Akademik Veri Yönetim Sistemi
International Conference on Engineering Technologies (ICENTE’17), Konya, Türkiye, 07 Aralık 2017, ss.74
There are many ways to reduce product quality faults and improve quality levels in production processes. Response Surface Methodology is an Experimental Design Method, is also an important tool for predicting response variables. In this study, a quality fault is considered as a response variable. The patenting process is a high added value production process in which zinc and phosphate coated steel wire productions are made. Quality faults were recorded in steel wires and priorities were given by Pareto analysis method. It was observed that 53% of the total faults occurred due to the wire tensile strengths. In this study, it was aimed to reduce the tensile strengths faults to a minimum. Within the scope of the study, 5 independent variables which can be measured at the level that can affect the response variable were evaluated. These variables were determined as the diameter of the raw material used for the wire, the amount of carbon, the amount of manganese, the uncoated wire diameter and the final product wire diameter. From the investigated independent variables, two formulation was obtained for two different patent wire groups with the aid of the full quadritic Response Surface Methodology model. The processes were monitored for some time by the resulting formulations. Also it was seen that the independent variables explained the response variable by 93%(R2). Tensile strength faults have been reduced by 62%. In addition to this, with the help of these theoretical models, more accurate and efficient use of the raw materials has been achieved.