Analytical Simulation of 3D Wind-Induced Vibrations of Rectangular Tall Buildings in Time Domain

Huergo, Iván F; Hernández-Barrios, Hugo; Gómez-Martíne, Roberto

doi:https://doi.org/10.1155/2022/7283610

Shock and Vibration

On this page

Abstract Introduction Conclusions Appendix Data Availability Conflicts of Interest Acknowledgments References Copyright Related Articles

Research Article | Open Access

Volume 2022 | Article ID 7283610 | https://doi.org/10.1155/2022/7283610

Analytical Simulation of 3D Wind-Induced Vibrations of Rectangular Tall Buildings in Time Domain

Iván F Huergo,¹Hugo Hernández-Barrios,²and Roberto Gómez-Martíne³

Academic Editor: Antonio Batista

Received16 May 2022

Revised15 Aug 2022

Accepted16 Aug 2022

Published07 Oct 2022

Abstract

The Monte Carlo simulation method for along-wind loads on tall buildings performed in the time and space domain may be the only analytical viable option for specific problems such as nonlinearity behavior, nonclassically damping, and detailed structural models in commercial software. However, both across-wind and torsional-wind loads due to vortex shedding are not usually simulated in time domain because the vertical decay constants are unknown or the empirical coherence functions cannot be applied in Monte Carlo simulation methods. In this paper, the spectral representation (SR) method is used to simulate in time domain the along-wind, across-wind, and torsional-wind loads on rectangular tall buildings considering the vertical correlation between the signals. The Krenk root-coherence function is used for the normalized cross-spectrum on the along-wind direction, whereas the Davenport root-coherence function is used for the other two types of wind loads. For both across-wind and torsional-wind loads, the Davenport root-coherence function was assessed at the vortex shedding frequency by changing the vertical decay constants until converge between the Davenport model and the Liang empirical coherence model was achieved. Based on a three‐dimensional model with two translational and one torsional degree of freedom for each floor, the proposed vertical decay constants were validated by comparing the elastic response between frequency domain and time domain approaches. Generally speaking, the results show that peak displacements are significantly underestimated for both across-wind and torsional directions when vertical correlation is neglected. In addition, the advantages of time domain simulation were shown by performing a nonlinear time history analysis considering a bilinear isotropic material hardening model in both translational directions.

1. Introduction

Tall buildings are usually susceptible to wind-induced vibrations, which are usually computed analytically in the frequency domain by using empirical cross-spectrum density functions of wind loads and simplified continuous beam models [1, 2]. However, some particular problems hinder the use of a frequency domain approach to compute the wind-induced response such as nonlinearity behavior (hysteric loops), nonclassical damping (soil-structure interaction and passive energy dissipation devices) [3–6], and detailed structural models in commercial software where the user can only define time history functions as dynamic loads. Therefore, the temporal response analysis has to be performed in such cases rather than the analysis in the frequency domain.

The autoregressive moving average (ARMA) algorithm [7] and the spectral representation (SR) model [8] are the most commonly used methods to simulate along-wind loads on tall buildings in time domain. The essential feature of the SR method [8] is that a random process can be simulated by a series of cosine functions with random frequency; then, the density function of the random frequency is derived from the specified cross-spectral density matrix for multivariate process or from the specified spectral density function for a multidimensional process. It is worth mentioning that the summation of a large set of trigonometric terms involved in the SR method renders the approach computational less efficient. On the other hand, the parameters associated with ARMA models are determined in such a manner that the spectral description of the system response to white noise approximates the target spectral characteristics in an optimum sense. However, for the multivariate ARMA models such as the wind field simulation, a large number of nonlinear equations must be solved, which is why the SR method is usually preferred.

For along-wind loads, the SR method is based on the assumption that fluctuating components of the wind velocity field can be idealized as a mean-zero multivariate and multidimensional Gaussian process, where the normalized cross-spectrum describes the statistical dependence between the turbulent components at two points at a given frequency. The real part of the normalized cross-spectrum is called the normalized cospectrum, and the root-coherence function is defined as the absolute value of the normalized cross-spectrum. On a purely empirical basis, Davenport [9] originally suggested an exponential expression as the root-coherence function, where the vertical decay constant controls the vertical correlation between turbulence at two points. The vertical decay constant usually ranges from 7 to 11.5 [9–12].

The empirical root-coherence function proposed by Davenport [9] has the advantage of simplicity, but incorporates two inconsistencies [11]: (1) the function is positive for any separation, which is in conflict with the definition of the longitudinal turbulence component with a zero mean; (2) the normalized cospectrum approaches unity for small frequencies, which is not true for separations of the same order of magnitude or even larger than the average size of gusts, where the wind structure is characterized by a lack of correlation even at low frequencies. Consequently, Krenk [13] derived a simple modified exponential format not encumbered by the two inconsistencies mentioned above. Using full-scale measurements with vertical separations of the order of 10–20 m, Hansen and Krenk [14] determined that a vertical decay constant of 5 corresponds to a vertical decay constant of 7.5 in the exponential format proposed by Davenport [9], i.e., close to the vertical decay constant of 8 used by some building codes. Different power-spectral density (PSD) functions have been successfully used to simulate the turbulent along-wind velocity component [15–18]. However, these wind time series are usually simulated using the exponential format [9] instead of the modified exponential format [13]. It is worth mentioning that there are other models for the root-coherence function of the along-wind turbulent component, such as the Frøya model [19] and the IEC model [20]; however, both models are usually used to simulate along-wind loads on offshore wind turbines.

Based on the above, different kinds of power-spectral density (PSD) functions can be used to simulate along-wind loads of line-like structures by using the SR method. However, Bojórquez et al. [21] demonstrated that the PSD functions of Von Kármán, Von Kármán Harris, and Solari generated wind time series with better characteristic of turbulence intensity and scale length, compared with the models of Davenport, Kaimal, and modified Kaimal.

For tall buildings with aspect ratios of over 3, their across-wind responses usually exceed along-wind responses and can even reach them several times. Across-wind dynamic load on tall buildings is induced by three mechanisms: (1) along-wind turbulence, (2) across-wind turbulence, and (3) wake excitation (vortex shedding). Vortex shedding is the main contributor to the across-wind response of a tall building, particularly when either of its lowest translational natural frequencies approaches the vortex shedding frequency. In addition to transmission lines, bridge pylons, and bridge deck sections, potential super slim and tall buildings can also be subjected to self-excited forces in the across-wind direction; that is, galloping. In fact, the galloping effect can be measured in a wind tunnel by using different methodologies like the hybrid aeroelastic-pressure balance technique [22].

Generally speaking, wind-induced vibrations can be reduced in three different ways: (1) by changing the stiffness or mass, (2) by increasing the damping with passive or active control devices [3–6, 23–26], and (3) by reducing wind loads with aerodynamic shapes such as tapering. The fact that a tapered tall building might spread the vortex shedding over a broad range of frequencies makes it more effective for reducing across-wind responses has been established [27]. However, Chen et al. [28] discovered a new phenomenon called partial reattachment, which suppresses vortex shedding in reattached regions and forms a separation envelope. Based on the partial reattachment phenomenon, tapering promotes the unsteady effect, and galloping interaction near the free end, making the prism more susceptible to wind-induced vibrations. Overall, tapering promotes a structure’s susceptibility to wind-induced vibrations and may cause unanticipated perils in engineering practice [28].

For tapering cylinder-like structures, Vickery and Clark [29] proposed a formula for the normalized across-wind load, whereas Gu and Quan [30] proposed new formulas for the power spectra of the base moment induced by across-wind loads for 15 typical tall building models. However, these power spectra are not useful to simulate across-wind loads along the height of the building. In fact, the literature features some empirical formulas for the power spectra of across-wind loads on square cross section buildings [31, 32] and rectangular tall buildings [1, 33] as a result of wind tunnel tests.

Tsukagoshi et al. [15] simulated across-wind loads on a square tall building using the power spectrum proposed by Ohkuma and Kanaya [31], the empirical root-coherence function proposed by Davenport [9], and a vertical decay constant equal to 3. However, Tsukagoshi et al. [15] did not justify the value of the vertical decay constant used in the simulation process. On the other hand, Slooten [34] simulated across-wind loads on a rectangular tall building by using the power spectrum proposed by Liang et al. [1]; however, Slooten [34] did not consider the vertical correlation between the signals, which is physically wrong.

Wind-induced torsional vibration of tall buildings can enlarge the dynamic response near the peripheries of their cross section, especially when the side faces of a rectangular tall building are wider, and/or it is asymmetric, and/or its lowest torsional natural frequency approaches either of its lowest translational natural frequencies. Choi [35] and Carini [36] studied the wind-induced torsional response on tall buildings using experimental data, whereas Liang et al. [2] proposed an empirical formula for the power spectra of torsional-wind loads on rectangular tall buildings.

The most widely used empirical formulas for across-wind force spectra and torque spectra are those proposed by Liang et al. [1, 2], which are valid for rectangular buildings with various side ratios at normal attack angles. Based on the experimental research, Liang et al. [1, 2] also proposed empirical formulas for the root-coherence function of across-wind loads and torque between two different levels within the spectral peaks. However, these empirical formulas of vertical correlation cannot be used directly by the SR method because they do not depend on the frequency content of the spectra and lead to a negative definite coherence matrix that cannot be factorized through the Cholesky decomposition.

In this paper, a review of the parameters involved in simulating wind-induced loads on rectangular tall buildings was carried out in order to unify the simulation process by using the SR method. The vertical decay constants for both across-wind and torsional-wind loads are proposed based on a comparative analysis between Liang coherence function [1, 2] and Davenport coherence function [9], In this way, both correlated and uncorrelated simulated signals are used in time domain dynamic analysis in order to study the effect of vertical correlation on 3D wind-induced vibrations of rectangular tall buildings, particularly for across-wind and torsional responses. Furthermore, a nonlinear time history analysis was performed to demonstrate the advantages of performing a time domain simulation against the frequency domain approach.

2. Wind Loads on Tall Buildings

Under the action of turbulent wind, tall buildings are loaded simultaneously in the along-wind, across-wind, and torsional directions as shown in Figure 1. When a vortex forms on the side of a building, it creates a suction force that can induce large amplitude vibrations in the plane normal to the wind when the vortex shedding is in resonance with one of the natural frequencies of vibration of the building. In a similar way, torque is mainly induced by the asymmetric distribution of wind pressure caused by vortex shedding.

Based on Figure 1, wind loads on a uniform N-story building are given bywhere j = 1,2,3, …, N; is the along-wind force at height at time t; is the across-wind force at height at time t; is the wind-induced torque at height at time t; is the density of air, which depends on the barometric pressure and the average annual temperature; is the windward pressure; is the leeward pressure; is the wind velocity in the longitudinal direction at height at time t; is the mean wind velocity at height , that is, the mean value of ; is the turbulent component in the wind direction at height at time t, that is, the fluctuating component of ; is the projected area of the jth story, which is taken perpendicular to ; is the width of the windward side; is the depth of the cross section of the building; is the total height of the building; is the drag coefficient (see Appendix A); is the nondimensional lift force at height at time t; and is the nondimensional torque at height at time t.

2.1. Mean Wind Velocity

In practical wind load codes and standards, the power law and the logarithmic law are used to describe the wind profile in the atmospheric boundary layer. However, neither law is valid at very high altitudes above ground [11]. A more precise expression based on the mathematical model developed by Harris and Deaves [37] is the corrected logarithmic law, which fits the experimental data accurately and covers surface roughness changes. According to Harris and Deaves [37], the mean wind velocity is given bywhere is the friction velocity, is the Von Kármán´s constant (∼ 0.4), is the surface roughness length, is the gradient height, is the reference height (), is the Coriolis parameter, is the angular velocity of the Earth (), and is the latitude. According to Simiu and Scanlan [38], typical values of for various types of built-up terrain are shown in Table 1.

2.2. Power Spectral Density Functions

The power-spectral density (PSD) function of a time series describes the distribution of power into the frequency components composing that signal, that is, the variations (energy) as a function of frequency. The unit of a PSD function is energy (variance) per frequency (width), and the energy within a specific frequency range can be obtained by integrating the PSD function within that frequency range. Generally speaking, the PSD function of a stationary stochastic process can be defined aswhere the subscript is referred to the stationary stochastic process , that is, , ,or ; is the nondimensional PSD function of ; and is the standard deviation of . For rectangular tall buildings, the PSD functions of , , and are shown in Appendices A, B, and C, respectively.

2.3. Vertical Correlation between Two Points

Tall buildings are usually modeled as line-like structures, which means that wind loads are simulated along the z-axis. The statistical dependence between the power spectra at two points and at frequency is given by [11]where is the root-coherence function between the two points; is the cross-spectrum of the two signals at points and , respectively; is the vertical separation between the two points; is the PSD function of the stochastic process at point , and is the PSD function of the stochastic process at point .

2.4. Strouhal Number

In fluid dynamics, vortex shedding is an oscillating flow that takes place when a fluid such as air or water flows past a bluff body at certain velocities, depending on the size and shape of the body. In this flow, vortices are created at the back of the body and detach periodically from either side of the body forming a Von Kármán vortex street. The fluid flow past the object creates alternating low-pressure vortices on the downstream side of the object (see Figure 1). The object will tend to move toward the low-pressure zone. If the frequency of vortex shedding matches the resonance frequency of the structure, then the structure may begin to resonate in the across-wind and torsional-wind directions. Accordingly, the vortex shedding frequency is given bywhere is the Strouhal number, which is a nondimensional quantity describing oscillating flow mechanisms. When , the Strouhal number for across-wind loads on rectangular tall buildings is given by [1]and the Strouhal number for wind-induced torque is given by [2]

2.5. Stochastic Simulation by Using the Spectral Representation Method

In this paper, the bandwidth of the power-spectral density (PSD) function is ranged from to in order to include the wind gust frequencies and the structural frequencies of any tall building. Based on the Nyquist theorem, the sampling interval for the signals is given byand the sampling frequency for the PSD function is given bywhere is the total duration of the signals, which is usually equal to 600 s.

Let us consider a set of N homogeneous Gaussian multidimensional processes (j = 1,2,3, …, N) with mean-zero and with the cross-spectral density matrix defined bywhere is the root-coherence function between the two points at frequency , is the PSD function at point at frequency , is the PSD function at point at frequency , and . Through the Cholesky decomposition, the cross-spectral density matrix can be decomposed into the following format:whereand the superscript T represents matrix transpose.

The power spectra of both the across-wind loads and wind-induced torque may lead to a negative definite Hermitian cross-spectral density matrix that cannot be decomposed through the Cholesky factorization. The SR method, based on the Cholesky decomposition of the lagged coherency matrix [39], is another form of the SR method that can solve this problem if certain types of empirical root-coherence functions are used. Alternatively, the cross-spectral density matrix can be expressed as follows [39]:where

For certain types of empirical root-coherence functions, the coherence matrix is a nonnegative definite Hermitian matrix and can be factorized through the Cholesky decompositionwhere is a lower triangular matrix expressed as

then, the stationary stochastic process (j = 1,2,3, …, N) can be simulated by the following equation (8):whereand are independent random phase angles uniformly distributed between 0 and 2π.

Based on (19), the longitudinal turbulent component of wind velocity , the across-wind load , and the wind-induced torque , can be simulated with prior knowledge of the PSD functions and the empirical root-coherence functions. For the along-wind direction, the simulated turbulent component at height must be filtered by multiplying the aerodynamic admittance function (see Appendix A) by the discrete Fourier transform of ; subsequently, must be retrieved into the time domain through the inverse discrete Fourier transform in order to compute the along-wind load using Equation (1).

3. Vertical Decay Constants

3.1. Along-Wind Loads

On a purely empirical basis, Davenport [9] originally suggested an exponential expression as the root-coherence function of the along-wind turbulence, which is given bywhere is the nth frequency, is the vertical separation between points and , and is the vertical decay constant that controls the vertical correlation between turbulence at the two points. Some building codes propose a vertical decay constant of 8; whereas Davenport [9], Strømmen [10], Dyrbye and Hansen [11], and Solari [12] propose a vertical decay constant equal to 7, 9, 10, and 11.5, respectively.

As mentioned earlier, (21) has the advantage of simplicity, but incorporates two inconsistencies [11]: (1) the function is positive for any separation, which is in conflict with the definition of the longitudinal turbulence component with a zero mean; and (2) the function approaches unity for small frequencies, which is not true for separations of the same order of magnitude or even larger than the average size of gusts. To solve the two inconsistencies mentioned above, Krenk [13] suggested a modified exponential expression as the root-coherence function, which is given bywhereand is the integral length scale at height , which is defined in Appendix A.

On the other hand, the IEC root-coherence function [20] is usually used for offshore wind turbines, and is given bywhere is a coherence scale parameter approximately equal to 340.2 m.

For , , and , Figure 2 shows a comparison among the three different models for the root-coherence function at three different points , , and ; where different vertical decay constants ranging from 5 to 11.5 were used in the Davenport root-coherence function (21), whereas a vertical decay constant of 5 was used in the Krenk root-coherence function (22).

(a)

(b)

(c)

(d)

tUsing full-scale measurements with vertical separations of the order of 10–20 m, Hansen and Krenk [14] determined that a vertical decay constant of 5 in the modified exponential format proposed by Krenk [13] corresponds to a vertical decay constant of 7.5 in the exponential format proposed by Davenport [9], i.e., close to the vertical decay constant of 8 used by some building codes. For separations greater than 20 m, Figure 2 also shows the same pattern, whereas the IEC model [20] corresponds to a vertical decay constant approximately equal to 11.5 in (21). Figure 2 shows that different values of do not significantly modify the vertical correlation discussed above. Therefore, it is highly recommended to correlate along-wind loads by using in (22).

3.2. Across-Wind Loads

According to Liang et al. [1], the root-coherence function for across-wind loads on rectangular tall buildings within the spectral peaks can be approximated by the following empirical equation:where is the vortex shedding frequency, is the vertical separation between points and , is the width of the windward side, and is a parameter related to the side ratio of the building: for and , whereas for and . However, substitution of (25) into (16) leads to a negative definite Hermitian coherence matrix that cannot be decomposed through the Cholesky factorization.

As Tsukagoshi et al. [15] suggested, the root-coherence function proposed by Davenport [9] could be used to correlate across-wind loads. However, they did not justify the value of the vertical decay constant used in the simulation process. Accordingly, a comparison of the vertical coherence function between (21) and (25) within the spectral peaks could solve the problem. Unlike Equation (30), Equation (26) depends on the frequency content of the spectra; however, the values of the vertical decay constants that fit (25) are unknown.

Assuming that , , , and , Figure 3 compares (25) with the Pearson correlation coefficients , for different values of the vertical decay constant, which were computed once the across-wind loads were simulated by using (19) and (21).

(a)

(b)

The lack of a single value of that perfectly matches (25) is shown in Figure 3, perhaps because both are processes of different nature. For this reason, a proposal of vertical decay constants not only for different values of but also for different values of seems to be a better option.

(25) is valid within spectral peaks. Therefore, the average vortex shedding frequency between points and is given bywhere is the Strouhal number. If (26) is substituted into (21), what follows is that the root-coherence function at the vortex shedding frequency is given by

The values of in (27) that fit (25) were determined by assuming 6 different values of ratio and 6 different values of , that is, a total of 36 comparative graphs were generated for across-wind loads, 4 of which are shown in Figures 4 and 5. Based on the 36 comparative graphs, the values of shown in Figure 6 are proposed to simulate across-wind loads on rectangular tall buildings by using (21) in the SR method based on the Cholesky decomposition of lagged coherency matrix.

(a)

(b)

(a)

(b)

(a)

(b)

3.3. Wind-Induced Torque

According to Liang et al. [2], the root-coherence function for wind-induced torque on rectangular tall buildings within the spectral peaks can be approximated by the following empirical equation:where is the vortex shedding frequency, is the vertical separation between points and , is the width of the windward side, whereas and are the parameters related to the side ratio of the building (see Table 2). However, substitution of (28) into (16) leads to a negative definite Hermitian coherence matrix that cannot be decomposed through the Cholesky factorization.

Assuming that , , , and , Figure 7 compares (28) with the Pearson correlation coefficients , for different values of the vertical decay constant, which were computed once the wind-induced torque was simulated by using (19) and (21).

(a)

(b)

Similarly to across-wind loads, the root-coherence function at the vortex shedding frequency for wind-induced torque is given by (27). The values of in (27) that fit (28) were determined by assuming 7 different values of ratio and 6 different values of , that is, a total of 42 comparative graphs were generated for wind-induced torque, 10 of which are shown in Figure 8 9 10 11 to 12. For , it is assumed that , in this way, (27) can be fitted into (28) because the cosine term is removed. Based on the 42 comparative graphs, the values of shown in Figure 13, 14 are proposed to simulate the wind-induced torque on rectangular tall buildings by using (21) in the SR method based on the Cholesky decomposition of lagged coherency matrix.

(a)

(b)

(a)

(b)

(a)

(b)

(a)

(b)

(a)

(b)

(a)

(b)

(a)

(b)

Based on Figure 15, Table 3 shows the values of , , and for the nth shear frame. On the other hand, Table 4 shows the first period of vibration of the buildings in each direction.

It was assumed that the building is located in the city of Cancun, México, where ,, and . According to the Mexican wind load code [45], for a 10-year return period, which is related to the serviceability limit state of the building. For the buffeting response, the aerodynamic admittance function proposed by Castro et al. [16] was used to attenuate the power spectra of the along-wind loads (see Appendix A). The aerodynamic coefficients and vertical decay constants are shown in Table 5, whereas the Strouhal numbers are and for across-wind loads and wind-induced torque, respectively.

For H, Figures 16, 17, 18, 19, and 20 show a well-fitting simulation for the standard deviations, PSD functions, Pearson correlation coefficients, and wind-induced loads, among other parameters.

(a)

(b)

(c)

4. Wind-Induced Vibrations in Time Domain

Let us consider a three-dimensional N-story building, where the diaphragm is assumed to be rigid and therefore there will be three degrees of freedom per floor. Assuming a Wilson-Penzien damping [40], the equations of motion can be grouped in the following matrix system [41]:where is the total number of frames of the building; ; ; is the lumped mass at the jth story of the building; is the mass moment of inertia of the jth floor about the vertical axis passing through the center of mass (C.M.); is the ith modal damping ratio; is the angular frequency of the ith mode of vibration; is the generalized mass corresponding to the ith mode of vibration; is the mode shape vector corresponding to the ith translational mode in x-direction; is the mode shape vector corresponding to the ith translational mode in y-direction; is the mode shape vector corresponding to the ith torsional mode; is the condensed stiffness matrix of the nth frame corresponding to the lateral degrees of freedom; is the projected distance from the nth frame to the center of the mass of the rigid diaphragm; are the Cartesian coordinates of the center of the mass of the nth frame, assuming that the origin of the Cartesian coordinate system is the center of the mass of the rigid diaphragm; is the angle between the nth frame and the x-axis; is a vector containing the lateral displacements of the rigid floor diaphragms in the x-direction; is a vector containing the lateral displacements of the rigid floor diaphragms in the y-direction; is a vector containing the angular displacements of the rigid floor diaphragms about the z-axis; is a vector containing the wind-induced forces in the x-direction; is a vector containing the wind-induced forces in the y-direction; and is a vector containing the wind-induced torque about the z-axis. The matrix equation (34) can be solved using a numerical integration method such as the state space method [42].

For standard structures, the across-wind and torsional responses are generally less significant than the along-wind response. However, for structures with significant height, the across-wind and torsional responses can become more significant [43]. According to the National Research Council of Italy [43], the effects of across-wind and torsional actions should be assessed if . In this paper, the analyses were performed for a square plan building () and 4 different values of . Thereby, the effects of vertical correlation on along-wind, across-wind, and torsional vibrations can be assessed for different aspect ratios.

The mass per unit length of the building is equal to , which means that , , , and are , , , and , respectively. Based on Figure 15, the lateral resisting system in each translational direction consists of 4 uniform shear frames, that is, . In this idealization, the beams and floor systems are rigid (infinitely stiff) in flexure, and several factors are neglected: the axial deformation of the beams and columns, and the effect of axial force on the stiffness of the columns. Assuming a story height equal to 4 m, the lateral stiffness of each column is ; therefore, the story stiffness for each translational direction is equal to . In wind engineering, the total damping ratio is composed of the structural part and the aerodynamic part, where the aerodynamic damping is caused by the interaction between the motion of a structure and motion of the air around it. In this numerical example, it was assumed that is equal to 0.01 [44].

(a)

(b)

(c)

(a)

(b)

(c)

(a)

(b)

(c)

(a)

(b)

(c)

Figures 21, 22, and 23 compare the co related rooftop response in time domain with the uncorrelated rooftop response, that is, if the vertical coherence function is neglected by reducing (19) to . Accordingly, Table 6 shows the root-mean-square (RMS) and peak values of the rooftop dynamic response for both uncorrelated and correlated wind loads.

(a)

(b)

(c)

(d)

(a)

(b)

(c)

(d)

(a)

(b)

(c)

(d)

Peak displacements are frequently used to assess wind-induced damage, whereas peak and root-mean-square (RMS) accelerations are commonly used for human comfort assessment to wind-induced motion [46]. Therefore, Table 6 shows the effects of the vertical correlation on the wind-induced responses for different aspect ratios. Generally speaking, Table 6 shows that the along-wind peak displacements are slightly modified when vertical correlation is neglected, that is, in the range of 2 to 15%; on the contrary, the peak displacements of both across-wind and torsional directions decrease 67–81% for all aspect ratios. When the vertical is neglected, RMS values of along-wind accelerations increase 62–109%, whereas peak values of along-wind accelerations increase 25–42% for all aspect ratios. When the vertical correlation is neglected, both RMS and peak values of across-wind accelerations increase 15–51% for aspect ratios in the range of 2 to 4, whereas they decrease 25–36% for aspect ratios in the range of 6 to 8. On the other hand, RMS torsional accelerations increase 72–243% when the vertical correlation is neglected, whereas peak torsional accelerations increase 97–236% for all aspect ratios.

5. Wind-Induced Vibrations in Frequency Domain

According to Newland [47], the frequency response function matrix between the generalized structural displacements and the external forces of a multi degree of freedom (MDOF) system can be obtained using the following equation:where ,, and are the stiffness, mass, and damping matrices, respectively; is the angular frequency and is the imaginary unit. Then, the displacement and acceleration power-spectral density (PSD) matrices are defined, respectively, aswhere the superscript denotes complex matrix conjugation and is the cross-spectral density matrix of the wind forces, which is given in (11). Next, the variance of the response is obtained by integrating the response auto-spectra populating the main diagonal elements of the response PSD matrix on the frequency axis up to a maximum (cut-off) frequency above which the energy of the underlying processes is negligible. In this way, the RMS displacement and RMS acceleration of the jth floor are obtained, respectively, aswhere and are the jth elements of the main diagonal of and matrices, respectively. For both across-wind and torsional-wind loads, the peak displacement and peak acceleration of the jth floor can be estimated, respectively, by the expressionswhere is the peak factor estimated by the widely used empirical formula [48] as follows:where is the fundamental period of vibration of the structure in the direction of analysis and is the time duration of exposure to the wind action, which is usually equal to 600 s.

As already mentioned above, both across-wind and torsional-wind loads cannot be simulated in time domain by using the empirical root-coherence functions proposed by Liang et al. [1, 2] because these functions lead to a negative definite Hermitian coherence matrix that cannot be decomposed through the Cholesky factorization. However, the empirical root-coherence functions proposed by Liang et al. [1, 2] can be effectively used to compute the elastic wind-induced vibrations in the frequency domain by using Equations (50) and (34). In this way, the vertical decay constants proposed in this paper (see Figures 6, 13, and 14) can be validated by comparing the correlated time domain response with the correlated frequency domain response resulting from using the empirical coherence functions proposed by Liang et al [1, 2]. Accordingly, Table 7 shows a close approximation between the time domain and frequency domain responses.

Generally speaking, Table 7 shows that RMS elastic response decreases in the range of 0.3 to 20% when the time domain analysis is used in conjunction with the proposed vertical decay constants (see Figures 6, 13, and 14). On the other hand, the peak elastic response decreases in the range of 3 to 44% because the peak factor shown in (35) is just an approximation based on along-wind response; therefore, the RMS elastic response is the recommended one to be taken as a validation parameter for the vertical decay constants proposed in this paper. In this way, a time domain analysis has a wider field of application because it is not only useful for nonlinearity behavior, nonclassically damping, and detailed structural models in commercial software but also to compute precisely the peak elastic response compared to an analysis in the frequency domain where the peak factor may not be suitable for across-wind and torsional-wind responses.

6. Nonlinear Time History Analysis

As already mentioned above, the nonlinear behavior of a structure hinders the use of a frequency domain approach. Therefore, a nonlinear time history analysis will show the advantages of the vertical decay constants proposed in this paper to simulate the wind-induced vibrations considering the correlation along the height of the building.

In this section, the 60-story shear building (see Figure 15) is subjected simultaneously to the along-wind and across-wind loads shown in Figure 20, which are correlated along the height of the structure. For the nonlinear behavior, it was assumed a bilinear isotropic hardening ASTM A992 steel model considering a Young´s modulus equal to 199 GPA; a yield stress equal to 344.7 MPA; and a postyield story stiffness equal to 304.4 MN/m, that is, 20% of elastic story stiffness. Assuming a story height equal to 4 m, the yield story force and yield story displacement are 10.5 MN and 6.9 mm, respectively.

The nonlinear matrix system was solved in MATLAB by using the Newmark β-method [42], in this way, the nonlinear displacements and hysteretic loops of the 60^th floor and 5^th floor are shown in Figures 24 and 25, respectively. In addition, Figure 26 shows the energy balance in order to check that the total energy (kinetic + dissipated + elastic strain + hysteresis) is equal to the input energy.

(a)

(b)

(c)

(a)

(b)

(c)

(a)

(b)

For the along-wind response, the structural material yields from 1^st story to the 31^st story, whereas it remains elastic from story the 32^nd story to the rooftop. On the other hand, for the across-wind response, the structural material yields from 1^st story to the 50^th story, whereas it remains elastic from the 51^st story to the rooftop. In fact, Figure 26 clearly shows that the input energy related to the across-wind loads is much greater than the input energy related to the along-wind loads. Furthermore, the hysteresis energy in Figure 26(a) remains approximately constant over time because the along-wind loads are strictly “positive” most of the time. Therefore, the hysteresis behavior usually remains in the loading branches (see Figure 25(b)). On the contrary, Figure 25(c) shows multiple hysteresis loops related to the across-wind vibrations, which means that both hysteresis energy and dissipated energy increase over time (see Figure 26(b)).

For the along-wind response at the rooftop, Figure 24(a) shows that the peak displacement is equal to 68.7 cm, which is 1.37 times greater than that computed by a linear time history analysis (see Table 6). On the contrary, Figure 24(a) shows that the across-wind peak displacement at the rooftop is equal to 51.8 cm, that is, it decreases 30.6% compared to that computed by a linear time history analysis (see Table 6).

7. Conclusions

The spectral representation (SR) method is frequently used to simulate along-wind loads on tall buildings. However, it can be used to simulate any mean-zero multivariate and multidimensional process such as across-wind loads and wind-induced torque. However, scientific literature reveals that across-wind and wind-induced torque have only been simulated in time domain without considering the vertical correlation between the signals, which is physically incorrect. Therefore, it has not been possible to simulate both linear and nonlinear response induced by vortex shedding (across and torsional directions). For the particular case of nonlinear response, it is very important to compute the damage induced by the wind in the 3D space.

As the first goal of this research, a review of the parameters involved to simulate along-wind, across-wind, and torsional-wind loads on rectangular tall buildings was carried out to unify the simulation process in order to compute the 3D wind-induced vibrations. In this way, the most important empirical models for the parameters of wind simulations are fully discussed. Accordingly, the following conclusions were obtained:(a)For along-wind loads, the power-spectral density function proposed in ESDU adequately represents the power of the turbulent component in all the frequency content. Furthermore, it is highly recommended to correlate along-wind loads by using a vertical decay constant equal to 5 in the empirical coherence function proposed by Krenk, which is not encumbered by the two inconsistencies related to the empirical function proposed by Davenport.(b)For across-wind and torsional-wind loads, the power-spectral density functions proposed by Liang adequately represent the power of both types of wind loads on rectangular tall buildings with various side ratios. However, the empirical coherence functions proposed by Liang cannot be used in the SR method because they do not depend on the frequency content of the spectra and lead to a negative definite coherence matrix that cannot be factorized through the Cholesky decomposition. As the second goal of this research, a comparative analysis between the Liang coherence model and the Davenport coherence model was carried out to propose vertical decay constants for various side ratios at normal attack angle, which is the most important contribution of this paper because it solves the problem of vertical correlation in the simulation of across-wind and torsional-wind loads in time domain. Accordingly, the following conclusions were obtained:(c)There is not a single value of the vertical decay constant in Davenport’s empirical model that perfectly matches Liang coherence function, particularly when H decreases and B increases.(d)The Davenport coherence function fits well the Liang coherence function as H increases and B decreases. As the third goal of this research, the effect of vertical correlation on along-wind, across-wind, and wind-induced torsional responses were assessed in the time domain for shear buildings with different aspect ratios by using a three‐dimensional model with two translational and one torsional degree of freedom for each floor. Furthermore, the proposed vertical decay constants were validated by comparing the elastic response between frequency domain and time domain approaches. Accordingly, the following conclusions were obtained:(e)Generally speaking, the results show that the peak displacements are significantly underestimated for both across-wind and torsional directions when the vertical correlation is neglected. On the contrary, the along-wind peak displacements are slightly modified.(f)When the vertical correlation is neglected, across-wind acceleration increases for low-rise buildings, whereas it decreases for high-rise buildings. On the other hand, torsional accelerations increase significantly when the vertical correlation is neglected.(g)Generally speaking, the RMS elastic response decreases in the range of 0.3 to 20% when the time domain analysis is used in conjunction with the proposed vertical decay constants for both across-wind and torsional directions.(h)For the elastic RMS response considering vertical correlation, the difference between time domain and frequency domain approaches was mostly negligible.(i)The peak factor in the frequency domain may not be suitable for across-wind and torsional-wind elastic responses. Therefore, a time domain analysis has a wider field of application. As the final goal of this research, the advantages of time domain simulation were shown by performing a nonlinear time history analysis for a 60-story shear building considering a bilinear isotropic material hardening model in both translational directions. Accordingly, the following conclusions were obtained:(j)For the along-wind direction, the structural material yields from 1st story to the 31st story, whereas it yields from 1st story to the 50th story for the across-wind direction.(k)The along-wind peak displacement at rooftop increases 37.3% when a nonlinear time history analysis is performed. On the contrary, the across-wind peak displacement decreases 30.6% compared to that computed by a linear time history analysis.

Based on contributions of this research, we believe that structural engineers will be able to compute accurately the 3D wind-induced response of rectangular tall buildings for specific problems where frequency domain analysis is not applicable (nonlinearity behavior, nonclassically damping, and detailed structural models in commercial software), particularly for across-wind and torsional responses. For future work, the 3D wind-induced vibration of high-rise buildings will be studied in time domain considering nonlinear behavior and passive energy dissipation devices.

Appendix

A. Power-spectral density function of along-wind loads

Throughout the literature, many semiempirical expressions for the power spectrum of the turbulent component in the wind direction have been proposed based on a number of physical parameters. These expressions [49–58] can be divided into two families: the so-called Kaimal spectra [54–57], which provide a good agreement with the natural observed turbulence; and the Von Kármán spectra [52, 53], which are more in agreement with the turbulence of a wind tunnel. The modified Von Kármán spectrum proposed in the ESDU [58] offers both a fair agreement with empirical data together with simple correlation implementations. In addition, this model is the only one that adequately represents the power spectrum of in all the frequency content. Based on the ESDU [58], the power-spectral density function of in (m/s)²/Hz is given bywhere the nondimensional power spectrum and the standard deviation in m/s are defined, respectively, aswhere is the integral length scale of , is the Richardson number, is the friction velocity, is the gradient height, is the Coriolis parameter, and is the surface roughness length.

Based on Equations (1) and (A-1), the power-spectral density function of in N²/Hz is given bywhere j = 1,2,3, …, N; is the aerodynamic admittance function at height , which represents the relationship between the gust frequency and its area of influence; is the density of air; is the projected area of the jth story; is the mean wind velocity at height ; and is the drag coefficient. The empirical formula commonly used for the aerodynamic admittance function in rectangular tall buildings is given by [59]

However, Castro et al. [16] slightly modified Equation (A-19) by changing the exponent -2 with the exponent -7/6. According to Castro et al. [16], this modified expression for the aerodynamic admittance function adequately attenuates the power spectrum of the along-wind loads for two structures widely studied by other authors: a 100-meter steel tower [60] and the CAARC standard tall building [61, 62]. Therefore, in this paper the aerodynamic admittance function modified by Castro et al. [16] is recommended to attenuate the power spectrum of the along-wind loads.

Based on wind tunnel tests or computational fluid dynamics, practical wind load codes and standards suggest values of the aerodynamic coefficients for a variety of situations. Based on the Brazilian wind code [63], Patlán [64] performed a surface fitting to compute the drag coefficient for a rectangular tall building immersed in a turbulent flow at normal attack angle, which is given by

When and .

B. Power-spectral density function of across-wind loads

According to Liang et al. [1], the power-spectral density function of in N²/Hz is given bywhere the standard deviation in N is defined asand the lift coefficient is given bywhen at normal attack angle of wind flow.

When , the nondimensional power spectrum of the across-wind force for a rectangular tall building (see Figure 1) is given by the following equation [1]:where is the vortex shedding frequency and is the Strouhal number.

When , the nondimensional power spectrum of the across-wind force for a rectangular tall building (see Figure 1) is given by the following equation [1]:where is the turbulence intensity of , (s) is the parameter defined in Equation (B-6) and is the nondimensional frequency defined in Equation (B-11).

C. Power-spectral density function of wind-induced torque

According to Liang et al. [2], the power-spectral density function of in (N m)²/Hz is given bywhere the standard deviation in N m is defined asand the torque coefficient is given bywhen at normal attack angle of wind flow.

When , the nondimensional power spectrum of the wind-induced torque for a rectangular tall building (see Figure 1) is given by the following equation [2]:where is the vortex shedding frequency and is the Strouhal number.

When , the non-dimensional power spectrum of the wind-induced torque for a rectangular tall building (see Figure 1) is given by the following equation [2]:where is the turbulence intensity of , (s) is the parameter defined in Equation (C-7) and is the nondimensional frequency defined in Equation (C-11).

Data Availability

The authors confirm that the data supporting the findings of this study are available within the article.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

The authors would like to acknowledge CONACYT-Becas México for financially supporting the first author in conducting doctoral studies at Universidad Nacional Autónoma de México under the supervision of the second author. The authors also wish to thank and Universidad Michoacana de San Nicolás de Hidalgo.

References

S. Liang, S. Liu, Q. S. Li, L. Zhang, and M. Gu, “Mathematical model of acrosswind dynamic loads on rectangular tall buildings,” Journal of Wind Engineering and Industrial Aerodynamics, vol. 90, no. 12-15, pp. 1757–1770, 2002.
View at: Publisher Site | Google Scholar
S. Liang, Q. S. Li, S. Liu, L. Zhang, and M. Gu, “Torsional dynamic wind loads on rectangular tall buildings,” Engineering Structures, vol. 26, no. 1, pp. 129–137, 2004.
View at: Publisher Site | Google Scholar
I. F. Huergo and H. Hernández, “Coupled shear-flexural model for dynamic analysis of fixed-base tall buildings with tuned mass dampers,” The Structural Design of Tall and Special Buildings, vol. 28, no. 17, Article ID e1671, 2019.
View at: Publisher Site | Google Scholar
I. F. Huergo and H. Hernández, “Coupled-two-beam discrete model for dynamic analysis of tall buildings with tuned mass dampers including soil-structure interaction,” The Structural Design of Tall and Special Buildings, vol. 29, no. 1, Article ID e1683, 2020.
View at: Publisher Site | Google Scholar
I. F. Huergo, H. Hernández-Barrios, and C. M. Patlán, “A continuous-discrete approach for pre-design of flexible-base tall buildings with fluid viscous dampers,” Soil Dynamics and Earthquake Engineering, vol. 131, Article ID 106042, 2020.
View at: Publisher Site | Google Scholar
H. Hernández-Barrios, I. F. Huergo, C. Arce-León, and C. M. Patlán, “Unified shear-flexural model for vibration control of buildings using passive dynamic absorbers,” Shock and Vibration, vol. 2020, pp. 1–25, 2020.
View at: Publisher Site | Google Scholar
M. P. Mignolet and P. D. Spanos, “MA to arma modeling of wind,” Journal of Wind Engineering and Industrial Aerodynamics, vol. 36, no. 1, pp. 429–438, 1990.
View at: Publisher Site | Google Scholar
M. Shinozuka, C. B. Yun, and H. Seya, “Stochastic methods in wind engineering,” Journal of Wind Engineering and Industrial Aerodynamics, vol. 36, no. 2, pp. 829–843, 1990.
View at: Publisher Site | Google Scholar
A. G. Davenport, “The response of slender line-like structures to a gusty Wind,” Proceedings - Institution of Civil Engineers, vol. 23, no. 3, pp. 389–408, 1962.
View at: Publisher Site | Google Scholar
E. N. Strømmen, Theory of Bridge Aerodynamics, Springer, Berlin Heidelberg, 2010.
C. Dyrbye and S. O. Hansen, Wind Loads on Structures, John Wiley & Sons, Hoboken New Jersey, 1997.
G. Solari, “Gust buffeting. I: peak wind velocity and equivalent pressure,” Journal of Structural Engineering, vol. 119, no. 2, pp. 365–382, 1993.
View at: Publisher Site | Google Scholar
S. Krenk, “Wind field coherence and dynamic wind forces,” IUTAM Symposium on Advances in Nonlinear Stochastic Mechanics, vol. 47, pp. 269–278, 1996.
View at: Publisher Site | Google Scholar
S. O. Hansen and S. Krenk, “Dynamic along-wind response of simple structures,” Journal of Wind Engineering and Industrial Aerodynamics, vol. 82, no. 1-3, pp. 147–171, 1999.
View at: Publisher Site | Google Scholar
H. Tsukagoshi, Y. Tamura, A. Sasaki, and H. Kanai, “Response analyses on along-wind and across-wind vibrations of tall buildings in time domain,” Journal of Wind Engineering and Industrial Aerodynamics, vol. 46-47, pp. 497–506, 1993.
View at: Publisher Site | Google Scholar
H. G. Castro, M. E. De Bortoli, R. R. Paz, and J. O. Marighetti, “Una metodología de cálculo para la determinación de la respuesta dinámica longitudinal de estructuras altas bajo la acción del viento,” Revista Internacional de Métodos Numéricos para Cálculo y Diseño en Ingeniería, vol. 31, no. 4, pp. 235–245, 2015.
View at: Publisher Site | Google Scholar
Y. Li, C. Yu, X. Chen, X. Xu, K. Togbenou, and H. Xiang, “An efficient Cholesky decomposition and applications for the simulation of large-scale random wind velocity fields,” Advances in Structural Engineering, vol. 22, no. 6, pp. 1255–1265, 2019.
View at: Publisher Site | Google Scholar
N. Zhao and G. Huang, “Wind velocity field simulation based on enhanced closed-form solution of Cholesky decomposition,” Journal of Engineering Mechanics, vol. 146, no. 2, Article ID 04019128, 2020.
View at: Publisher Site | Google Scholar
O. J. Andersen and J. Løvseth, “The Frøya database and maritime boundary layer wind description,” Marine Structures, vol. 19, no. 2-3, pp. 173–192, 2006.
View at: Publisher Site | Google Scholar
IEC 61400-1, Wind Turbines Part 1: Design Requirements, IEC, Geneva Switzerland, 2005.
E. Bojórquez, O. Payán-Serrano, A. Reyes-Salazar, and A.. Pozos, “Comparison of spectral density models to simulate wind records,” KSCE Journal of Civil Engineering, vol. 21, no. 4, pp. 1299–1306, 2016.
View at: Publisher Site | Google Scholar
Z. Chen, K. T. Tse, K. C. S. Kwok, A. Kareem, and B. Kim, “Measurement of unsteady aerodynamic force on a galloping prism in a turbulent flow: a hybrid aeroelastic-pressure balance,” Journal of Fluids and Structures, vol. 102, Article ID 103232, 2021.
View at: Publisher Site | Google Scholar
Z. Chen, X. Fu, Y. Xu, C. Y. Li, B. Kim, and K. T. Tse, “A perspective on the aerodynamics and aeroelasticity of tapering: partial reattachment,” Journal of Wind Engineering and Industrial Aerodynamics, vol. 212, Article ID 104590, 2021.
View at: Publisher Site | Google Scholar
A. E. Taha, “Vibration control of a tall benchmark building under wind and earthquake excitation,” Practice Periodical on Structural Design and Construction, vol. 26, no. 2, 2021.
View at: Publisher Site | Google Scholar
A. M. Aly and S. Chapain, “Large-scale open-jet testing for cladding design in high-rise buildings: higher peak pressures compared to wind tunnels,” Practice Periodical on Structural Design and Construction, vol. 27, no. 2, 2022.
View at: Publisher Site | Google Scholar
J. S. Love, B. Morava, J. K. Robinson, and T. C. Haskett, “Tuned sloshing dampers in tall buildings: a practical performance-based design approach,” Practice Periodical on Structural Design and Construction, vol. 26, no. 3, 2021.
View at: Publisher Site | Google Scholar
M. H. Stanikzai, S. Elias, and Y. Chae, “Recent advances in hybrid vibration-control systems,” Practice Periodical on Structural Design and Construction, vol. 27, no. 3, 2022.
View at: Publisher Site | Google Scholar
Y. M. Kim and K. P. You, “Dynamic responses of a tapered tall building to wind loads,” Journal of Wind Engineering and Industrial Aerodynamics, vol. 90, no. 12-15, pp. 1771–1782, 2002.
View at: Publisher Site | Google Scholar
B. J. Vickery and A. W. Clark, “Lift or across-wind response of tapered stacks,” Journal of the Structural Division, vol. 98, no. 1, pp. 1–20, 1972.
View at: Publisher Site | Google Scholar
M. Gu and Y. Quan, “Across-wind loads of typical tall buildings,” Journal of Wind Engineering and Industrial Aerodynamics, vol. 92, no. 13, pp. 1147–1165, 2004.
View at: Publisher Site | Google Scholar
T. Ohkuma and A. Kanaya, “On the correlation between the shape of rectangular cylinders and characteristics of fluctuating lifts of them,” Proceedings of 5th Symposium on Wind Effects on Structures, pp. 147–154, 1978.
View at: Google Scholar
A. Kareem, “Model for predicting the acrosswind response of buildings,” Engineering Structures, vol. 6, no. 2, pp. 136–141, 1984.
View at: Publisher Site | Google Scholar
M. Gu, “Study on wind loads and responses of tall buildings and structures,” in Proceedings of the The Seventh Asia-Pacific Conference on Wind Engineering, Taiwan, November 2009.
View at: Google Scholar
E. C. Slooten, “Feasibility Study of a wood-concrete Hybrid Super Tall Building and Optimization of its Wind-Induced Behaviour: A Case Study on a Skyscraper in the City-centre of Rotterdam,” Master´s Thesis, Delft University of Technology, Delft Netherlands, 2018, http://resolver.tudelft.nl/uuid:01da1849-6478-46c9-8a59-e6c7c2e4fb47.
View at: Google Scholar
H. Choi, “Study On Fluctuating Wind Force for 3-dimensional Response Analyses of Tall Buildings with Rectangular Cross-Section,” Tokyo University, Tokyo Japan, 1990, Doctoral Dissertation, Tokyo University.
View at: Google Scholar
M. R. Carini, “Resposta dinâmica em torção de edifícios sob ação do vento,” Master´s thesis, Universidade Federal do Rio Grande, Rio Grande Brazil, 2017, http://hdl.handle.net/10183/157820.
View at: Google Scholar
R. I. Harris and D. M. Deaves, “The structure of strong winds,” Wind Engineering in the Eighties: Proceedings of the CIRIA Conference, Construction Industry Research and Information Association, London, 1981.
View at: Google Scholar
E. Simiu and R. H. Scanlan, Wind Effects on Structures: Fundamentals and Applications to Design, John Wiley & Sons, Hoboken New Jersey, 1996.
Y. Gao, Y. Wu, D. Li, H. Liu, and N. Zhang, “An improved approximation for the spectral representation method in the simulation of spatially varying ground motions,” Probabilistic Engineering Mechanics, vol. 29, pp. 7–15, 2012.
View at: Publisher Site | Google Scholar
E. L. Wilson and J. Penzien, “Evaluation of orthogonal damping matrices,” International Journal for Numerical Methods in Engineering, vol. 4, no. 1, pp. 5–10, 1972.
View at: Publisher Site | Google Scholar
A. Tena, Análisis de estructuras con métodos matriciales, Limusa, Johannesburg South Africa, 2007.
G. C. Hart and K. Wong, Structural Dynamics for Structural Engineers, John Wiley & Sons, Hoboken New Jersey, 1999.
D. T. CNR, Guide for the Assessment of Wind Actions and Effects on Structures, National Research Council of Italy, Italy, 2008.
G. Tamura and A. Kareem, Advanced Structural Wind Engineering, Springer, Berlin Heidelberg Germany, 2013.
CFE, Manual de Diseño de Obras Civiles: Diseño por Viento, CFE, Mexico, 2008.
F. A. Johann, M. E. Carlos, and F. L. Ricardo, “Wind-induced motion on tall buildings: a comfort criteria overview,” Journal of Wind Engineering and Industrial Aerodynamics, vol. 142, pp. 26–42, 2015.
View at: Publisher Site | Google Scholar
D. E. Newland, An Introduction to Random Vibrations, Spectral & Wavelet Analysis, Prentice-Hall, Hoboken New Jersey United States, 1996.
A. G. Davenport, “Note on the distribution of the largest value of a random function with application to gust loading,” Proceedings - Institution of Civil Engineers, vol. 28, no. 2, pp. 187–196, 1964.
View at: Publisher Site | Google Scholar
A. N. Kolmogorov, “The local structure of turbulence in an incompressible viscous fluid for very large Reynolds numbers,” Proceedings of the Royal Society A: Mathematical, Physical & Engineering Sciences, vol. 434, no. 1890, pp. 9–13, 1991.
View at: Publisher Site | Google Scholar
T. Von Kármán, “Progress in the statistical theory of turbulence,” Proceedings of the National Academy of Sciences, vol. 34, no. 11, pp. 530–539, 1948.
View at: Publisher Site | Google Scholar
A. G. Davenport, “The spectrum of horizontal gustiness near the ground in high winds,” Quarterly Journal of the Royal Meteorological Society, vol. 87, no. 372, pp. 194–211, 1961.
View at: Publisher Site | Google Scholar
R. I. Harris, On the Spectrum and Auto-Correlation Function of Gustiness in High Winds, Electrical Research Association, Leatherhead United Kingdom, 1968.
R. I. Harris, “The nature of the wind,” Seminar on Modern Design of Wind-Sensitive Structures, Construction Industry Research & Information CIRIA, London, pp. 29–55, 1970.
View at: Google Scholar
J. C. Kaimal, J. C. Wyngaard, Y. Izumi, and O. R. Coté, “Spectral characteristics of surface-layer turbulence,” Quarterly Journal of the Royal Meteorological Society, vol. 98, no. 417, pp. 563–589, 1972.
View at: Publisher Site | Google Scholar
E. Simiu, “Wind spectra and dynamic along-wind response,” Journal of the Structural Division, vol. 100, no. 9, pp. 1897–1910, 1974.
View at: Publisher Site | Google Scholar
U. Högström, H. Bergström, and H. Alexandersson, “Turbulence characteristics in a near neutrally stratified urban atmosphere,” Boundary-Layer Meteorology, vol. 23, no. 4, pp. 449–472, 1982.
View at: Publisher Site | Google Scholar
D. Hiriart, J. L. Ochoa, and B. Garcia, “Wind power spectrum measured at the san pedro mártir sierra,” Revista Mexicana de Astronomía y Astrofísica, vol. 37, no. 2, pp. 213–220, 2001.
View at: Google Scholar
ESDU, “Characteristic of atmospheric turbulence near the ground, part II: single point data for strong winds (neutral atmosphere),” Technical Reports Series, vol. 85020, 2001.
View at: Google Scholar
B. J. Vickery, “Load fluctuations in turbulent flow,” Journal of the Engineering Mechanics Division, vol. 94, no. 1, pp. 31–46, 1968.
View at: Publisher Site | Google Scholar
G. Solari, “Wind response spectrum,” Journal of Engineering Mechanics, vol. 115, no. 9, pp. 2057–2073, 1989.
View at: Publisher Site | Google Scholar
W. H. Melbourne, “Comparison of measurements on the CAARC standard tall building model in simulated model wind flows,” Journal of Wind Engineering and Industrial Aerodynamics, vol. 6, no. 1-2, pp. 73–88, 1980.
View at: Publisher Site | Google Scholar
L. Chen and C. W. Letchford, “Parametric study on the along-wind response of the CAARC building to downbursts in the time domain,” Journal of Wind Engineering and Industrial Aerodynamics, vol. 92, no. 9, pp. 703–724, 2004.
View at: Publisher Site | Google Scholar
NBR-6123, Forças Devidas Ao Vento Em Edificações, Associação Brasileira de Normas Técnicas ABNT, Río de Janeiro Brasil, 1998.
C. M. Patlán, “Control de vibraciones sísmicas y eólicas en una estructura esbelta,” Universidad Michoacana de San Nicolás de Hidalgo, Morelia Mexico, 2020, Master´s thesis].
View at: Google Scholar

Copyright

Copyright © 2022 Iván F Huergo et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PDF Download Citation

Download other formats

Order printed copies

Views

614

Downloads

498

Citations