Issue |
Mechanics & Industry
Volume 17, Number 2, 2016
|
|
---|---|---|
Article Number | 206 | |
Number of page(s) | 15 | |
DOI | https://doi.org/10.1051/meca/2015057 | |
Published online | 01 February 2016 |
A constant phase approach for the frequency response of stochastic linear oscillators
1 LMA, Centrale Marseille, CNRS, UPR
7051, Aix-Marseille University, 13451
Marseille,
France
2 LOFiMS, INSA de Rouen, BP 8,
76801
Saint-Étienne du Rouvray,
France
a Corresponding author:
emmanuelle.sarrouy@centrale-marseille.fr
Received:
29
October
2014
Accepted:
10
May
2015
When studying a mechanical structure, evaluation of its frequency response function (FRF) over a given frequency range is one of the main interests. Computational cost aside, evaluating FRFs presents no methodological difficulty in the deterministic case. Doing this when the model includes some uncertain parameters may however be more difficult as multimodality and discontinuity can arise around resonances. Indeed, even for a single degree of freedom system, it can be shown that usual methods of the probabilistic frame such as generalized Polynomial Chaos may fail to properly describe the probability density function of the response amplitude. This study proposes another approach which involves a shift in the usual quantities used to draw FRFs. Instead of computing the stochastic response for a given excitation frequency, this work adopts a constant response phase point of view. For each phase value of the oscillator response, the uncertainty over some parameters is propagated to the corresponding uncertain amplitudes and excitation frequencies. This provides much smoother variations of the involved quantities which are much easier to describe using a simple Polynomial Chaos approach. Both analytical and numerical results will be exposed for a single degree of freedom oscillator whose stiffness follows a uniform law.
Key words: Structural dynamics / frequency response function / random vibrations / uncertainty propagation / polynomial chaos
© AFM, EDP Sciences 2016
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.