Corrosion of historic grey cast irons: indicative rates , significance, and protection

Abstract

In Europe many buildings and machinery in industrial sites are recognised as cultural heritage. These structures, often made from various types of irons or historic steels, have been for decades or centuries exposed to aggressive atmospheric environments and suffered from corrosion attack. The contribution discusses corrosion rates, the effects of corrosion on structural reliability, and the efficiency of surface treatments. The model for corrosion rates of historic metals cannot be based on the degradation model for mild steels even though specific features of historic alloys such as increased content of carbon and different chemical composition would be taken into account. realistic estimates of corrosion rates need additionally account for different micro-structure with inputs and different surface properties of historic alloys. This is why the presented model is based on a limited experimental data, considering the corrosivity of environment. The model assumes no corrosion during first seven years of service life and the same type of regression function for the progress period as is provided in ISO 9224 for mild steels and other metals. The effects of repeated applications of paintings are discussed. four principal strategies to the corrosion protection of industrial heritage structures include 'leave as it is', apply temporary protection to reduce degradation progress, apply long term protection, or undertake a complex restoration with replacement of damaged elements. Numerical example indicates that corrosion is normally insignificant for load-bearing iron structures, but may lead to severe problems for thin secondary structural and non-structural members such as railing or decorative elements. The proposed model estimates degradation progress in a mid-term perspective and supports decisions on maintenance of industrial heritage structures.

Probabilistic Analysis of Corrosion Rates and Degradation of Weathering Steel Bridges

“…In the absence of statistical data, a lognormal distribution can be assumed for the model uncertainty [1], [3]. Considering the variability of the exponent b for carbon steels according to ISO 9224, Vθ ≈ 12-13% was derived for long-term exposures, texp = 20-50 years [4].…”

“…βactual = [ln δ(texp) + αR β VR] / (αR VR,corr) (5) Considering available knowledge about corrosion progress and related total losses [4], it is expected that uncertainties of corrosion loss D may be insignificant for large crosssections members (for example, main bridge girders with large web thickness and flanges) and, in contrast, significant for small cross-sections members (side load-bearing members, truss members bridges). In the following three I-profiles are analysed:…”

“…Due to large uncertainties, the effect of corrosion on the reliability of the structure should be analysed using probabilistic methods [1], [4], [8], taking into account the randomness of resistances and load effects. In preliminary investigations aimed to provide first insights, a simplified reliability analysis based on a semi-probabilistic approach is conducted.…”

“…Initially focusing on a cross-section unaffected by corrosion, the design value of resistance Rd should take into account the variability of material strength, geometric variables and the uncertainty of the resistance model [4]:…”