Direct measurements of the forces induced by the buffeting action of wind on a motionless bridge deck have been carried out at the Danish Maritime Institute. The wind tunnel experiments aimed at defining the spatial distribution of the wind loading as a function of the deck width and the scales of the incident turbulence.This paper presents the main findings of the experiments and illustrates the impetus behind this research which was to evaluate, for closed-box girder bridge decks, the error margin of the wind load predictions based on the strip assumption.

This state-of-the-art has been compiled by RILEM TC 109 TSA, Behaviour of Timber Structures under Seismic Actions. The Technical Committee completed this work during 1989-1993 and met in Italy, Portugal, Denmark, and the United Kingdom.Experience from seismic areas shows that some timber houses behave extremely well during earthquakes, but the mechanism for the good behaviour has not been fully analysed. Code writers have thus not found evidence for timber structures to be characterized in general as behaving well. Therefore, the main aim of the work of the Committee was to improve the information on wood and wood-based materials for the design of structures exposed to seismic loading. This includes the identification of adequate test methods.To achieve the main objectives it was necessary to look into the wide spectrum of requirements related to the seismic design of timber houses: the loads, the design method including their material input parameters, and test methods to determine these parameters. This is a complex spectrum and it was natural to coordinate this with related activities. One Committee meeting (in Florence) was therefore arranged in connection with a workshop for the presentation and discussion of the Eurocode on structures in seismic regions. One Committee meeting (in Lisbon) was arranged jointly with CIB W18, Timber Structures and, finally, one meeting (London) was arranged in connection with a workshop on the Full-scale Behaviour of Wood-framed Buildings according to an initiative by Professor A. Gupta.The meetings of the RILEM Committee have been used to identify specific tasks of interest and to discuss drafts presented to the Committee. Volunteers were found to draft the different chapters. This means that the Committee has influenced the scope and the specific papers, but that the different authors have kept the responsibility for their papers, being passed on in their final versions to the Editor. The individual chapters are thus papers which can be read individually (and minor overlaps or repetitions can be found), but which together give the background for recommending test methods.

The along-wind response of line-like structures to a turbulent wind can be estimated theoretically by methods originally proposed by Davenport in the beginning of the sixties.

The paper presents three essential modifications of the current analysis of the dynamic along-wind response of simple structures. It is demonstrated that the consistency requirement of zero mean turbulent flow leads to a reduction of the high-frequency response. A simple representation of the joint acceptance function via an accurate explicit approximation to the two-dimensional area of the joint acceptance integral of the cross-spectrum is presented. The total response is obtained from superposition of a quasistatic and a modal part. The new proposals are presented as an integral part of a critical appraisal of basis of analysis of dynamic along-wind response, suggesting that the turbulence length-scale is adequately represented by a simple power function of height. The results are summarised in the form of a design procedure, illustrated by examples.

Post mortem vibration measurements on one human tibia during gradual transection reveal the vibration modes and frequencies of a tibia during a simulated healing. The modes are identified in a tibia in an above knee amputation specimen with the leg in two positions: hanging down with the knee flexed (90 degrees) and supported in a special designed bone clamping splint (knee flexed 45 degrees). The vibration measurements are analysed using Modal Analysis and are translated to mechanical stiffness by mathematical modelling. The single bending 'free-free' mode turned out to be more sensitive to weakening of one cross-section than the 'rigid body' and single bending 'hinged-spring' modes. The error on the assessed value of the stiffness is a multiple of the error on the measured frequencies. This multiplication factor decreases for more sensitive modes. In this experiment, the results are accurate enough to reflect the asymmetric weakening imposed upon the tibia. Attempts are made towards automatization of the measurement and analysis in order to get a system for clinical use. The actual system is still too cumbersome and time consuming for standard clinical use.

The influence of soft tissues and joints on the vibration of the human tibia was examined by modal analysis on amputated lower limbs, where the soft tissues and the fibula were dissected gradually. Measurements were made in two different set ups, IFR and BRA, which were both designed to monitor fracture healing. In IFR, vibrations are generated by hammer impact on a relaxed hanging lower leg, with the knee flexed. Resonant frequencies are determined by a computer Fourier transform procedure. In BRA, a steady state vibration is induced in a lower leg, supported near the ankle and the tibial tuberosity, using an electromagnetic shaker. Resonant frequencies are determined from the maxima in vibration amplitudes. In both set ups the soft tissues have a similar influence on the vibration of the tibia: the skin hardly influences the determined modal parameter. The mass of the muscles inflences both the resonant frequency and the damping. The fibula has a stiffening effect on the tibia. The influence of the joints is small in the IFR-set up: the tibia vibrates in conditions close to those for the free-free vibration. In the BRA-set up, the supports determine the boundary conditions.

Vibrational methods to monitor fracture healing, the BRA and the IFR, are compared under different supporting conditions, excitation technique and signal processing. Mode shapes are identified by modal analysis. A wet excised human tibia and an amputation specimen are investigated. Excitation technique and signal processing caused only minor differences in the resonance frequencies. The supporting conditions had an important influence on the single bending modes changing both mode-shapes and frequencies. Thus the BRA-splint imposed a node at the malleolus. Modal analysis revealed the following modes in the two supporting conditions: BRA-splint: A 'rigid body' mode of 165 Hz in the sagittal plane. A single bending mode of 315 Hz close to the sagittal plane. IFR-hanging leg: A 'rigid-body' mode of 167 Hz close to the sagittal plane. Two single bending modes ('free-free'), a mode of 303 Hz close to the frontal plane and a mode of 470 Hz in the sagittal plane.

Driving point impedance technique was used for in vivo determination of the lowest frequency of resonance (Fa) in the human tibia. Optimum conditions for measurement were investigated. The precision of the method was 4.7 per cent and the greatest source of variation was the positioning of the leg and muscular tension. Fa was investigatyed during transection of the human tibia post-mortem and was found to decrease as stiffness was reduced by the transection. Accordingly Fa was decreased in four patients with crural fractures. The experiments indicate that the method can be used for the determination of fracture healing.