The noise and vibration behaviour of structures at higher frequencies has been a strong area of research activity of the Dynamics Group for many years. Applications typically concern noise and vibration transmission in land, marine and aerospace structures, buildings etc. The engineer needs tools to model and analyse such structures, as well as to measure and control their vibrational behaviour.
Special problems arise at higher frequencies for two reasons. First, wavelengths get shorter as the frequency increases, so that more detail is required in a finite element model. The computational cost thus grows enormously. Secondly, uncertainty and variability inevitably exist in the properties of structures - mass-produced commodities are never identical. At higher frequencies vibrational behaviour becomes sensitive to structural detail, so that the responses of different realisations of a structure become very different.
Our research concerns all aspects of this field, ranging from fundamental theoretical studies, through the development of numerical and analytical methods, to the application of these methods to a wide range of engineering problems. Approaches typically describe the vibrational behaviour of the structure in terms of energy and the flow of energy through the structure from source to receiver. They include statistical energy analysis (SEA) and wave-based methods, together with the development of hybrid SEA/FEA/wave approaches for mid-frequency and other applications. Specific engineering applications of our research include noise and vibration transmission in automotive and rail structures, pipework and tyres, together with measurement of energy flow through machinery mounts (using mobility techniques) and structural members.
Statistical Energy Analysis: SEA is a method for modelling the noise and vibration behaviour of complex systems at higher frequencies. It is particularly suited to those problems where many vibrational and/or acoustic modes contribute to the response, and to systems which are built up from subsystems joined together. It aims to predict the distribution of vibrational energy in the structure, and how this energy flows through the structure from source to receiver. Examples of applications include Automotive Noise and Vibration and Railway Noise and Vibration together with noise and vibration transmission in ship and building structures, the response of space satellite structures to launch noise and that of aircraft structures to jet noise.
While SEA is an established high frequency modelling technique, there are still many unanswered questions. Our research aims to extend the applicability and accuracy of SEA models and includes:
Vibrations of uncertain structures: the mid-frequency region: The properties of engineering structures are uncertain. At high enough frequencies the uncertainties are large enough so that some statistical approach is required, but neither FEA nor SEA are satisfactory. This is the mid-frequency range. A central concern is the development of numerically efficient modelling techniques to predict the flow of vibrational energy through the structure and to predict the response, together with its statistics. Our research includes:
Vibration transmission, waves and energy flow:
The transmission and control of vibrations in structures is often best described in terms of waves and energy flow. Wave techniques have been developed and applied to the modelling of vibration and shock transmission in structures, with applications ranging from aerospace structures, pipework systems, to vehicle bodies and tyres.
Energy flow is also a useful way to describe the power input to structures from sources, or the power that flows between coupled substructures. Our research interests concern simple, approximate, engineering methods for describing this energy flow for source characterisation, for response prediction and so on. Specific examples include a multi-pole method and model-reduction approaches using power-modes.
Measurement techniques have been developed to measure power in structures and previous work has resulted in several strategies and sensor systems. One broad aim is to measure the energy flow through isolator systems, from the machine to the surrounding structure, and to develop design rules, so that the engineer can assess the effectiveness of an isolation system. These are based on mobility approaches. A second is the measurement of structural intensity, i.e. the energy flow through structural components and methods have been developed for the bending and in-plane vibrations of beams and plates and for liquid-filled pipes. For these special techniques and algorithms are required.
Other techniques are also being developed and applied. These include extending spectral finite element methods to allow for distributed loading, modelling noise transmission in buildings using component mode synthesis and vibration transmission using strip plate theory and dynamic stiffness methods.
Contact:
Dr Neil Ferguson
Dynamics Group
Institute of Sound and Vibration Research
University of Southampton
Southampton SO17 1BJ
Telephone: +44 (0)23 8049 3225
Facsimilie: +44 (0)23 8059 3190
E-mail: nsf@isvr.soton.ac.uk