Part Two: DIY Drones in DM&F
By Ben Clifford, Student Ambassador for Digital Modeling & Fabrication Program at IYRS
This is part-two of the DIY Drones blog by Ben Clifford. Read part one.
Back in the electronics section of the project, we began to use the SBC and other new pieces of circuitry to create a GPS tracking unit for the drone. The SBC we used is called a Raspberry Pi. It is a surprisingly powerful miniature computer that uses a Linux operating system, and can connect to almost any other type of electronics with the right convertors and connections. It can even connect via Wifi.
We used a wiring diagram from the Adafruit website to connect the Raspberry Pi to a small Lithium Ion battery through a power convertor circuit board. This lowered the voltage coming from the battery so that the Pi did not burn out when we turned it on. We also got a basic introduction to the coding language Python, and a programing interface software called PiBakery. These allowed us to create more programs to run the GPS system, although we are still working on getting the code running.
To further prepare for the final assembly stage, we diagramed out where all the circuit boards and other electronics would go inside the drone chassis. As you can see, it was quite a tight fit. These additions also helped us account for any additional space we needed to add to our drone shell models before we began to create the final files to mill them out.
We used the software MasterCAM to prep the shell models for the Tormach mills. We had been learning the basics of this program in a separate class, so we had a good foundation to build on when we began the process. As you can see below, the process involved exporting the model from Rhino and opening it in MasterCAM. We repositioned the model, and rejoined all the surfaces into a solid format the CAM software could recognize. After this was done, we selected the profile for the Tormach milling machines and chose our tools (a Ball Endmill).
While the first mold I milled came out great, there were issues with how I designed it that would affect the vacuum forming process. The placement of the fin and the straight walls of the nose cap would make it almost impossible to get a decent pull off of this mold, as well as making the plastic grip the vertical surfaces and require excessive amounts of force to separate them. This might damage the mold, since we didn’t have time to prep and strengthen the molds with resin.
I re-oriented the model in Rhino, as well as making several adjustments to turn vertical surfaces into slanted ones. These slanted sides allow for easier removal of the mold from the shell. I created new toolpaths in MasterCAM and ran the final files for my molds.
While I forgot to get a video of it, the vacuum forming was a success! I spent a few hours drilling holes in the molds to allow for all the detail to be captured by the vacuum, which was the key to success. The vacuum can only pull air down and over the molds, so any detail that would form closed pockets to trap air must have relief hole drilled down into the vacuum chamber. The only issue with these molds was that the fin was still too flimsy and close to the other parts, and didn’t transfer to the plastic properly. I also cracked off some small details from the molds getting them out.
The final step for the electronics was to wire all the basic flight components together and set up the drones flight control programing. We began by wiring the ESCs (Electronic Speed Controllers) to the motors. This required us to cut off the heat shrink material covering their circuit boards and re-solder the wire connections. This was easier than expected, since there was a large blob of solder to heat up and poke the wires back into. They had to be wired in an alternating sequence so that the motors would all spin in opposite directions to provide a balancing downward force.
Once we had rewired and connected the ESCs to the motors, we soldered the positive and negative wires to the Distribution Board. This board connects the ESCs to the battery, and allows for multiple motors to draw power at an even rate. The ESCs also have a separate set of wires that connect to the flight controller Arduino, which sends the commands to move the motors, which spin the propellers.
Once everything was wired up and the chassis was screwed back on, we used a program called LibrePilot to program and calibrate the flight controller. It uses a guided setup wizard to walk you through the process and make sure you don’t skip any vital steps. My calibration process was cut short, as I discovered that one of my ESCs was faulty. Even though I couldn’t get any further with the setup on the final day of class, I still had one final step to complete this project: attaching the shell.
I used a box cutter to cut the shell away from the rest of the plastic, and a Dremel tool to clean up and smooth out the edges. After this was done, I checked the propeller clearance on the laser-cut chassis. It fit perfectly, and the only thing left to do was to attach it to main drone chassis with all the electronics.
While the GPS components are still waiting to be completed, and my drone has yet to fly (no, the drone in the beginning was not mine), I believe this project was a success. We learned many new skills and applied some of the CNC machining and 3D modeling techniques we had been practicing up to this point. And with the electronics demystified, we can move onto bigger and more complex projects. Especially our Capstone. I now plan on including LED circuits and a voice changer into my Doctor Doom suit build. Stay tuned for an update into my research for this massive project, as well as a 3D modeling adventure to recreate a rather complex adjustable chair.
Next time, we design butt-shaped impressions in an office chair! ONWARD! 01/26/18 On the backbone, framing, and planking
By Daniel Mollet, Student Ambassador for the
Boatbuilding & Restoration Program at IYRS
Well, after disappearing from the blog for far too long, I’m back. I’m sure that my legions of fans are rejoicing already. Unfortunately, right when things started getting the most interesting in shop, my work obligations just about tripled, and I spent a long while snowed under both literally, by snow, and figuratively, by grading responsibilities. Such is the life of the online English professor learning to build boats in Newport, Rhode Island. But the snow (both literal and figurative) is now gone and I hope I’ll be able to keep up a more regular schedule of updates in 2018.
Since my last post a lot has happened in Restoration Hall. I’ll try to get you up to speed.
The backbone was the first significant construction project of the year, in which we had to fabricate and assemble the various components of the backbone of the boat.
[Our boat, with the garboards removed and the old backbone fully exposed.] [Here I am rough cutting the keel from a slab of oak.]
The keel was cut from a thick slab of green white oak still so wet that we had to override the safety feature on the Sawstop while rough-cutting the rabbet, for fear of activating the finger-saving, saw-stopping mechanism for which this table saw is named. We laid a pattern on the board, traced around it using high-tech marking tools (a washer and a Sharpie), and cut it out with a circular saw just outside of the line. Then we planed down to the line and cut the rabbet out on the table saw using an incredibly safe operation that in no way went against every single rule of shop safety we had been taught in the preceding weeks. Then we cooked the whole thing in the steambox and clamped it overnight onto a bending jig, and then ultimately to the boat.
[Dana, Hans, Phredd, and Kate preparing to clamp their stem laminates onto the jig.] [The stems are glued up and clamped to the molds.]
We assembled the stem after steam-bending 18 thin strips of mahogany onto a jig and leaving them clamped overnight. The stem assembly involved more clamping and lots and lots of epoxy mixed thickly with silica. After twenty-four hours of curing, one of my intrepid teammates painstakingly shaped the rough stem into the fine curve that will slice through the waves of Newport Harbor on Launch Day.
I assembled the centerboard trunk from two pieces of pine and two oak staves, called standards. The centerboard trunk is a long, narrow box that holds the centerboard on a pin so that the sailor may raise and lower the board as necessary. The assembly was pretty straightforward, but shaping the bottom of the box into a convex curve that is the mirror image of the concavity of the keel was a long slow process of trial and error and blaming my tools for my mistakes.
[The centerboard trunk is dry-fit to the keel. Seems like it’s going to work.]
The transom, stern post, and skeg were all made by tracing plywood patterns onto thick slabs of white oak. The Beetle Cat is a truly overbuilt vessel. The skeg and sternpost are pretty straightforward processes–it comes down to creating flat or fair joint surfaces and little else, but they took time, patience, and exactitude that I did not possess after shaping the bottom of the centerboard trunk, so my teammates took the lead on them while I laid out the transom and plotted the points where we ought to drill so as to insert the three bronze drifts (rods) that would run through all three boards, working in concert with epoxy to keep it all together. When it was all pre-drilled and ready to go we glued the three pieces together, somehow managing to put the middle piece in upside down. I expected this mistake to be a world-ending catastrophe, but in the end it was not and my team and I learned a cheap lesson on the benefits of moving a hair more slowly.
When all the pieces were finished we used Elmer’s glue and Scotch tape
* to slap them all together and we stuck the thing in the boat. With the backbone finished, we were ready for framing. [The boat’s got a new backbone.] On framing, or: Breaking frames
You can’t frame a boat until you’ve cut the frames, so my team and I, along with another team that had finished their backbone as well, spent the better part of a day milling ¾" square frames. (Again: the Beetle Cat is impressively overbuilt.) We planed and sawed and routered and planed and sawed and routered and planed and sawed and routered until we’d gone through all of the framing stock we had–most of an oak tree.
[The boat, with the new backbone installed, with every other plank removed, and with ribbands firmly holding what’s left of the hull to shape.]
At ¾" thick, each frame needs to cook in the steambox for about 45 minutes before we can hope to thread it through the old planking (held in place by ribbands) and bend it into shape, using the old hull as a jig. We found that, in many cases, even after cooking the frames for that long we still were unable to bend it into place without the pliant oak separating along the grain. I believe my team holds the record for Most Frames Snapped in a single day, at 13.
[The frames are in.]
Framing is a group effort that is most successfully performed by at least six sets of hands, so everybody had a hand in framing all the boats. no one team could have gotten their boat framed without volunteers from other teams, and as a result our class felt like a much more tightly-knit group after the week of se we spent framing up all five boats. What’s more, our boats were now ready for planking.
I won’t dive into planking too deeply here as we’ve only just started the process. My team’s boat has the garboards on and one of the first broad planks. The other first broad is spiled but not yet drawn on cedar because upon returning from the holidays the first years walked right past our boats and headed upstairs to complete a second lofting and drafting exercise.
[Phil and Jihong laid the battens for our plank lines and did a damn good job of it.]
But, the first step from a framed boat toward planking is fairing the keel rabbet and joining it up with the stem rabbet. A rabbet is a fancy word for a groove in one piece of wood that accepts another piece. The keel rabbet accepts the garboard along its length, and the stem rabbet accepts the hood end (front) of every plank. The long edges of the planks are carefully fit to match their adjoining surfaces (like the bottom of the centerboard trunk or the top of the skeg). When the edges fit, the 9/16" plank is steamed for just over a half hour, and clamped onto the boat overnight. The next day the edges are examined again with a block plane, the plank is lightly sanded, the inside gets a coat of primer, and then the fasteners go in.
[The garboards are fastened. Next up: first broads!]
For the last two weeks we’ve been lofting and drafting, so progress on the boats has slowed, but this week or the following we’ll be getting back down onto the shop floor and the hulls will start to take shape. We’re hoping to achieve our mid-March deadline for rolling over the boats, and my fingers are crossed that we can hold to it.
In future posts I hope to look at planking in a little more detail, go through the process of a dutchman repair, talk about the tool chest that we’re building in addition to our beetle cats, and take a look at first year side projects. Hopefully there won’t be any more two-month gaps in my publication schedule.
*Not really. I promise we used bolts, bedding compound, and screws. 01/25/18