This research investigated the galvanic corrosion of the magnesium alloy AZ91D coupled to steel. A systematic research program was carried out, which aimed to understand the galvanic corrosion of an idealized component. The approach combined experimental measurements with numerical simulation. Numerical methods have been used for the past two decades in galvanic corrosion research. However, the numerical methods have not consistently provided predictions that were in agreement with experimental measurements.
The use of magnesium alloys in automotive applications has increased rapidly in the last decades. However, exposure to a corrosive environment can result in galvanic corrosion whenever a magnesium alloy is in contact with any other engineering metal, such as steel or aluminium. As a consequence, galvanic corrosion is a major concern for the application of magnesium alloys in industry. The accurate prediction of the galvanic current density distribution is important in the design and optimization process. but has scarcely been investigated in previous studies. This research makes contributions in: (1) the experimental measurement of the galvanic current density distribution of AZ91 D coupled to steel; (2) the prediction of the galvanic current density distribution using a numerical method and (3) the prediction of the galvanic corrosion behavior for an idealized auto component.
The research started with the evaluation of the commercial computer package BEASY. BEASY is widely used in the evaluation of galvanic corrosion. BEASY is based on the boundary element method (BEM). The BEASY predictions were compared with literature results for different electrolytes, various geometries and different solution film thickness. This evaluation indicated that the galvanic current distribution calculated by BEASY was similar in each case to the previous published results.
The properties of the electrolyte play an important role in determining galvanic corrosion. This study investigated the galvanic current distribution for an idealized lD AZ91D-steelgalvanic couple exposed to (1) a severe corrosive environment represented by a 5% NaCl solution, (2) a weakly corrosive environment represented by ASTM corrosive water, and (3) an inhibited environment represented by auto coolant. The galvanic current density distribution was measured using a specially designed lD galvanic corrosion assembly. The BEASY model used the measured polarization curves as input boundary conditions. The BEASY model predictions were in good agreement with the experimental measurements for the 5% NaCI solution, when there were steady state measurements for polarization curves as input to the BEASY model. The geometry of the galvanic couple also influences the galvanic corrosion. This study investigated the influence of area ratio of anode/cathode, solution film thicknesses, and insulation distances between anode and cathode. The BEM model gave predictions, which agreed in all cases with the measured galvanic current density distribution. Also investigated was the influence of two independent interacting galvanic couples. This showed that the galvanic current density on the AZ91D from the interaction of two independent galvanic couples was equal to the sum of the current densities from the independent galvanic couples.
The idealized component was an AZ91 D plate, with a central circular steel insert, exposed to a 5% NaCI solution. Steel inserts of various diameters were investigated. The BEM model and the experimental measurements gave a similar distribution of the current density distribution: a maximum at the interface with the current density decreasing to zero within I to 2 cm from the interface. The total corrosion was interpreted as being due to galvanic corrosion plus self-corrosion. The self-corrosion was evaluated on the basis that the BEM model provided a good evaluation of the galvanic corrosion. On this basis, the self-corrosion rate was evaluated to be typically -230 mm/y for the area surrounding the interface and to a distance of about 1 cm from the interface. Self-corrosion rates may be controlled by localized microgalvanic corrosion processes. SKPFM confirmed the presence of microgalvanic cells on the surface of AZ91D, which appear to have potential differences in the order of 100 m V between phases in the alloy.