Development of a robotic running foot for footwear testing

• Main Author: Nguyen, Thanh Luan
• Other Authors: Sam Allen
• Format: Theses and Dissertations
• Language: English
• Published: 2017
• Subjects: DRNTU::Engineering::Mechanical engineering::Robots
• Online Access: http://hdl.handle.net/10356/72697
• Institution: Nanyang Technological University

In footwear testing, the human wear participation method requires a long period and high cost to conduct the experiments and may expose the participants to injury risks. The virtual testing method, which uses computer simulations, faces significant challenges in modeling the complex geometry and realistic testing conditions. Thus, with the ability to stably continuously repeat thousands of testing cycles to reduce the testing time, a third method which uses mechanical devices and automatic systems is preferred in footwear testing. However, available systems do not possess a realistic representation of the human foot, cannot reproduce the wear conditions accurately, and do not mimic properly human running gaits. This thesis addresses the development process of a robotic running foot (RRF) for footwear testing. The first goal was to design and manufacture a realistic robotic foot which possesses multiple powered degrees of freedom (DOFs); and has sufficient torques, speeds, and ranges of movement to mimic human running gaits and replicate wear conditions. The second goal was to develop novel control strategies for driving the foot joints to mimic human running gaits with very high loads, high frequency, and discrete loading conditions in different phases of the gait cycle. Firstly, based on reviews of contemporary footwear testing systems, innovative robotic feet, and human foot biomechanics in running, some essential requirements for a robotic foot for footwear testing were investigated, and various possible solutions were proposed and analyzed. Then, a robotic foot with two powered joints (i.e., the ankle and metatarsophalangeal [MTP] joints) actuated by two pairs of cable-conduit mechanisms (CCMs) was constructed. The foot joints possess sufficient power to mimic human running gaits up to 14 km/h. Secondly, three studies on torque tracking control for robotic joints actuated by the CCMs were carried out to develop the most effective strategy for the foot joints. Each study was conducted under different conditions and references. Each proposed a torque transmission model and developed an adaptive controller to compensate for the nonlinear hysteresis of the torque transmission profile. Both theoretical proof and experimental validation were carried out to evaluate their torque tracking performance. These control methods are important for applications actuated by CCMs. They can be adopted directly if the applications require torque/force control and have available output feedback. Finally, an overall design for an integrated footwear testing system (IFTS) was developed to mimic human running gaits. This IFTS consists of two DOFs of an upper-leg mechanism, two DOFs of the robotic foot, and two DOFs of a contact platform. The final IFTS was equipped with abundant feedbacks allows for applying effective closed-loop control. Meanwhile, many human running and walking trials of eight healthy adults were conducted to collect a reference data pool. Then, a hybrid control design for the foot joints and a motion planning algorithm for four other actuators were developed and implemented to control the IFTS mimic the ground reaction forces and human foot profile of a running trial at about 12 km/h in the reference data pool. Then, the RRF and IFTS can be used for footwear durability tests including the tests for lifespan of individual components or of the whole shoe (the weakest part). However, there are many other tests can be done by the systems such as friction (slip resistance), abrasion of the outsole, or the tests for surface performance under effects of running, walking gaits.