Diffusion bonding modeling and ductile fracture in materials with distributed porosity.

Abstract

The ductile fracture of porous materials depends not only on the average volume fracture of porosity, but also the distribution of porosity. To this date, there has been no available experimental data for three dimensional porous materials describing the effect of distribution of porosity on ductile fracture. In this research, a novel approach was proposed to model 3D porous low-carbon steel with controlled distributions of porosity using diffusion bonding techniques. An extensive study on diffusion bonding of carbon steel has been conducted through experimental work to reach optimal bonding conditions. Macro mechanical behaviors and microstructures of diffusion-bonded joints show that the bonding sections are homogeneous and identical to their parent material. To examine the effects of distribution of voids on porous materials, three patterns of distribution of cylindrical holes were made: uniform, center dispersed and center dense distributions. Their local porosities were calculated based on the Dirichlet method. A distribution function was defined to describe the mass distribution associated with the local porosity. Applying Haghi and Anand's model, a framework was presented for the constitutive equations and failure criteria for materials whose properties are not adequately representable as average, but where the distribution of material properties must be taken into account. The framework is used to describe the evolution of the void growth with deformation and allowed the failure criteria to be formulated based on a much more complete knowledge of the state of the material. The framework can also be used as a tool to probe the actual interactions between parts of a body with distributed properties. Uniaxial tensile tests of the diffusion-bonded specimens with different distributed voids were conducted to examine their ductile fracture patterns and compare with the results of simulation. Experimental data show that the ductility exponentially decreases as the standard deviation of distribution function increases. The results of simulation are in good agreement with the experimental data.