Multi-scale and multiphase modeling of abrasive flow machining process with experimental validation

• Main Author: Mohamed Adel Abdelhakem Amen
• Other Authors: Li Hua
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Manufacturing::CAD/CAM systems
• Online Access: https://hdl.handle.net/10356/169891
• Institution: Nanyang Technological University

Abrasive flow machining (AFM) is an interesting process with abilities to smoothen complex internal, external, macro-size, or micro-size workpieces down to nanoscale roughness values. The AFM tool is a flexible mixture between abrasive particles and high viscoelastic material while the abrasion action happens upon extruding this mixture through the channel walls. The flexibility and nature of the AFM tool introduce considerable complexity to evaluate the abrasion action, its material removal (MR), the particle forces, and the detailed rules controlling this particle-surface interaction. The process gets more complex for irregular shapes to the workpiece surface and the abrasive particle which is the real case. In addition, the AFM process has different nature of multiscale (macroscale fluid flow and microscale abrasion action in the channel surface) and multiphase (viscoelastic fluid and solid abrasive particle), and even multiphysics (flow and cutting/deforming physics) to consider. Probably because of these complexities, the AFM studies manipulated the process in a quantitative manner in terms that it lumps the whole microscale particle-surface interaction and its related variables either in an empirical equation or assuming simple particle/surface shapes and simple force model. Despite that AFM process is considered as multiples of a single particle-surface interaction, it is observed that there is a lack of understanding of this particle-surface interaction especially, in case of irregular shape to either particle or surface. Accordingly, the current study chooses its novelties as to reveal the details of this particle-surface interaction, to better represent reality through more complex shapes to surface/particle, and to study this interaction experimentally by evaluating the MR per particle. Achieving such novelties means clarifying the abrasion action mechanism, its related force, and the corresponding MR amount. The current thesis proposes a multiscale multiphase simulation of the AFM process with more emphasis on the micro-abrasion action. Two software are employed for each scale, namely ANSYS-FLUENT for the macroscale parameters related to fluid flowing (channel shape/size) and Molecular Dynamics-LAMMPS for microscale parameters related to the abrasion action. The simulation is validated by experiments with the same concern to investigate the abrasion action of a single part as in the simulation. To connect between LAMMPS and experiment scales, dimensionless terms are obtained with help of dimensional analysis and similitude. The study successfully achieved the proposed novelties with help of 3D simulation with video data; details about the achieved novelties are as follows: • To reveal the particle-surface interaction details: The main challenges to revealing these are the flexible nature of the AFM tool and the complex shapes of particles and the surface. AFM studies commonly lump this abrasion action either by empirical model or simplified shapes with simplified force-indentation equation therefore, the abrasion action details are missing. Here, the study finds out the particle forces, surface deformation, and MR mechanisms. Different from the common thought, the study proves that MR mechanism mainly comes from drag and impulse forces activated when the particle is blocked by surface asperity/protrusion. • For a better representation to reality: Typical unreal points in AFM models are simple particle/surface shapes, pre-impose MR equation, and the limitation of representing this complex abrasion action in mathematical or empirical ways. Here, the study established all the simulations using complex shape for surface and basic to complex shapes for the particle. Also, no need to pre-impose MR model since the simulation adjusts the force-MR relation smartly. • To study experimentally the particle-surface interaction: From open literature, no study is dictated to track the MR amount of individual particles in AFM process. The challenge is the small particle size (microscale range) and hence, the trivial scratch dimension for measurement. The current study scaled up the process, developed a methodology, and then design the corresponding test rig to obtain the MR amount for a single tested abrasive particle. Finally, it is observed that the experimental behavior agrees with the behavior predicted through simulation. In overall, the study presented an inclusive model for the AFM process and improves the understanding of it either quantitatively or qualitatively.