Vibratory disassembly - a novel robotic dismantling strategy for remanufacturing

Micheli, Marcel Massimo (2023). Vibratory disassembly - a novel robotic dismantling strategy for remanufacturing. University of Birmingham. Ph.D.

Micheli2023PhD.pdf
Text - Accepted Version
Restricted to Repository staff only until 26 May 2028.
Available under License All rights reserved.
Download (57MB) | Request a copy

Abstract

Climate change is currently one of the most significant challenges in human history. Transitioning from a linear to a circular economic model is necessary and affords an opportunity to reduce environmental damage. Remanufacturing is an important element of a circular economy. The first and arguably most critical process in remanufacturing is disassembly which, at present, is labour-intensive. This research proposes a novel disassembly technique that facilitates the unfastening of bolts and screws which is the most common operation in the dismantling of mechanical products. The technique involves using vibrations to cause multiple bolted joints simultaneously to self-loosen, and its application should enable the operation to be automated.
First, the ability of the self-loosening phenomenon to disengage multiple joints simultaneously was assessed, and factors such as the number and order of the bolts and the material of the bolted parts were analysed. To accomplish this, detailed 3D finite element models were developed, and simulations were conducted to determine the effect of the aforementioned variables on the loosening trajectory and time. These results were verified experimentally using a purpose-designed test stand. The trends of the simulations and experimental findings were in good agreement, and it was discovered that these effects must be accounted for in disassembly planning when vibrations are used for automated disassembly.
Second, potential damage caused by vibratory disassembly on the interfaces of bolted joints was evaluated, and influence factors such as preload level and material influence were analysed. To achieve this, subroutine-based 3D wear simulations were devised to predict the maximum wear depth and wear profiles at the interface between the flanged nut and the top surface. These results were experimentally confirmed using a purpose-built test stand. For the maximum wear depth, the tendencies of the simulations and the experimental findings were in good agreement. However, the wear width varied substantially, and it was determined that this was due to the influence of multiple asperities, which were not present in the simulation environment with flat surfaces. Even for extremely high preloads, it was discovered that the wear in steel components is relatively low and can be simply eliminated with common postprocessing techniques. Third, a robotic prototype instrument for vibratory disassembly of multiple bolted connections was created. Evaluations were conducted on loosening trajectories and time, as well as the overall performance and control of the novel device. To accomplish this, three-dimensional FEA simulations were developed to predict the loosening time and trajectories for multi-bolted joint configurations with varying preload levels. These results were then compared to the experimental results obtained using the novel robotic instrument. Errors of less than 10% were found, with results in outstanding agreement for the loosening time. For the loosening trajectories, there was good agreement with minor differences in the shape of the curve during the first quarter of the procedure. This robotic tool should help achieve fully automated disassembly through vibrations, which is advantageous over conventional methods in terms of speed and simplicity.
Overall, it was found that vibrational loosening is an effective and quick disassembly method with the potential to serve as a disassembly strategy in remanufacturing to reduce disassembly time and increase its adoption in the long run. The developed tool was a first step to prove this, and further development could lead to commercialisation of this method.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):