Variation Propagation in Multistage Machining Processes Using Dual Quaternions

Abstract

Abstract The application of rigid transformations matrices in variation propagation has a long tradition in manufacturing community. However, the matrix-based modeling of variation propagation in multistage machining processes is complicated. Moreover, there is a need to improve the computational efficiency of manipulation of geometrical models for variation analysis purposes. This paper introduces the representation of rigid transformation by dual quaternions, which have a computational advantage and mathematical elegance compared to matrices. In comparison to a commercial tool, the implementation of the purposed method predicted parallelism with an average error of 4.3 % in a hypothetical two stage machining process. The proposed approach has a potential to be an alternative to matrix based rigid transformation practices in variation and tolerance analysis.