An Improved Expansion Process for Guyan Reduced Models: Technique for Improved Guyan Expansion Reconstruction (TIGER)

摘要

While many model reduction techniques exist, often times Guyan reduction is still performed mainly due to the fact that this is generally the only model reduction scheme available in all of the commercial finite element codes. As long as a sufficient number of active DOF are utilized, the Guyan reduction can produce reasonable results. However, there are errors due to the reduction and subsequent expansion of the reduced model back to the full space of the finite element model. In order to compensate for these, an alternate expansion methodology is proposed that makes two adjustments to the set of vectors obtained from the Guyan reduced model. First, the modal vectors are adjusted in reduced space and second, the Guyan transformation matrix is adjusted to properly account for the true modes of the system from the full space model. These adjustments essentially morph the Guyan reduced model and Guyan reduced transformation to a more appropriate transformation to achieve a better representation of the modal vectors associated with the full space finite element model. Several models are presented to illustrate the usefulness of the proposed methodology and technique. Advantages are discussed and limitations are presented for the cases studied.