Development of a framework for the reduction of manufacturing defects in a composite material process

• Main Author: Muhammad Iqbal Muhammad Hussain
• Format: Thesis
• Language: English
• Published: Universiti Malaysia Perlis 2009
• Subjects: Composite materials; Engineering design; Manufacturing defects > Analysis; Manufacturing defects > Reduction; Sandwich construction; Composite construction > Fatigue; Manufactures > Engineering
• Online Access: http://dspace.unimap.edu.my/xmlui/handle/123456789/4910
• Institution: Universiti Malaysia Perlis

Reduction of defects is a critical issue in manufacturing operations. It goes without saying that defect reduction leads to manufacturing cost reduction, and this translates to increased profitability for the organization. Company A is a manufacturer of high technology composite materials. Of late, it has been experiencing high levels of defects from its manual hand lay-up and autoclave processes that generate products coded ‘L’and ‘T’in this thesis. Thus, a study that integrates the use of ‘statistical design of experiments’(SDE), ‘failure mode and effect analysis’(FMEA), several side experiments, control charts, and certain process controls is carried out. The study combines time-tested industrial problem-solving and process-improvement methods in a way that is both regimented as well as flexible, in line with the numerous uncertainties that inevitably present themselves in any live manufacturing environment. This culminates to the development of a generic framework, of which its execution enables the determination of the best process set-up that gives the minimum number of defects in the final product. Taking into account the circumstances under which the processes operate, fractional factorial design (2^ (4-1) resolution IV design) is used in production line ‘L’, and , a two-level factorial blocked design with 24 runs and eight center points is used in production line ‘T’. These designs give much insight into this line’s defect-causing variables, and enables the examination of important process parameters such as geometry of core, temperature, pressure, and cooling rates, to name a few. Consequently, after the entire research process is carried out, it is seen that the number of defects is greatly reduced (from 30 panels/month to 3 panels/month for wrinkles, 18 panels/ month to 4 panels/month for delamination in production line ‘L’, and from 25 © This item is protected by original copyright xxviii panels/month to 5 panels/month for delamination in production line ‘T’), leading to tremendous cost savings on the shop floor.