Conditional Simulation of a Class of Nonstationary Space-Time Random Fields

Heredia‐Zavoni, Ernesto; Santa-Cruz, Sandra

doi:10.1061/(asce)0733-9399(2000)126:4(398)

Cited by 31 publications

(14 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Later, Heredia-Zavoni and Santa-Cruz [11] extended the nonstationary (in both amplitude and frequency content) model proposed by Yeh and Wen [41] for simulations at a single location on the ground surface to conditionally simulate spatially variable ones. To estimate the parametric forms of the time-domain amplitude and frequency modulating functions of the time series in di erent frequency bands, Heredia-Zavoni and Santa-Cruz [11] used frequency domain segmentation of the prescribed accelerogram through direct and inverse Fourier transforms. However, the local energy in the prescribed time history tends to smear to the entire time window through the Fourier transform, which can a ect the shape of the estimated modulating functions for each frequency band.…”

Section: Conditional Simulations Resulting In Zero Residual Displacemmentioning

confidence: 97%

“…The length of the weighting function is selected such that it covers a few oscillations of the motions so that it will lead to fairly smooth connections of the acceleration time histories in the transition regions [8]. Later, Heredia-Zavoni and Santa-Cruz [11] extended the nonstationary (in both amplitude and frequency content) model proposed by Yeh and Wen [41] for simulations at a single location on the ground surface to conditionally simulate spatially variable ones. To estimate the parametric forms of the time-domain amplitude and frequency modulating functions of the time series in di erent frequency bands, Heredia-Zavoni and Santa-Cruz [11] used frequency domain segmentation of the prescribed accelerogram through direct and inverse Fourier transforms.…”

Section: Conditional Simulations Resulting In Zero Residual Displacemmentioning

confidence: 98%

“…Hoshiya [10] also improved the conventional kriging method by directly evaluating the simulation estimation error. Heredia-Zavoni and Santa-Cruz [11] further extended the MLP method to nonstationary, in amplitude and frequency content, random ÿelds. Kameda and Morikawa [12] presented a conditional probability density function (CPDF) method, which provides the closed-form conditional probability function of the Fourier coe cients of the time histories to be simulated.…”

Section: Introductionmentioning

confidence: 97%

See 2 more Smart Citations

Physically compliant, conditionally simulated spatially variable seismic ground motions for performance-based design

Liao

Zerva

2006

Earthquake Engng Struct. Dyn.

View full text Add to dashboard Cite

SUMMARYThe performance-based design of lifeline systems requires spatially variable seismic excitations at the structures' supports that are consistent with prescribed seismic ground motion characteristics and an appropriate spatial variability model-such motions can be obtained through conditional simulation.This work revisits the concept of conditional simulation and critically examines the conformity of the generated motions with the characteristics of the target random ÿeld and observations from data recorded at dense instrument arrays. Baseline adjustment processing techniques for recorded earthquake accelerograms are extended to ÿt the requirements of simulated and conditionally simulated spatially variable ground motions. Emphasis is placed on the use of causal vs acausal ÿltering in the data processing. Acceleration, velocity and displacement time histories are evaluated in two example applications of the approach. The ÿrst application deals with a prescribed synthetic time history that incorporates nonstationarity in the amplitude and frequency content of the motions and depends on earthquake magnitude, source-site distance and local soil conditions; this example results in zero residual displacements. The second application considers as prescribed time history a recording in the vicinity of a fault and yields nonzero residual displacements. It is shown that the conditionally simulated time histories preserve the characteristics of the prescribed ones and are consistent with the target random ÿeld. The results of this analysis suggest that the presented methodology provides a useful tool for the generation of spatially variable ground motions to be used in the performance-based design of lifeline systems.

show abstract

Section: Conditional Simulations Resulting In Zero Residual Displacemmentioning

confidence: 97%

Section: Conditional Simulations Resulting In Zero Residual Displacemmentioning

confidence: 98%

Section: Introductionmentioning

confidence: 97%

See 1 more Smart Citation

Physically compliant, conditionally simulated spatially variable seismic ground motions for performance-based design

Liao

Zerva

2006

Earthquake Engng Struct. Dyn.

View full text Add to dashboard Cite

show abstract

“…Hoshiya [1995] also improved the conventional Kriging method by directly evaluating the simulation estimation error. Heredia-Zavoni and Santa-Cruz [1999] further extended the MLP method to nonstationary, in amplitude and frequency content, random fields, and Hu et al [2012] extended the Kriging method to simulate nonstationary spatially variable seismic ground motions based on evolutionary spectra. Kameda and Morikawa [1994] presented a conditional probability density function (CPDF) method, which provides the closed-form conditional probability function of the Fourier coefficients of the time histories to be simulated.…”

Section: Introductionmentioning

confidence: 99%

Simulation of Spatially Varying Ground Motions in V-shaped Symmetric Canyons

Gao

Zhang

et al. 2016

Journal of Earthquake Engineering

View full text Add to dashboard Cite

A method of simulating spatially varying ground motions in V-shaped symmetric canyon is presented. The topographic amplification effect is taken into account by considering a 2-dimension wave propagation model. Two approaches are developed to simulate spatially varying ground motions. First, the spatially varying ground motions are simulated by using the power spectral density and an empirical coherency loss function. Second, the spatially varying ground motions are simulated by using given ground motions and an empirical coherency loss function. These two proposed methods are validated by comparing statistical characteristics of the synthetic motions with target theoretical models.

show abstract

“…Application of the conditional simulation of ground motion records marked by nonstationary behavior are available in the literature, e.g., Vanmarcke and Fenton (1991), Hoshiya (1995), Shinozuka and Zhang (1996), Kameda and Morikawa (1994) and Heredia-Zavoni and Santa-Cruz (2000). These techniques are based on Kriging, conditional probability density function or their combination.…”

Section: Non-stationary Processesmentioning

confidence: 99%

Numerical simulation of wind effects: A probabilistic perspective

Kareem

2008

Journal of Wind Engineering and Industrial Aerodynamics

130

View full text Add to dashboard Cite

Numerical simulations of wind loads and their effects are critical in the design of structures to ensure their safety under winds. The simulations range from generation of time histories of wind velocity, pressure and force fluctuations to structural response and assessment of attendant functionality and safety under service and design loads, respectively. Typically these schemes employ Monte Carlo based approaches that encompass model based simulations or information derived from observed data. The scope of simulations spans uni-variate to multivariate processes; uni-dimensional to multi-dimensional fields; Gaussian to non-Gaussian; stationary to non-stationary; conditional and unconditional cases. In order to accomplish these tasks, methods based on the time, frequency, or time-frequency domains are employed. This paper summarizes a historical perspective, recent developments and future challenges. Also included in the discussion are computational tools employed for data analysis, response analysis and its management, and prediction. Examples are presented to illustrate some of the topics discussed.

show abstract

Conditional Simulation of a Class of Nonstationary Space-Time Random Fields

Cited by 31 publications

References 12 publications

Physically compliant, conditionally simulated spatially variable seismic ground motions for performance-based design

Physically compliant, conditionally simulated spatially variable seismic ground motions for performance-based design

Simulation of Spatially Varying Ground Motions in V-shaped Symmetric Canyons

Numerical simulation of wind effects: A probabilistic perspective

Contact Info

Product

Resources

About