February 4th, 2019
Since September of 2017, AERO McGill has been working on creating a solar-powered UAV. This week, the technological aspect of our design that we would like to highlight is our use of innovative manufacturing techniques. Not being limited by competition objectives has allowed this project to explore the capabilities of a variety of manufacturing techniques. The three unique techniques we would like to focus on are as follows:
1. CNC hot wire cutting: for the manufacture of the wings and tail, we opted to use wire cut foam to allow for airfoil customizability, rapid prototyping, and durability. A major issue we had noticed with balsa designs, as highlighted by frequent damage to past SAE advanced aircraft, was the fragility of balsa wood components. We opted to sacrifice slightly on the higher weight of the wing and tail for the sake of their durability.
2. 3D printing of all the joints: In manufacturing the joints in the aircraft, we opted to use both stereolithography and fused deposition modeling 3D printers, depending on the desired physical characteristics of the part. The decision to make ubiquitous use of 3D printing, like CNC wire cutting, has been hugely beneficial by allowing for rapid prototyping.
3. Polycarbonate thermoforming of the fuselage: currently under development is a polycarbonate thermoformed fuselage to improve on the previous iteration of the fuselage which was cut from foam. The new design will provide greater fuselage rigidity to allow for belly landings and a removable landing gear. Additionally, the transparent material will allow for mounting of a camera within the fuselage, providing greater protection and reduced drag.
We continue to adopt new manufacturing solutions to our design problems. We hope that by experimenting with these manufacturing methods in a non-competition setting, the long-term manufacturing capabilities of the competition teams will be improved.
January 29th, 2018
Micro team wing construction and materials - In order to assess the manufacturability of our design, we tried to manufacture various materials, designs, and manufacturing steps prior to downselecting our options. By providing ample time for manufacture, we had the time and resources to trial different designs and downselect on the basis of empirical manufacturing results. One example of this is our wing construction methodology
Foam wing: This was the method of choice for the team in the year prior, but was a heavy option. This was the earliest and quickest manufacturing trial - while heavy, the wing will be used for the first flights because a foam wing is less likely to break after flight.
Balsa ribs & foam trailing edge: This was recommended by one of the professors we talked to, who was concerned about the weak trailing edge of the airfoil. However, upon attempting to manufacture the foam trailing edges, because of the thin dimensions, the hot wire burned through the foam and created inaccurate shapes.
Balsa ribs, leading and trailing edge reinforcements, & monokote: The third option was improved for monokote application because 2 extra balsa sheets were added for better application of monokote: on the rear of the front section, and at the front of the ailerons.
January 22nd, 2018
Automating the CAD process was driven by a need for a time and labor efficient design process as iterative design is required for optimization. For AERO specifically, many aircraft parameters can generally only be attained via computational analysis after certain design decisions are made. The results from this process can then be used to alter aspects of design, to ensure the aircraft fits within design constraints, or to improve function.
Complex systems comprise of subsystems or subcomponents have dependencies or interdependencies with one another. In the interest of time, the design of these various components or systems should occur simultaneously. For our aircraft, one of the most critical aspects of design is the shape of the aircraft: this affects stability, control surfaces, payload and internal space management, structures, etc. Redesign of the aircraft shape is inevitable, due to problems encountered in subsystem design, such as needing to internalize payload to reduce drag, or due to results from more intensive aerodynamic analysis. Thus, it was imperative to automate the CAD process for quick and effortless generation of aircraft geometry, to prevent bottlenecks in aircraft development.
In the first [and current] iteration of code, function is split in two: a MATLAB script reads user inputs of airfoil type, size, position, and orientation for predefined ‘critical’ cross-sections, and outputs the geometric shape of the airfoil in a format readable by SolidWorks. This output is then read into the SolidWorks API, which generates the CAD model, lofting between airfoil cross-sections, defined by a combination of linear and user-defined splines. Note that this code is written in Visual Basic, to ensure compatibility with the default user’s SolidWorks program – C++ was originally used but lacks compatibility as the interpreter must be selected for installation during the SolidWorks installation process, a step that many of us disregard.
Future developments seek to expand the utility of the code to generate other aircraft components, and increase the number of generated features. As well, integration of the function provided by the MATLAB script into the SolidWorks API to further increase ease of access is planned.
January 15th, 2018
AERO McGill uses simulation in SITL (software-in-the-loop) mode to conduct tests on our UAS. Simulated missions provide exercise in mission planning and vehicle command. This helps new members get familiarized with autonomous UAS operation.
More importantly, SITL simulation testing is vital for evaluating our customizations to the Ardupilot code. This involves removing many sections of code to provide a reduced set of capabilities, just enough for success in Unmanned Systems Canada.
Refer to the Ardupilot documentation site for how to setup SITL simulation: http://ardupilot.org/dev/docs/sitl-native-on-windows.html.
We give all our thanks to the Ardupilot development team!