ABSTRACT
This chapter discusses the concept of response surface methodology and reviews some response surface designs. Sometimes it is known that the linear response surface model is not quite adequate because a few factors are thought to interact with each other. This is handled in the model by adding a term consisting of the product of the two factors that interact. Of course, the design must be modified to be able to handle this interaction. The Box-Behnken designs are constructed by combining two level factorial designs on some of the factors with center points on the remaining factors. At least one center points for all factors should be added to the design. For three, four, or five factor designs, two factors are chosen to be at the extreme values, and the remaining factors are fixed at their center values.
