Date of Award


Publication Type

Master Thesis


Industrial and Manufacturing Systems Engineering




We develop a new analytic approximation method to replace a set of parallel machines by an equivalent machine in a series-parallel flow line with finite buffer. We develop our method based on discrete state Markov chain. The proposed technique replaces a set of parallel machines at a work centre by an equivalent machine in order to obtain a traditional flow line with machines in series separated by intermediate buffers. We derive equations for the parameters of the equivalent machine when it operates in isolation as well as in flow line. The existing analytic methods for series-parallel systems can tract only lines with a maximum of two machines in series and a buffer in-between them. The method we propose in this thesis can be used in conjunction with an approximation method or simulation to solve flow lines of any length. We also model and evaluate the performance of series-parallel systems manufacturing more than one product types with predefined sequence and lot size. We address this issue for a considerable longer flow line system with finite buffer which is common in industry. We consider the set-up time of the machines as the product type changes, deterministic processing times and operation dependent failures of the machines. We analyze the effects of buffer and number of machines in parallel on the performance of series-parallel systems.