Additive manufacturing (AM) and welding are transformative technologies extensively used across various industries due to their capability to fabricate complex, high-performance components. However, challenges such as thermal distortion, residual stresses, and defects like cracks and porosity frequently arise due to the
inherent high thermal gradients and repeated thermal cycles during these processes. Traditional numerical methods, such as the Finite Element Method, often encounter difficulties in effectively addressing crack discontinuities in AM and welding processes due to their reliance on continuity assumptions in classical continuum
mechanics. This thesis has developed a peridynamics-based numerical modelling tool for simulating mechanical, thermal, thermo-mechanical, and fluid behaviours, suitable for the numerical investigation of AM and welding processes. Peridynamics, a nonlocal integral-based continuum theory, is capable of modelling discontinuities such as cracks without the need for remeshing, providing a promising alternative to conventional numerical methods. The research includes systematic investigations into optimal horizon size (a length scale parameter determining the level of nonlocal interactions) selection criteria across different peridynamic formulations, including bond-based, ordinary state-based, and non-ordinary state-based approaches. A dual-horizon peridynamic formulation is developed and validated to effectively handle non-uniform discretisation issues, improving accuracy and computational efficiency in mechanical and thermal diffusion
analyses. Furthermore, a coupled thermomechanical peridynamic model incorporating phase-change phenomena is formulated to simulate the structural deformation during welding and AM processes. To further expand peridynamic capabilities, the
Peridynamic Differential Operator is utilised for modelling multiphase flow behaviours, including wetting dynamics and thermo-capillary (Marangoni) effects, which are closely related to AM and welding scenarios involving surface tensiondriven fluid motion in the molten pool. Results demonstrated that the models developed consistently produced reliable and accurate predictions of deformation, thermal diffusion characteristics, phase transitions, and multiphase flow dynamics when benchmarked against reference data. This thesis advances peridynamic modelling capabilities for AM and welding applications by offering recommendations for horizon size selection and demonstrating the method’s suitability for simulating mechanical deformation, heat
conduction with phase change, and multiphase flow interactions. Overall, the work contributes to bridging fundamental peridynamic research with industrial practice, providing modelling tools and clear methodological guidelines to substantially enhance process reliability, component quality, and manufacturing efficiency.
| Date of Award | 1 Oct 2025 |
|---|
| Original language | English |
|---|
| Awarding Institution | - University Of Strathclyde
|
|---|
| Sponsors | University of Strathclyde |
|---|
| Supervisor | Selda Oterkus (Supervisor) & Erkan Oterkus (Supervisor) |
|---|