The replacement of copperlbrass radiators in the automotive industry with radiators made from aluminium components provided the basis of this research. Since aluminium is more susceptible to corrosion than either copper or brass, factors that contribute to its corrosion are of major interest and importance, and have been investigated. Three different aluminium alloys were selected for study because of a special interest in their corrosive behaviour by the automotive industry. These are the aluminium alloy AA 3003 (samples A and B) and two supplier specific alloys (sample D containing Zn and sample E containing Cu and Mg). The various joining operations used in the automotive manufacturing process dictated the preparation of the aluminium alloys used for corrosion studies. Mechanically Assembled (MA) aluminium radiators use alloy samples as supplied by the aluminium industry and hence suitable experiments were carried out on the 'as-supplied' (AS) samples used for both finstock and tubestock material. The development of Composite Deposition (CD) Technology to braze together finstock and tubestock material introduced new challenges to corrosion research. To gain an insight into the corrosion of a Brazed aluminium radiator, all samples were subjected to a thermal profile identical to that experienced industrially under a Controlled Atmosphere Brazing (CAB) furnace. Two cases of interest emerged. Firstly the 'heat-treated' (HT) samples were used to evaluate the effect ofheat treatment on the alloy's resistance to corrosion. Secondly, alloy samples treated with a Composite Powder Coating (CPC) and then subjected to the thermal profile provided a surface of an AI-Si melt which represented the brazed joint. Experiments on these samples yielded information on the AI-Si melt and the likely corrosion in a brazed joint. The resulting corrosion of the AS, HT and CPC samples immersed in various corrosive electrolyte solutions for 60 minutes was examined using two microscopic techniques. Firstly, the actual surface pitting was examined using a Scanning Electron Microscope (SEM), and secondly, cross-sections of the samples mounted in a resin, then suitably polished and etched were examined using an optical microscope to further reveal the nature of corrosion of the samples. The nature of corrosion was best revealed in an acidified chloride solution. The AS samples showed delocalised crystallographic pitting consisting of coalesced pits at localised regions of the surface. The HT samples showed IV localised crystallographic pIttIng consIstIng of many individual pits and intergranular corrosion both at and below the surface. Intergranular corrosion was most severe for HT sample E containing Cu and Mg. The CPC samples showed total corrosion of the surface layer and eutectic AI-Si melt, some crystallographic pitting of the a-AI filler metal, and crystallographic pitting including intergranular corrosion of the base alloy. The extent of corrosion was found to depend on the chemical composition of the aluminium alloys, the presence of Zn, Cu and Mg causing more severe corrosion of the aluminium alloys, with the effect ofZn being most severe. The electrochemical investigation involved the measurement of two fundamentally important parameters. Firstly, the open circuit potentials (OCP) of the alloy samples immersed in the various corrosive electrolyte solutions were measured as a function of time. Secondly, the pitting potentials (Bp) of the alloy samples were measured using anodic polarisation techniques by extrapolation of the resulting log i vs E plots. The OCP and Bp of the AS samples were found to be influenced by the chemical composition of the aluminium alloys. Heat treatment of the AS samples was found to change their microstructure and solid solution composition which in turn affected the electrochemical results. The effect of the Composite melt layer on the electrochemistry of the CPC samples is discussed. Micrographic and electrochemical results were used to assess the best combination of finstock and tubestock material that would yield an aluminium radiator most resistant to corrosion. The likely corrosion of the components in these combinations was assessed and these results were compared with the actual results obtained industrially using the SWAAT exposure test. / Thesis (M.Sc.)-University of Natal, Pietermaritzburg, 1999.