Oxford University, Civil Engineering: Research: Structural Dynamics: Projects

Structural Dynamics Research Projects

Current

The SERIES project at Oxford University

SERIES (Seismic Engineering Research Infrastructures for European Synergies) is a large, pan-European collaborative project funded under the EC Framework 7 programme. It is coordinated by Professor M Fardis of the University of Patras and involves 23 partners, including the University of Oxford. The project runs for four years from 1 March 2009 and the total funding provided by the EC is approximately 8.7M Euros, of which 320k Euros is allocated to Oxford.

The aim of SERIES is optimise use of European experimental facilities for earthquake engineering by taking advantage of advances in computational and communications technologies, providing access to large laboratories and developing improved testing techniques, hardware, instrumentation and control systems. SERIES has many aspects in common with the US NEES programme and the aim of SERIES is to improve collaboration both within the EC and with other centres of excellence such as the US, Japan and Taiwan.

Pdf with further information about SERIES

Past

Vibrations of Precast Concrete Floor Systems

Andreas Ehland, Martin Williams, Tony Blakeborough

Modern structures are often built with a high slenderness ratio. Especially floor-systems show liability to vibrations induced by life loads. To prove the serviceability of floor systems it is necessary to find out about the stiffness and damping of the structure. The effects of geometric aspects, design values and mechanical and material properties have to be analysed. It is also necessary to understand the dynamic loading which may give rise to the serviceability problems. The focus of this work is on floor systems built with precast and prestressed concrete elements under dynamic life loads.

In the precast plant an element is produced with two prestressed webs and a thin slab (about 60mm), called 'double-tee', using a stressing bed method and high-performance concrete. A reinforced layer of cast in-situ concrete is added after assembling the precast elements on the building site. The cast in-situ concrete is usually of normal strength.

To get accurate estimates for the stiffness and damping of the system, it is necessary to find out about several influences:

Firstly geometric aspects, like span length, depth and width of precast slab girder elements and depth of cast-in-situ concrete.
Secondly design aspects, like percentage of reinforcement and prestressing force.
Thirdly mechanical and material properties, like modulus of elasticity, effects of creep and shrinkage, bond stress and strength as well as cracks.

The structure will be analysed by using computer-based models and simulations. The results of these calculations should be verified by field testing. To find out about loads and dynamic action caused by fork-lift trucks it will be necessary to do some experiments in the field to get data for a statistical analysis. This data can be used to obtain a load model that can be used in the computer-based structure model.

Human-Structure Interaction In Cantilever Grandstands

Jackie Sim, Tony Blakeborough, Martin Williams

Problems of excessive vibration on cantilever grandstands have been reported both in the UK and overseas. There are two major factors contributing to this problem: increasingly lively human-induced loadings in pop concerts and greater flexibility of cantilever tiers. Excessive vibration on cantilever grandstands can cause human discomfort, crowd panic or collapse of the structure.

The present state of knowledge of the dynamics of grandstands is not sufficiently advanced to allow a sensible analysis to be performed. The reason for this is that current knowledge and practice are deficient in four main areas, (i) defining the human-induced load; (ii) modelling the human/structural system; (iii) analysing the systems and finally (iv) assessing the resultant vibration level prediction for serviceability criteria. This research project addresses the first three areas that are identified as deficient above. The first two areas involve the study of human-structure interaction which can be broadly classified into active and passive crowds interaction with the structure.

Active crowd exerts dynamic loads on the structure by jumping and bobbing. Experimental tests were performed to measure the vertical loads due to individuals jumping and bobbing in time to audio prompts set at certain discrete frequencies. Statistical analysis was carried out on the test results to model the jumping process of individuals in terms of the impulse shape and timing. The statistical model captures the random variation in timing one's jumps and bobs between individuals and enables the simulation of crowd jumping and bobbing loads.

A passive crowd (standing and seated) on the structure is modelled as a two-degree-of-freedom system, added to the structural system which is modelled as a single-degree-of-freedom system. Frequency domain analysis was performed to quantify the changes in the natural frequency and structural response of the occupied structure compared to when it is empty.

The final stage of this project involves analysing the dynamic response of the passive crowd-structure system subjected to active crowd loadings. A parametric study will be conducted to quantify the structural response as the ratio of active and passive crowds occupying the structure changes.

The findings of this project will enable designers to estimate the dynamic response of a structure for various ratios of active and passive crowds.

This project is now being extended through large-scale experimental work - see below.

Photo of stadium

Human Excitation of Cantilever Grandstands - Large Scale Experimentation

Anthony Comer, Tony Blakeborough, Martin Williams

This EPSRC-funded project addresses the problem of human excitation of cantilever grandstands. There is currently a lack of information on the loads crowds impose on grandstands and the effects that occupants have on the properties of grandstands.

To this end, extensive experimental work will be performed in the Structural Dynamics Laboratory at Oxford University. Some of the main objectives of the programme are summarised below:

To build a full-scale section of a raked grandstand.
To perform detailed tests to determine the dynamic loading that individuals in a crowd impose in a setting approximating the space constraints existing in a grandstand, both with and without the coordinating effect of external stimuli (visual, aural, grandstand motion).
To develop a numerical model of the mechanism by which individuals in a crowd synchronise their motions with each other and the various stimuli, calibrated using data from the tests above.

This project will provide important data which will be of direct interest not only to structural engineers involved in grandstand design but also to the authorities responsible for licensing sports stadia and stadium owners. Ultimately, the project will greatly benefit the safety of the many people who attend sports and music events in modern stadia each year.

Click the link for a fuller description of our work on cantilever grandstands.

Seismic Analysis of Guyed Masts

Matthew Grey, Tony Blakeborough, Martin Williams

With recent advances and expansion in the telecommunications industry, the need for taller and more reliable antenna supporting structures has increased. As design heights increase above 100m, the use of guyed masts becomes increasingly more cost effective relative to other free standing options.
The analysis of these structures is complex, with the structure exhibiting non-linear characteristics even under working conditions. This non-linearity is due to the fact that the stiffness of the guy cable changes with changes in the tension of the cable, leading to a non-linear force/displacement relationship in the structure.

Current codes and guidelines relate predominately to the design of these structures under wind loading conditions with allowances made for the dynamic response of the structure by a number of methods. It is normally assumed that this is the dominant load case and a detailed seismic analysis is not undertaken.

Following a number of recent collapses and permanent deflections beyond acceptable limits (leaving the antenna with reduced capabilities), a closer investigation is needed in order to ensure that communication networks are fully operational following seismic events to facilitate rescue efforts etc.

This project undertakes to model the response of a number of different sized guyed masts to seismic loading as stipulated by Eurocode 8, using SAP2000, and aims to provide design recommendations as well as an evaluation of EC8 and current standards.

Real-Time Substructure Testing

Paul Bonnet, Martin Williams, Tony Blakeborough

Real-time substructure (RTS) is a novel method for the dynamic testing of structures. It is hybrid by definition, as it involves sharing a complex model between a physical model and a computer model. It aims at combining the advantages of physical testing with those of computer simulations. The structure of interest is divided into two entities:

On one hand, parts that can easily be numerically modelled, because they have a simple behaviour or because they are not crucial for the analysis conducted.
On the other hand, the more crucial parts, physically replicated, for example exhibiting high non-linearity.

These two categories are exactly complementary and their combination forms the complete structure of interest: the emulated structure. The first category is numerically simulated and its dynamics are solved through time integration. This is the numerical substructure. The second category is physically modelled and subject to dynamic loading through hydraulic actuators. This is the physical substructure. Both substructures need to interact with each other at their interface in order to represent the emulated structure. Thus, their dynamics can only be studied simultaneously and in real-time.

Real-time substructure (RTS) testing is especially convenient to study the behaviour of structures that contain regions difficult to model computationally. It allows one to concentrate on the behaviour of a specific part, while having the rest of the structure modelled with an infinite repeatability.

Pdf with further information on real-time substructure testing.

Dissipative Devices in Seismic Design

Martin Williams, Tony Blakeborough

The group has a long-standing interest in the improvement of the seismic performance of buildings through the use of dampers and other dissipative devices. We have worked on the development of a variety of hysteretic devices and visco-elastic dampers, using large-scale cyclic testing, real-time substructure testing and finite element modelling. The results have been used to develop detailed design guidelines to optimise device performance. Work in progress aims to evaluate the applicability of displacement-based design procedures for frames incorporating dissipators.