Weighting analysis of pellet quality attributes using Multi Response Signal to Noise (MRSN) method

Abstract

Quality of pellets, one form of animal feed, is not only measured by the nutritional content but also by its physical form. The physical strength of the pellet is determined from crushing and not easily moldy. Both quality characteristics are measured by reliability (pellet durability index) and resistance (water content percentage). In order to improve the quality of pellet, this study applied Multi Response Signal to Noise (MRSN) method. The weight of product quality attributes used will influence the method in determining the selected alternatives. To accommodate the weighting of dynamic product quality attributes, this study also ran weighting sensitivity analysis of product quality attributes. The results showed that the combination of factor level that produced the optimal pellet is A2, B1, C1, D1, E1, F1, G2 or combination of production process run with vapor pressure 1.9 bar, temperature conditioner 80 ° C, 3.5mm pellet diameter mold, cooler temperature 30 ° C, time in cooler 2 minutes, roller distance 1.5 cm, mixing time 175 seconds. This optimum combination can increase PDI percentage by 2.132% and decrease difference to target of water content by 0.234%. The optimum factor level combination will change if the weight for % PDI rises to be more than 0.77228 or decreases to be less than 0.00561, or in other words, the optimum combination will not change if the weight for % PDI is in the range 0.00561 - 0.77228.