PSI - Issue 42

Darko Pastorcic et al. / Procedia Structural Integrity 42 (2022) 374–381 Darko Pastorcic et al./ Structural Integrity Procedia 00 (2019) 000 – 000

377

4

3. Numerical analysis 3.1. Modelling pitting corrosion

Pitting corrosion process can be considered as a complex physico-chemical process which consists of: pit generation and pit growth, Hong (1999), where pitting origination could be modelled by Poisson's process and pit growth in depth by Markov process, Kosareviych et al. (2016). More realistic models for pit nucleation consider also the pit repassivation, which occurs over time, Tarantseva (2010), and non-homogenous Poisson process. The spatial point pattern, i.e. distribution of the pits over surface is usually estimated using non homogenous spatial Poisson process, López De La Cruz and Gutiérrez (2008), López De La Cruz et al.(2008), with density function  . Besides Markov process, the pitting corrosion depth can also be modelled by gamma process, Velázquez et al.(2014). For large gamma distribution shape parameter, normal distribution closely approximates gamma distribution, Fig. 5.

Fig. 5. Normal and gamma distr. comparison, WM pit depth in the sea slush after 6 months

Diameters of the circular shape of pits, Paik et al.(2004), follow normal distribution. All simulated pit parameters were exported to the 3-D CAD program, where VBA ( Visual Basic for Applications ) routine was developed to obtain the 3-D model of corrosion pits on the surface of the test specimens. 3.2. Fracture criterion There are several ductile failure criteria which are used to solve engineering problems, such as Johnson-Cook, Gurson, SMCS (stress modified critical strain) by Hancock and Mackenzie(1976). The SMCS model includes the dependence of the strain limit to the triaxiality, which is the ratio of mean (hydrostatic) to equivalent (von Mieses) stress. The SMCS model predicts fracture when equivalent plastic strain in any point exceeds critical value over critical region, i.e. a length scale which characterizes critical volume: ̅ > ̅ = ∙ (−1.5∙ ) (1) where  is the material toughnes sparameter, T-triaxiality. The length scale parameter is based on the dimple size on the fracture surface formed as a result of void coalescence , Chi et al.(2006), and it is indicator of void colonies that generate macro crack formation, Shin et al. (2021). Based on the measurements, Ikram et al.(2020), the value of the length scale parameter is ≤ 0.1mm. The introduction of an average value of triaxiality, Bao and Wierzbicki (2004), into the formula (1) results with more reliable results, because it takes into account the previous conditions of the stress. Equation (1) can be rewritten in the following form: ̅ − ∙ (−1.5∙ ) ≥ 0 (2)