ABSTRACT

The commercial pressure to reduce design costs and cycle times has driven dramatic progress in the virtual world of simulation. The field of structural dynamics is no exception and it is common nowadays, through computeraided design (CAD) and subsequent finite element (FE) analysis, to obtain predictions of the forced response of complex, linear structures with relative ease. These models necessarily incorporate assumptions regarding boundary conditions and element types, as well as parameter estimates such as material properties, joint stiffnesses and damping factors. At various stages in a product’s design and development, measurements may become necessary to provide partial experimental validation for a structural dynamic model such that it can be exercised with some degree of confidence to predict that which is unmeasured. For this purpose, frequency response functions (FRFs), which are transfer functions pertaining to the steady-state vibration of linear systems, are most commonly sought by experimental means. The FRF, which is more generally referred to as a transfer function by the experimental community, can be defined and named in a number of ways depending on the input and output quantities to which it relates, as listed in Table 9.1. Natural frequencies, mode shapes and modal damping values may be subsequently inferred through the application of experimental modal analysis techniques. An experimental modal model of a structure is also an invaluable aid per se in diagnosing noise and vibration problems of existing systems for which palliative vibration control solutions are required.