Boundary element model for electrochemical dissolution under externally applied low level stress

Abstract

Abstract The effects of low levels of stress on the dissolution rate of type 304 stainless steel in seawater are determined, and these effects are incorporated into a boundary element method (BEM) code which was written to predict long-term changes in geometry, including those due to the stress-modified dissolution rates. Corrosion in the absence of stress effects is thoroughly documented, while the effects of micromechanical damage caused by strains in the plastic region are also well recognized. However, very little is known regarding the effects of low levels of stress (in the elastic region) on the behavior of dissolution rates of metals in general. To quantify this effect, a system consisting of stainless steel in seawater was chosen as the subject of this investigation. An initial set of controlled experiments using nearly pure copper with NHOH electrolyte was used to test the experimental methods developed for this study and to verify the functionality of the numerical code in predicting large changes in geometry due to long duration dissolution. The numerical code is based on the BEM to predict the electrochemical dissolution activity in 2D and in 3D-axisymmetric geometries with nonlinearities in the response to stress and the boundary conditions given by the highly non-linear polarization response of the specimen. A Newton–Raphson iterative procedure is used to solve for equilibrium at each solution step. In the BEM code, a nodal optimization routine dynamically modifies the number of nodes and their location on the boundary, which is required by the large changes in geometry experienced during long duration dissolution. New SE-elements are developed to model sections of the boundary where nodes are dynamically located, defined by a curvilinear fit using orthogonal Chebyshev polynomials through previous nodal locations. The code links stress and potential type corrosion formulations to generate geometrical changes due to stress and corrosion. Polarization curves were measured and input into the BEM code and recession profiles were predicted. Comparison between experiment and predictions reveal that, given the polarization curves measured in the lab, the BEM code predicts accurate recession profiles. Once the laboratory methods and computer program were verified, a second electrochemical system is adopted to study the effects of stress in the linear range upon recession rates. This system consists of type 304 stainless steel in simulated seawater subjected to compressive and tensile stresses up to 20% of yield. Comparison between numerical predictions using polarization curves determined by experiment for the copper/ammonium system reveals that the BEM code developed to model recession of corroding surfaces faithfully reproduces the recession fronts measured in the experiments. Furthermore, it is shown in a series of repeatable laboratory tests, in the stainless-steel/saline system, that stress in the linear range indeed affects the polarization curves for different levels of stress and, furthermore, it is found that the shift in the polarization curve depends on stress rate.