Browsing by Subject "Manufacturing processes"
Now showing 1 - 7 of 7
Results Per Page
Sort Options
Item A synergistic methodology and adaptive model for material management in a dynamic environment(Texas Tech University, 1998-05) Carrigo, E. AllenThis dissertation and the associated research compares the three main manufacturing philosophies of Manufacturing Resource Planning (MRP), Just In Time (JIT) and the Theory of Constraints (TOC) with a new manufacturing concept termed the Adaptive Model. The Adaptive Model uses the synergistic effect of combining selected aspects of the JIT and TOC philosophies into a new manufacturing philosophy. Specifically, the adaptive model identified a second constraint in addition to the primary constraint. This secondary constraint was both optimized and given an increased buffer size. It was envisioned before conducting the research that the adaptive model would be superior to the other models. Three different manufacturing lines of five, nine and fifteen workstations were developed and modeled. Each of these models with different buffer sizes was simulated for approximately ten time periods of six months. Performance factors were identified and employed for evaluating the four philosophies. These factors included Throughput levels. Work in Process (WIP) quantities, Time-in-the-system periods, the maximum amount of idle time at the main constraint and the utilization rate for the main constraint. Data for each of the performance factors were collected and analyzed. Additionally, hypothesis tests were developed and conducted. Differences existed between the various manufacturing philosophies, according to the tests. Observations of the data suggest that the adaptive model produces sporadic superior results when compared with the other philosophies. However, the adaptive model exhibits higher costs of larger WIP levels, increased time-in-the-system lengths and increased costs of operating at an accelerated pace. The research conclusion is that the TOC model for a flow shop manufacturing environment is the best overall manufacturing philosophy.Item Automation of semiconductor processing equipment(Texas Tech University, 1999-08) Vuppaladadium, VijayThis thesis describes a methodology to Automate Semiconductor Fabrication Equipment. The current processing Industry makes use of stand-alone equipment with built-in Microcontrollers, which are hardcoded or programmed using EPROMs to accomplish that particular process. The cost of building such systems is expensive. Automation of these systems is time consuming and difficult. These systems require a lot of user intervention during processing. This thesis presents a comprehensive insight into a generic approach of Automation of a process. A methodology has been discussed to automate a machine using DAQ (Data Acquisition Boards) and Extemal Interface Boards controlled by LabVTEW, a graphical programming language tool. This approach of automation is implemented and verified on two processing systems, namely a Plasma Etcher and a Plasma Deposition System. The project explains the specifications of the hardware needed and describes a modular approach to design the LabVIEW control program. It explains how this approach can achieve improved process performance by efficient monitoring and controlling of the process parameters for increased yield and productivity. The advantages of this methodology of automation are discussed along with applications.Item Development of a novel stopping technique for optimization(Texas Tech University, 1997-12) Iyer, Mahesh SubramaniamNeural networks are being used widely in areas of process control, pattern recognition, etc. The possibility of improving the efficiency of data utilization in neural network training and automating the decision to stop training, using a novel Steady-state Identifier (SSID) algorithm, have been investigated. One conclusion is that complete automation of the decision criterion to stop training is probably beyond the realm of possibility and human judgment seems unavoidable. However, as a beneficial outcome of this study, a technique has been developed to determine the number of neural network training repetitions to guarantee the convergence of the training algorithm within a certain vicinity ofthe global optimum of the objective function, with a desired level of confidence. The concept used is the weakest-link-in-the chain analysis. As another outcome, a novel approach of stopping neural network training has been developed. In this technique, a random fraction ofthe training set data is sampled at each epoch. The error on the random fraction is tested for its attainment of steady-state or otherwise using either a novel Steady-State Identifier or equivalently by visual observation by a human operator. Training is stopped when the error on the random fraction attains Steady-State. This technique, in general, is more cost effective than cross-validation. The overall developments are perfectly general and can also be applied to optimization problems other than neural network training.Item Experimental designs and analyses to measure process variation capabilities(Texas Tech University, 1986-12) Raju, PNot availableItem Part and tool scheduling rules for a flexible manufacturing system(Texas Tech University, 1983-12) Acree, Elaine StrongScheduling rules for a specific general purpose Flexible Manufacturing System were investigated. The system consists of six general purpose machines with local storage at each machine, a work-in-process queue and a material handling cart. The primary purpose of this research was to investigate effects and interactions of three types of scheduling rules on the FMS performance. The rules included part scheduling on the machines and two resource allocation rules for tool scheduling and cart scheduling. In addition the part scheduling rule was modified with a tool look ahead rule to minimize tool delay. Cue to the numerous similarities in the scheduling and resource allocation problems of computer operating systems and the FMS, some techniques from operating systems were applied to the FMS tc reduce thrashing, prevent deadlocks and increase cart utilization efficiency, A simulation model was developed to investigate the scheduling rules. SLAM was the language used to simulate this system. The main performance criteria used in this model were machine utilization, cart utilization, total time in system, number of finished parts and number of completed stages. The most important result of this research was the discovery that individual tool allocation is superior to total tool allocation. For the system studied, the performance measurements were insensitive to the part and cart scheduling rules. However a severe reduction in cart speed caused the system to thrash. The reduction of cart speed provided the proper experimental conditions to prove that SDTF can be superior to ECFS for this system operating under thrashing conditions. The tool look ahead feature provided a gross improvement in machine utilization and a reduction in time waiting for tools. However this improvement was not statistically significant in all cases.Item Program for a ceramic tile manufacturing plant: Monarch tile manufacturing, Inc. San Angelo, Texas(1969-01) Stiles, Marvin E.Item Support free infiltration of selective laser sintered (SLS) silicon carbide preforms(2006-12) Stevinson, Brooke Young; Bourell, David LeeIndirect selective laser sintered (SLS) silicon carbide objects have been successfully infiltrated with silicon. Selective laser sintering has several advantages over conventional manufacturing methods of silicon carbide parts including more greater geometrical freedom, lower process temperatures and pressures, cheaper starting materials, and higher geometrical tolerances. Post-processing of selective laser sintered silicon carbide objects entails binder conversion to carbon and subsequent silicon infiltration to create fully dense parts that include reaction-bonded silicon carbide. These objects exhibit an overfilling of silicon onto all surfaces, especially on the bottom and in corners. Geometrical studies and microscopy were used to characterize overfilling patterns. A metrology study and generation of an Ashby Densification map proved that shrinkage of the silicon carbide structure was not the cause of the overfilling. Further research and calculations determined that the origin of overfilling was the volume expansion on freezing of silicon. An attempt to infiltrate silicon carbide preforms with an alternate infiltrant was unsuccessful due to incompatible wetting characteristics. This method is advantageous for silicon carbide applications that require complex geometries and for which composite material is acceptable.