HEAT FUSION MACHINE PARAMETER SETTING OPTIMIZATION BY IMPLEMENTING RESPONSE SURFACE METHODOLOGY OF EXPERIMENTAL DESIGN TO REDUCE UNLOCK HAIR DEFECT

Abstract

In PT M, several serious issues are caused by machine parameter setting not adjusted well with the production. The issue affects high product variation which will cause high number of defect. One of the issue is unlock hair which caused by heat fusion machine, there were 2,685,963 products that must be scrapped because of the machine failure. To overcome this issue, a problem solving step is conducted. There are 3 machine parameter settings that often to be changed or directly affect to the machine parameter setting, which are Blowing Temperature (will be depicted as factor A, 370-390 oC), Air Flow Rate (will be depicted as factor B, 70-80 mm3/s), and Blowing Time (will be depicted as factor C, 5-11 seconds). Response Surface Methodology is chosen as the experimental design method. This method is chosen because it is able to analyze the curvature of the factors and depict the curvature into a model, which makes it easier to be analyzed and defined. To be able to perform Response Surface Methodology, first model is created using 2k full factorial design. To check the curvature by adding the center points based on the replication of the design. If the curvature is significant, the second order model needs to be created. Second order model is created by using Box-Behnken design, linear and full quadratic depiction which will test the model for the lack-of-fit. The last step is to optimize the result by using response optimizer. The result are Blowing Temperature is set to 388oC, Air Flow Rate is set to 79 mm3/s, and Blowing Time is set to 9s with 93.38% of the model fit to the data.