Abstract
This dissertation is a study of the effect of various design factors on the performance of production lines. A review of past research on production lines reveals that the majority of production line models reported in the literature assume single server production stages with limited storage, and that although some analytical studies provide closed-form solutions, the study of manufacturing systems and the complex interactions between the stochastic design variables necessitates the use of simulation techniques. Historically, there have been few studies reported on the behavior of actual production lines. The availability of a new simulation technique known as GEMS (Generalized Manufacturing Simulator), recently developed at Texas A&M University, in part facilitated this study of production lines. A brief review of GEMS concepts and notation essential to the understanding of this work is included in this presentation. A balanced, multi-server production line with an equal number of servers per stage, equal finite interstage storage capacity and line length varying from five to twenty-five production stages is studied to observe the effects of line length, in-process inventory capacity, number of servers per stage and worktime variability on the performance of the line. Worktime variability is measured by the coefficient of variation for the probability distribution governing worktimes. The effects of these factors on performance measures such as server utilization, average blocked time per server and average in-process inventory per stage are observed using a combination of simulation and statistical experimental design techniques. Regression predictor equations are developed to provide quantitative estimates of the performance measures in terms of the significant design factors. Next, for purposes of comparison between actual and ideal production systems, a GEMS model of an operating production line is developed using actual data. Additional experimentation on constituent lines, within the main production line, shows the sensitivity of the production lines to changes in various parameters...
Dharia, Vijay Kalaram (1980). A study of the effect of various design factors on the performance of complex production lines. Texas A&M University. Texas A&M University. Libraries. Available electronically from
https : / /hdl .handle .net /1969 .1 /DISSERTATIONS -675439.