Modified Taylor Impact Tests with Profiled Copper Cylinders: Experiment and Optimization of Dislocation Plasticity...

Area of Science:

• Materials Science
• Computational Mechanics
• Solid Mechanics

Background:

• Advanced numerical simulations and machine learning enable complex loading conditions for plasticity model parameter identification.
• This expands the range of examined deformation states, aligning models more closely with engineering practice.
• Dynamic plasticity of metals requires accurate models for applications involving high strain rates.

Purpose of the Study:

• To develop and apply a combined experimental-numerical approach for identifying parameters in dynamic plasticity models.
• To investigate the dynamic plasticity of cold-rolled OFHC copper using novel experimental techniques.
• To optimize dislocation plasticity models using advanced computational and statistical methods.

Main Methods:

• Utilized profiled projectiles (reduced cylinders/cones) in Taylor impact tests for material characterization, achieving large plastic deformations (true strain up to 1.3) at strain rates up to 10^5 s^-1.
• Implemented a 3D dislocation plasticity model with the smoothed particle hydrodynamics (SPH) numerical scheme.
• Employed a Bayesian statistical method combined with an artificial neural network (ANN) as an SPH emulator for parameter optimization.

Main Results:

• Profiled projectiles offer superior optimization of plasticity model parameters compared to classical Taylor cylinders.
• Combining different projectile shapes and increasing experimental data significantly improves optimization quality.
• The optimized model was successfully validated against shock wave profiles from flyer plate experiments.
• The model was extended to estimate grain refinement and volume fractions of weakened areas.

Conclusions:

• A cost-effective, simple, and efficient method for optimizing dynamic plasticity models has been proposed.
• The developed approach enhances the consistency of plasticity models with engineering practice.
• The study demonstrates the effectiveness of profiled projectiles and machine learning in advancing material modeling for dynamic events.