Introduction

Summary

Thus, for a complete investigation of dynamical systems, we require not only a computer and the direct integration methods. These provide no more than an ideal computer laboratory in which an arbitrary number of experiments can be performed, yielding an immense data flow. We require, in addition, certain principles according to which the data may be evaluated and displayed, thus giving an insight into the astonishing variety of responses of dynamical systems.

J. Argyris and H.-P. Mlejnek

Suppose the existence of a very powerful computer, so powerful that it can execute any given command such as build a cantilever beam, excite the beam with this force history, record the velocity histories, and the like. Suppose, further, that it cannot answer questions such as what is elasticity? why is resonance relevant to vibrations? how are vibration and stiffness related? Here then is the interesting question: With the aid of this powerful computer, how long would it take a novice to discover the law governing the vibration of structures? The answer, it would seem, is never, unless the novice is a modern Galileo.

Now suppose we add a feature to the computer; namely, access to a vast bibliographic database (in the spirit of Google or Wikipedia) that can respond to such library search commands as find every reference to vibrations, sort the find according to the type of structure, report only those citations that combine experiment with analysis, and so on.