Application of Taguchi experimental design in iron and steel industry

The iron and steel industry provides raw materials to many industries such as construction, automotive and white goods, and plays a major role in the industrialization and development of countries. The success of companies operating in the iron and steel industry highly depends on the quality of the products produced. Mechanical properties such as yield strength, tensile strength, hardness etc. are very important especially in the quality of steel products and these properties are affected by many factors such as chemical composition and heat treatment procedures. Therefore, examining the factors affecting mechanical properties of steel products is an important issue in terms of quality. Experimental design is a statistical method in which factors that have significant effects on the response variable, which is the performance characteristic of a product or process, and the optimum levels of these factors can be determined. However, as the number of factors and factor levels increase, the number of experiments to be conducted also increases. In this concern, the Taguchi experimental design method, which uses orthogonal arrays, allows the investigation of the entire experimental region with a smaller number of experiments and it provides significant savings in time and cost. In this dissertation, a company that operates in the iron and steel industry is considered and it is aimed to find the factor levels that optimize mechanical properties of S355J2 quality structural steel. To this aim, five factors (%C, %Mn, %V, N(ppm) and rolling temperature) are defined and effects of these factors on yield and tensile strength are evaluated using the Taguchi experimental design method. In this context, experiments are conducted based on L8 orthogonal array and the obtained results are analyzed by using MINITAB. In the analyses, the effects of the factors on yield and tensile strength are examined separately and optimum factor levels are determined. Additionally, the problem is expressed as a multi-response surface problem in which yield and tensile strength are evaluated together, and factor levels are optimized using the response surface method. Finally, validation experiments regarding optimum factor levels are carried out in the company and the results are compared with both MINITAB outputs and the current situation. The results show that the optimized factor levels give better results in terms of yield and tensile strength.