When a fleet of trains is nearing the end of its service life, the city transit authorities have to decide whether to replace the aging train inventory or apply life extension remediation to the existing fleet. Replacement of the entire inventory comes at a significant cost. Therefore, evaluation of the remaining service life and possible life extension methods for the current fleet is worth investigating.
The objective of this study is the evaluation of remaining service life of existing municipal trains and assessment of risk associated with life extension in order to help authorities in the decision making process of applying life extension measures or purchasing new equipment.
The main factors limiting service life is exceedance of material fatigue resistance and fracture. To this end, finite element models (FEM) of the train cars in question were created in ABAQUS® modeling software to find locations of stress concentration under prescribed service loads. Locations of concern from the FEM were then used as the basis for instrumentation planning for on-site testing. A single train unit was instrumented with strain gages and accelerometers and then loaded and tested under regular operating conditions. Collected data from the on-site test was then used for calibration and validation of the FEM. Load-time history was constructed based on the calibrated FEM and the data from testing. Calculation of fatigue damage accumulation was done in accordance to the Rain-flow counting algorithm and Palmgren-Miner rule. The remaining fatigue life was evaluated based on available S-N curves from test data.
Loads, as well as material resistance are random variables; therefore a reliability approach has to be applied in the calculation of risk associated with service life extension in order to make a sound recommendation as to the risk of continued service beyond the prescribed service life. Statistical parameters of load and resistance where gathered and used in calculations for this recommendation.
This paper presents the approach for service life evaluation/extension, reliability and stochastic methods for risk assessment in a real world example.