Development of Frameworks for Data-led Predictive Modelling of Complex Compounds and Structures at Different Scales

Qing, X (2025) Development of Frameworks for Data-led Predictive Modelling of Complex Compounds and Structures at Different Scales. Doctoral thesis, Liverpool John Moores University.

Text 2025xiaoliangqingphd.pdf - Published Version Access Restricted until 2 May 2026. Available under License Creative Commons Attribution Non-commercial. Download (11MB)

Abstract

The main work reported in this thesis is focused on first-principles calculation for establishing the effects of point defects and doping on the structures and properties of Mo2C and molecular dynamics (MD) simulation of different materials in Fe-Cr-Ni system, including single crystals and polycrystal structures. The first principles method was adopted to investigate the structure and energy of point defects and defect complexes in molybdenum β-Mo2C. Systematic data (Mulliken atomic charge, bond population, DOS, Band structures, magnetic moments) is developed, and the effects of a range of substitutional doping elements on the magnetic properties of Mo₂C is analysed. Based on the energy and structural data, the SV-Mo model (substituting Mo with V) is a preferable point defect in Mo2C. The dopant (Fe, Cr, Co, Ni, W)-specific local magnetic moments data indicate that dopants with higher unpaired d-electrons, such as Fe, Co, are more effective at enhancing the magnetic properties of Mo₂C and the doping elements have a similar effect on the magnetic moment for 3D and 2D Mo2C structures.

The work on molecular dynamics modelling is focused on program development to predict the melting process, phase changes, and mechanical behaviour of the Fe-Ni-Cr system. The data covered single crystals and polycrystalline grain structures of Fe, Fe-Ni or Fe-Ni-Cr compositions. Python program for extracting common neighbour analysis (CNA) and Radial Distribution Function (RDF) is presented and used to develop systematic structural and phase data related to melting simulation and different mechanical deformation. User interfaces have been developed for MD simulation of single crystals and polycrystals. The use of the models to parametrically study the phase changes in samples of different shapes and boundary conditions is presented and analysed. The key issues such as compositional effects, potential energy and phase changes under the more complex boundaries are discussed and future work is highlighted.

Item Type:	Thesis (Doctoral)
Uncontrolled Keywords:	Data-led; Materials science; First principles; Molecular dynamics
Subjects:	T Technology > TA Engineering (General). Civil engineering (General)
Divisions:	Engineering
Date of acceptance:	8 April 2025
Date Deposited:	06 May 2025 09:31
Last Modified:	06 May 2025 09:32
DOI or ID number:	10.24377/LJMU.t.00026243
Supervisors:	Ren, J and Liu, X
URI:	https://researchonline.ljmu.ac.uk/id/eprint/26243

View Item