High Efficiency Valve Design by Robust Design of Experiments

When we design a valve system there are many design factors and their values to decide. Generally, design factors which affect the whole performance of compressors are strongly related to each other so we can't decide the value of each design factors respectively. In order to decide the optimum value (among levels) of those factors simultaneously and reduce the number of experiments we applied robust design of experiment in this paper. Thereby we performed the experiment with great efficiency and found out that the degree of contribution of each factors to the whole performance of compressors quantitatively. By applying this experimental method (robust design of experiment) we developed highly efficient valve system for refrigeration compressors relatively short time and with low cost compare to conventional development process. INTRODUCTION Compressors are core part of refrigeration cycle and refrigeration capacity and performance of whole system are affected by them. In this respect, studies about improvement of efficiency of compressor have continuously performed. And because the green house effect of the earth has been accelerated, demands for highly efficient electric home appliances have been increased. In household refrigerators, power consumption of compressors occupy over 80% of whole power consumption of refrigerators. So it is evident that development of efficient compressor is essential for making refrigerator which have low power consumption. For improving efficiency of compressor, we have to decide many values of design factors. At this stage, not only are simulation(theoretical) methods used, but also experimental methods are used. But it is very difficult to find the exact value of design factors by theoretical method only. Because, there are many assumptions to simulate real systems, we usually use theoretical methods to find trends and predict design values qualitatively. And finally, experimental mer:hods are used to find exact value of design factors. In this paper, we focused on experimental method.