Evolution of the corrosion layer morphology on analogues after several wet/dry cycle in a climatic chamber (transverse sections).
Since several years in the frame of the French CEA COCON program, studies try to predict and model the atmospheric corrosion behaviour of iron for long duration (i.e. several hundred centuries). For this purpose, it is necessary to couple phenomenological modelling to investigations on archaeological analogues. The mechanisms occurring inside the rust layer during a wet-dry cycle are considered as well as the characteristics of the rust layer formed on archaeological artefacts during these processes and a related mechanistic modelling has been developed. In order to obtain experimental values of damage rates, ferrous archaeological artefacts have been aged in climatic chambers. Damage and corrosion rates have been determined from the weight gains measured during the ageing treatments. Moreover, these rates have been compared with those obtained on a low-alloy steel. The comparison between the values determinate according to the mechanistic modelling and these experimental values obtained on old rust layer gives a relative good agreement. The similar results, that have been obtained both on artefacts and contemporary low-alloy steel, seem to corroborate the modelling calculations which indicate that most of the corrosion phenomena take place at the surface of the rust layer where the lepidocrocite is mainly observed. Considering that the most important parameter controlling the corrosion rate seems to be the lepidocrocite content in the rust layer, this amount of lepidocrocite will govern the damage caused during the wetting stage. Furthermore, the rust layer ageing tests provide good information about the morphology and the composition in term of phases of the rust layer versus the number of cycles. * CEA/DEN/DPC/SCCME/LECA ** CEA