Mega-CNC Router Carves Styrofoam Into A Full-Size Flying Delorean

December 23, 2023 by Dan Maloney 18 Comments

When you own an enormous CNC router, you’ve got to find projects that justify it. So why not shoot for the sky — literally — and build the 1980s-est possible thing: a full-scale flying Delorean.

Attentive readers will no doubt remember [Brian Brocken] from his recent attempt to bring a welding robot out of retirement. That worked quite well, and equipped with a high-speed spindle, the giant ABB robot is now one of the biggest CNC routers we’ve ever seen. As for the flying Delorean, short of the well-known Mr. Fusion mod, [Brian] had to settle for less fictional approaches. The project is still in its early phase, but it appears that the flying car will basically be a huge quadcopter, with motors and propellers hidden under the chassis. That of course means eschewing the stainless steel of the OEM design for something lighter: expanded polystyrene foam (EPS).

The video below shows the fabrication of most of the body, which starts as large blocks of EPS and ends up as shaped panels and an unthinkable amount of dust. Individual pieces are glued together with what looks like plain old PVA adhesive. The standard Delorean “frunk” has been replaced by a louvered assembly that will act as an air intake; we presume the rear engine cover will get the same treatment. Interestingly, the weight of the finished model is almost exactly what Fusion 360 predicted based on the 3D model — a mere 13.9 kg.

[Brian] is currently thrust-testing motors and propellers and has some interesting details on that process in his write-up. There’s obviously a lot of work left on this project, and a lot more dust to be made, and we’ll be eagerly following along. Continue reading “Mega-CNC Router Carves Styrofoam Into A Full-Size Flying Delorean” →

Drone Motion Capture, The Open Source Way

November 23, 2023 by Lewin Day 4 Comments

If you want to do some really advanced flying with drones, you typically need to be able to track them in space. [Joshua Bird] has whipped up a drone tracking system that can do the job for as little as $20 with millimeter-scale precision.

The system uses four PS3 Eye cameras which can be had second-hand at a cost of just $5 each. They’re modified by removing their IR cut filter, and putting in an IR-passing filter in the form of a cut-up slice of floppy disk. The system tracks the drones via their infrared indicators and the known locations of the four cameras themselves, which the system is capable of mapping out automatically. By using four cameras, the system is robust in the event the view of a camera is occluded. The system can track multiple drones at the same time, with [Joshua] demonstrating it working with two drones each carrying three infrared markers. He has the system set up to send positional updates to ESP32 microcontrollers on the drones themselves, which command the drones to hold them in set positions.

Code is available on GitHub for the curious. We’ve seen other similar work before, too.

Continue reading “Drone Motion Capture, The Open Source Way” →

Variable-Nozzle Ducted Fan Provides Fluid Dynamics Lessons

October 19, 2023 by Bryan Cockfield 15 Comments

Any student new to the principles of fluid dynamics will be familiar with Bernoulli’s principle and the Venturi effect, where the speed of a liquid or gas increases when the size of the conduit it flows through decreases. When applying this principle to real-world applications, though, it can get a bit more complex than a student may learn about at first, mostly due to the shortcomings of tangible objects when compared to their textbook ideals. [Mech Ninja] discovered this while developing a ducted fan based around an RC motor.

The ducted fan is meant to be a stand-in for a model jet engine, based around a high-powered motor generally designed for drone racing. Most of the build is 3D printed including duct system, but in order to improve the efficiency and thrust beyond simple ducting, [Mech Ninja] designed and built a variable nozzle to more finely control the “exhaust” of his engine. This system is also 3D printed and can restrict or open up the outflow of the ducted fan, much like a real jet engine would. It uses two servos connected to collars on the outside of the engine. When the servos move the collars, a set of flaps linked to the collars can choke or expand the opening at the rear of the engine.

This is where some of the complexity of real-life designs comes into play, though. After testing the system with a load cell under a few different scenarios, the efficiency and thrust weren’t always better than the original design without the variable nozzle. [Mech Ninja] suspects that this is due to the gaps between the flaps, allowing air to escape and disrupting the efficient laminar flow of the air leaving the fan, and plans to build an improved version in the future. Fluid dynamics can be a fairly complex arena to design within, sometimes going in surprising directions like this ducted fan that turned out better than the theory would have predicted, at least until they accounted for all the variables in the design.

Continue reading “Variable-Nozzle Ducted Fan Provides Fluid Dynamics Lessons” →

Stretching The Flight Time On A Compressed Air Plane

October 15, 2023 by Danie Conradie 23 Comments

[Tom Stanton] has been experimenting with compressed air motors on model aircraft for a good few years, but keeping them aloft (and intact) for more than a few seconds has proven a tough nut to crack. His latest design represents a breakthrough — pulling off an impressive 1 minute and 26 seconds flight on 4 liters of compressed air.

The model incorporates an enhanced engine design featuring an expanding seal on the piston, a concept inspired by the old Air Hogs toy plane. For the airframe, he constructed lightweight wings using 3D printed ABS ribs on a carbon spar and reinforcing rods, all of which were wrapped in heat shrink film. Additionally, [Tom] incorporated a thin balsa former along the leading edge of the wing to help maintain its shape. The fuselage is also composed of a carbon fiber tube, and is outfitted with printed fittings to install the wings, V-tail, RC electronics, and soda/air bottles. A hollow nylon bolt holds the two bottles together end-to-end while allowing the motor to be screwed directly onto the front bottle. To conserve weight, each of the two V-tail control surfaces are actuated by single cables linked to servos, with piano wire torsion springs in the hinges to maintain tension

Despite successful flights, [Tom]’s trials were not without challenges. One crash threatened severe damage to his airframe, but thanks to a central 3D printed bracket that absorbed most of the impact, total destruction was avoided. Similarly, a printed shaft saved his expensive carbon fiber propeller from being damaged during multiple landings, an outcome that led [Tom] to devise a readily replaceable consumable connector.

A second video after the break offers a behind-the-scenes insights into this project including some fascinating technical details. For more on this project’s history, take a look at the initial diaphragm engines and his attempts to make them fly.

Continue reading “Stretching The Flight Time On A Compressed Air Plane” →

3D Printed Mini Drone Test Gimbal

October 9, 2023 by Richard Baguley 4 Comments

Drones are a pain, especially mini ones. When you are designing, building (or even reviewing) them, they inevitably fly off in some random direction, inevitably towards your long-suffering dog, hit him in the butt and send him scuttling off in search of a quieter spot for a nap.

[Tristan Dijkstra] and [Suryansh Sharma] have a solution: a mini-drone test gimbal. The two are in the the Networked Systems group and the Biomorphic Intelligence Lab who use CrazyFlie drones in their work, which require regular calibration and testing. This excellent design allows the drone to rotate in three dimensions, while still remaining safely contained. That means I could test the flight characteristics of a drone without endangering my dogs important napping schedule.

Efforts involved attaching a light tether that restricts the drone until we know how the it flies, but what usually happens is that the tether gets trapped in a rotor, or the tether gets tight and the drone freaks out and crashes into the ground.

Using a gimbal is far more elegant, because it allows the drone to rotate freely in three dimensions, so the basic features of the drone can be established before you let it loose in the skies.

The gimbal was designed with the CrazyFlie in mind, but as there’s nothing more exotic holding the craft down than a zip tie, it should work with similarly sized quadcopters.

Continue reading “3D Printed Mini Drone Test Gimbal” →

Control Tricks For Tailsitters

October 7, 2023 by Danie Conradie 5 Comments

An RC VTOL aircraft always makes for a compelling project, but ensuring the transition between hover and forward flight can be quite challenging. In the video after the break, [Nicholas Rehm] demystifies of the flight control algorithm required for a VTOL tailsitter.

Tailsitters are one of the simplest VTOL arrangements, the testbed here being a simple foam KF airfoil wing with two motors and two servo-controlled elevons. As with almost all his projects [Nicholas], uses of his open-source dRehmFlight flight controller to demonstrate the practical implementation of the control algorithm.

Three major factors that need to be simultaneously taken into account when transitioning a tailsitter VTOL. First off, yaw becomes roll, and vice versa. This implies that in hover mode, elevons have to move in opposite directions to control yaw; however, this same action will make it roll in forward flight. The same applies for differential thrust from motors — it controls roll in hover and yaw in forward flight. Nevertheless, this change of control scheme only works if the flight controller also alters its reference frame for “level” flight (i.e., flips forward 90°). As [Nicholas] demonstrates, failing to do so results in a quick and chaotic encounter with the ground.

With these adjustments made, the aircraft can transition to forward flight but will oscillate pitch-wise as it overcorrects while trying to maintain stable flight; this is due to PID gains – 3rd factor. The deflection required by control surfaces is much more aggressive during hover mode; thus PID gains need to be reduced during forward flight. A final improvement involves adding a brief delay when switching modes for smoother rotation.

For more interesting VTOL configurations, check out [Tom Stanton]’s RC V-22 Osprey, and this solar recharging trimotor

Continue reading “Control Tricks For Tailsitters” →

Autonomous Racing Drones Are Starting To Beat Human Pilots

September 28, 2023 by Danie Conradie 32 Comments

Even with all the technological advancements in recent years, autonomous systems have never been able to keep up with top-level human racing drone pilots. However, it looks like that gap has been closed with Swift – an autonomous system developed by the University of Zurich’s Robotics and Perception Group.

Previous research projects have come close, but they relied on optical motion capture settings in a tightly controlled environment. In contrast, Swift is completely independent of remote inputs and utilizes only an onboard computer, IMU, and camera for real-time for navigation and control. It does however require a pretrained machine learning model for the specific track, which maps the drone’s estimated position/velocity/orientation directly to control inputs. The details of how the system works is well explained in the video after the break.

The paper linked above contains a few more interesting details. Swift was able to win 60% of the time, and it’s lap times were significantly more consistent than those of the human pilots. While human pilots were often faster on certain sections of the course, Swift was faster overall. It picked more efficient trajectories over multiple gates, where the human pilots seemed to plan one gate in advance at most. On the other hand human pilots could recover quickly from a minor crash, where Swift did not include crash recovery.

The final results are impressive, especially given that all the processing and sensing comes from the drone. However, it still requires a well mapped track, so a human pilot should still come out on top given limited information about a new track. It would also be interesting to see how it handles large courses with gates that are much further apart.

Continue reading “Autonomous Racing Drones Are Starting To Beat Human Pilots” →