Cost = 0.05$
Rubber bands (like, regular ones?)
Price = 8.60/500 bands
Used = 6
Cost = 0.02$
Tool time cost:
WaterJet: (4.5 hours) at (~35$/hr varies depending on what abrasive you use) = 157.50$
Sandpaper and file: (0$/hr they don’t get more expensive the more you use them, and the price of the used piece of sandpaper is negligible) = 0$
Laser Cutter: (6 hours) at (~12.70$) = 76.20$
LabouR Hours cost:
The average pay for a workshop worker (similar conditions to the OEDK, not a factory) seems to be 25$/hr
I spent about 42 hours actively working on the pieces and assembly, which totals to 1050$.
The average pay for a mechanical engineer (Think up and design the engine model) seems to be 41$/hr
I spent around 12 hours designing the model and updating it as needed, which totals to 492$.
TOTAL:
Adding up all the individual costs brings the total up to:
1869.93$
If we ignore the graphic engineer’s pay, as it is a fixed cost, not a variable cost, (The design is only designed one time, after that we can make it indefiniteley) it brings the final cost of producing a single diamond to:
1377.00$
That is one. Hella. Expensive. Toy. That being said, it does look very cool. What are the chances I get to keep it after this semester is over? Pretty please? 😀
