Search

Accurate modelling of the behaviour of crack propagation has important implications in the design and risks assessment of glazing facade when subject to extreme load conditions. This paper presents an investigation into the propagation of cracks in annealed glass. The investigation involved physical experimentation using a high-speed camera to capture the propagation of cracks in glass panels in response to static and impact loading.The physical experimentation revealed that cracks initiated from impact loading did not propagate continuously as in static conditions. Instead, the crack could stop propagating and then restart following a short pause. This phenomenon was investigated numerically through simulation of crack propagation by incorporating fracture mechanics theory into the finite element model. The finite element analyses were based on a discrete crack model which employed an advance re-meshing technique. The re-meshing modelling technique involved the following steps: (i) calculation of stress intensities surrounding the tip of the crack; (ii) identification of the location of the propagating "crack tip"; (iii) re-meshing of the finite element model. Results from the analyses are presented and interpreted in this paper. © 2011 Taylor & Francis Group, London.

Point fixed glass facade system (PFGFS), also known as the spider glass system, is gaining popularity as it is the most elegant facade option among architects particularly compared to the conventional framed glass systems. The PFGFS system is fixed to the support structure at a few points only using bolts and metal clamps. PFGFS are used mostly at the ground floor level of buildings in Australia. Generally, the racking performance of these systems is not considered at the design stage. If the system does not have sufficient in-plane drift capacity it will be vulnerable to racking actions mainly during earthquakes. In this paper, previous research findings on PFGFS and connections are reviewed. An expression is developed to estimate the in-plane drift capacity of PFGFS due to articulation from rigid body movement and rotation. The equation is then benchmarked against analytical modelling using ANSYS finite element software. © 2011 Taylor & Francis Group, London.

A parallel iterative matrix solver is proposed. The solver takes advantage of the structure of matrices arising from nonlinear steady state analysis problems to partition the workload. Experimental results show that the efficiency of the proposed method increases with the size of the circuit.

This paper presents a new technique for the analysis of interconnects characterized by frequency dependent per-unit-length (FD-PUL) parameters. The proposed technique is based on using the idea of integrated congruence transform (ICT) to construct a reduced model in the form of ordinary differential equations (ODEs). The passivity of the reduced model is guaranteed. Numerical results demonstrate the validity of the proposed technique.

In the recent past, method of characteristics (MoC) has emerged as an efficient macromodeling and simulation technique for long high-speed transmission lines with low loss. However, one of the major limitations of this technique is that, the resulting interconnect macromodels are not guaranteed to be passive, which can lead to non-physical unstable transient responses during simulation. In order to address this problem, we present a new passivity compensation algorithm for non-passive MoC-based multiconductor transmission line (MTL) macromodels. The algorithm is based on the first-order perturbation of the non-passive macromodel and can handle single as well as coupled interconnects. A sample numerical example is presented to validate the efficiency of the algorithm.

This paper presents a new approach to develop a delay-based macromodel for nonuniform transmission line algorithmically. The proposed technique is based on employing results from the Lie group theory to represent the matrix of a nonuniform transmission line with arbitrarily shaped per-unit-length parameter matrices using closed-form expressions. The key advantages of these expressions is that they lend themselves naturally to delay representation.

This paper presents an outline of a new integration method and describes the results of its application in simulating electric circuits in the time-domain. The proposed method does not suffer from the stability vs. order limitation of classical linear multistep methods. Hence, it enables using arbitrarily high- order approximations to the circuit waveforms while maintaining the stability over the entire left side of the complex-plane (A-stability).

This paper describes a new algorithm to obtain reduced-order models for large networks with delay elements. The proposed algorithm can be used in situations where delay extraction based modeling approaches have been used to model portions of interconnects with low losses, while other portions are modeled with large networks of lumped components. It is shown that the reduced-order model is passive by construction.

This paper presents a new algorithm for verifying passivity in macromodels of lossy multiconductor interconnects. The macromodels considered in this paper are constructed from the method of characteristics and include delay element. The new algorithm is derived from the results of a new theory that describes the necessary and sufficient conditions of positive-realness in dynamical system described by delay differential equations.

This paper describes an algorithm for computing time-domain sensitivities via the adjoint analysis principal. The key feature in the proposed algorithm stems from enabling the application of the adjoint sensitivity in computing sensitivities of large linear systems represented through model-order reduction techniques.

This work presents a new technique for computing large-change sensitivity (LCS) of steady-state operating point in nonlinear circuits. The basic idea underlying the algorithm is the construction of a reduced system of nonlinear equations that preserves the derivatives of steady-state response with respect to the desired network parameters. Large change variations are then obtained by solving the reduced systems instead of the original one.

Piecewise-linear (PWL) neural networks are networks with piecewise-linear node functions. They have attractive features, such as speed of training and amenability to digital VISI implementation. The paper presents an algorithm for training PWL networks. The algorithm is general in that it can be used for the optimization of general PWL functions. It is based on moving from one linear region to the next. This is achieved by exploring 2N specific directions along the boundaries between the linear regions (N is the dimension), and moving along the direction that achieves a descent in the objective function. Convergence to the local minimum is proved, and simulation results confirm the computational efficiency of the algorithm.

Summary form only given. Two approaches are presented for transient analysis of multiport distributed interconnects with nonlinear terminations. The first approach is based on macromodeling of linear subnetworks. The second approach is based on direct time-domain reduction of entire nonlinear circuit equations.

Computing the steady state response of large nonlinear circuits is becoming a key simulation requirement due to the rapid market growth of RF silicon ICs. In this paper we describe a nonlinear circuit reduction algorithm for finding the steady state response. The proposed algorithm uses a congruent transformation-based technique to reduce the harmonic balance equations into a much smaller set of equations. The main feature of the reduced circuit is that it shares with the original one a certain number of the derivatives w.r.t. the RF input power. Steady state analysis is then done on the reduced circuit instead of the original circuit.

Certain types of damage generated by the impact of a solid object are correlated with the amount of force that is localized around the area of contact between the impactor and the surface of the target. The stiffer the impactor, the higher the contact force when the amount of impulse delivered by the impact is held constant. Thus, realistic simulations of the contact force by finite element analysis (FEA) require representative, and detailed, information of the dynamic compressive properties of both the impactor and the target. In situations where relevant properties of the impactor material have not been documented, the estimation of contact force is filled with uncertainties. In addressing this challenge an innovative experimental–simulation calibration procedure involving the use of a custom made (inexpensive) tubular device is presented in this paper for measuring the dynamic compressive stiffness properties of impactor objects. Given the calibrated values the amount of contact force generated by the same impactor material in projected scenarios could be simulated with good accuracies for predicting damage.

Piecewise-linear (PWL) neural networks are widely known for their amenability to digital implementation. This paper presents a new algorithm for learning in PWL networks consisting of a single hidden layer. The approach adopted is based upon constructing a continuous PWL error function and developing an efficient algorithm to minimize it. The algorithm consists of two basic stages in searching the weight space. The first stage of the optimization algorithm is used to locate a point in the weight space representing the intersection of N linearly independent hyperplanes, with N being the number of weights in the network. The second stage is then called to use this point as a starting point in order to continue searching by moving along the single-dimension boundaries between the different linear regions of the error function, hopping from one point (representing the intersection of N hyperplanes) to another. The proposed algorithm exhibits significantly accelerated convergence, as compared to standard algorithms such as back-propagation and improved versions of it, such as the conjugate gradient algorithm. In addition, it has the distinct advantage that there are no parameters to adjust, and therefore there is no time-consuming parameters tuning step. The new algorithm is expected to find applications in function approximation, time series prediction and binary classification problems.

This paper is concerned with a simplified approach of estimating the effects of the impact of a projectile on a circular dome. The procedure to be introduced involves simplifying the impactor (projectile) and the target (the dome) by a two-degree-of-freedom (2DOF) system which is made up of two lumped masses connected by elastic springs. This modeling approach has only been adapted for analyzing the impact response behavior of beams and plates. The original contributions of this paper is the development of equations and charts for estimating the value of the lumped mass and spring stiffness in the 2DOF lumped mass model to emulate the response behavior of circular domes. Linear elastic behavior of the dome is assumed but nonlinear behavior of the impactor has been taken into account. The developed calculation procedure has been validated and illustrated by case studies.

Hollow sections have become increasingly popular in structural applications due to their structural efficiency. However, their use in low rise structures is hampered by the limited means of providing fully bolted site connections. This limitation can be overcome by using blind bolting system which requires access to one side only. This eliminates both the problem of lack of access and also the need to weld various accessories to the hollow sections. This paper is part of an ongoing research project which focuses on the development of a range of connections to hollow sections using blind bolts. The paper presents the results of an experimental program for blind bolted channel side plate connections to unfilled hollow section columns under static loading. The channel side plate connection was found to be much stiffer than a typical face connection and has the potential to achieve a rigid connection for braced frames according to the EC3 classifications. © 2011 Taylor & Francis Group, London.

This paper presents results from an experimental study investigating the seismic performance of lightly reinforced concrete columns. The drift capacity of concrete columns is made up of flexural, yield penetration, and shear components, in which those components are affected by aspect ratio, axial load ratio, longitudinal and lateral reinforcement ratio. Very low reinforcement is prohibited by design standards, but such columns are prevalent in many old buildings, particularly in developing countries. A collaborative research project has been undertaken to investigate the collapse behaviour of such columns. The paper presents a summary of the laboratory results and the implications for seismic performance of building structures with such low level detailing. © 2011 Taylor & Francis Group, London.

This paper presents a practical method for estimation of the roof and maximum inter-storey drift ratios which can be utilized in the approximate seismic assessment of existing building structures. The height-dependent building models employed can represent the existing buildings of different height which are laterally supported by walls, moment resisting frames or a combination of walls and frames. In the employed macro model, realistic estimations of the distribution of floor lumped masses and height-wise lateral stiffness of existing buildings were considered. Extensive dynamic analyses were conducted to investigate the generic modal displacement hypothesis and to model drift parameters of interest considering the higher mode effects. The suggested drift predictors has the capability to be used in conjunction with site specific spectral displacement for constructing the drift spectra which is a plot of the roof and maximum inter-storey drift ratios versus building first mode period or height. © 2011 Taylor & Francis Group, London.

Impact by debris is a major cause of damage to buildings in hailstorms, windstorms and rockfalls. The effects of the impact need to be quantified so that protective installations including building facades, roof coverings and other forms of built installations that are exposed to the hazard, can be designed or retrofitted to withstand projected hazards in a rational and economical manner. This research covers this type of hazards in order to address the lack of information on physical properties of common debris materials, lack of guidelines on how to obtain reliable predictions of the impact, and the poor understanding of the fundamentals. The new knowledge and methodology to be developed in this poorly informed and understood area of technology should instill a new perspective to every engineer in their daily practices. Ultimately, a better understanding and quantification of impact forces would enable the development of innovative protective systems.

This paper presents results obtained from a recent study that is aimed at assessing the drift demand on buildings for a range of projected earthquake scenarios in Australia. Parameters considered that may affect the response of buildings included building height, structural systems, and mass and stiffness distributions. It has been found that, for the range of buildings studied, the maximum angle of drift is 2.6-4.4 times the maximum response spectral displacement of the earthquake divided by the building height. This can be checked against the limiting drift capacity of the building to enable various levels of damage to be predicted for given earthquake scenarios.

Experimental field testing of a soft-storey building in Melbourne has been undertaken. The soft-storey open ground floor was a precast concrete frame with connections significantly weaker than the members they connected. Four unique tests were conducted with combination between load directions (strong and weak) and restraints of ground slab (with or without ground slab). Displacement measurements undertaken using a number of contemporary techniques are described together with results from ambient vibration testing of the building. The experimental results showed that soft-storey columns had significant displacement capacity controlled by the column width irrespective of strength degradation. An analytical model developed to predict the force-displacement relationship of the tested frame including the influences of connection strength at column ends, gravity rocking strength and ground slab restraint are briefly described and compared with the experimental results.

A new algorithm based on projection techniques is proposed for constructing parameterized reduced order systems of nonlinear equations. The resulting reduced system can be used to trace the variation in the steady-state solution in response to variations in any arbitrary circuit or physical parameter. Experimental results show up to two orders-of-magnitude reduction in the problem size required in conventional harmonic balance technique.

In areas of low to moderate seismicity, displacement controlled behaviour is a phenomenon in which the displacement demand on the structurewould not increase indefinitely with increasing structural period. Deficiencies that are inherent in the conventional force-based seismic design method could result in undesirable and illogical consequences for long period structures in particular. This paper is concerned with the adaptation of the displacement based seismic design methodology to cable stayed bridges which are characterised by high horizontal flexibility and low natural frequencies. The model that has been developed using program Opensees can be used for investigating the dynamic characteristics of structures with a cable-pylon system. © 2011 Taylor & Francis Group, London.