Conclusions

Aiming at developing useful tools to quantify continued structural reliability of railroad tank cars as a function of service history, reliability analysis methodologies for tank cars subjected to fatigue and corrosion, the two commonly-encountered deterioration mechanisms for tank car structures, have been developed and demonstrated. Specifically, the following has been accomplished.

(1) A phenomenon-based tank car general corrosion model is proposed, which includes the effect of coating or lining of the tank metal to reduce corrosion. The model has the versatility to incorporate specific corrosion mechanisms through characterization of corrosion rate and corrosion initiation time using applicable kinetic laws. Corrosion reliability analysis performed has illustrated the usefulness of the reliability method in quantifying and predicting failure probability as a function of usage for given operating conditions.

(2) A methodology for tank car fatigue reliability analysis has been developed that uses commercial reliability analysis software, STRUREL, for the reliability analysis, while incorporating FORTRAN routines to perform fatigue crack growth analysis with the Walker crack growth law for a three degrees-of-freedom surface crack. The FORTRAN routines can be extended to include various other crack configurations and crack growth laws as more tank car material data becomes available. The methodology is versatile in dealing with a variety of random variables (with or without correlation), and capable of considering various features involved in tank car fatigue crack growth analysis, such as spectrum loading, residual stresses, and asymmetric, bi-variant stress field.

(3) Incorporating the general corrosion model with the fatigue crack growth model, a component reliability analysis model for corrosion-accelerated fatigue crack growth has been developed. Reliability analysis of corrosion-accelerated fatigue crack growth is performed with a time-dependent tank wall thickness reduction due to corrosion and a fatigue crack growth model that incorporates a three-degrees-of-freedom surface crack obeying Walker crack growth law under tank car spectrum loading. Illustrative examples are given for quantifying the effects of corrosion rates and corrosion initiation time on the probability of failure caused by corrosion-accelerated fatigue crack growth.

(4) A system reliability analysis methodology has been developed for tank cars with multiple corrosion sites. Important features of failure probability for a series system in relation to failure probabilities of individual corrosion sites are illustrated. The results show that the tank car reliability may be lower than that predicted for the most severe site, confirming the importance of system reliability analysis in developing a reliability-centered maintenance program.

(5) A methodology for the reliability analysis of a tank car with fatigue damage at multiple locations has been developed. It can be used to quantify the reliability of the tank car with fatigue damage at multiple locations and to assess the effect of damage at different individual locations on the reliability of a tank car as a whole, thus, providing necessary information for prioritizing the inspection, maintenance and repair of tank car structures.

(6) A methodology for system reliability analysis for tank cars having both fatigue and corrosion at multiple locations has been developed and demonstrated. Based on the state-of-the-art system reliability theory and solution methods, as well as taking into account of important parameters in fatigue and corrosion of tank car fleet, the methodology can be used to quantify railroad tank carís structural reliability as a function of usage, and to assess the effects of various parameters, various damage locations, and different failure modes on the overall reliability of the tank cars.

In summary, the accomplishments achieved through this project provides useful tools to quantify continued structural reliability of railroad tank cars under most of the commonly-encountered practical situations that occur in the tank car fleet. Although the current focus is on structural issues, the reliability theory and methods are universal and applicable to other aspects of tank cars, such as loading and unloading devices, safety devices, brake systems, trucks and wheels, as well as evaluating risks related to homeland security issues.

Full Paper