Industry case study: rapid prototype of mountain bike frame section

The purpose of this study is to detail a virtual and physical prototyping process to overcome a design constraint in the mountain bike industry. Through a series of techniques, 3D scanning, developing detailed CAD models, then through additive manufacturing processes, a solution wasdeveloped. The challenge in the industry is the constant geometrical changes of components; the trend has been that bike cranks are becoming narrower due to biomechanical factors and tyres are becoming wider due to rider preferences and increased grip. This change in geometry results in metal tubes that can no longer be deformed without exceeding the minimum bend radius for the material. As such exceeding the minimum bend radius will induce early performance failure and geometrical (aesthetic) defects. The solution is an additivemanufactured part that can be substituted into the process without disrupting the entire conventional build process of a customised bike build.

Evolution of 3D printed soft actuators

Developing soft actuators and sensors by means of 3D printing has become an exciting research area. Compared to conventional methods, 3D printing enables rapid prototyping, custom design, and single-step fabrication of actuators and sensors that have complex structure and high resolution. While 3D printed sensors have been widely reviewed in the literature, 3D printed actuators, on the other hand, have not been adequately reviewed thus far. This paper presents a comprehensive review of the existing 3D printed actuators. First, the common processes used in 3D printing of actuators are reviewed. Next, the existing mechanisms used for stimulating the printed actuators are described. In addition, the materials used to print the actuators are compared. Then, the applications of the printed actuators including soft-manipulation of tissues and organs in biomedicine and fragile agricultural products, regenerative design, smart valves, microfluidic systems, electromechanical switches, smart textiles, and minimally invasive surgical instruments are explained. After that, the reviewed 3D printed actuators are discussed in terms of their advantages and disadvantages considering power density, elasticity, strain, stress, operation voltage, weight, size, response time, controllability, and biocompatibility. Finally, the future directions of 3D printed actuators are discussed.

MULTISCALE MODELING OF LIQUID CRYSTALLINE/NANOTUBE COMPOSITES

The objective of this research is to synthesize structural composites designed with particular areas defined with custom modulus, strength and toughness values in order to improve the overall mechanical behavior of the composite. Such composites are defined and referred to as 3D-designer composites. These composites will be formed from liquid crystalline polymers and carbon nanotubes. The fabrication process is a variation of rapid prototyping process, which is a layered, additive-manufacturing approach. Composites formed using this process can be custom designed by apt modeling methods for superior performance in advanced applications. The focus of this research is on enhancement of Young's modulus in order to make the final composite stiffer. Strength and toughness of the final composite with respect to various applications is also discussed. We have taken into consideration the mechanical properties of final composite at different fiber volume content as well as at different orientations and lengths of the fibers. The orientation of the LC monomers is supposed to be carried out using electric or magnetic fields. A computer program is modeled incorporating the Mori-Tanaka modeling scheme to generate the stiffness matrix of the final composite. The final properties are then deduced from the stiffness matrix using composite micromechanics. Eshelby's tensor, required to calculate the stiffness tensor using Mori-Tanaka method, is calculated using a numerical scheme that determines the components of the Eshelby's tensor (Gavazzi and Lagoudas 1990). The numerical integration is solved using Gaussian Quadrature scheme and is worked out using MATLAB as well. . MATLAB provides a good deal of commands and algorithms that can be used efficiently to elaborate the continuum of the formula to its extents. Graphs are plotted using different combinations of results and parameters involved in finding these results

REAL TIME FUZZY CONTROLLER FOR QUADROTOR STABILITY CONTROL

In this report, we develop an intelligent adaptive neuro-fuzzy controller by using adaptive neuro fuzzy inference system (ANFIS) techniques. We begin by starting with a standard proportional-derivative (PD) controller and use the PD controller data to train the ANFIS system to develop a fuzzy controller. We then propose and validate a method to implement this control strategy on commercial off-the-shelf (COTS) hardware. An analysis is made into the choice of filters for attitude estimation. These choices are limited by the complexity of the filter and the computing ability and memory constraints of the micro-controller. Simplified Kalman filters are found to be good at estimation of attitude given the above constraints. Using model based design techniques, the models are implemented on an embedded system. This enables the deployment of fuzzy controllers on enthusiast-grade controllers. We evaluate the feasibility of the proposed control strategy in a model-in-the-loop simulation. We then propose a rapid prototyping strategy, allowing us to deploy these control algorithms on a system consisting of a combination of an ARM-based microcontroller and two Arduino-based controllers. We then use a combination of the code generation capabilities within MATLAB/Simulink in combination with multiple open-source projects in order to deploy code to an ARM CortexM4 based controller board. We also evaluate this strategy on an ARM-A8 based board, and a much less powerful Arduino based flight controller. We conclude by proving the feasibility of fuzzy controllers on Commercial-off the shelf (COTS) hardware, we also point out the limitations in the current hardware and make suggestions for hardware that we think would be better suited for memory heavy controllers.

Computation and fabrication of scaled prototypes

The formal, functional, and material attributes of design are routinely investigated through the construction of physical models and scaled prototypes. With the increasing adoption of computational workflows, the digital to physical translation process is central to the construction of scaled prototypes. However, the choice of methods, tools and materials for computational prototyping is a developing area. Therefore a systematic body of knowledge on the benefits and costs of multiple methods of computational prototyping for the construction of physical prototypes need to be identified. This paper addresses the prototyping process through the comparison of three computational methods of fabrication through the modelling, analysis and construction of a Gaussian Vault. It reports on the process of digital to physical construction using additive manufacturing, surface fabrication and structural component models. The Gaussian Vault offers a unique set of geometric, structural and physical characteristics for testing all three methods of prototyping. The size, shape and proportion of vault prototypes are rapidly generated and tested. The design geometry, material properties and physical construction of the Gaussian Vault are realised using commonly used practice workflows comprising parametric modelling and analysis of geometry, model rationalisation with material characteristics and finally the use of digital fabrication methods. Comparison of the results identifies the characteristics, benefits and limitations of the three approaches. Finally the paper discusses the digital to physical translation processes and summarises the characteristics, benefits and issues encountered in each.

Implémentation sur FPGA de l’algorithme MUSIC sur antenne-réseau expérimentale à 10 GHz

Les techniques des directions d’arrivée (DOA) sont une voie prometteuse pour accroitre la capacité des systèmes et les services de télécommunications en permettant de mieux estimer le canal radio-mobile. Elles permettent aussi de suivre précisément des usagers cellulaires pour orienter les faisceaux d’antennes dans leur direction. S’inscrivant dans ce contexte, ce présent mémoire décrit étape par étape l’implémentation de l’algorithme de haut niveau MUSIC (MUltiple SIgnal Classification) sur une plateforme FPGA afin de déterminer en temps réel l’angle d’arrivée d’une ou des sources incidentes à un réseau d’antennes. Le concept du prototypage rapide des lois de commande (RCP) avec les outils de XilinxTM System generator (XSG) et du MBDK (Model Based Design Kit) de NutaqTM est le concept de développement utilisé. Ce concept se base sur une programmation de code haut niveau à travers des modèles, pour générer automatiquement un code de bas niveau. Une attention particulière est portée sur la méthode choisie pour résoudre le problème de la décomposition en valeurs et vecteurs propres de la matrice complexe de covariance par l’algorithme de Jacobi. L’architecture mise en place implémentant cette dernière dans le FPGA (Field Programmable Gate Array) est détaillée. Par ailleurs, il est prouvé que MUSIC ne peut effectuer une estimation intéressante de la position des sources sans une calibration préalable du réseau d’antennes. Ainsi, la technique de calibration par matrice G utilisée dans ce projet est présentée, en plus de son modèle d’implémentation. Enfin, les résultats expérimentaux du système mis à l’épreuve dans un environnement réel en présence d’une source puis de deux sources fortement corrélées sont illustrés et analysés.