The University of Strathclyde is pleased to be able to offer a highly cross-disciplinary engineering project on the process optimisation of high-value manufacturing. The position is hosted by the Department of Design, Manufacturing & Engineering Management (DMEM) and is offered as part of the Strathclyde Centre for Doctoral Training (SCDT) in AI-enabled Digital High-Value Manufacturing. The SCDT studentships are partly supported by the National Manufacturing Institute Scotland (NMIS) and Advanced Forming Research Centre (AFRC), and students will benefit from alignment with multiple ongoing EPSRC research grants. Research undertaken as part of this SCDT capitalises on new infrastructure investment made by the UK government into world-class forging capabilities at the AFRC, and collaboration with the Center for Ultrasonic Engineering (CUE), the LiFi Research and Development Centre (LRDC), and with strategic partner Fraunhofer UK.
This PhD will focus on the “in-process monitoring of microstructure evolution during ingot-to-billet conversion of the aerospace grade Ti-6Al-4V (Ti64) material in support of the ongoing development of Digital Twin infrastructure for high-value manufacturing”.
Forging remains the state-of-the-art manufacturing route for high-value components that demand reliable structural integrity. Starting with vacuum arc re-melted ingots at elevated temperatures, tightly controlled thermomechanical processing is performed. Several open-die forging hits and reheats may be required to refine the inhomogeneous as-cast microstructure. Recent improvements in control, sensor and material testing technology have been introduced as part of industry 4.0 and digital-twin requirements, but to take full advantage these must be coupled with robust and verified models of the physical deformation mechanisms and paths and in-process optimisation algorithms/methodologies for direct operation of equipment, control of process and ultimately decision making with minimum to no uncertainties. This is only achievable through new and data-centric approaches that brings together knowledge and know-how of materials behaviour, simulation and modelling, sensing, data analysis and optimisation, and most importantly artificial intelligence with decision making capability.
An infrastructure for DT of forging is currently under development at the AFRC. This includes a platform containing a comprehensive database in its heart, with connection with different FE software packages, forging equipment, sensors, robotic arms, and materials models. However, there are still gaps in understanding material behaviours during forging, particularly during ingot-to-billet conversion, which needs to be well understood before being implemented into the DT platform. The main objective is to create a fully operational DT of forging titanium alloys with decision making abilities and minimised human interactions. Currently, the infrastructure still requires the development and integration of robust materials models to sufficiently capture the physical microstructure evolution mechanisms.
Thus, the PhD project will have the following tasks:
a. Select sufficient number of samples, with appropriate size, from an aerospace grade Ti64 billet in as-cast condition.
b. Identify and determine the critical process parameters for an onset of recrystallisation:
i. When the as-cast material is fully recrystallised.
ii. Determine the strain/strain rate for the start and completion of recrystallisation.
c. Define DoE for (upstream) processing route including heat-treatments
i. Upsetting vs. cogging trials
ii. Processing conditions and scheduling
d. Run open-die and heat-treatment simulations and execute selected trials to generate desired wrought products and microstructures/mechanical properties
i. Extract key information from simulations
ii. Test and improve/expand experimental set-up – die temperature, thermal camera, DIC
iii. Collect, record and analyse live process data
iv. Characterise (fully) as-received and forged/heat-treated parts
e. Built process optimisation tools for forging
i. Start from single step processes (upsetting) and move towards multi-step incremental processing routes (cogging)
ii. Use existing optimisation tools based on Data Analytics, Artificial Intelligence, Neural Networks etc. to optimise design
iii. Integrate modelling data, live process data, experimental data and processing window information to optimise current step and advise on how to perform next step
iv. Introduce uncertainty and methodology to account for it
v. Decision making tool for further processing or discard current part and start a new
f. Develop purpose-built algorithms for the optimisation of manufacturing based on the requirements of Design for Manufacturing and the holistic view around manufacturing
g. Built a GUI/software to be the platform to control and run the optimisation of the Digital Twin for forging.