Projects
This work is part of EPSRC funded project, Doing More With Less, in collaboration with Sheffield Titanium Alloy Research (STAR) group. It aims to develop a digital threading framework that allows linking process sensor data to M2i2 simulator platform to enable optimisation of Field Assisted Sintering Technology (FAST). This will be achieved within a digital model ecosystem where users have remote access to materials models to act on FAST sensor data (temperature and force profile) and generate simulated microstructure and mechanical property data via Avrami-based method for solid state phase transition, Cellular Automata method for grain growth simulation, mean field description for the evolution (sintering) of process-induced voids, and microstructure-sensitive constitutive model for the flow stress.
ModSim of Novel Laser Welding Techniques in partnership with MTC: microstructure prediction
This project is part of a Manufacturing Technology Centre (MTC) program called ModSim aiming to determine a modelling framework to predict microstructure and material properties of dissimilar-alloy welds. The M2i2 is providing remote access to advanced modelling techniques within the M2i2 web-platform materials simulator. Thanks to the M2i2 web-platform, MTC users can upload thermo field data to the platform and predict the grain evolution during solidification and heat treatment.
The Materials Made Smarter Centre (MMSC) is a consortium grant funded by EPSRC that serves as a hub for researchers spanning the entire materials manufacturing value chain. This collaborative community is dedicated to pioneering breakthroughs in digital research, focusing on cutting-edge technologies such as edge-AI, sensors, and advanced physics-based modelling.
M2i2 is leading the activities in Beacon 1: Efficient Fused Modelling Lead. This Beacon will deliver hybrid modelling tools for prediction of material behaviour, structure and performance during processing and link this to in-service performance prediction, testing and characterisation. Our hybrid modelling approach will combine data-driven modelling and AI to connect physics-based models, expert knowledge and manufacturing process data, creating computational tools that can be employed through Edge computing or used offline, to search for unexplored manufacturing process windows, enabling our goal of Performance on Demand.
Completed Projects
Laser Metal Deposition wire-based processing is known for high deposition rates and suitability for large-scale manufacturing. However, the large deposition rate also leads to challenges like residual stress and cracking. This ATI(DAM) funded study with GKN as industrial partner introduces a multi-scale process model that links process variables to microstructure and properties, aiding in certification and lifing methods for these additively manufactured components. The approach includes accurate thermal field representation using FEA followed by models for grain growth and phase transitions based on Cellular Automaton and JMAK. This multi-scale model enables the generation of realistic microstructures and the implementation of location-specific mechanical behaviour predictions.
The focus of this project is the development of a theoretical framework for understanding oxide-controlled dwell fatigue crack growth in γ'-strengthened nickel-based superalloys. Specifically, the study explores the interplay between externally applied loads and variations in the dispersion of γ' particles, examining their impact on grain boundary oxide growth kinetics. To model the stress state evolution near a crack at elevated temperatures, a dislocation-based viscoplastic constitutive description for high-temperature deformation is employed. A multicomponent mass transport formulation is utilised to simulate the formation and evolution of an oxide wedge ahead of the crack tip, assuming the operation of stress-assisted vacancy diffusion. Simulation results highlight the significant influence of a fine γ' size distribution on the predicted flow stress of the material, thereby impacting the relaxation in the vicinity of the crack-tip/oxide wedge. Furthermore, the simulations reveal that the presence of a fine γ' size distribution reduced oxide growth rates with a parabolic trend compared to a bimodal dispersion.
The overall aim of DRAMA was to accelerate the uptake of metal powder bed additive manufacturing (AM) by the UK aerospace additive supply-chain. A multi-scale AM modelling capability was developed to push the AM toward informed process optimisation route by introduction of AM-modelling digital thread. This work streamlined AM modeling through:
-
Development of meso-scale representative volume element (RVE) coupled thermo-mechanical FEA models for the accurate laser energy mapping;
-
Introduction of surrogate models from series of RVEs to expedite component level simulations;
-
Finally, developing component level models that predicts the AM residuals stresses/temperature.
The results demonstrated notable improvements in prediction of AM induced residual stresses, thereby enhancing geometrical tolerance of reversed-engineered components compared to conventional CAD-based as well as commercial platforms-based designs.
The M2i2 research team has participated in Additive Manufacturing Benchmarks 2018 and 2022 (AM-Bench 2018 and AM-Bench 2022). These challenges are a first of a series of benchmark tests focusing on validating the predictability of computational methods for additive manufacture (AM). Experiments carried out include measurements of microstructure variations, residual stresses, part distortions, 3D grain structures and melt-pool geometry of an AM build and simple single track passes. The AM-Bench 2018 organising committee received a total of 46 challenge submissions from 19 groups: 9 from North America, 6 from Europe and 4 from Asia. The 2022 AM Bench Organizing Committee received a total of 138 Benchmark Challenge submissions from 19 groups around the world. The geographical distribution of the groups, as determined by the home institution of the submitting author, included 15 from North America, 1 from Europe, and 3 from Asia.
M2i2 has been awarded:
NIST AM-Bench Challenge 2022 – 1st prize: “Modeling results predicting phase evolution during post-build heat treatments of IN718 test artifacts produced using laser powder bed fusion”.
NIST AM-Bench Challenge 2022 – 2nd prize: “Modeling results predicting residual elastic strain components at select locations internal to an as-built IN718 bridge structure”.
NIST AM-Bench Challenge 2018 – 1st prize: “Best modeling results predicting the phase evolution during residual stress annealing of an as-built IN625 bridge structure”.
Pressure quenching of carburised-austenitic steels to achieve high-strength and wear resistance through martensitic transformation is a vital process-route for gearbox component longevity. However, the process induced high thermal and transformational strains introduces cracks in the material. To mitigate this, a multi-scale framework that considers carbon content's impact on phase transformation and associated residual stresses (across the process's broad thermal spectrum) has been created. The framework facilitated predictions of trans-granular and intra-granular residual stresses, as well as dislocation dynamics, through crystal plasticity models. These insights facilitate the route for developing dislocation density-based damage models for void initiation.