Corrosion Modeling Using Electrochemistry and Computational Fluid Dynamics

Corrosion modeling using electrochemistry and computational fluid dynamics

The Butler-Volmer equation for surface electrochemistry potential and Laplace equation for electrolyte solutions have been available in CFD code for a while, but the challenge remains to find a numerically affordable way to model the mass-transfer limited corrosion rates. Mass-transfer coefficient using the Sherwood number (Nesic et al., Corr Sci., 1996) approach seems promising, but it is limited by the use of average pipe velocity values, which is unable to differentiate local flow changes in geometries like pipe bend and welding locations. Conventional methods use the wall shear stress as the surface mass transfer parameter, but unfortunately it is extraordinarily expensive to solve wall shear stress accurately even for flows in a straight pipe.  A better method is proposed to use the free stream velocity as the input parameter to the Sherwood number. However, the free stream velocity value requires the estimation of the boundary layer thickness in turbulent flows and its accuracy depends strongly on the turbulence models. Interestingly enough it was found that the free stream velocity value is not sensitive to the accuracy of boundary-layer thickness predictions and can be reliably used to identify flow velocity variations due to local geometry changes. This method effectively addresses the drawbacks of the Sherwood number approach, and avoids the costly computation needed to accurately predict the wall shear stress. The proposed model was applied to a pipe bend experiment conducted by Zhang et al. (Corr. Sci., 2013) and the results show that the predicted variations of corrosion rate are with 10 % range from the experimental results.