Summary of Research
There is a huge need for efficient and practical tools for predicting
vibration in complex structures, from automotive brakes to oilwell
drills.
A host of simplified methods exist for studying linear systems (usually
valid when the motion of the structure is sufficiently small), but
linear theory can only go so far. Real structures often have regions
that are significantly non-linear, such as friction forces, requiring
less efficient methods of computer simulation to investigate. This is a
problem that is exaggerated when trying to account for variations and
uncertainties such as those resulting from a production line.
The aim of this research is to find ways of modelling non-linear
behaviour efficiently by exploiting the fact that many examples of
non-linearity occur in localised regions of the structure. This allows
computational effort to be focussed on these regions while retaining the
efficiency of linear analysis for the majority of the structure. A
core objective of the research is practical relevance. Therefore the
new methods developed will be assessed against experiments designed to
exemplify the key features of complex structures with localised
non-linearities that make their study traditionally difficult.
The benefits of a successful approach have the potential to impact a
wide range of areas such as brake noise, drilling and machining, turbine
blade vibration, offshore structures, MEMS devices and 'buzz, squeak
and rattle'
within the automotive industry, and potentially within medical contexts
such as drilling in orthopaedic surgery and understanding tinnitus
(spontaneous vibration in the cochlea). Such goals are long-term and
require expertise in other fields. The role of this research is to
develop novel tools that may open new opportunities for research in
these areas.
The project will be
in collaboration with the Bristol Laboratory for Advanced Dynamics
Engineering (BLADE) and the Structural Dynamics and Vibration Laboratory at
McGill University, Montreal.
