Since its inception in the 1980s, abrasive waterjet machining (AWJM) technology has gained consistent and increasing adoption across several industrial sectors. This can be ascribed to its distinct advantages, including versatility, economic feasibility, and the absence of mechanical, thermal, or microstructural distortions. Nonetheless, this technology is still in development due to the inherent complexity of its physical nature, attributed to highly dynamic, nonlinear, and stochastic fluid-solid interactions (FSI). This underscores the necessity for understanding the underlying AWJM mechanisms to unleash its full potential. Within this scope, extensive research has been conducted employing several approaches, with numerical modeling emerging as one of the most effective. Numerous AWJM numerical models have been developed, successfully uncovering a range of obscure phenomena spanning from jet formation to material removal. Yet, some research gaps remain unresolved, which should be addressed to improve the modeling accuracy of AWJM. The present thesis proposes a suite of numerical models based on Lagrangian formulations, implemented through the finite element method (FEM) and smoothed particle hydrodynamics (SPH). The ultimate aim is to establish a high-fidelity model that can realistically capture the physical phenomena associated with AWJM and accurately predict the erosion process in order to reduce exhaustive trial and error experiments. Four research stages were established to achieve this aim, including analyzing the potential discrepancies of existing models, revealing the complex interplay between waterjet dynamics, abrasive particle motion, and erosion, then utilizing findings to develop accurate and efficient numerical simulation models, and verifying the numerical results of these models. The proposed models were developed and simulated using two software: Abaqus CAE and MATLAB. All investigations were performed on a high-strength aluminum alloy (AL 7075-T6), and the selected abrasive material was garnet. Numerical results demonstrate that the Johnson-Holmquist (JH-2) model can provide a more effective representation of garnet's mechanical behavior compared to existing material models. Furthermore, the axial spatial distribution of abrasive particles insignificantly affects the erosion rate, unlike the radial distribution, which critically controls both erosion results and kerf geometry. Other results also reveal that hydrodynamic effects due to the jet impact, such as the stagnation zone and drag forces, substantially affect the motion, trajectory, and kinetic energy of the abrasive particle, and consequently workpiece erosion. The comparison between the coupled SPH-FEM model and experimental depth of cut (DOC) results showed an average absolute error of 5.82%. Overall, this thesis contributes to AWJM simulation and optimization with more than four hundred virtual experiments, demonstrating the value of numerical modeling in process analysis and development.

Abrasive waterjet turning (AWJT) is a relatively new machining process that is still under investigation and not commercialized yet. In this process, cylindrical parts can be turned on an ordinary abrasive waterjet (AWJ) machine by using an additional device (turning fixture) for holding the rotating workpiece. Whereas, the AWJ nozzle is fed along the workpiece surface and the material is removed from its surface as a result of the erosive action. AWJT is a promising process surpassing the traditional turning by its ability to turn difficult-to-cut materials without severe cutting forces or thermal distortion. However, controlling its performance is very challenging as being dominated by several parameters in addition to the complex physical nature of material removal. In the present work, two numerical models based on the finite element method (FEM) were proposed to predict the radial depth of cut (DOC) in AWJT. All finite element (FE) simulations were conducted using the explicit dynamic solver of the commercial FE software ABAQUS‏. In the first model, the multi-particle impact approach was introduced to simulate the impact of abrasive particles over limited distance on the workpiece surface. The main objective was to predict the DOC under three levels of waterjet pressure and abrasive flow rate. Whereas, the values of DOC were estimated by including the model results of the crater area in a mathematical equation, and then were compared with the experimental data acquired from literature. The model prediction accuracy was relatively good compared to the previous FE models with a mean absolute error of 18.37%. Referring to the first model, too much time and effort were spent to complete the modeling procedures in each run. Consequently, a second one was developed to extend that model to become more user-friendly using ABAQUS Scripting. Furthermore, the involved process variables, including water pressure, traverse speed, abrasive flow rate, and surface speed, were investigated under a larger scale. The combined effect of these variables was examined at five levels where a standard Taguchi orthogonal array (L25) was employed to design the test plan. In order to validate this model, the turning experiments were conducted on AISI 4140 alloy steel while using #80 garnet abrasives. A tremendous improvement regarding the spent time in the modeling procedures was achieved where that time was reduced from days to 150 seconds only. In addition, the accuracy of DOC prediction was remarkably enhanced as the model validation showed an average absolute error of 9.74%. Eventually, a further modification was done to the second model reducing the simulation time from 28 hours to 7 hours.