I've drawn up a very rough diagram for what my initial test track concept may look like:
My concept involves a track made from aluminium extrusion and a carriage with forward and rear-facing ducted fan thrusters. The aim is to configure the trust output so that the alignment arrow stops in line with various marked goal positions accurately and in a reproducible manner.
Sensors (such as a laser range finder) will be added at a later time and the controller programmed to dynamically calculate the required thrust control, once I have developed an understanding of the thrust characteristics and implemented an algorithm capable of reaching hard-coded points.
Many consumer off-the-shelf electric ducted fan modules exist for use in hobbyist aircraft and similar projects. These come in various common formfactors: 50mm, 70mm, 90mm, 105mm and 120mm.
These modules consist of a brushless motor, propeller/impeller and an enclosure, and are pre-balenced for safety and efficiency at high rates of rotation. Being designed for model aircraft, they are often made from lightweight materials such as carbon fibre, nylon and fibreglass.
Module prices vary from less than £20 all the way to £300 or more for premium models, with the above module being £33.00 at time of writing.
These modules produce varying amounts of thrust, with larger diameters capable of producing higher amounts. This 70mm module has a thrust rating of 1.810kg thrust-equivalent (17.75N)
They are often controlled by connecting them to a standard Electronic Speed Controller (ESC) that allows the speed of the motor and therefore the thrust output to be adjusted. As part of my project, I will need to calculate how adjusting the speed of the motor impacts the thrust produced and then work out how to use those calculations to automatically produce the desired movements.
My current search methodology to find this existing research is to search Goolgle Scholar using search terms such as “electric ducted fan thrust vector control”
Much of the existing research into ducted fans is targeted at drones, UAVs and experimental electric aircraft.
Here are some papers I found that seemed the most relevant to my exploration into ducted fan thrust vecotr control systems for use in robotics:
Design and Full-Pose Control of a Ducted Fan Aerial Robot with Dual-Axis Gimbaled Thrust-Vectoring Mechanisms describes an aerial robot with gimbaled ducted fans instead of classic rotors to create a 6DoF thrust vector control system. The paper describes the use of “Lyapunov-based control” which looks complex and may be beyond the scope of my 12 week project, from my current understanding this refers to a mechanism used to find a stable point that may only be relevant to flying craft/robots (although this video by Professor Giordano Scarciotti on YouTube does provide a seemingly good visualisation that I may return to if understanding of this becomes relevant to me).
Experimental Investigation and Performance Evaluation of an Electric Ducted Fan (EDF) for Distributed Propulsion Systems Looks like it may be relevent to me as it mentions the building of a test rig “was used to characterize the relationships between the Pulse Width Modulation (PWM) control input and the primary performance outputs, including fan rotational speed (RPM), static thrust, and electrical power consumption.” However, I only have access to the abstract for this paper, I'll need to check with the university to see if I can get access to it.
# Development of a One Axis Thrust Vectoring Control System Demonstrator with Electric Ducted Fan Seems like a good candidate to be looked into more deeply, this paper describes a single degree of freedom test rig for a ducted fan system being used to model thrust vector control for spacecraft in an atmospheric environment. The test rig described in this paper specifically tests the pitch rotational axis, somewhat different to the initial idea I have in mind for a linear test rig
Jet-HR2: A Flying Bipedal Robot Based on Thrust Vector Control Describes the use of a ducted fan to enhance the capabilities of a disaster response robot, allowing it to perform long or high jumps. This probably isn't all that relevant to my research but it is cool
