Degree Name

Doctor of Philosophy


Department of Metallurgy and Materials Engineering


An electrochemical test system of computer programs has been written that controls electrochemical instrumentation to carry out galvanic current, impedance and corrosion current tests, and processes the data. This test system has been applied to the study of corrosion mechanisms for painted and unpainted metals.

Laboratory simulated atmospheric conditions in a pollution gas box, PGB, shown that mass loss of uncoupled steel increases markedly with increase in [SO2] up to 1 ppm, but that Zn, Al are relatively unaffected, whereas Zn-55%Al alloy shows a modest increase. Sacrificial protection of a steel substrate may therefore be difficult at the higher [SO2]. However, ACM tests reveal the opposite is true when steel is galvanically coupled to Zn-55%Al alloy, which has difficulty providing sacrificial protection at zero and [SO2], and a mechanism is proposed. Corrosion of Fe, Zn-55%Al and Zn in the presence of up to 1 ppm [SO2] appears to be under mixed control, the first because increases in [SO2] cause increases in corrosion brought about by depolarization of anodic and cathodic reactions. Random partial short circuits in Zn and Fe ACMs in third and fourth PGB cycles due to electron conducting corrosion products can make interpretation of data difficult. This has been modelled by the inclusion of an interfacial corrosion product resistance, Rcp in the metal/solution equivalent circuit. Several methods are presented in order to determine the effect of Rcp . Mass losses for Fe, Zn-55%Al and Zn have been calculated from ACM galvanic current, impedance, and corrosion current data over the [SO2] range and compared with weighed mass loss.

The impedance test has been applied to study the permeation of chloride ions through paint films (painted Zn-55%Al coated steel), the rate of which appears to be rapid for the films tested, and about 50% complete within several minutes. Solution uptake has been modelled by an equivalent circuit and paint film degradation is shown to occur predominantly at direct or D-type areas where the film resistance follows the resistance of the solution. Comparisons between free and attached film results, both here and below, provide evidence that the high frequency Nyquist semicircle contains paint film information and not metal substrate information.

Impedance tests have also been applied to study the effect of paint film flatting agent particles, which was found to be minimal, on solution uptake. Corrosion resistance of the substrate is controlled by the rate of degradation of the paint film, both of which decrease from the moment of sample immersion. Water initially enters the film but after a few hours this rate slows down and paint film degradation, probably resulting from hydrolysis of the silicone modified polyester top coat portion of the paint film, plus ingress of chloride ions, increase so that substrate corrosion also increases. Sample failure time decreases as solution chloride ion concentration increases, in agreement with outdoor exposure results. This is opposite to some other paint films where performance life is controlled by the solution osmotic pressure. Water uptake over the first few days is greater for attached than for free films, thought due to some special influence of the substrate. However, attached films perform better for a number of reasons. The capacitance method for calculating water uptake is more convenient experimentally but not as meaningful as the gravimetric method due to the superimposed effect of paint film degradation occurring after a few hours with these films. A comparison of single frequency and wide frequency range impedance methods indicates that the steeply rising, falling sections of 1 kHz resistive, capacitive/time trends can be simply explained by the rate of decline of paint film resistance.