Carlex Glass Laminated Side Glass Process Optimization Jake

Carlex Glass Laminated Side Glass Process Optimization Jake Massingill, Project Coordinator Marshal Turner, Process Optimization Specialist Mike Mund, Project Manager Systems and Industrial Engineering Dr. Adeel Khalid, Mentor Methods and Tools Introduction Carlex Glass manufactures laminated side glass for automobiles. Laminated glass offers increased safety, reduced road noise, and improved theft resistance. The manufacturing process for this type of glass consists of five stages: Cutting and Grinding, Bending and Tempering, Lamination, Autoclave, and Finishing. This project focuses on the last three phases of this process. Currently, this line has a daily throughput of 4, 320 pieces of finished glass per day, ignoring losses. There is no clear value of how many racks of glass to insert into the autoclave per cycle. Two shifts of four workers conduct final trimming of the vinyl from the edge of the glass and inspect the pieces before they are shipped. Key Questions to Answer • How many racks of glass should be inserted into the autoclave before a cycle is initiated? • The queue sizes before and after the autoclave phase will be profoundly impacted; if this value is too high, production may back up before this phase. If it is too low, glass will build up around the finishing area. • In general, higher values are preferred as this reduces cost because the autoclave will run fewer cycles per day. • How many work stations should be utilized at finishing? • There iis space for up to four finishing stations, each with a trimmer and an inspector. There also twelve hour shifts per day. It is important to determine the minimum number of workstations required to account for the total capacity of the process. Lamination Autoclave Finishing • Vinyl is applied between two bent and tempered glass pieces, then trimmed to shape • Glass pairs and vinyl pass through a furnace to remove air bubbles and provide adhesion • Glass is placed in racks of 75 pieces, then inserted into the autoclave for about three hours • High temperatures and pressures created by jets of steam permanently seal the glass and vinyl into one unit • Excess vinyl is trimmed from around the edges of the glass • The finished glass is inspected for defects, such as chips, debris between the glass panes, optical distortions, and other undesirable features Figure 1: Simplified System Overview Diagram Analysis Questions Answered Figure 2: Arena Simulation of Autoclave and Finishing Processes Minimum Success Criteria • Resource utilization (time busy/12 hour shift) shall be improved to 90% or better. • Expenses of this line shall be reduced by $8, 000 per month. • Consistency shall be brought to the finishing process such that the queue length is rarely zero until the end of the shift. • The finishing process will be able to adequately handle the capacity of all prior phases; the throughput of finishing will be greater than or equal to the throughput of previous phases. • Implementation cost and time will be zero dollars and less than one week, respectively. • Design alternatives will be verified by simulation and presented to the project manager. • Arena Simulation served as the most important tool utilized in this project. This tool was used to determine the number of racks to insert into the autoclave per batch, decided how many work stations are necessary at finishing, and the values of the queue sizes and resource utilization. • Time studies were conducted in order to acquire process times for this simulation. Where time studies would not be feasible, theoretical specification data was acquired from the project manager. • A force field diagram was produced in order to model the resistance to changes in autoclave batch size and number of workers at finishing versus the forces seeking to make these changes happen. • This diagram was developed from anecdotes from Carlex employees given during interviews and also from the personal experiences of the project manager and project coordinator. • Maintainability, forecasting, and economic analysis were also utilized to determine future improvements that could be made to the system as well as to provide guidance to determining the value of autoclave batch size (a larger value is preferred based on economic consideration, so long as it is also optimal in the simulation). • Twelve racks should be inserted into the autoclave per cycle. • The total number of autoclave cycles should be minimized in order to minimize cost. • This number of racks does not create a buildup of glass before or after the autoclave • Only one work station is required per shift. • In the simulation, it was determined that two shifts of one work station each can handle the capacity of 4, 320 per day • By economic analysis, it is most beneficial to maintain the minimum number of work station in order to achieve maximum profit. Conclusions and Results Figure 3: Original Model Worker Activity versus Time Figure 4: Improved Model Worker Activity versus Time Figure 5: Optimized Model Worker Activity versus Time Three models were created in Arena: the Original Model, the Improved Model, and the Optimized Model. The Original Model uses an arbitrary autoclave batch size used to represent the fact that this value is currently not constant. It also utilizes the current number of workers per shift, which is four. The Improved Model holds the autoclave batch size constant at twelve racks per cycle and also cuts the workers at finishing down to two per shift. The Optimized Model uses the same batch size and workers per shift as the Improved Model, but increases the capacity of the line to 4, 800 pieces per day, which is theoretically feasible. As one might expect, the Optimized Model is the superior choice, though the Improved Model can be implemented without any adjustments to machinery and is worth consideration if this proves to be difficult. • In the Improved Model, resource utilization was found to be 85. 28%. In the Optimized Model, it was found to be 93. 55%. • Expenses were reduced by over $16, 000 per month. • Consistency has been brought to the line because the wait time for workers is easily predictable and constant • All three models are capable of handling their respective capacities. • Improvements have no dollar cost and can be implemented within a day for both the Improved and Optimized Models. • All three models have been verified in Arena and will be presented to Carlex management this weekend. It can be concluded that the Optimized Model meets all Minimum Success Criteria, while the Improved Model meets 5/6.