THE LATEST
Check out some updates as they pertain to some of my recent and most significant projects!
APRIL
Y-axis print bed assembly- I've yet to mount the rollers.
Lining up the X-axis on the Z rails.
X-axis mounted- properly tensioning the rollers was done via. elliptical bolts on the innermost rollers.
Z-axis screw placeholders with the X-axis mounted up.
SPOKE
Finalizing assemblies:
This includes mechanical assemblies, such as landing gear, the rotating EDF blocks, control surfaces, and more.
SPOKE
Heat set inserts-
Behind almost every important attachment point there is now a threaded heat set insert.
Assuming I'll be able to get a good stable fit between these and the LW-PLA I intend to use,
these inserts will serve a dual purpose: they will make the attachment points stronger, last longer, and they will also make the entire airframe far easier to assemble and disassemble. Because reliability is one of my main goals, these are all very important qualities.
I plan to press the inserts in by use of a soldering iron.
MARCH
METER
Y-Axis belt drive-
Although it is arguably easier to design and assemble, a disadvantage of my cartesian design is the weight (in the form of the print bed and the print itself) that must be carried along the Y-axis. To facilitate this weight, as of now I'm using two 40-42 NEMA stepper motors in a belt drive configuration. I haven't mounted the print bed yet, but my current design appears to be sufficient.
METER
Development of a custom direct drive extruder-
An important aspect of my large format printer will be print speeds- I intend to print primarily from LW-PLA, a doped version of PLA with heat-released foaming components that allow volume expansion as a function of the heat block temperature. In order to reliably print using an expanding filament, not to mention at a respectable rate, the backpressure on the nozzle must be fairly high. The direct drive's proximity to the extruder ensures this feed pressure is maintained- a Bowden tube, in comparison, seems to dampen this pressure.
This is my first crack at a direct drive system- there are plenty of commercially available designs out there, and I may switch to one of those depending on how successful I am with this version.
METER PRINTER (SPOKE)
By and large, when 3D printing an assembly, more parts means more attachment points, more attachment points means more surfaces, more surfaces mean greater volume, and finally, greater volume means greater weight. Therefore, the less parts, the lighter the aircraft, and the more control I have over weight distribution. Due to this, I have designed and recently begun constructing a printer capable of producing parts in the range of 960x600x960mm. Printers of this size are often coreXY as opposed to cartesian due to the efficiency disadvantages of mobilizing large and heavy parts on the Y axis. I went for a cartesian design for cost and construction reasons. CoreXY printers require more advanced control boards and the frames generally require a great deal more slotted framing.
An assembly of the main parts that will take advantage of the larger printing space.
Rear base, previously roughly 20 parts.
Rotating bulkhead mount, previously about 12 individual parts.
The completed base frame section of the printer, currently on the floor of my shed. I will need to get creative to house this thing.
The extent of what I've put together so far- the rest relies on printed parts, and my printers are tied up making rocket components at the moment.
FEBUARY
SPOKE
This is the print progress as of now. My main goal at the moment is to print enough parts to begin testing the engine rotate system.
SPOKE
The size of SPOKE, compared with my outline.
My longer term vision (I'd have to figure out how to get in a room with a printer up for the job) is to print a variant of this in 3-5 major parts. This would decrease assembly time, attachment points, weight, and cost, all while improving on structural integrity. Plus it would be incredibly cool. And more scalable. The list goes on.
JANUARY
SPOKE
I'm just getting to some of the more complex mechanical systems I've wanted to implement on SPOKE- all is relative, however, because this system is a lot more simple than the landing gear I designed for FWD-II. This system is nice, because it works on one servo, requires minimal suspension, and easily locks in both positions.
SPOKE
These are the latest images on the progress with SPOKE. The main compromise I am continuously having to make with this design is on the size of individual parts- I am currently in the process of modifying my 3D printer in order to be able to access 500mm in the z axis, which will be critical for producing some of these parts. Fewer parts mean fewer connection points, which ultimately means lighter weight and more versatility in terms of center of gravity placement and modification.
SPOKE
(earlier version of rotate system)
I'm essentially trying to answer two questions with this design:
First, what is the minimum amount of added volume needed to make a delivery box fly at ~25m/s?
Second, is this added
volume enough to house a VTOL system in some capacity?
I don't have a precise answer as of yet, but I have come some way in developing a response. A staggered tilt-EDF system would be able to, in my estimates, be compact enough to fit in my lifting body design, versatile enough to provide thrust in a horizontal and vertical position, and have the ability to suppress and control noise via blade pitch. Each side of the body will have two inline EDFS, each with a similar configuration as my pictured design. One reason this is traditionally not done with EDFs is because of diminishing returns on the second inline EDF- the high airflow makes the blade pitch ineffective. One solution I'm working on involves a variable pitch system allowing sufficient thrust at both attitudes. This is, as this was written, still very in progress.
ROCKET
This design extends upon the same mentality of getting as much testing done in a launch as possible. In order to help facilitate this, I've integrated variable extension flaps into the design. Instead of needing to adjust a nosecone or center of gravity to adjust for altitude relative to weather conditions, the flaps can be fixed to a range of positions. The flap mechanism itself is incredibly compact, and can be added or removed to the rocket chassis.
SPOKE
As I write this, spoke is a very current and early design- I don't have a mockup
yet. The main idea: an aircraft doesn't inherently have to look like an aircraft. Conventional wings snap, get in the way, and fit awkwardly in boxes. They provide lift and control in flight, which is obviously critical, but in every other
respect they are an entity which must be accounted for, not the other way around. Trying to improve E-Commerce supply
chains by making more efficient ground routes is like beating a dead horse, and so everybody is clambering to get their hands on the promised 'delivery drone'. So many of these designs are caught in wing limbo- either they remove them completely, compromise their effectiveness, become inconvenient, or
some combination of these possibilities. This is a really fun and thought provoking problem to design for, because so many typical design conventions are thrown out. SPOKE is a design I'm working on to embrace the dissolution of these conventions.
DECEMBER
ROCKET
One of my most recent TARC designs- everything on this new rocket is built around the idea of modularity. There are only four main body components in my latest design, all of which print faster and are way stronger than previous versions. Screw in mounts for every part allow on-site swapping of components, which allows for easier testing and on site adjustment for windspeed, humidity, and other variable factors. The rocket is shorter with a larger body diameter, allowing in part for the quicker print times, and the wide but incredibly low infill walls allow far more strength for a similar weight. Electronics are also more easily accessible here. It is realistic to print at least 6 of these per launch at 2-week intervals. Swappable motor mounts and easy center of gravity adjustment combined with all of the other advancements specific to this model mean we'll be able to accomplish in one launch what may take three or four for other types of rockets, regardless of the site conditions.
FWD-II
This is the CAD assembly of another one of my recent projects- a larger and less modular drone capable of more payload versatility. Much of the work that goes into my projects often starts with comprehensive CAD assemblies that integrate every single part I intend to produce. I then virtually assemble and reassemble the design to try and iron out the inevitable errors. This render shows the result of this CAD planning, and hopefully will be insightful when contrasted with the physical 3D printed objects. I'm still in the stage right now where I'm editing and reconfiguring parts, so the number changes frequently, but at the moment this design consists of 102 printed parts. From this number alone, it's pretty clear that I went for a different design doctrine here than some other designs meant to be easily and inexpensively manufactured.
FWD-II
Pictured here is what I have printed so far for my high fixed wing drone prototype. So far, everything appears to be coming
together properly, save for a few screw sink holes I forgot. Ultimately, this specific model will serve as more of a draft than anything- I intend to eventually print it in LW-PLA. LW-PLA is deceptively expensive per kilogram- the low density allows for large volumes to be printed off of a relatively small amount of material. It's still costly to mess up a print, however, hence my draft prints in normal PLA to diagnose any issues before switching to the real stuff. I went for cubic subdivision as opposed to gyroid here, as can be seen clearly on the right side images of the pre-sprayed wing junction. This infill provides better support for a few of the internal components, which was a consideration when optimizing GCODE
FWD-II
Large, fixed wing drones typically don't compete for the same bandwidth as quadcopters. They're meant to stay aloft for longer, carry larger loads, and have a certain degree of versatility. With this in mind, it's not unreasonable to expect
landing gear on the High Fixed Wing Drone II (pending more creative name). I tried to go for the simplest designs I could come up with- as one does- and ended up with what is seen on the left. For the rear landing gear, I took inspiration in large part from gear designs that originated from the US Navy's early WW-II requirements. At this point in time, simplicity was still chief among concerns, and the second most important factor was ability to survive hard landings on carrier decks/
short runways. See 'PBY-5a landing gear'. A rectilinear mechanism was the way to go for me as well. The front gear is
based on my own requirements- the important part is that they lock in the down position, destressing the servo on takeoff/ landing.
FWD-II
I put together a testbed to demonstrate my gear system. As I'm
currently working through the design, I'm learning that screws
as joints and 9g servos aren't exactly the best fit. While this
joint method is easy, I'm not sure yet if it will cut it as far as
reliability goes. It's just a bit on the stiff side at the moment.
FWD-II
Mechanization of the printed front landing gear, as shown in
the diagram on slide 20, is accomplished here via the laziest
Arduino code the world has ever seen. I ditched my
convoluted button system for a program where I simply type in
the corresponding position and upload it to the board.
LW-PLA
LW-PLA has the potential to change the way things are done in the 3D printing sphere. Baking soda works on a chemical
level because, when it reacts with an acid, it releases CO2, expands in volume, and makes the whole item less dense. LW-PLA releases CO2 on a temperature basis. 200c will result in a material essentially identical to normal PLA, whereas 270c will create a foam. This volume expansion is not confined to
outside of the nozzle- flowrates and specific heat change from temperature to temperature, and this must be compensated additionally by adjusting printing pace. TL:DR- this stuff is
complicated to get right. When you do though, it's not only lighter- the low volume and high surface area means incredible layer adhesion, so you also get strength. I'm at the point now where I can produce parts of about 55% less weight.
The only issue is that my current expertise results in a print success rate of a similar percentage.
Improved Wing Structure
The heavier a plane is, the harder it is to get it flying. Once it's already flying, there's not as much of an issue. Taking off in short order, however, is one of my main goals in this endeavor. Therefore, I set about finding ways to shave weight. With a fixed filament density, this is largely a matter of reducing internal surface area while maintaining internal support proximity. Evolution luckily figured this out before me - the gyroid is a mathematically minimal 3D structure that can, relevantly, be seen in the wings of butterflies. The virtuous gyroid certainly beat my next favorite structure, the cubic subdivision. One initially unconsidered bonus is what happens when you hold the plane up to the light.