The Simple project aims to integrate a range of machining, fabrication and inspection operations onto a single manufacturing platform.
Single platform manufacturing reduces the need to move large components between work areas, and helps ensure accuracy and quality control throughout the manufacturing process.
By doing more operations on a single platform, and automating common tasks such as weld inspection, the Simple project aims to significantly improve productivity for large nuclear components measuring at least two metres. These include pressure vessel sections, large valve casings and decommissioning waste containers. The technology could also be used to reduce the risk of manufacturing error and cut cost and time for smaller high-value components for the nuclear island, and be deployed in other sectors such as energy, oil and gas, marine and aerospace.
Developing a wider selection of machining, cladding and inspection heads which could be deployed on a single large manufacturing platform, and using Industry 4.0 techniques to analyse and act on large amounts of real-time data, could achieve cost and time savings of at least 50 per cent for a range of large complex fabrications.
Simple (Single Manufacturing Platform Environment) is supported by £1.35 million funding from the Department for Business, Energy & Industry Strategy (BEIS) through the Small Business Research Initiative (SBRI) managed by Innovate UK.
The Nuclear AMRC is leading the Simple research consortium, which includes two of our sister centres within the High Value Manufacturing Catapult – the Advanced Forming Research Centre at the University of Strathclyde and the Advanced Manufacturing Research Centre at the University of Sheffield – plus the University of Sheffield physics department, TWI and Peak NDT.
Simple is supported by a range of nuclear industry partners – including reactor developers and operators, and decommissioning site owners – who will ensure the research is addressing industry challenges. The results will be shared with UK industry, including the Fit For Nuclear network of companies from along the supply chain.
As a proof of concept for the Simple approach, the project’s first phase is focused on developing an integrated welding and monitoring tool. The 20-month programme started in January 2018, and is on track for completion in August 2019.
The tool combines a range of sensors and testing tools with a mechanised arc welding head. This will allow automated in-process inspection of welds, improving quality, and reducing the risk of weld failure leading to costly rework.
Following initial work including process selection, systems engineering and sensor development, the Nuclear AMRC team started bead welding trials in June 2018 using a Polysoude narrow-groove welding torch, with full-depth welds by the end of the year.
Demonstration trials were carried out in early 2019 – see the video below to watch the technology in action.
Nuclear AMRC researchers have developed a variety of analytical and modelling techniques to allow full automation of the welding and inspection tool. These include a laser sensor for measuring weld geometry; electronic speckle pattern interferometry techniques to characterise the weld surface; and tools for acoustic analysis of the welding process. The team have also developed new tools to predict weld bead geometries using an artificial neural network, and machine learning to model the effects of multiple weld passes.
Other project partners have worked on a variety of additional inspection and monitoring technologies.
The University of Strathclyde AFRC and Peak NDT collaborated on an ultrasonic inspection system, with a prototype system installed on a welding rig at Strathclyde.
TWI developed a visual inspection system, and has recorded a series of videos of different welding processes to train and test a neural network for real-time monitoring.
Physicists from the University of Sheffield have developed a process monitoring system which can be integrated with the Polysoude to record and analyse the welding parameters and optimise the process.
And the University of Sheffield AMRC has developed a system to integrate the data from the all the sensors in real time, drawing on technology used in the automotive industry for real-time vehicle testing.