Reliability Modelling of Train Set Motor Coaches for the Passenger Rail Agency of South Africa

Transcription

1 Reliability Modelling of Train Set Motor Coaches for the Passenger Rail Agency of South Africa Project by Kristov Trollip Presented by Pieter Conradie and Wyhan Jooste 4 6 June

2 AGENDA Introduction Methodology Model Development Model Application Conclusions, Recommendations and Future Work

3 Introduction Strategic position of the PRASA Engineering Research Chair Importance of Academic-Industry partnerships Based at Stellenbosch University since 2011 One of the biggest rail research institutes in Southern Africa Focus of the Chair Supply academic and scientific rail engineering solutions Assist PRASA to exploit assets Seamless integration of new technologies Origin of Reliability Modelling project Train sets are periodically maintained Static maintenance schedules No reliability based decision making model for maintenance

4 Introduction Motor coach Powered vehicle with ability to transport passengers as well as pull passenger trailers which is not powered Motor coach 4-7 trailers Motor coach 4-7 trailers Motor coach Train set A combination of 3 motor coaches with varying number of passenger trailers

5 Introduction 279 Train Sets 2228 kilometersof open track across South Africa 471 stations

6 Introduction Requirement stage Reliability Engineering Maintenance tactics Feasibility Conceptual design QFD Reliability Modelling Reliability HALT DOE, Taguchi, ANOVA FRACAS Data Analysis RCM SPC FRACAS Data Analysis RCM Manufacture Detail design Modifications Testing Construct / Install Acceptance Operation and maintenance Procure CAD, CAE FMECA Risk Assessment FME(C)A FTA HAZOPS DOE, Taguchi, ANOVA Monte Carlo Simulation RCM Reliability growth Data The probability that an item will perform a required function without failure under state conditions for a stated period of time Source: Smith (2007); O Connor and Kleyner (2011)

7 Introduction Objectives Apply knowledge/techniques to prevent failures Identify and correct causes of failures Find ways to cope with failure Estimate and analyse reliability

8 Methodology 1. Construct Reliability Block Diagram (RBD) graphical analysis technique made up of blocks block represents working state Interruptions lead to failed system Brake Valve 1 Brake Valve 2 Brake Valve 3 Exhauster 1 Exhauster 2 Exhauster 3 Exhauster cut out valve 1 Exhauster cut out valve 2 Exhauster cut out valve 3 2/3 2. Data Analysis Ascher and Feingold approach Chronological failure data Non-repairable systems Weibull distribution Repairable systems Non-Homogenous Poisson Process Trend? No Yes Repairable systems theory 3. Decision Making Model (DMM) model in constructed in Excel Dependence? No Yes Branching Poisson Non-repairable systems theory

9 Reliability Block Diagram Typical train set schematic with redundancy: Motor coach 1 Motor coach 2 Motor coach 3 Functional subsystems can be defined to system groupings, same end goal. Drive Subsystem Brake Subsystem Operational Subsystem Safety Subsystem

10 Reliability Block Diagram Drive Subsystem Brake Subsystem Operational Subsystem Safety Subsystem

11 Data Analysis On train set: MC13128 Failure inter-arrival data for period from 2008 to 2015 between pre- and post-2008 asset portfolios Component Failure Data Analysis in Excel

12 DMM Development

13 DMM Application Drive Subsystem Propulsion Subsystem Motor Controller 1 Line Switches 1 Pantograph 1 Motor 1.1 Motor 1.3 Motor Blower 1.1 Motor Blower 1.2 Motor 1.2 Motor 1.4

14 DMM Development Propulsion Subsystem Motor Controller 1 Line Switches 1 Pantograph 1 Motor 1.1 Motor 1.3 Motor Blower 1.1 Motor Blower 1.2 Motor 1.2 Motor 1.4

15 DMM Application

16 Conclusions & Recommendations DMM integrates with RBD allows for user decision making Model can be used for: Production Planning Determining Risk Maintenance scheduling Limitations: One train set at a time Failure data cleansing and configuration is laborious Recommendations: Failure data accuracy DMM needs to be expanded Incorporate maintenance and spare part models Study the effect of human action and technical skills

17 Future Work

18 Pieter Conradie, Pr.Eng PRASA Engineering Research Chair Department of Industrial Engineering Stellenbosch University Wyhan Jooste, PhD, Pr.Eng Asset Care Research Group (ACRG) Department of Industrial Engineering Stellenbosch University For more information and to apply visit: