Smoothed Particle Hydrodynamic Modeling of Solid State Additively Manufactured Aluminum Alloys

Abstract

Additive friction stir-deposition (AFS-D) is a nascent additive manufacturing process shown to have better mechanical properties of deposited material than conventional techniques. While significant experimental research has been conducted on AFS-D, relatively little computational research exists for AFS-D. Simulating AFS-D is challenging because traditional finite element approaches fail to accommodate severe deformation. One solution is to use a meshfree framework, such as smoothed particle hydrodynamics (SPH), which better handles large deformation processes. This work aims to create a meshfree framework, improve it, and utilize it to better understand AFS-D and provide predictive power to improve AFS-D processing.Firstly, a meshfree framework was laid out to describe the underlying mechanics and SPH equations. Several AFS-depositions were created while monitoring substrate temperature for use in model calibration. The meshfree framework showed good agreement with the substrate temperature and build profile results. Previously unforeseen phenomena, such as the temperature dip under the stir zone, were revealed in the simulations. Simulations also revealed the relationship between actuator feed rate and processing temperature and plastic strain. To inform future AFS-D research and developments with the meshfree framework, a study was undertaken to compare constitutive models for AFS-D simulations. Two different AA6061 tempers were considered for this study: T6 and O. The constitutive models were calibrated against experimental torsion data at a variety of strain rates and temperatures. Constitutive model selection was found to have a major impact on simulation peak values, temperature, stress, strain, and build profiles. Finally, the meshfree framework was then applied for particle tracking analysis. Two types of depositions were created: one using an anodized feedstock to track oxide distribution in the deposition, which is analogous to material flow from the outside of the feedstock, and one using a copper wire core feedstock to track copper distribution in the deposition, which is analogous to material flow from the inside of the feedstock. Results revealed the tendency of oxides to flow to the retreating side. The copper wire was mainly deposited in a clear line on the advancing side, with some fragments scattered through the deposition. Unique insight into material flow behavior was illustrated with the meshfree framework.