University of Sheffield
Smart manufacturing, a broad category of manufacturing that employs digital information, flexible skilled workforce training, and computer-integrated systems, has become the primary pathway for transforming and upgrading manufacturing industry in the coming decades. Real time monitoring of cutting tool conditions is the key to the realization of smart manufacturing required machining processes. Cutting tools are one of the keys and essential components for cutting, shaping and removing materials from workpieces in machining. The failure of cutting tools usually damages workpieces, causing production disruption/shutdown, and economic loss, which often amounts to plenty of times of the materials’ cost. Therefore, tool condition monitoring (TCM) is of critical importance for reduction of machine tool downtime, ensuring product quality, and increasing manufacturing productivity.
Existing TCM technologies can be classified into either a direct or an indirect approach. When applying a direct approach, machines have to be stopped for tool wear inspections, which often causes unnecessary downtime. Indirect approaches use sensor measurements to deduce tool wear via an empirically determined correlation between the tool wear and sensor data. This has lower complexity and is more suitable for real-time TCM. The most widely reported indirect TCM approaches are the signal analysis-based techniques where the features of signals from different sensing devices are used to continuously monitor the status of cutting tools. Various signal feature extraction methods have been used and, more recently, the application of AI to the interpretation of signal features for TCM has also been widely studied. However, there are two fundamental drawbacks and limitations with existing indirect approach based TCM technologies. First, signal features have low adaptability to overcome the impact of varying working environments. Secondly, signal features are often too complicated because the number of candidate features is theoretically infinite, making the feature selection is extremely difficult tasks. Consequently, only a very few of the signal based indirect approaches have now been implemented for TCM in real-world applications.
In order to address these challenges, supported by EPSRC grant EP/T024291/1, the supervisors’ team has recently developed a model feature-based TCM technique based on an innovative data and systems science approach uniquely derived at Sheffield, successfully completed a comprehensive range of industrial scale experimental studies, and demonstrated research outcomes to key players in advanced manufacturing including BAE Systems, Sandvick, and Mitsubishi. The PhD projects are motivated by advices from these key players, concerned with engineering application oriented research studies, and involve close collaborations with the engineers in AMRC at Sheffield and their industrial partners. The objectives are to advance the innovative model feature-based TCM technique to higher TRL levels to satisfy industrial needs for TCM in various complicated machining scenarios as required by the production of sophisticated workpieces of different complexities.
The novelty is the introduction and exploitation of a new concept known as machine tool dynamics-based digital twins. Digital twins have been widely used in engineering practice. However, the application of conventional digital twins in TCM is challenging due to the complexity of machine tools and limited understanding of the mechanisms of tool degradation. The new concept is proposed to address this difficulty. The concept will, for the first time, integrate the data-driven machine tool dynamics modelling, physically interpretable model feature extraction, as well as the model feature and machine learning based TCM into the framework of digital twins. This allows building up and updating a digital twin of machine tool dynamics via a completely data driven approach. Then, the unique data and system science approach proposed at Sheffield and a machine learning method can be applied to extract and analyse the physically interpretable features of the digital twin and realize TCM in real time.
Currently, effective TCM systems are urgently needed. For example, one of AMRC industrial partners wants to increase the productivity of one of their manufacturing processes by 10 times but realizes that the bottle neck is the tool wear induced manufacturing down time, and labour costs needed for evaluation of cutting tool conditions. The project outcomes would fundamentally resolve these challenges, reducing manufacturing costs, ensuring product qualities, and timely meeting the urgent need for real time TCM in smart manufacturing.
We require applicants to have either an undergraduate honours degree (1st) or MSc (Merit or Distinction) in a relevant science or engineering subject from a reputable institution. The research works are currently funded by EPSRC and supervisor team’s industrial partner. The successful candidates will therefore be involved in a range of world class research studies to apply innovative data/systems science and AI approaches uniquely derived at Sheffield to address significant challenges in smart manufacturing.
Interested candidates are strongly encouraged to contact the project supervisors Prof Zi-Qiang Lang and Dr H Laalej to discuss your interest in and suitability for the project prior to submitting your application.
Please refer to the EPSRC DTP webpage for detailed information about the EPSRC DTP and how to apply.
Application Form Please use the following project code: ACS-99-OPEN when making your application
This is a funded PhD studentship. The award will fund the full (UK or Overseas) tuition fee and UKRI stipend (currently £18,622 per annum) for 3.5 years, as well as a research grant to support costs associated with the project.
View or Apply
To help us track our recruitment effort, please indicate in your cover/motivation letter where (vacanciesin.eu) you saw this job posting.