1992 Research Reports No R648-R668

Abstract
A theoretical method for predicting the post-buckling behaviour of thin-walled sections is applied to study the influence of the interaction between local and distortional buckling modes for test channel columns described in an earlier report. The theoretical predictions are based on a nonlinear elastic spline finite strip method for thin­walled sections which has recently been developed. The nonlinear elastic spline finite strip method takes account of geometric imperfections, residual stresses and non-simple boundary conditions at the ends of the section under study. The method can handle sections buckling simultaneously in local and distortional buckles at different half-wavelengths.

The investigation includes convergence studies convergence with different convergence criteria, strip subdivision and longitudinal subsections so as to determine reliable convergence criteria and the required discretisations for use in the nonlinear elastic spline finite strip method.

Abstract
This paper describes the results of a series of laboratory experiments carried out to study the behaviour of piles in expansive clay.

Measurements are presented for the progress of swelling with time, at various points in the clay after it is subjected to a controlled change in moisture content. Measurements of pile head movement and tensile force in the pile are then presented with particular attention being paid to their relationship with soil surface movement.

The influence of a number of variables is then examined including:

• compressive axial load on the pile,
• an applied overburden pressure on the soil surface,
• the presence of a stable sand layer below the expansive clay layer.

Finally, comparisons are made between the observed behaviour and that predicted from a theoretical boundary element analysis. The general trends of behaviour are adequately reproduced by the theory, but the detailed variations of pile movement and pile force with time are not always predicted accurately by the theory employed. Suggestions are made for refinements to the theory to reproduce more closely the real behaviour.

Abstract
A two year study into the effects of coal mining induced subsidence on cliff lines was undertaken by the New South Wales Department of Mineral Resources with the support of a grant under the National Energy Research, Development and Demonstration Program. A new monitoring technique, using an Electronic Distance Meter and acrylic reflectors attached to the cliffs, provided safety for the surveyors and extensive and accurate three dimensional displacement data on the movement of the cliffs as they were being undermined.

The final report of the study included many graphs showing the measured vertical and horizontal displacements, differential subsidence, tilts and strains and included analysis of these movements with various parameters, such as seam thickness, depth of cover and distances between the reflectors, longwall face and the edge of the chain pillar. Some of these graphs are presented in this paper and could be used, with caution, to predict subsidence movements around other cliff lines, after allowing for differences in geology and mining layout.

A mechanism is suggested as an explanation for the observed high horizontal movements around the cliff lines. Sixteen percent of the overall lengths of cliffs that were undermined during the monitoring study fell and the contributions of various factors that appeared to influence the cliff falls are discussed. It is concluded that no single factor dominated as the major cause of the cliff falls and, often. many factors combined to cause cliff face instability.

Abstract
The size and position of intermediate stiffeners in the compression flanges of thin-walled profiled steel decks exerts a strong influence on the dominant buckling mode of the flange. The ability of the deck to provide both high load carrying capacity before the onset of elastic buckling and a high ultimate load capacity may therefore be affected.

A programme of tests to determine the effectiveness of intermediate stiffeners in controlling buckling modes has been undertaken. A series of specimens were loaded in pure bending resulting in various buckling waveforms prior to ultimate failure through a plastic collapse mechanism. All specimens were brake-pressed to designs that were chosen to ensure that the primary buckling took place in the elastic range and was not affected by plasticity.

The experimentally determined buckling stresses were found to be comparable with studies performed using a computer analysis based on the finite strip method in which both local and distortional buckling modes were predicted. The test rig developed for the programme successfully approximated the boundary conditions implicit in the numerical model and prevented premature failure due to localised stressing at the specimen ends.

A simplified design procedure for distortional buckling is proposed. The existing design procedure for local buckling in the Australian Standard AS 1538-1988 is confirmed.

Abstract
A series of wind tunnel model tests were performed to investigate the torsional response of a tall square building and interference effects from neighbouring buildings with different sizes and shapes. The tall square building model, called the principal model, was tested by using an aeroelastic test rig designed for pure torsion vibration. Four types of interfering models were used to provide interference by individually locating each of them at positions upstream and downstream of the principal model. The aeroelastic test results of the isolated building model were compared with those obtained by pressure measurements and multi-degree aeroelastic tests, and used as a reference for estimating interference effects. It was found that significant torsional response occurred when the principal building was under interference. The dynamic torsional response increased by up to 2.2 times that of the isolated building. However, in some positions, the proximity effect provided by the interfering models reduced the torsional response. The critical relative locations of the interfering models were identified and possible torsional excitation mechanisms were discussed.

Abstract
This paper critically examines the use of the modified Cam clay stress-strain model in predicting the thermo-mechanical behaviour of soft clays. The equations governing the thermo-mechanical behaviour of a saturated soil are summarised and their methods of solution are briefly discussed. The observed thermo-mechanical soil behaviour reported in the literature has been compared with the predictions made using the modified Cam clay model. In making these comparisons, two extensions of the well known modified Cam clay model have been considered: one proposed by Britto et al in which the heating induces thermal stresses and strains in the soil but has no direct effect on the work hardening, and the other proposed by Hueckel and Borsetto in which a change in temperature also affects the yield surface. The comparisons are confined to the behaviour of normally and lightly overconsolidated clays, where the modified Cam clay is known to perform well. Apart from the effect of a single heating-cooling loop, cyclic behaviour is not considered. It is concluded that both models provide reasonable predictions under isotropic stress conditions. Although exhaustive comparisons have not been made for deviatoric stress excursions (because of the lack of experimental data), it appears from preliminary studies that neither model performs particularly well for this form of loading.

Abstract
These tests were part Of an On-going study aimed at extending knowledge on the performance of liquid-type auxiliary systems which could be used for possible reduction of vibrations of buildings or other large structures which may be subject to excitation by wind or other causes.

Earlier laboratory studies on various one-way systems using a Liquid Column Vibration Absorber’ or LCVA’ have been found very effective. That has led to the present study of a two-way LCVA, which contains a body of liquid whose particles have the capability of moving to absorb vibrations of a main body when it is excited in two different directions.

Tests have been made with the model two-way system with a number of variations to the LCVA and for various combinations of exciting forces parallel to two axes.

Abstract
In this paper a boundary element formulation is developed for anisotropic rock masses. The modelling of jointed rock masses, the boundary element treatment in which the fundamental solutions of Lekhnitskii have been incorporated, and the numerical evaluation of integrals with singularities are discussed. The good agreement found between the numerical and analytical solutions for several example problems demonstrates the capability, accuracy and efficiency of the present formulation. The problem of a deep circular tunnel excavated in a variety of jointed rock masses has also been analysed using the present formulation. The effect of the jointing on the behaviour of the rock mass around the tunnel is evaluated.

Abstract
The tests described in this report are the second in a programme carried out to determine the effectiveness of intermediate stiffeners in controlling buckling modes of profiled steel decks. A series of specimens with ‘flat-hat’ intermediate stiffeners were brake-pressed to designs that were chosen to ensure that the primary buckling mode took place in the elastic range and was not affected by plasticity. The specimens were loaded in pure bending, resulting in various buckling waveforms prior to ultimate failure through a plastic collapse mechanism. Significant interaction between local and distortional buckling modes was observed in the tests.

The experimentally determined buckling stresses are compared with studies performed using a computer analysis based on the finite strip method in which local and distortional buckling modes were predicted. A test rig was developed to approximate the boundary conditions implicit in the numerical model and prevent premature failure due to localised stressing at the specimen ends.

The existing design procedure for local buckling in Australian Standard AS1538-1988 and various methods involving distortional buckling are compared with the test results. A method of design accounting for the interaction of buckling modes and its effect on the ultimate moment capacity is proposed and compared with the test results.

Abstract
The report describes a test programme performed at the University of Sydney on small-scale (15.2 m span and 4.62 m high) but complete single stressed-arch frames subjected to simulated gravity (vertical) loading. The analytical techniques employed in, and the results of, finite element nonlinear analyses of the frame behaviour are also reported.

The distinguishing feature of stressed-arch frames compared to conventional steel truss frames is the way the arches are erected using a post-tensioning process rather than through the traditional techniques involving the use of cranes and scaffolding. As a result of the post-tensioning erection process, the top chord becomes curved and often strained into the plastic range.

The purpose of the tests on small-scale frames and the subsequent theoretical analyses was to determine the behaviour and strength of the highly curved and yielded top chord in complete frames, as opposed to the isolated subassemblages of stressed-arch frames investigated in previous studies. The finite element nonlinear analyses employed the plastic zone approach to model material inelasticity, and the updated Lagrangian formulation to model the gross geometrical changes associated with the erection procedure and the geometric nonlinearity under the subsequent vertical loading.

Abstract
Thin-walled lip-stiffened channel Columns composed of high strength steel may fail in a distortional mode involving movement of the lip stiffener perpendicular to the flange plate it supports. In this report, test results for a simple lipped channel section (CH1) and an intermediate stiffened channel section (CH2) of thickness of approximately 1.1 mm (0.043 in) and 550 MPa (79.8 ksi) steel and undergoing distortional or mixed local-distortional buckling failure are compared with different design methods. The design methods compared are Australian Standard AS1538, EC3 Part l/Annexe A, the ECCS Recommendations and the AISI Specification. In addition, new methods based on a modified Winter formula for effective width are presented and compared with the tests.

Abstract
This paper reviews available procedures for estimating the settlement of pile foundations, and classifies these procedures into three categories: empirical, simplified hand calculation methods based on sound theory, and methods employing computer analyses.

The geotechnical parameters required for the various approaches, and methods of estimating these parameters, are reviewed.

For single piles and pile groups, hand and computer methods are compared for three cases, and conclusions reached regarding the level of analysis necessary to obtain an adequate settlement prediction.

Abstract
A series of direct shear tests have been performed to investigate the effects of different interface roughnesses for a dry silica sand. The implications of these results for the skin friction developed on pile shafts are considered. For each surface roughness tests have been performed for three normal stresses and two different soil densities. The interface results are compared with the responses from conventional shear tests on the sand. It has been found that both the maximum stress ratio and the volume changes are significantly affected by the interface roughness, and also by the boundary conditions imposed by the equipment.

Abstract
This paper presents some previously unpublished test data on the engineering properties of Ashfield shale and investigates the relationships between them. Also included is an examination of the important influence of moisture content and the degree of weathering on the mechanical properties, particularly the strength. ‘ne present findings are compared with the results of previous studies on the mechanical properties of these shales.

Abstract
This paper outlines a procedure for seismic microzonation for environmental planning and management, in the broadest sense, to assist in the planning, construction and management of projects. The procedure involves the following steps:

1. evaluation of the regional geology and typical subsurface soil profiles in the area or region being studied;
2. estimation of the dynamic stiffness and damping characteristics of the subsurface soils;
3. estimation of design histories of bedrock acceleration versus time;carrying out seismic response analyses of typical soil profiles subjected to these earthquake histories;
4. assessment of site liquefaction potential;
5. assessment of seismic effects on stability of slopes and retaining structures;
6. calculation of structural response spectra and assessment of potential for structural damage;
7. the appropriate portrayal of the results of the above assessments, e.g. via maps identifying areas in which there may be a high risk of liquefaction, or of high seismic forces induced by an earthquake on existing or proposed structures.

Each of these steps is discussed in detail, and applications of this approach are illustrated with respect to both a discrete site and to a large regional area.

Abstract

Model tests on pairs of piles jacked into reconstituted calcareous sand have been performed to study the interaction behaviour of these piles during jacking, static and cyclic loading. The tests have been performed on medium-dense and dense samples of calcareous sand which have been consolidated under two different overburden pressures. Results of the tests are presented for the following aspects of behaviour: a) influence of a newly-installed pile on the force and displacement within an already-installed instrumented pile and b) the displacements and loads caused in the instrumented pile by static and cyclic loading of an adjacent pile.

The tests reveal that the interaction between the piles is dependent on the initial density of the soil and the spacing between the piles. During jacking, the residual loads developed in the adjacent pile increase as the sand density increases, while the direction of movement can change from downward to upward, depending on the soil density.

Abstract
This report describes an experimental investigation in a wind tunnel of torsional excitation of a tall square building with structural eccentricity and torsional interference effects on the eccentric building response from neighbouring buildings of different sizes. The eccentric building model was an aeroelastic model of pure torsional motion and the interfering building models were rigid. Tests were carried out in 1:400 scale wind models of natural wind flow over open country and city centre terrains developed in the wind tunnel. The probability distribution of peak response, aerodynamic damping and torsional excitation spectrum were examined to determine torsional excitation and response characteristics. It was found that the tall square building with a 10% geometric eccentricity ratio was more susceptible to torsional motion and vibration, compared with one without eccentricity. There was a significant increase of maximum mean responses, and an unstable torsional vibration due to vortex shedding was also observed in which large amplitude twist angular displacement occurred. With interference effects from a small square interfering building, the torsional response of the eccentric building was enhanced by a factor of up to 4.6 in the open country terrain due to resonant buffeting. The critical location of the interfering building was found to be dependent on the size of the interfering building and the eccentricity position of the eccentric principal building relative to wind incidence.

Abstract
In this report the role of modern geotechnical analysis in foundation design is reviewed, and analysis methods are classified into three categories, depending on their level of sophistication. The benefits of using a soundly-based theory to better understand foundation behaviour are emphasised, with particular reference to the case of a laterally loaded pile.

Two case histories are then studied in detail, in order to examine the sensitivity of performance predictions to the method of analysis used, the soil and pile parameters selected, and the idealisation of the soil profile adopted. It is shown that the choice of the method of analysis is less significant than the selection of the strength and deformation parameters of the soil and their distribution with depth.

Abstract
Stressed-arch frames are characterised by a post-tensioning procedure which is used to erect the frames from an assembled configuration at ground level to a final erected shape. During the erection process, the top chord becomes curved and is usually stressed into the inelastic range. Consequently, the strength of the erected frame under service loading cannot be assessed rationally using conventional elastic or plastic methods for the analysis and design of steel structures.

The purpose of this report is to examine in detail the provisions of specifications and standards worldwide for the design of the top chord of stressed-arch frames, and to propose and validate a simple design procedure. In particular, the irrationality of design based on conventional elastic and plastic methods, and, alternatively, the rationality of design based on advanced analysis as described in the Australian Standard for Steel Structures, AS 4100-1990, is identified. Aside from the Australian Standard, the other standards and specifications considered in the report are the American Institute of Steel Construction and American Iron and Steel Institute Load and Resistance Factor Design Specifications, the Canadian Standard for Steel Structures and the proposed Eurocode 3.