PROJECT VISION
Modeling Localized Corrosion
The Project Vision for the Corrosion / CSP II Consortium is divided into the following sections:
Objectives
Approach
Benefits
References
The objective of this project is to develop a simulation system that will make it possible to predict the occurrence of localized corrosion of engineering metals in complex aqueous systems. The system will help the user to understand the effects of environmental variables (e.g., composition of the aqueous stream, temperature, flow conditions, etc.) on the localized corrosion and assess the probability of failure. The proposed system will rely on the model for general corrosion, which is a part of CSP I
The computational system will consist of modules for predicting the initiation and propagation of localized corrosion. Also, it will include a module for assessing the probability of failure of a metal in a given environment. In this proposal, we briefly summarize the ideas that will be used as a starting point for developing these modules.
Initiation of localized corrosion
Various theories have been advanced to understand the processes that lead to the initiation of localized corrosion (cf. reviews by Szklarska-Smialowska, 1986 and Marcus and Oudar, 1995). In particular, the initiation of localized corrosion has been interpreted in terms of local acidification in pits (Galvele et al., 1978), thermodynamic equilibrium between passive films and salts (Vetter, 1965) or movement of point defects within a passive film (Lin et al., 1981, Macdonald and Urquidi-Macdonald, 1986, 1989). In some cases (e.g., in CO2 corrosion), localized corrosion may be initiated by fluid flow (Schmitt et al., 1999, Halvorsen and Sontvedt, 1999). The various theories that are available in the literature reflect the large number of possible initiation mechanisms.
Despite the multitude of possible mechanisms, the initiation of localized corrosion can be rationalized in terms of characteristic parameters, i.e., the pitting and repassivation potentials. Pitting initiates when the corrosion potential of the metal exceeds the pitting potential whereas existing pits or crevices cease to grow at potentials below the repassivation potential. In particular, the repassivation potential has been shown to be a conservative parameter for the long-term prediction of the occurrence of localized corrosion (Sridhar and Cragnolino, 1993). The pitting and repassivation potentials are strong functions of solution chemistry, temperature and alloy composition (Szklarska-Smialowska, 1986). Therefore, it is necessary to develop a method for predicting them with sufficient accuracy. Such a method should reproduce the available experimental data for simple systems and make it possible to predict the potentials for complex systems. The predicted pitting and repassivation potentials can be utilized in conjunction with the corrosion potential, which can be obtained from the model for general corrosion (Anderko and Young, 1999a,b). A comparison between the corrosion potential and the pitting and repassivation potentials will make it possible to predict whether localized corrosion is likely to initiate.
Pit growth
To compute the rate of pit growth, it is necessary to simulate the electrochemical processes that occur inside the pit and in the external environment. The solution inside the pit usually contains concentrated aggressive ions (e.g., halides) and is acidified in comparison with the external environment. In particular, the behavior of this solution is influenced by the formation of salt films. The dissolution of the salt film and transport of the metal-anion complexes may control the crack growth rate (Sridhar and Dunn, 1997, Brossia et al., 1998). At the same time, the external environment usually has a different composition, which corresponds to the bulk solution chemistry. It should be noted that the recently developed general corrosion model (Anderko and Young, 1999a,b) is applicable to modeling the polarization behavior of concentrated halide solutions in wide pH ranges and should serve as a convenient starting point for modeling the solutions within the pits.
Crack growth
In the case of stress corrosion cracking, a satisfactory model must account for the observed relationships between stress corrosion susceptibility and crack velocity and various environmental and mechanical parameters that characterize the system. Mechanical effects due to applied stress influence the crack velocity. Several alternative mechanisms have been proposed for the propagation of cracks (cf. a review by Newman, 1986). For example, it can be assumed that crack propagation occurs by film rupture and slip dissolution mechanism at the tip of the crack (Staehle et al., 1969, Scully et al., 1971, Swann et al., 1977, Macdonald and Urquidi-Macdonald, 1991). The kinetics of dissolution processes at the crack tip and in the external environment can be represented by the techniques that have been developed for modeling general corrosion. Because of charge conservation, the metal dissolution processes within the crack are coupled with the processes that occur on surfaces external to the crack.
Probabilistic aspects of localized corrosion
Localized corrosion is inherently stochastic in nature. Therefore, it is convenient to consider probabilistic measures of localized corrosion such as the damage function. The damage function gives the distribution of events (e.g., the number of pits per unit area) versus the depth of penetration. It can be obtained using deterministic models for pit initiation, growth and repassivation (Engelhardt and Macdonald, 1998). The damage function can be used to calculate the probability of failure for a specific wall thickness and estimated design life.
Effect of inhibitors
Inhibitors play an important role in both general and localized corrosion. The recently developed model for general corrosion takes into account the effect of inhibitors by considering adsorption equilibria in the active dissolution range and surface reactions between passive films and active ions from the solution. Currently, a limited number of inorganic corrosion inhibitors has been incorporated into the general corrosion model. In the future, a more extensive data bank of kinetic parameters will be developed for the computation of the effects of inhibitors. This will be especially challenging in the case of organic compounds. In view of the large number of possible organic inhibitors, an attempt will be made to develop a technique for relating the chemical structure of organic molecules to their tendency for adsorption and, hence, inhibition. Several molecular descriptors of the inhibiting capabilities of molecules have been investigated in the literature (Rozenfeld, 1981, Lukovits et al., 1997). They include the hydrophobicity, multipolar moments and molecular orbital energies (e.g., the difference between the highest occupied and lowest unoccupied molecular orbital energies). Alternatively, the adsorptive properties of molecules can be rationalized in terms of group contributions (Lukovits et al., 1995).
Anticipated Benefits of the Proposed Research
The computational system that will be developed as a result of this research will make it possible to investigate the effect of environmental or process conditions on the initiation and propagation of localized corrosion. It will provide a tool for simulating how the probability of localized corrosion of engineering metals will change when the composition of the aqueous phase, temperature, pressure or flow conditions are varied. Also, it will make it possible to assess the probability of failure due to localized corrosion. Thus, it can be anticipated that it will serve as a design tool as well as for understanding the causes of localized corrosion.
Anderko, A., Young, R. D., Computer Modeling of Corrosion in Absorption Cooling Cycles, CORROSION’99, San Antonio, TX, paper no. 243 (1999a).
Anderko, A., Young, R. D., Simulation of CO2/H2S Corrosion Using Thermodynamic and Electrochemical Models, CORROSION’99, San Antonio, TX, paper no. 243 (1999b).
Brossia, C.S., Dunn, D.S., Sridhar, N., in Critical Factors in Localized Corrosion III, Electrochemical Society, Pennigton, NJ, 1998.
Engelhardt, G., Macdonald, D.D., Corrosion, 54, 469 (1998).
Galvele, J.R. Lumsden, J.B., Staehle, R.W., J. Electrochem. Soc., 125, 1204 (1978).
Halvorsen A.M., Sontvedt, T., CO2 Corrosion Model for Carbon Steel Including Wall Sheer Stress Model for Multiphase Flow and Limits for Production Rate to Avoid Mesa Attack, CORROSION/99, San Antonio, TX, paper no. 42 (1999).
Lin, L.F., Chao, C.Y., Macdonald, D.D., J. Electrochem. Soc., 128, 1194 (1981).
Lukovits, I., Kalman, E., Palinkas, G., Corrosion, 51, 201 (1995).
Lukovits, I., Palfi, K., Bako, I., Kalman, E., Corrosion, 53, 915 (1997).
Macdonald, D.D., Urquidi-Macdonald, M., Electrochim. Acta, 31, 1079 (1986).
Macdonald, D.D., Urquidi-Macdonald, M., Corrosion Sci., 32, 51 (1991).
Macdonald, D.D., Urquidi-Macdonald, J. Electrochim. Soc., 139, 961 (1989).
Marcus, P., Oudar, J., Corrosion Mechanism in Theory and Practice, M. Dekker, New York, 1995.
Newman, R.C., in Corrosion Mechanisms in Theory and Practice (ed. by P. Marcus and J. Oudar), M. Dekker, New York (1995).
Rozenfeld, I.L., Corrosion Inhibitors, McGraw-Hill, New York, 1981.
Schmitt, G., Mueller, M., Papenfuss, M., Understanding Localized CO2 Corrosion of Carbon Steel from Physical Properties of Carbonate Scales, CORROSION/99, San Antonio, TX, paper no. 38 (1999).
Scully, J.C. (ed.), Theory of Stress Corrosion Cracking, NATO, Brussels (1971).
Sridhar, N., Cragnolino, G., Corrosion, 49, 885 (1993).
Sridhar, N., Dunn, D.S., J. Electrochem. Soc., 144, 4243 (1997).
Staehle, R.W., Forty, A.J., Van Rooyen, D. (eds.), Fundamental Aspects of Stress Corrosion Cracking, NACE, Houston, TX (1969).
Swann, P.R., Ford, F.P., Westwood, A.R.C. (eds.) Mechanisms of Environment Sensitive Cracking of Materials, The Metals Society, London (1977).
Szklarska-Smialowska, Z., Pitting Corrosion of Metals, NACE, Houston, TX, 1986.
Vetter, K.J., Ber. Bunsenges. Phys. Chem., 69, 589 and 683 (1965).
OLI, always committed to delivering technology