Abstract
The corrosion of mild-steel tubes, through which chloride solutions are flowing at velocities up to 400 ft/min., has been studied at temperatures up to 55° and test periods from 3 days to 10 weeks. Corrosion was assessed from the change in weight of tube specimens and the products characterised by X-ray examinations and bulk density determinations.
Where the corrosivity of the solution is relatively low the weight loss due to corrosion w, is a rectilinear function of period of test, t, i.e., constant corrosion rate, k1. In more corrosive solutions the corrosion rate decreases with time owing to the restraining influence of the corrosion product and also by restriction in flow which under certain conditions leads to complete blockage. Both forms of corrosion/time curve may be represented by an equation of the form found previously.
w = K1 t/1 + K 2 t
where k1 is the initial corrosion rate and k2 is a ‘blocking factor’ describing the retardation in corrosion rate.
A relation is developed for the initial rate, k1 (which varies from about 20 to over 2000 mg/dm2/day according to the chloride concentration, c), involving temperature, t, and flow velocity, ν. This equation gives values for k1 Which agree with the experimental values to within ± 30%. Optimum values of k2 are found in dilute chloride solutions at medium and high flow velocities.
At slow flow velocities, attack is in the form of longitudinal bands of etched metal while in more rapidly moving corrosive solutions a coarse deposit with nodules up to 6 mm high may be formed. The corrosion product consists of a mixture of magnetite and hydrated ferric oxide, the latter being the major constituent in the outer layers. The hydrated oxide is mainly lepidocrocite, γ-FeOOH, but goethite, α-FeOOH, and haematite, α-Fe203, are often present in small amounts. From bulk density determinations the product is shown to be spongy in character With voids occupying up to 50% of the total volume.
Get full access to this article
View all access options for this article.