Computational & Technology Resources
an online resource for computational,
engineering & technology publications 

Computational Science, Engineering & Technology Series
ISSN 17593158 CSETS: 14
INNOVATION IN COMPUTATIONAL STRUCTURES TECHNOLOGY Edited by: B.H.V. Topping, G. Montero, R. Montenegro
Chapter 20
Recent Advances in Modal Reduction of Vibrating Substructures D. Givoli*, P.E. Barbone+ and I. Patlashenko#
*Department of Aerospace Engineering, Technion  Israel Institute of Technology, Haifa, Israel D. Givoli, P.E. Barbone, I. Patlashenko, "Recent Advances in Modal Reduction of Vibrating Substructures", in B.H.V. Topping, G. Montero, R. Montenegro, (Editors), "Innovation in Computational Structures Technology", SaxeCoburg Publications, Stirlingshire, UK, Chapter 20, pp 417438, 2006. doi:10.4203/csets.14.20
Keywords: substructure, model reduction, modal reduction, DirichlettoNeumann map, vibration, linear dynamic systems, finite elements.
Summary
In structural dynamics, one often wishes to approximate a given
discrete model of a structure by another model which involves a
much smaller number of degrees of freedom. Such a reduction is
very important in cases where the computational effort associated
with the direct analysis of the given system is prohibitive.
There is a very large volume of literature on the
subject, often called "model order reduction;" . For example the
review papers [1,2,3,4]. Some of this work is
related to robust control, where there is a need for the
repetitive realtime solution of large structural dynamic
problems [].
Modal reduction is a special case of model reduction. The original linear system is first decomposed into its eigenmodes. Then a small number of these eigenmodes is retained to represent the system, whereas all the other modes are discarded. The following question then arises: "Which of the modes should be retained?" In structural dynamics it is traditional to retain those modes associated with the lowest frequencies [6]. In control theory, a common procedure is the balanced realisation method proposed by Moore [7], where a special mode truncation is used to obtain a reduced system with equal amounts of controllability and observability. Both these approaches are simple and easy to code, but they are not based on any optimality criterion. Hence, although in many cases they produce very good approximations, they are not guaranteed to do so. In this paper, we consider a linear substructure "attached" to a main structure. Such a substructure may represent a piece of equipment such as an antenna, connected to the main structure. On the other hand, if an engineer is interested in the dynamics of different antenna designs, the "substructure" may be most of the structure, while the antenna may be treated as the "main" structure. We wish to reduce the substructure alone, without modifying the main structure. When doing this, we are not interested in the accurate representation of the dynamics of the substructure itself, but in accurately representing the effect this substructure has on the dynamic behaviour of the main structure. Here we ask a question similar to the one mentioned above: In the reduction process, which of the substructure's modes should be retained? In contrast to the lowfrequency rule dominating structural dynamics, we shall obtain a new criterion for "modal importance." We shall show that the most important modes of the substructure are those whose coupling matrices, to be defined in a particular way, have the highest norm. This will lead to a simple and effective algorithm for optimal modal reduction. Recently [8,9] we have developed such a modal reduction scheme for substructures with linear dampingfree timedependent behaviour. In the present paper we first review the general subject of structure reduction. Then we extend the scheme mentioned above to substructures which experience Rayleigh damping; this is an important extension which turns out to be quite involved technically. The approach we take here involves the DirichlettoNeumann map as the main analysis tool. We show the resulting scheme to be approximately optimal in a certain sense. We compare the scheme via a numerical experiment to standard modal reduction. References
purchase the fulltext of this chapter (price £20)
go to the previous chapter 
