Error-propagation Based Geometrical Quality Prediction and Control Strategy for Complex Manufacturing Processes Using Parallel Kinematic Machines

UK manufacture accounts for 13% of GDP, 50% of exports and directly employs 2.5 million people. Parallel Kinematic Machines (PKM) are a new type of machine tools and have been identified as a key technology that fills in the gap between computer numerical controlled machines and industrial robots due to their superior dynamic performance, flexibility and versatility to large-scaled parts machining. The use of PKMs creates more flexibility and dexterity in manufacturing processes while achieving high precision and high speed. This contributes significantly to the economy by improving efficiency, reducing product defects, and saving time/money/energy.

The PKM integrated manufacturing system would inevitably introduce errors due to stiffness and motion of the components in the system. These errors will be accumulated through the production chain, and influence the geometrical quality of the machined parts. Predicting part quality based on error propagation in the PKM manufacturing processes represents a step change in managing production processes, as it removes the current cumbersome trial-and-error processes and enables rapid reconfiguration of production systems. Other benefits would include 20% reduction of part defects and rework, leading to a significant cost saving.

Part quality resulted from interaction of manufacturing systems and machining processes, with intertwined machining errors and their propagation through multiple operations, machine tools, and fixtures and jigs. At the moment, there is no robust industrial or international standard to evaluate machining capability of PKM tools with these errors. Current trial-and-error based approach that requires a large amount of time, materials and energy, is not sustainable and suitable for future smart factories to meet frequent changes with reconfigurability. Therefore new analytical methods are urgently needed.

The proposed research is adventurous in creating a new quality prediction capability for PKM based flexible manufacturing processes by revealing the relationship between manufacturing system errors and part or assembly quality. This leads to an effective error discrimination control strategy to achieve a better process control while ensuring the required product quality.

Error propagation in a production process is to be explored by investigating the role of stiffness characteristics of a PKM in influencing the machining process. This will lead to the development of machining load-models in both milling and drilling on a specific machining process. Experiments are to be implemented at QUB's PKM laboratory and KCL PKM laboratory, and a map between errors and part quality is to be created through modeling and testing. This will deliver an enhanced understanding of errors and their propagation mechanism thereby leading to the identification of potential strategies for reducing individual, propagated, and residual errors.

An integrated validation system that consists of a kinematic/dynamic analysis module, kinetostatic model, CAD module, and FEM module will be implemented in a virtual environment and in a manufacturing site. The project will access expertise from world-leading groups in advanced PKM machining processes.

The research is highly transformative in its nature of connecting academic cutting-edge research to the practical issues encountered in complex PKM manufacture processes. Key results are to be generated and fundamental science is to be revealed in the collaborative work, training and workshops with support of AMRC, MTC and Tianjin University. The research will benefit the academic community in manufacture and robotics, and industrial sectors who will gain knowledge for reduction of errors particularly propagated errors in manufacturing processes integrated with PKMs.

Knowledge impact: The investigators have established worldwide network which will warrant for knowledge dissemination to a much wider community, including researchers in flexible manufacture, parallel kinematic machines, robotics, machine tools, machining, process control, and smart manufacturing. Collaborations with project partners are integral part of this proposal and will contribute to the advance of understanding of machine tool user requirements as well as accelerating the knowledge transfer from academia to industry. A plan has already been in place to ensure the maximum dissemination of the project results, via internet, social media, open-access publications, special issue journals, topic symposium at international conferences, and dissemination events.

Economic impact: The economic impact will be resulted from the utilisation of the new error minimization strategy in manufacturing processes, which leads to efficiency improvement, product defects reduction, and time/money/energy savings. Collaboration with the two UK catapult centres (AMRC and MTC) in high value manufacturing will ensure the fast transformation of the project outcome to industry through tailored workshops at the two centres open to their industrial partners. The investigators will also work closely with the commercial development teams at QUB and KCL to protect and exploit any intellectual property resulted from the project. The expected impact would be 50% rework reduction, 12% reduction on non-conformance management activities, and 20% of cost reduction.

People development: The proposed project will create forefront knowledge in flexible manufacturing technologies, which will be transferred to next generation engineers, through teaching, training, and outreach activities. The researchers appointed to the project will gain not only the cutting-edge knowledge in PKM machine tools and manufacturing processes, but also soft skills such as project management and team working throughout the project. The theoretical advances, models and methods will be integrated into Dr Jin's and Prof. Dai's teaching modules, which will be delivered to their undergraduates and postgraduates. A generic case study will also be developed for both education and industrial training. Collaboration with Tianjin University China and organizing workshop and symposium overseas will create further impact on people at international level.

Societal impact: Using the proposed technology to control error propagation and optimize manufacturing process will lead to energy and materials savings, as well as reduction of carbon emissions. In addition, quality prediction method will help to reduce the cumbersome trail-and-error processes, which will improve the working conditions of shop floor workers. The investigators have rich experience of publication engagement, and activities will be conducted by the investigators and researchers of this project to engage the public through school talks, public lectures, and science festivals, as shown in the pathways to impact.