Journal of Engineering Mechanics

A number of simplified design methods have been developed to predict composite slab capacities in fire. Most of these extend ambient-temperature large-deflection slab behavior to the elevated-temperature phase by reducing the strengths of fire-exposed concrete and reinforcement while neglecting the effects of thermal expansion and thermal bowing of the slab. Experiments have shown that there are significant differences between the predictions from these methods and the actual behavior and failure modes of ambient- and elevated-temperature concrete slabs in tensile membrane action. Therefore, this paper describes the development of a new analytical method that incorporates both thermal and mechanical effects into the prediction of slab behavior in fire conditions. It uses the variational Rayleigh-Ritz approach to classical large-deflection plate theory. The method is found to produce accurate predictions of deflections and membrane tractions; however, it requires further refinement for accuracy of stresses. The results are compared with numerical modeling using VULCAN, a specialist finite-element (FE) program for structural fire engineering.

There is a huge interest in developing new performance metrics, based on monitoring information and considering uncertainties in the resistance and loading effects, to be used by stakeholders when determining maintenance and rehabilitation strategies for existing bridges. The cost of structural health monitoring is generally a criterion in the duration of the monitoring program, and determining when, and how long, monitoring can be interrupted is still a challenge. Providing an approach that quantifies the benefits of monitoring is crucial for stakeholders. The objective of this paper is threefold: build a performance indicator based on monitoring information that considers uncertainties and correlation in recorded data; use this indicator to check and predict if serviceability and safety thresholds are reached; and analyze the impact of short monitoring interruptions on the performance assessment accuracy by balancing cost of monitoring with the efficiency of the results. The originality of the proposed approach consists in the introduction of a performance indicator in the structural reliability analysis, based on monitoring programs. The regression methodology enables one to apply the reliability analysis when monitoring is interrupted during short-term periods and to assess the impact of such interruptions on the structural analysis accuracy. The proposed approach is applied on an existing bridge.

The paper proposes an energy-based damage feature to detect the presence of earthquake-induced structural damage in instrumented buildings. The proposed feature is a function of the ratio between the hysteretic dissipated energy and the total dissipated energy. The main contribution of the paper consists in deriving an expression to estimate the dissipated energy ratio from a measured vibration response at a limited number of degrees of freedom. The proposed methodology is successfully implemented in the context of simulated multistory shear buildings and in a real instrumented building during two independent seismic events; one where the building did not sustain damage and the other were significant damage occurred. In both cases the proposed methodology performed the correct diagnosis.

A crack-propagation criterion is proposed in this paper for mixed-mode I–II fracture in concrete. In this criterion, crack propagation is initiated when the difference between the stress-intensity factor at the crack tip caused by the external force and that by the cohesive stress satisfies the crack-initiation equation. On the basis of this criterion, a numerical method is developed to simulate mixed-mode I–II crack propagation in concrete. To verify the criterion, three sets of experimental data are selected: one is obtained from the self-conducted test and the other two are collected from the literature. The numerical results show that the calculated load versus crack-mouth opening and sliding displacement curves and crack trajectories are in good agreement with experimental results. Therefore, once the mode I initial cracking toughness, elastic modulus, fracture energy, possion’s ratio, and tensile strength of concrete are available, the whole mixed-mode I–II fracture process in concrete can be predicted with reasonable accuracy.

The topic of the present paper is a damage detection technique based on the theory of proper orthogonal decomposition (POD). Numerical examples and experimental verification are conducted on an aluminum plate. The numerical examples provide useful information on the effect of several parameters, such as damage severity, type of excitation, noise level, and grid size of the sensors and on the reliability of the technique. The experimental verification using the POD method on the acceleration data show that it is possible to locate some defects in a cantilever plate; however, the analysis of the experimental data highlights the sensitivity of the method to the modification of the boundary conditions. Finally, the application of this technique to the detection of delamination in composite materials is conducted numerically on a cantilever beam. A three-dimensional model was used to investigate how the position of the damage along the length and thickness of the beam is detected and to demonstrate the sensitivity of the method to the location of the sensors with respect to the position of the damage.

This paper presents a numerical procedure based on smoothed particle hydrodynamics (SPH) to analyze the failure process of a rock medium by predicting the initiation of microcracks. The subsequent propagation of cracks has also been analyzed without any special treatment or assumption regarding fracturing process. The procedure for implementing softening elastoplastic model has been discussed in the SPH framework. For failure of a rock specimen under uniaxial compression, the Drucker-Prager yield criterion is used in the elastic-plastic constitutive model by considering associative and nonassociative plastic flow rules. The Rankine maximum tensile failure criterion is implemented to model the tensile failure of a circular rock specimen. The results obtained from this study have been compared with laboratory tests and existing analytical solutions. It is found that the developed procedure has the potential to provide useful information to understand the key physical phenomena that occur in the failure process.

The objective of this paper is to provide the theoretical background and illustrate the capabilities of the constrained finite strip method (cFSM) for thin-walled members with general end boundary conditions. Based on the conventional finite strip method (FSM), cFSM provides a mechanical methodology to separate the deformations of a thin-walled member into those consistent with global, distortional, local, and other (e.g., shear and transverse extension) modes. For elastic buckling analysis, this enables isolation of any given mode (modal decomposition) or quantitative measures of the interactions within a given general eigenmode (modal identification). Existing cFSM is only applicable to simply supported end boundary conditions. In this paper, FSM is first extended to general end boundary conditions, including simply–simply, clamped–clamped, simply–clamped, clamped–guided, and clamped–free. Next, with the conventional FSM for general end boundary conditions in place, the derivation of the constraint matrices for global, distortional, local, and other modes that play a central role in cFSM are summarized. Several bases (i.e., the constraint matrices) are presented for general end boundary conditions involving, in particular, different orthogonalization conditions. For modal identification, normalization schemes for the base vectors as well as the summation method employed for the modal participation calculation are also provided. Numerical examples of modal decomposition and identification are illustrated for a thin-walled member with general end boundary conditions. Recommendations on the choice of basis, orthogonalization, and normalization are provided.

A two-stage viscoelastic-viscoplastic damage constitutive model and the corresponding parameters calibration method were briefly given. Then, the numerical algorithms of the proposed constitutive model were implemented based on a radial return mapping algorithm and realized using the user material subroutine in the well-known finite-element code. A rutting prediction method of asphalt pavement was proposed based on the two-stage viscoelastic-viscoplastic damage constitutive model along with a loading equivalence method. Laboratory wheel tracking tests under different temperatures, different loading levels, once variable amplitude loads, and step variable amplitude loads were conducted to evaluate the efficiency of rutting prediction. Finally, the test results were compared with finite-element simulated results. All the results indicate that the constitutive model and the corresponding rutting prediction method proposed in this paper can effectively represent the rutting developing rules under standard loading and complicated variable amplitude loading with reliable prediction accuracy. Meanwhile, the mechanism of rutting development can be characterized through analyzing the evolution rules of internal state variables based on the proposed constitutive model.

An experimental program consisting of triaxial compression tests and torsion shear tests was designed and carried out to study the variation of shear strength in cross-anisotropic deposits of fine Nevada sand under two- and three-dimensional (3D) conditions using triaxial compression and a recently developed torsion shear apparatus. Triaxial compression tests were performed on vertical and horizontal specimens with cross-anisotropic fabric to obtain the shear strength and its variation with mean normal stress and with testing direction. The results formed the basis for characterizing the variation of curvature in meridian planes with mean normal stress. Furthermore, 44 drained torsion shear tests were performed at constant mean confining stress ${\sigma }_m$; constant intermediate principal stress ratios, as indicated by $b=\left({\sigma }_2-{\sigma }_3\right)/\left({\sigma }_1-{\sigma }_3\right)$; and principal stress directions $\alpha$. The experiments were performed on large hollow cylinder specimens deposited by dry pluviation and tested in an automated torsion shear apparatus. The specimens had a height of 40 cm, an average diameter of 20 cm, and a wall thickness of 2 cm. The 3D failure surface of the fine Nevada sand is presented with discrete combinations of principal stress direction $\alpha$ and the intermediate principal stress, and the effects of these two variables on the shape of the failure surface are presented.

In many cases, an arch may have in-plane elastic end restraints provided by the connected structures or elastic foundations, which may influence the elastic lateral-torsional buckling of the arch. However, little research of the lateral-torsional buckling of arches with elastic end restraints has been reported in the open literature. This paper analytically investigates the lateral-torsional buckling of pin-ended circular arches having in-plane elastic rotational end restraints under a uniform radial load. The analytical solutions for the prebuckling behavior of such an arch show that the uniform radial load produces combined axial compressive and bending actions in the arch and that the axial compressive force produced by the uniform radial load is approximately uniform along the arch axis. It is found that the stiffness of elastic rotational end restraints has significant effects on the magnitude and distribution of the axial compressive forces and bending moments. The axial compressive force decreases with an increase of the stiffness of rotational end restraints. The analytical solution for the lateral-torsional buckling load of the arch is derived, which accounts for the effects of rotational end restraints. It is found that the effects of the stiffness of rotational end restraints on the lateral-torsional buckling load are profound. The buckling load increases with an increase of the stiffness of the end restraints. It is demonstrated by comparisons with the finite-element results that the analytical solution provides good predictions for the lateral-torsional buckling load of both shallow and deep arches having in-plane rotational end restraints.

The subject of objective rates in continuum damage mechanics is under consideration. It is shown that by changing the definition of the objective rate in constitutive theory, one obtains different material behavior and different final fracture modes as causes of intrinsic microdamage evolution. Theoretical considerations and a numerical example include the comparison of two special cases of objective rates for which the symmetric part of the spatial velocity gradient defines a true rate of the assumed strain rates, namely Lie and Logarithmic. The problem is described in the framework of the recently proposed Perzyna-type viscoplasticity, including an anisotropic damage description; nevertheless, final conclusions are general.

Permeability tests on limestone and sandstone during the course of deformation and failure are carried out with a servocontrolled testing machine. The permeability evolution of the two kinds of rocks with axial, lateral, and volumetric strains are investigated. The test results show that the permeability has an obvious correlation with the strain. The permeability mechanisms during the course of deformation and failure are discussed. According to the feature variations of permeability and strain of rock specimens, a conceptual model for the permeability evolution with the strain is proposed. The test results of sandstone are used to validate this model, and the relevant permeability-strain equation in the process of deformation and failure is fitted by using the mathematical method. The fitting precision is good. The proposed equation may provide another important control equation for the stress-seepage coupling analysis.

This study presents an interface fracture mechanics analysis of delamination failure in a multilayered beam subject to general mechanical loads and residual stresses. Analytical solutions of the energy release rate and its phase angle are obtained through a crack-tip-element analysis. Both the transverse shear and thermal residual stresses, as well as their coupling effect, are included in these solutions. Representative interface fracture specimens are analyzed using these new solutions as well as an existing one. Excellent agreement with finite-element analysis has been achieved by the present solutions, whereas the existing solution underestimates the total energy release rate significantly, because it neglects the transverse shear forces. This suggests that the present solutions can significantly enhance the accuracy of the total energy release rate and its phase angle for an interface crack in multilayered structures.

This paper addresses the mechanism of coupled deck-stay motions in cable-stayed bridges. Field observations of deck-stay interactions are summarized and an analytical model is developed to simulate such interactions. Using this model, a dynamic analysis of an example stay is conducted to illustrate the mechanism underlying the coupled motions. Three-dimensional finite-element models of a cable-stayed bridge are presented to confirm that coupling of deck-stay motions can occur in actual structures. A comparison between the results from the analytical and numerical models indicates that the proposed analytical model can reasonably explain the observed deck-stay interactions. In particular, it is shown that coupled motions are associated with the veering of natural frequency loci and mode localization and that the distributions of generalized masses between decks and cables are useful indexes for assessing the degree of coupling.

In geotechnical practice, it is important to quantify the response of laterally loaded piles in layered soil systems. In this paper, semianalytical solutions were developed to assess the static behavior of vertical piles embedded in multiple-layered soil systems and subjected to lateral loads. The behavior of the surrounding soil was idealized as an elastic-perfectly plastic model. The basic idea of the subgrade reaction theory was used to obtain the governing differential equations in terms of pile lateral deflection. Combining with continuity conditions at the adjacent pile segments, the response of the whole pile was expressed through the deflections and lateral forces at the pile head in a matrix form. The validation of the proposed solutions was verified by other theoretical solutions and measured data. The layering effect of subgrade soils on the response of the laterally loaded pile was also discussed.

The literature provides an analytical solution to the state of stress in an infinite plate with a circular notch and biaxial tensile stress. This solution requires that a plastic region fully surrounds the notch. In an existing paper, a minor reshaping mistake at the very end of the derivation occurred that is corrected herein. The purpose of this work is mainly to document literature and to provide an amendment to the analytical solution. Numerical simulations using finite elements prove the validity of the corrected equation.