Time dependent reliability analysis for cast iron pipes subjected to pitting corrosion. (August 2019)

• Record Type: Journal Article
• Authors: Wang, Weigang; Shi, Wenhai; Li, Chun-Qing
• Is Part Of: International journal of pressure vessels and piping. Volume 175(2019)
• Publication Date: 2019-08
• Subjects: Fracture -- Corrosion modeling -- Pipe failure -- Fracture -- Upcrossing method -- Stress intensity; Pressure vessels -- Periodicals; Pipe -- Periodicals; Récipients sous pression -- Périodiques; Tuyaux -- Périodiques; Pipe; Pressure vessels; Periodicals; 681.76041
• DOI: 10.1016/j.ijpvp.2019.103935
• Languages: English
• ISSNs: 0308-0161

Abstract

Abstract: In this paper, a method is developed to evaluate the probability of fracture failure of pressurized cast iron pipes subjected to pitting corrosion. A new model for corrosion pit depth is developed which considers the correlation between model parameters. The first passage probability method is employed to quantify the probability of fracture failure where the failure criterion is established based on fracture mechanics and the stress intensity factor is modeled as a nonstationary lognormal process. A case study is presented to illustrate the proposed method, followed by a sensitivity analysis to investigate the effects of contributing factors on the probability of fracture failure. It is found in the paper that the developed model for corrosion pit depth is able to reproduce the evolution of the pit depth growth in cast iron pipes in a short period of corrosion time (<10 years) and predict the growth of pit depth for a longer corrosion time (>10 years). It is also found that the risk of pipe fracture failure increases with the increase of corrosion pit depth and the decrease of fracture toughness of pipe material and that the corrosion depth and internal pressure have the most influence on the probability of fracture failure compared with other parameters. Furthermore, the contribution of each variable to the failure probability of pipe varies with exposure time. The method presented in this paper can assist pipe engineers and asset managers in developing a risk-informed strategy for pipe maintenance, repair, and replacement. Highlights: A new model for corrosion pit depth is developed which considers the correlation between model parameters. SIF is modeled as a nonstationary lognormal process with the up-crossing method employed for failure prediction. The developed corrosion model is able to predict the evolution of the pit depth growth well. The auto-correlation of the stress intensity factor between two points in time is found significant in failure prediction.