The optimisation of train make-up and train handling-simulating longitudinal train dynamics.

Abstract

The South African rail industry is undergoing a phase of restructuring and much focus is concentrated on re-engineering i.e. optimising the utilisation of available assets and using existing technology in order to improve efficiency; attention shifts to improved heavy haul asset management through train performance models. The computer programs presented in this thesis have been developed to calculate longitudinal in-train forces accruing in long heavy haul trains and their effect on train operations. The model of the train is implemented by dedicated differential equations for the movements of each vehicle. The simulation is menu driven for all input and output decisions using Microsoft Excel while the engine for the dynamic analysis is ACSL (Advanced Continuous Simulation Language). The main program is capable of simulating the operation of any train configuration over any route, including remote operation. The thesis comprises: - • a discussion on the need for alternative train configurations based on the current fleet and the potential of such operating changes; • the comparison of the dynamic response of trains operating with only head-end locomotives, trains operating with both head-end locomotives and remote locomotives and trains operating with different class locomotives in one locomotive consist; • the investigation of the lateral effects in the different train consists as a function of the longitudinal in-train force in the simulation environment; • the advantages of operating with remote locomotives in terms of increased train length, reduced force spectrum on vehicle components and improved energy consumption; • the implications of the optimum position of the in-train locomotive consist on loading and unloading operations; • the implications of different train configurations on driver technique or train handling and the need for an optimum driving strategy to gain maximum benefit from the locomotives.