“…Among several approaches proposed in literature, normalization with a characteristic length (Khan & Angeles, 2006;Krefft & Hesselbach, 2005b) or a nominal link length (Lee et al, 2001), and partitioning the matrices into translational and rotational parts (Krefft & Hesselbach, 2005b;Lee & Kim, 2006) are the most popular choices. Normalization is not necessary for the sample problem presented in this chapter, as it possesses only a translational workspace.…”