Quantitative Law of Diffusion Induced Stress and Fracture

TL;DR

This paper develops an analytical model to accurately predict maximum diffusion-induced stress and fracture in solid materials, validated by experiments, providing a practical tool for engineers and material scientists.

Contribution

It introduces a nearly analytical formula for maximum diffusion-induced stress, including effects of stress-dependent diffusion, and predicts crack spacing independent of specimen thickness.

Findings

The model accurately predicts maximum DIS for various geometries.

Experimental validation confirms constant crack spacing regardless of thickness.

The model explains hierarchical crack patterns observed in experiments.

Abstract

In diffusion processes of solid materials, such as in the classical thermal shock problem and the recent lithium ion battery, the maximum diffusion induced stress (DIS) is a very important quantity. However a widely accepted, accurate and easy-to-use quantitative formula on it still lacks. In this paper, by normalizing the governing equations, an almost analytical model is developed, except a single-variable function of the dimensionless Biot number which cannot be determined analytically and is then given by a curve. Formulae for various typical geometries and working conditions are presented. If the stress and the diffusion process are fully coupled (i.e. stress-dependent diffusion), as in lithium ion diffusion, the normalized maximum DIS can be characterized by a two-variable function of a dimensionless coupling parameter and the Biot number, which is obtained numerically and presented in contour plots. Moreover, it is interesting to note that these two parameters, within a wide range, can be further approximately combined into a single dimensionless parameter to characterize the maximum DIS. These formulae together with curves/contours provide engineers and materialists a simple and easy way to quickly obtain the stress and verify the reliability of materials under various typical diffusion conditions. Via energy balance analysis, the model of diffusion induced fracture is also developed. It interestingly predicts that the spacing of diffusion induced cracks is constant, independent of the thickness of specimen and the concentration difference. Our thermal shock experiments on alumina plates validate these qualitative and quantitative theoretical predictions, such as the constant crack spacing and the predicted critical temperature difference at which the cracks initiate. Furthermore, the proposed model can interpret the observed hierarchical crack patterns.