Search results

Parameter identification is an important issue in structural health monitoring and damage identification for concrete dams. The purpose of this paper is to introduce a novel adaptive fireworks algorithm (AFWA) into inverse analysis of parameter identification.

Swarm intelligence algorithms and finite element analysis are integrated to identify parameters of hydraulic structures. Three swarm intelligence algorithms: AFWA, standard particle swarm optimization (SPSO) and artificial bee colony algorithm (ABC) are adopted to make a comparative study. These algorithms are introduced briefly and then tested by four standard benchmark functions. Inverse analysis methods based on AFWA, SPSO and ABC are adopted to identify Young’s modulus of a concrete gravity dam and a concrete arch dam.

Numerical results show that swarm intelligence algorithms are powerful tools for parameter identification of concrete structures. The proposed AFWA-based inverse analysis algorithm for concrete dams is promising in terms of accuracy and efficiency.

Fireworks algorithm is applied for inverse analysis of hydraulic structures for the first time, and the problem of parameter selection in AFWA is studied.

For gravity dams built on foundations with directional joint sets, the seepage in the foundation possesses anisotropic characteristics and may have adverse effects on the foundation stability. A methodology for system reliability analysis of gravity dam foundations considering anisotropic seepage and multiple sliding surfaces is proposed in this paper.

Anisotropic seepages in dam foundations are simulated using finite element method (FEM) with the equivalent continuum model (ECM), and their effect on dam foundation stability is involved by uplift pressures acting on the potential sliding surfaces. The system failure probability of the dam foundation is efficiently estimated using Monte Carlo method (MCM) combined with response surface method (RSM).

The case study shows that it is necessary to consider the possibly adverse effect of anisotropic seepage on foundation stability of gravity dams and the deterministic analysis of the foundation stability may be misleading. The system reliability analysis of the dam foundation is justified, as the uncertainties in shear strength parameters of the foundation rocks and joint sets as well as aperture, connectivity and spacing of the joint sets are quantified and the system effect of the multiple potential sliding surfaces on the foundation reliability is reasonably considered.

(1) A methodology is proposed for efficient system reliability analysis of foundation stability of gravity dams considering anisotropic seepage and multiple sliding surfaces (2) The influence of anisotropic seepage on the stability of gravity dam foundation  is revealed (3) The influence of estimation errors of RSMs on the system reliability assessment of dam foundation is investigated.

In a finite element model of typical dam—foundation—reservoir systems, the presence of heterogeneous material properties for the dam and the foundation produces a combined damping matrix that is non‐proportional to the mass and/or the stiffness matrices of the system. In this case, the undamped real free‐vibration modes cannot uncouple the damping forces such that the classical mode superposition method using real modes is not applicable. This paper presents comparative analyses of recent coordinate reduction procedures that have been developed to compute the response of linear systems with non‐proportional damping. The comparisons are based on the numerical efficiency and the accuracy of the displacement, acceleration and stress response, and on the distribution of the damping energy in the system.

This paper presents several methods for enhancing computational efficiency in both static and dynamic analysis of structural systems with localized non‐linear behaviour. A significant reduction of computational effort with respect to brute‐force non‐linear analysis is achieved in all cases at the insignificant (or no) loss of accuracy. The presented methodologies are easily incorporated into a standard computer program for linear analysis.

The response of the Pine Flat dam–water–foundation rock system is studied by a new described approach (i.e. FE-(FE-TE)-FE). The initial part of study is focused on the time harmonic analysis. In this part, it is possible to compare the transfer functions against corresponding responses obtained by the FE-(FE-HE)-FE approach (referred to as exact method which employs a rigorous fluid hyper-element). Subsequently, the transient analysis is carried out. In that part, it is only possible to compare the results for low and high normalized reservoir length cases. Therefore, the sensitivity of results is controlled due to normalized reservoir length values.

In the present study, dynamic analysis of a typical concrete gravity dam–water–foundation rock system is formulated by the FE-(FE-TE)-FE approach. In this technique, dam and foundation rock are discretized by plane solid finite elements while, water domain near-field region is discretized by plane fluid finite elements. Moreover, the H-W (i.e. Hagstrom–Warburton) high-order condition is imposed at the reservoir truncation boundary. This task is formulated by employing a truncation element at that boundary. It is emphasized that reservoir far-field is excluded from the discretized model.

High orders of H-W condition, such as O5-5 considered herein, generate highly accurate responses for both possible excitations under both types of full reflective and absorptive reservoir bottom conditions. It is such that transfer functions are hardly distinguishable from corresponding exact responses obtained through the FE-(FE-HE)-FE approach in time harmonic analyses. This is controlled for both low and high normalized reservoir length cases (L/H = 1 and 3). Moreover, it can be claimed that transient analysis leads practically to exact results (in numerical sense) when one is employing high order H-W truncation element. In other words, the results are not sensitive to reservoir normalized length under these circumstances.

Dynamic analysis of concrete gravity dam–water–foundation rock systems is formulated by a new method. The salient aspect of the technique is that it utilizes H-W high-order condition at the truncation boundary. The method is discussed for all types of excitation and reservoir bottom conditions.

The main purpose of this paper is to present and use the particle simulation method to explicitly simulate the spontaneous crack initiation phenomenon in brittle materials, and to compare the particle simulation results with experimental ones on the laboratory scale.

Using the particle simulation method, the brittle material is simulated as an assembly of particles so that the microscopic mechanism of inter‐ and intra‐particle crack initiation can be straightforwardly considered on the microscopic scale. A laboratory test has been conducted using a gypsum sample model to validate the particle simulation method for explicitly simulating the spontaneous crack initiation phenomenon.

The paper finds that in terms of simulating the macroscopic sliding surface along or around the contact plane between a block and its foundation, both the laboratory test and the particle simulation have produced consistent results. This indicated that the particle simulation method is capable of simulating macroscopic cracks through simulating conglomerations and accumulations of microscopic crack initiation within the brittle material. Compared with other numerical methods, the particle simulation method is more suitable for explicitly and effectively simulating spontaneous crack initiation problems on the microscopic scale in brittle materials.

The particle simulation method can be used to explicitly and effectively simulate the spontaneous crack initiation on the microscopic scale in brittle materials. It can be also used to simulate the macroscopic sliding surface along or around the contact plane between a block and its foundation. The experimental results of simulating the spontaneous crack initiation on the laboratory scale in brittle materials are very valuable for validating the numerical simulation results obtained not only from the particle simulation method, but also from other numerical simulation methods.

There is not a unified modelling approach to finite element failure analysis of concrete dams. Different behaviours of a dam predicted by different fracture methods with various material constitutive models may significantly influence on the dam safety evaluation. The purpose of this paper is to present a general comparative investigation to examine whether the nonlinear responses of concrete dams obtained from different fracture modelling approaches are comparable in terms of crack propagation and failure modes.

Three fracture modelling approaches, including the extended finite element method with a cohesive law (XFEM-COH), the crack band finite element method with a plastic-damage relation (FEPD), and the Drucker-Prager (DP) elasto-plastic model, are chosen to analyse damage and cracking behaviour of concrete gravity dams under overloading conditions. The failure process and loading capacity of a dam are compared.

The numerical results indicate that the three approaches are all applicable to predict loading capacity and safety factors of gravity dams. However, both XFEM-COH and FEPD give more reasonable crack propagation and failure modes in comparison with DP. Therefore, when cracking patterns are the major concern for safety evaluation of concrete dams, it is recommended that XFEM-COH and FEPD rather than DP be used.

The comparison of cracking behaviours of concrete dams obtained from different fracture modelling approaches is conducted. The applicability of the modelling approaches for failure analysis of concrete dams is discussed, and from the results presented in this work, it is significant to consider the suitability of the selected fracture modelling approach for dam safety evaluation.

The response of an idealized triangular concrete gravity dam is studied due to horizontal and vertical ground motions for both fully reflective and absorptive reservoir bottom conditions. For each combination, in this paper different orders of Givoli-Neta (G-N) high-order truncation condition are aimed to be evaluated from accuracy point of view by comparing the results against corresponding exact solutions which relies on utilizing a two-dimensional fluid hyper-element.

In present study, the dynamic analysis of concrete gravity dam-reservoir systems is formulated by Finite Element (FE)-(FE-TE) approach. In this technique, dam and reservoir are discretized by plane solid and fluid finite elements. Moreover, the G-N high-order condition imposed at the reservoir truncation boundary. This task is formulated by employing a truncation element at that boundary. It is emphasized that reservoir far-field is excluded from the discretized model.

It was observed that trend in gaining accuracy with increase in the order of G-N condition were basically the same for both horizontal and vertical ground motions under full reflective reservoir bottom condition. Moreover, convergence rate increases for absorptive reservoir bottom condition cases in comparison with fully reflective cases. It is also noticed that in certain cases, the responses are hardly distinguishable from corresponding exact responses. This reveals that proposed FE-(FE-TE) analysis technique based on G-N condition is quite successful, and one may fully rely on that for accurate and efficient analysis of concrete gravity dam-reservoir systems.

Dynamic analysis of concrete gravity dam-reservoir systems are formulated by a new method. The salient aspect of the technique is that it utilizes G-N high-order condition at the truncation boundary. This is achieved by developing a special truncation element which its generalized matrices are derived for Finite Element Method (FEM) programmers. The method is discussed for all types of excitation and reservoir bottom conditions. It must be emphasized that although time harmonic analysis is considered in the present study, the main part of formulation is explained in the context of time domain. Therefore, the approach can easily be extended for transient type of analysis.

Subsequently, the response of idealized Morrow Point arch dam is studied due to stream, vertical and cross-stream ground motions for reservoir bottom/sidewalls conditions of both fully reflective and absorptive. For each combination, different orders of Hagstrom–Warburton (HW) condition are evaluated from accuracy point of view by comparing them against exact solutions. It should be emphasized that normalized length of reservoir near-field region is taken as a very low value of L/H = 0.2 during this process which makes it a very challenging test for any kind of truncation boundary condition.

In present study, dynamic analysis of concrete arch dam-reservoir systems is formulated by FE-(FE-TE) approach [i.e. finite element-(finite element-truncation element)]. In this technique, dam and reservoir are discretized by solid and fluid finite elements. Moreover, the HW high-order condition imposed at the reservoir truncation boundary. This task is formulated by employing a truncation element at that boundary. It is emphasized that reservoir far-field is excluded from the discretized model. The formulation is initially explained in details.

The trend in gaining accuracy with increase in order of HW condition were basically the same for all three types of excitations under both full reflective and absorptive reservoir bottom/sidewalls conditions. The only exception was for cross-stream excitation response which was showing less accurate results near the first major peak for moderate orders of HW (e.g. O3-2) in comparison to what was observed for responses due to symmetric excitations (stream and vertical). This is mainly attributed to the selection of evanescent-type parameters of HW condition which is based on the first symmetric mode of reservoir. However, it is noted that error diminishes even for cross-stream excitation as order increases. High orders of HW condition, such as O5-5 considered herein, generate highly accurate responses for all three possible excitations under both types of full reflective and absorptive reservoir bottom/sidewalls conditions. It is such that responses are hardly distinguishable from corresponding exact responses. This reveals that proposed FE-(FE-TE) analysis technique based on HW condition is quite successful, and one may fully rely on that for accurate and efficient analysis of concrete arch dam-reservoir systems.

Dynamic analysis of concrete arch dam-reservoir system is formulated by new method. HW high-order condition is applied for a very low and challenging reservoir length. Different orders are evaluated against exact solution with excellent agreement. Generalized matrices of truncation element are derived for FEM programmers. The method is discussed for all types of excitation and reservoir base conditions.

The purpose of this paper is to propose a stable high-order absorbing boundary condition (ABC) based on new continued fraction for scalar wave propagation in 2D and 3D unbounded layers.

The ABC is obtained based on continued fraction (CF) expansion of the frequency-domain dynamic stiffness coefficient (DtN kernel) on the artificial boundary of a truncated infinite domain. The CF which has been used to the thin layer method in [69] will be applied to the DtN method to develop a time-domain high-order ABC for the transient scalar wave propagation in 2D. Furthermore, a new stable composite-CF is proposed in this study for 3D unbounded layers by nesting the above CF for 2D layer and another CF.

The ABS has been transformed from frequency to time domain by using the auxiliary variable technique. The high-order time-domain ABC can couple seamlessly with the finite element method. The instability of the ABC-FEM coupled system is discussed and cured.

This manuscript establishes a stable high-order time-domain ABC for the scalar wave equation in 2D and 3D unbounded layers, which is based on the new continued fraction. The high-order time-domain ABC can couple seamlessly with the finite element method. The instability of the coupled system is discussed and cured.