Journal of Engineering Mechanics

A solution is developed for the evolution of buildup, steady, and postarrest dissipative pore-fluid pressure fields that develop around a finite-radius conical penetrometer advanced in a saturated linearly elastic porous medium. The analog with cone penetrometer testing is direct and is used to enable continuous distributions of permeability and diffusivity to be determined with depth. This analysis reveals the direct dependence of penetration rate on the induced fluid pressure field magnitudes, and predicts that a penetration rate threshold limit exists with respect to pore-pressure generation. This represents the essence of a partially drained system. The developed pore-pressure field is determined to be a function of the dissipation rate of the material, the penetration rate, and the storage effects of the advecting medium. Analysis of the pore-pressure field under start-up conditions reveals that the time required to reach steady state is strongly influenced by the penetration rate and the pressure-dissipation properties of the material. Analysis of the developed stable pressure fields illustrates the inversely proportional relationship that exists between penetration rate and pore-pressure magnitudes at the cone surface; representing the influence of storage in the medium on stable pore-pressure magnitudes. Stable pressure fields below the penetration threshold limit, $U_D⩽{10}^{-1}$ , form a spherical response around the cone tip transitioning to an elongated radial response for penetration rates above this limit. Postarrest analysis indicates that the prearrest penetration rate strongly influences the dissipation rate and pattern of dissipation. The developed analysis can be correlated with CPTu-recovered data to independently evaluate permeability magnitudes during steady penetration.

Most structures exhibit some degrees of nonlinearity such as hysteretic behavior especially under damage. It is necessary to develop applicable methods that can be used to characterize these nonlinear behaviors in structures. In this paper, one such method based on the empirical mode decomposition (EMD) technique is proposed for identifying and quantifying nonlinearity in damaged structures using incomplete measurement. The method expresses nonlinear restoring forces in semireduced-order models in which a modal coordinate approach is used for the linear part while a physical coordinate representation is retained for the nonlinear part. The method allows the identification of parameters from nonlinear models through linear least-squares. It has been shown that the intrinsic mode functions (IMFs) obtained from the EMD of a response measured from a nonlinear structure are numerically close to its nonlinear modal responses. Hence, these IMFs can be used as modal coordinates as well as provide estimates for responses at unmeasured locations if the mode shapes of the structure are known. Two procedures are developed for identifying nonlinear damage in the form of nonhysteresis and hysteresis in a structure. A numerical study on a seven-story shear-beam building model with cubic stiffness and hysteretic nonlinearity and an experimental study on a three-story building model with frictional magnetoreological dampers are performed to illustrate the proposed method. Results show that the method can quite accurately identify the presence as well as the severity of different types of nonlinearity in the structure.

The paper presents an analysis of the coupled vibration of beams with arbitrary thin-walled open cross section, braced with identical transversal header beams uniformly distributed along their length. The explicit form of analytic solution is derived by directly solving the governing differential equations of motion. The development is based on Vlasov theory which includes the effect of flexural-torsion coupling, the constrained torsion warping, and rotary inertia. The governing differential equations for coupled bending-torsional vibrations are performed using the principle of virtual displacements. In the case of simply supported beam, exact explicit expressions are derived to predict the natural frequencies and the corresponding mode shapes. The frequency equation, given in determinantal form, is expanded in an explicit analytical form, and then solved using the symbolic computing package Mathcad. The expressions are concise and very simple and as such convenient to be used by a practicing engineer who does not need to go into detail of thin-walled beam theory. Also, the use of explicit expressions gives significant savings in computing time compared with the alternative numerical methods [finite-element method (FEM), finite strip method, differential transform method, etc.]. To demonstrate the validity of this method the natural frequencies of braced thin-walled beams, having coupled deformation modes, are evaluated and compared with FEM.

The paper focuses on the development of a general method for extreme response estimation of dynamical systems subjected to random excitations. One of the most important elements in structural safety is an assessment of the exceedance or failure probability. The frequently used Poisson assumption tends to overestimate the exceedance level if the response spectrum has narrow-banded features since the so-called clustering effects occur. The latter inaccuracy may be corrected by using an alternative approach studied in this paper, which enables accurate prediction of extremes irrespective of clustering effects.

In this paper, the dynamic response of a pounding oscillator subjected to pulse type excitations is revisited with dimensional analysis. The study adopts the concept of the energetic length scale which is a measure of the persistence of the distinguishable pulse of strong ground motions and subsequently presents the dimensionless $\Pi$ products that govern the response of the pounding oscillator. The introduction of Buckingham’s $\Pi$ theorem reduces the number of variables that govern the response of the system from 7 to 5. The proposed dimensionless $\Pi$ products are liberated from the response of an oscillator without impact and most importantly reveal remarkable order in the response. It is shown that, regardless the acceleration level and duration of the pulse, all response spectra become self-similar and, when expressed in the dimensionless $\Pi$ products, follow a single master curve. This is true despite the realization of contacts with increasing durations as the excitation level increases. All physically realizable contacts (impacts, continuous contacts, and detachment) are captured via a linear complementarity approach. The proposed analysis stresses the appreciable differences in the response due to the directivity of the excitation (toward or away the stationary wall) and confirms the existence of three spectral regions where the response of the pounding oscillator amplifies, deamplifies, and equals the response of the oscillator without pounding.

Monolayer cable net system supporting glass facades is structurally sensitive to wind excitations. At present, there are limited researches on its wind-induced vibration performance, therefore it appears imperative to understand the wind-resistant behavior of this type of cable net. The wind-induced response of the monolayer cable net subjected to fluctuating wind loads is investigated with frequency-domain method in this paper, when the cable net deforms to the balance position under the mean wind loads. Some critical factors to wind-induced response are highlighted, including participation property of the modes in the dynamic vibration, and coupling effect among modes. The response spectrum of the cable net is also intensively investigated. It is shown that the first mode dominates wind-induced response significantly in all the modes, and the modes contributing to the wind-induced responses prominently are distributed in a narrow band of low order modes. When some lower modes and coupling effects among these modes are considered, the results in frequency domain agree well with the corresponding results obtained from time domain method, which are adequate for engineering practice. The characteristics of response spectrum of the nodal displacements are similar to those of the cable forces. When the wind loads and structural parameters vary in practical ranges in engineering, the resonant component in the total response sometimes occupies larger part in the total fluctuating wind response of the cable net, while the background component dominates in the wind response more commonly. Nevertheless, the first mode makes the largest contributions, no matter the background or the resonant component dominates.

Concrete subjected to combined compressive stresses and temperature loading exhibits compressive strains, which are considerably greater than for concrete subjected to compressive stresses alone. This phenomenon is called transient thermal creep or load induced thermal strain and is usually modeled by macroscopic phenomenological constitutive laws which have only limited predictive capabilities. In the present study a mesoscale modeling approach is proposed in which the macroscopically observed transient thermal creep results from the mismatch of thermal expansions of the mesoscale constituents. The mesostructure of concrete is idealized as a two-dimensional three-phase material consisting of aggregates, matrix, and interfacial transition zones. The nonlinear material response of the phases is described by a plasticity interface model. The mesoscale approach was applied to analyze compressed concrete specimens subjected to uniform temperature histories and the analysis results were compared to experimental results reported in the literature.

This paper deals with a dynamic method for damage detection in beams. Under the assumption that the damaged beam is a perturbation of the undamaged one, it is shown that natural frequency and antiresonant frequency shifts induced by structural damage contain information on certain generalized Fourier coefficients of the stiffness variation caused by the degradation. A reconstruction method based on this property is proposed to solve the inverse problem. Cases with pseudoexperimental and experimental data are discussed. The results are in good agreement with the theory, provided that average frequency and antiresonant frequency shifts are bigger than modeling/measurement errors.

An analytical approach for the detection of a beam with multiple cracks is presented in this article. The method is based on the bending vibration theory of Euler-Bernoulli beam and the cracks are treated as massless rotational springs, by which the cracked beam is separated into a number of segments of perfect beams. By using the nontrivial solution condition of the vibration mode of the beam elements, specially using the transfer matrix method for the multiple cracks detection, the crack identification equation of the cracked beam is obtained explicitly, which is a function of natural frequencies, the locations, and depths of the cracks. Since the natural frequencies of a cracked beam can be measured through many of the structural testing methods, then the relations of the locations and the depths of the cracks can be determined explicitly from the identification equation of a cracked beam which geometrical and physical parameters as well as the boundary conditions are given. The results of some examples are shown and the present method is validated with the existing and measured experimental data. The detection for other types of beams with different number of cracks and various boundary conditions can also be obtained by a similar procedure.

The strength prediction of rounded dovetail connections (RDCs), a relatively new connection for structural timber members, is difficult due to the anisotropic and brittle nature of the material, the complex stress distribution as well as the uncertainties regarding the associated material resistance. Experimental investigations were carried out to provide input and benchmark data for developing a method to predict RDC strength. Numerical analyses confirmed that the experimentally observed failure location was also the highest stressed part of the model. A probabilistic method is presented to predict the strength of RDC. The method, rather than being stress-based, incorporates size effect for the combined action of tension perpendicular to grain and shear parallel to grain stresses in timber by comparing computed stress volume integrals to unit volume strength thresholds. Therefore not only the magnitude of the stress distributions is considered but also the volume over which they act. The capacities of RDC configurations were predicted and successfully validated with experimental tests. The presented strength prediction method has immediate application for the improvement of RDC design.

The dynamics of most civil engineering structures is affected by the ambient temperature. This raises the issue of discriminating changes in modal parameters due to damage from those due to such effects. A statistical parametric damage detection algorithm based on a null space residual associated with output-only subspace identification and a ${\chi }^2$ test built on that residual has been designed by some of the writers. The purpose of this paper is to propose two extensions of this detection method which account for the temperature effect. The first extension uses a thermal model for deriving a temperature-adjusted null space. The second extension exploits the thermal model together with a statistical nuisance rejection technique. Both methods are illustrated on a laboratory test case within a climatic chamber.

To understand the behavior of concrete material in ambient water, a series of triaxial compressive tests of concrete cylindrical specimens ( $\varphi$ $100\times 200\text{mm}$ ) was conducted on a large scale triaxial machine. The acting pattern of water, confining pressure, loading strain rate, and moisture content were chosen as test parameters. The water acting patterns on concrete were directly divided into mechanical loading and real water loading according to whether the specimens were directly exposed to water or not. The confining pressure ranged from 0–8 MPa and the strain rate included ${10}^{-5}/\text{s}$ , ${10}^{-3}/\text{s}$ , and ${10}^{-2}/\text{s}$ . By testing dry and saturated specimens, the effect of moisture on concrete strength was also examined. The test results indicated that the compressive strengths of both dry and saturated concrete increase obviously with the confining pressure under mechanical confining pressure. However, the effect on the strengthened dry concrete specimens is more significant. The strength of dry concrete under real water loading decreased remarkably, even less than its uniaxial strength, whereas the compressive strength of the saturated concrete specimen under real water loading is close to its uniaxial compressive strength. The strength of concrete increases with strain rate, and this phenomenon becomes more apparent under water loading.

This paper considers the steady-state vertical vibrations of a rigid, cylindrical massive foundation embedded in a poroelastic soil. The foundation is subjected to time-harmonic vertical loading and is perfectly bonded to the surrounding soil. The contact surface between the foundation and the soil is assumed to be smooth and fully permeable. Biot’s poroelastodynamic theory is used in the analysis. The soil underlying the foundation base is assumed to be a homogeneous poroelastic half space while the soil along the side of the foundation is assumed to consist of a series of infinitesimally thin layers. The dynamic interaction problem is solved by a simplified analytical method. The accuracy of the present solution is verified by comparisons with existing solutions for both elastodynamic and poroelastodynamic interaction problems. Selected numerical results for the vertical dynamic impedance and response factor of the rigid foundation are presented to demonstrate the influence of nondimensional frequency of excitation, depth ratio, mass ratio, shear modulus of the backfill, and poroelastic material properties on dynamic interaction between an embedded foundation and a poroelastic half space.

The Cartesian cut-cell method can be used to represent irregular and complex computational domains with less computational efforts by cutting the grid cells on the boundary surfaces in a background uniform Cartesian mesh. In this study, a modified Cartesian cut-cell grid technique is proposed to better represent complex physical geometries. A point shifting treatment was employed to determine the start and end points of a line segment in cut-cell grids. This led to an improved representation of sharply-shaped corners in surface polygons. Numerical simulation to solve a set of shallow-water equations was performed by incorporating a finite volume approach into the Cartesian cut-cell mesh. The advective fluxes at intercells were first estimated by a Harten, Lax and van Leer for contact wave approximate Riemann solver. In order to improve the model accuracy to the second order, a total variation diminishing-weighted average flux method was applied to work adaptively with the cut-cell mesh. The numerical model was then employed to simulate dam-break flow propagation in a small channel with a rectangular obstacle or a 45° bend. The numerical results show good agreement with available laboratory measurements.