1991 Research Reports No R629-R647

Abstract
This paper develops a mitre model for the shear strain distribution in steel members under uniform torsion.

The mitre model provides reasonable approximations for the shear stress distributions, except near reentrant corners, and accurate predictions of the fully plastic uniform torque. The elastic torque-­twist relationship is predicted with high accuracy, and the inelastic relationship with reasonable accuracy.

The mitre model uses a simple approximation to the shear strain distribution caused by uniform torsion. It can be expected to provide a relatively simple method of predicting the inelastic behaviour of steel members when uniform torsion acts in combination with axial force and bending moment.

Abstract
Theoretical investigation is presented for the wind-induced motion of tall/slender structures equipped with tuned mass dampers, taking into account the effect of soil compliancy under the footing. The superstructure is modelled as a lumped mass multi-degree-of-freedom cantilevered system and the soil behaviour, including footing embedment effect, is characterised by a known frequency-dependent compliant matrix. The along-wind turbulence and the cross-wind wake excitation are modelled as stochastic processes ~which are stationary in time and non-homogeneous in space. A transfer matrix formulation is used to analyse the soil-structure-mass damper interaction in the frequency domain and a Monte-Carlo simulation is employed to demonstrate the system response behaviour in the time domain. Numerical examples show that soil compliancy will affect structural responses and the effectiveness of tuned mass dampers, depending on the properties of the soil, the properties of the structure, the nature of the excitation and the type of the structural response.

Abstract
Artificially cemented samples of a carbonate soil have been manufactured with predetermined dry densities and cement contents. Tests to investigate the uniformity and repeatability of the specimens have shown that the manufacturing process is satisfactory. The variations of unconfined compressive and tensile strengths with density and cement content have been investigated. Polished sections of the specimens have been examined using electron microscopy and have shown the changing nature of the effects of the cement at different densities and cement contents.

Abstract
This Paper presents results of a limited series of model pile tests in Offshore calcareous soil subjected to three different storm loadings. Attention is concentrated on the accumulation of pile displacement induced by the storm loading, and on the post-storm load capacity of the pile.

A non-linear boundary element analysis is described and is used in an attempt to predict the model test results based on son parameters derived from earlier tests. The analysis is also employed to simulate the general cyclic behaviour of a field grouted pile test in calcareous soil. Some measure of agreement is found between predicted and measured behaviour. Finally, the analysis is used to investigate the behaviour of a hypothetical, but realistic, offshore pile subjected to a complete storm loading sequence.

Abstract
This paper presents a modified procedure for the analysis of the axial response of piles embedded in multi-layered soils. The results obtained by this procedure are compared with those computed by some previous methods and with a limited number of field test measurements.

In the determination of the group settlement interaction between piles embedded in multi-layered soils, an additional simple soil mass stiffness model is developed in order to include the horizontal non-­homogeneity of the soil due to soil disturbance caused by pile installation. The predictions by this model agree more closely with the observed field test group performance than do predictions by the conventional method which assumes lateral homogeneity of the soil.

Abstract
A closed-circuit turbulent boundary layer Snowdrift Wind Tunnel (S.W.T.) was built in the School of Civil and Mining Engineering, the University of Sydney for the study of Antarctic snow drifting. Scaled turbulent boundary layer winds’ representing prototype summer and winter conditions (coastal sites of Australian Antarctic Territory), were generated. Sodium bicarbonate was chosen as model snow out of a number of different materials tested. , The results of snowdrift simulation using the model snow were compared with the field measurement taken by Mitsuhashi (1982). Based on the comparison of experimental results and field study results, the significance and relevance of various similitude parameters to the snowdrift modelling are discussed. These parameters include flow and geometry, angle of repose, particle trajectory and time scale. In particular, the similitude requirements of scaled turbulent boundary layer wind, which is believed to be one of the important modelling requirements, are discussed in detail.

Abstract
This paper presents a derivation of the classical energy equation for the lateral buckling of doubly symmetric thin-walled beams which is based on the non-linear relationship between the longitudinal normal strain and the member deformations.

The classical equation is compared with an alternative energy equation, and found to be significantly different. The difference is attributed to the omission of some non-linear components from the longitudinal displacements used for the alternative equation.

Comparisons of finite element predictions based on the two energy equations demonstrate some substantial differences. Comparisons with available experimental evidence agree well with the classical predictions, and provide independent evidence that the alternative energy equation is incorrect.

Abstract
An analysis of a pile subjected to lateral soil movements has been outlined and used to investigate the behaviour of a pile in an unstable slope. Attention has been centred on the response of the pile and the modes in which the pile may be considered to have failed. A realistic example has been employed to illustrate the effects of a number of pile and soil parameters. The importance of assessing the likely deformation magnitude and pattern, which is a natural consequence of the analysis used here and is not available from limit state equilibrium analyses, is highlighted.

Abstract
This paper employs a modified non-linear boundary element approach to analyse the response of offshore piles subjected to external soil movements arising from submarine slides. 7leoretical solutions for a realistic hypothetical offshore pile computed by the approach are presented and discussed. The analysis is then used to analyse the behaviour of two full scale offshore piles in submarine slide areas, and reasonable agreement is found between observed and theoretical behaviour.

Abstract
The aim of this study was to investigate the possibility of modelling the most significant features of tornadoes with respect to their effects at near ground level. Tornadoes typically are local vortex-type phenomena with high-speed rotating cores. Due to their unpredictability there is a lack of data of the loadings they can impose on structures and hence it is desirable to provide a convenient and controllable model testing system for such cases.

After trying a number of possible arrangements, a satisfactory system has been evolved which is essentially a closed cylindrical chamber which has fluid movement initiated and maintained by a drive unit in its upper part, thereby leaving the lower part free for ground and structural modelling.

Abstract
This study was aimed at extending knowledge on the performance of various arrangements of ‘Liquid Column Vibration Absorber’ or LCVA for possible reduction of vibrations of buildings or other large structures which may be subject to excitation by wind or other causes.

An earlier laboratory study indicated the potential efficacy of one arrangement of LCVA, and the present studies were made on a number of possible variations of the basic system.

The first set of tests was made to see what difference in performance occurred with a tenfold difference in the mass ratio of absorber mass to main body mass. The second set was to observe the effects of adding orifice restrictors to a LCVA. The third set of tests was to compare the performances of a long absorber and of a short absorber where both had the same mass.

Abstract
Thin-walled channel sections may undergo a mode of buckling called distortional in which a lip stiffened flange of the section rotates about the flange-web junction. If the sections are composed of high strength steel, then there may be a significant post-buckling reserve of strength beyond the elastic distortional buckling stress in a similar manner to that which normally occurs for local buckling.

In this report, the results of compression tests of thin-walled channel section columns formed by brake-processing are described. Two different section geometries, a simple lipped channel and a lipped channel with an intermediate stiffener In the web, wore tested between fixed-ended boundary conditions. The geometries and yield strength were chosen to ensure that a substantial post buckling strength reserve occurred in the distortional mode and that local buckling might occur simultaneously at a shorter wavelength. The deformations were measured using a movable frame on which linear displacement transducers were mounted. The residual stresses and material properties wore also measured.

Distortional or mixed local distortional buckling stresses obtained from testing are compared with the theoretical buckling stresses. Design curves for the maximum strength of columns undergoing distortional or mixed modes of buckling are proposed and compared with the test results.

Abstract
A series of wind tunnel model tests and theoretical analyses were conducted to investigate the mechanism of torsional excitation, torsional response of tall buildings, sensitivity of the torsional response to eccentricity between centres of twist and building geometry, as well as the effects of tuned mass dampers on the torsional response of tall buildings. An aeroelastic model for pure torsion vibration was developed at the Fluids Laboratory of the University of Sydney. The aeroelastic test results were compared with those obtained by force balance technique, direct pressure measurement technique and multi-degree-of-freedom aeroelastic models. It is shown that this type of modelling technique can be a convenient and efficient way to explore the mechanism of torsional excitation and predict the torsional response of tall buildings to wind. It is also shown that tuned mass dampers were quite effective in suppressing the torsional vibration of the building if the parameters of the tuned mass damper were properly selected. A reliable and economical parametric study method of tuned mass dampers is presented, which is based on directly measured torsional excitation or response spectra. A mode shape correction factor was used to adjust the mis-match between the model and prototype torsional mode shapes.

Abstract
The analysis of soil-raft interaction is carried out by using the finite layer method to determine the behaviour of the soil, and the finite element method for the analysis of the raft. The method developed in this paper involves setting up influence matrices for the soil and raft and then combining these two matrices together to solve for the contact pressure. The behaviour of the raft is analysed by the use of thin plate theory while the response of the soil under action of the contact stresses is determined by a Fourier transform technique which reduces the three-dimensional problem to a two-dimensional one. This method may be used to obtain the distribution of contact pressure, differential settlement and bending and twisting moments in rafts of various shapes and stiffnesses.

Abstract
Variance-covariance estimates for satellite networks are estimated for single-session solutions and also for the signal components of the control points. The combination of these two estimates form an efficient expression for a simultaneous network adjustment for satellite networks. The second part of this work deals with the statistical analysis of the residuaIs using analysis of variance test and hypothesis test in least-squares collocation. The analysis of variance test depends on the discrepancies within each session of independently observed baselines and the discrepancies between the session of independently observed baselines. An independent test of the multi-session network adjustment was applied using least squares collocation. Here, the applications of the hypothesis tests to the estimated parameters, signals and covariance functions are discussed. A numerical example, using the Bass Strait survey is demonstrated.

Abstract
A new plastic mechanism model is developed in this paper for T-­joints in rectangular Hollow Sections (RHS) under concentrated force. This model includes the plastic hinges in the web, the membrane force in the flange and the strain hardening of the material. It can predict the yield load, the post-yield response and the ultimate load of a T-joint under concentrated force.

The model developed in this paper is compared with the CIDECT model, the Kato model and the modified Kato model. The yield load, the post-yield response and the ultimate load determined from the present model are compared with 6 T-joint tests performed by Zhao and Hancock (1991) and 20 T-Joint tests performed by Kato and Nishiyama (1979). The ratio (β) of the T-Joints tested varied from 0.291 to 0.890.

Abstract
When the ratio of the minor axis flexural stiffness to the major axis flexural stiffness is not small, classic analysis may lead to inaccurate predictions of the lateral buckling loads of beams and beam-columns because the effects of prebuckling deformations are not considered. The energy equation for the elastic lateral buckling of monosymmetric beam-columns including the effects of prebuckling deformations is derived, and the buckling differential equilibrium equations are obtained.

A finite element algorithm for the prediction of the lateral buckling loads of monosymmetric beam-columns including the effects of prebuckling deformations is developed from the energy equation. This includes the effects of second order moments due to the pre- buckling displacements and the axial loads. An iterative procedure for determining the lateral buckling loads is recommended.

The finite element results given in a companion paper show that the classic predictions of the lateral buckling loads of beam and beam-columns are generally conservative, but that the predictions by the linearized procedure are overestimated. The predictions by the recommended non-linear iteration procedure agree well with experimental results.

Abstract
In a companion paper it was pointed out that when the ratio of the out-of-plane flexural stiffness to the in-plane flexural stiffness is not small, classical analysis may lead to inaccurate predictions of the out-of-plane lateral buckling loads of beams and beam-columns, since the effects of the in-plane prebuckling deformations are not considered. This paper investigates numerically the effects of the in-­plane prebuckling deformations on the beam-columns by using the finite element lateral buckling procedure developed in the companion paper.

The numerical results show that the classical predictions of the lateral buckling loads of beams and beam-columns are generally conservative, but the predictions obtained by a linearized procedure which ignores second order terms are overestimated. The predictions by the recommended nonlinear iteration procedure agree well with the experimental results.

The effects of span length and the height of the load point have also been investigated. It is shown that lateral buckling is possible for sections where the out-of-plane flexural stiffness is equal to or greater than the in-plane flexural stiffness.

Abstract
This paper presents a finite element model for analysing the elastic ­plastic nonuniform torsion behaviour of thin-walled steel I-beams. The model uses Vlasovs warping strain model to represent warping torsion, and a mitre model for uniform torsion. The mitre model allows the representation at the tips of flanges of the transverse uniform torsion shear stresses.

For elastic-plastic analysis, the incremental theory of plasticity is adopted, employing the Prandtl-Reuss flow rule and the von Mises yield criterion. The material state determination algorithm and the variable arc-length method are also employed to improve the accuracy and reliability of the solutions. The numerical integration over the cross section is carried out by dividing it into a number of triangular areas and then applying the no-bias area-coordinate Gaussian numerical integration scheme.

The numerical examples provided show that good agreement is obtained with predictions of other models of inelastic non-uniform torsion and the experimental results by using the present finite element mode. It is also concluded that the present model can predict more realistic results at higher rotations than other models which ignore the transverse uniform torsion shear stresses.