Mathematical modelling and subsequent model-based optimisation algorithms are fundamental for advanced process control systems. Insufficient understanding of processes, process-material interaction & post-manufacture properties are currently a barrier to developing the mathematical models that relate process inputs to process outputs. Such models are a prerequisite to developing advanced control systems that exploit such process knowledge in order to improve performance.

We propose a cascaded controller architecture, whereby the localised feedback controllers handle set-point tracking and disturbance rejection at the sub-system level, whilst an outer, centralised, feedback controller determines the set-points to achieve system-wide performance and handle constraints. Our pragmatic approach will enable a step change in advanced multivariate control of film substrate manufacture, hence realising a key enabler for true process resilience.

For more detail about the individual tasks in this WP, please see the link below.

For idealised systems, film drying has traditionally been modelled as mass transfer limited, often with a spatially dependent evaporation rate . Flows in the film are driven by surface tension gradients with drying fronts propagating laterally. Current debate still surrounds the mechanism(s) generating the driving forces for particle arrangement with diffusion, binary interactions, diffusio-phoresis, and even air-water surface attraction being postulated Additionally, during commercial film drying, speed is often a dominant factor, due to commercial process efficiencies, meaning large particle concentrations are involved, providing minimal time for morphological evolution.

For more detail about the individual tasks in this WP, please see the link below.

We will build and deploy fast single-shot optical measurement systems to the manufactured films, acquiring function related data (as directed and informed by WP1 & 2) from moving substrates. We envisage the system will be based on Multi-wavelength Polarising Interferometry (MPI).

This approach is chosen due to its ultrafast data acquisition time (ms) and its consequent robustness to manufacturing environmental perturbations. The interferometer works by using polarising effects to shift the phase of the interference simultaneously and then using multiple wavelengths to generate colour interferograms to expand the range beyond λλ/4. Through such an approach, the system can measure track height, track width, thickness variation and film defects with single micron resolution in the x-y directions and nanometer resolution in the z direction.

Optical systems are the primary focus of the work due to their fast acquisition times and non-contact nature, providing spatial information as well as being able to handle non-uniform surfaces as well rapidly identifying defect locations.

For more detail about the individual tasks in this WP, please see the link below.

Lab/pilot scale R2R, slot die and bar coating equipped with fully computerised parameter control, with sensors developed in WP3, will be used to investigate a range of model and real material systems, chosen and based on the important material parameters identified by WP2 (size distribution, shape and density of the solid component, varying liquid phase properties (from simple Newtonian fluid to complex rheology of real-world formulations), substrate defects/spikes/pores).

A unique feature of our methodology is the ability to rapidly and automatically characterise the effect of varying a large number of coating process parameters (substrate speed, ink feed rate, die/bar to substrate spacing, substrate temperature, local atmosphere, drying speed) on process dynamics and final film quality. This will be enabled by full automation of parameter selection, and ability to perform on the fly characterisation of a continuous process, both to determine the effect of process parameters on final film quality and film formation dynamics. The throughput gain of this approach compared to a conventional make, test, make, test cycle will be key to provide the large volume of data required to develop robust process models via Machine Learning.

For more detail about the individual tasks in this WP, please see the link below.

We will realise a full demonstrator system within a manufacturing setting. We will install and integrate our responsive inspection and control systems developed across WP1-4 into a full scale processing line at CPI through three main tasks. This will cover aligning the sensors with respect to the position on the drying film and verify the performance of the metrology tools and machine learning approach to optimise the manufacturing process.

For more detail about the individual tasks in this WP, please see the link below.

Central to any responsive manufacturing process is ensuring that proposed changes lead to a more efficient and more sustainable processing system. Taking a cross-cutting and interdisciplinary perspective, WP6 aims to integrate the criteria for environmental and economic sustainability across the supply chain, from materials design to responsive manufacturing, thereby supporting sustainable manufacturing supply chain decision making. It is hypothesised that such integration will dramatically improve productivity and efficiency, leading to lower energy requirements, less waste and reduced costs. Opportunities for energy integration and industrial symbiosis will be explored to ensure maximum efficiency gains. This concept extends prior work and moves towards the development of a novel manufacturing supply chain, integrating predictive (ML), in-situ (sensor), pilot test intelligence with supply chain life cycle assessment (LCA) and techno-economic assessment (TEA) for environmental and economic sustainability optimisation in early technology development.

For more detail about the individual tasks in this WP, please see the link below.