Use this URL to cite or link to this record in EThOS:
Title: Dynamic performance of high frequency floors
Author: Middleton, Christopher James
Awarding Body: University of Sheffield
Current Institution: University of Sheffield
Date of Award: 2010
Availability of Full Text:
Access from EThOS:
Access from Institution:
This thesis presents an investigation into the dynamic performance of high frequency floors (HFFs) due to walking excitation. The rationale behind the research is the lack of knowledge of the dynamic characteristics of the HFFs and the poor guidance available in estimating the vibration response due to walking excitation. The most recent design guides published for the evaluation of HFF are by the Concrete Centre and the Steel Construction Institute. The design of HFF is based on similar principles used in the design of low frequency floors (LFFs). However, the dynamic characteristics of each floor type are very different, which leads to inaccurate response estimations. The thesis is split into three main sections: the classic source, path, receiver layout. The source, within the scope of this work, is excitation due to walking. HFFs are defined at which resonance will not occur from walking, and are designed for environments that require low levels of vibration. Traditionally, a floor's response to walking was analysed using the harmonic amplitudes of the footfall force. As resonance does not occur, this approach is no longer valid, and a number of different methods were developed, namely the 'kf method', Arup's 'effective impulse', and a polynomial method presented in a European Commission (EC) report. It was shown that each method has a number of flaws and characteristics for a new, improved, footfall model are defined. A new footfall model is created based on a cubic spline fit of 'key points' of the footfall force. The new model included intra-subject variability (i.e. natural variation between each pace rate). The new model was shown to be more accurate than current models recommended in the relevant guidance. Due to the inclusion of variation in the model, the spline force was also valid for LFFs and is therefore the first accurate universal force model. Assessment of the path consisted evaluating current deign methods for HFFs. It was found that for a transient response of floor (i.e. not resonant) a large number of modes contribute to the response. As such, the only simplified guidance suitable for analysis was the guide published by the Concrete Centre. The Concrete Centre guide was then compared with finite element analysis (FEA) and was found to give inaccurate response estimates due to poor estimates of modal mass and mode shapes. HFFs are often large structures containing many floor bays. These multi-bay structures have interesting characteristics, unique to this style of floor. A large parametric study considering the effects of the number of bays within the structure, the size of the floor bay and the stiffness of columns had on the characteristics of the floor. It was found that mode groupings of closely spaced modes exist due to the large number of bays. It was also found the column stiffness affects the modal mass of the floor. Due to the complexities of the large multi-bay floors, a number of methods were investigated to make the analysis process more efficient. A method of modal participation was developed to assess the importance of the large number of modes. The degree of modelling detail that was required in the model was investigated, and it was found that away from areas of interest the structure could be modelled very crudely. Wave propagation analysis was conducted on the floors using the spectral element method (SEM) applied to a grillage model of the floor. It was shown that the SEM had advantages over conventional FEA, including more efficient analysis and the use of semi-infinite elements. Assessment of the receiver consisted of an evaluation of the current generic vibration criteria for sensitive occupancies. The vibration criteria were assessed under a number of different types of responses. It was shown that if a criteria for a sensitive machine was developed using one type of excitation (e.g. pure-tone sine) it could not be compared with the response of another type of excitation (e.g. broadband random). Overall, it was shown that all aspects regarding response estimation of HFFs require further research. The work presented in this thesis adds to the current knowledge surrounding HFFs.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available