Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
A holistic equivalent linear method for site response analysis
Abstract In order to solve the problem of site response analysis under strong nonlinearity, this paper proposes a concept of a Holistic Equivalent Shear Strain (HESS) and establishes a Holistic Equivalent Linear (EQL-H) method. The key point of the HESS is that the seismic wave components that affect the nonlinear response should be selected to join the iterative calculation of the equivalent linearization. A simple formulation for calculating the HESS is constructed by the parameters related to the nonlinear characteristics of soils themselves, the intensity of the seismic excitation and all the contributing waves in the shear strain time history. The difference between the EQL-H method and the traditional Equivalent Linear (EQL) method such as SHAKE2000 lies in the equivalent shear strain solution and the traditional EQL method basically is a special solution of the EQL-H method under the weak nonlinear condition. Comparison of the EQL method, EQL-H method and the existing fully nonlinear (NL) method is made by using 1468 actual downhole records on the level sites, especially on site classes D and E. The comparison indicates that the EQL-H method show quite a noticeable improvement for site classes D and E and slight improvement for site class C than the traditional EQL method and the existing NL method. The mean relative error of the predicted PGA by the EQL-H method on different site classes under ground shaking intensities less than 1.0 g is within ±20%.
Highlights A new equivalent shear strain formula which can better estimate site response in cases of strongly nonlinear is presented. To construct the new equivalent shear strain formula, the principle of only including contributors to nonlinear reactions is proposed and can be achieved by setting an effective shear strain threshold. The threshold is determined by the soil nonlinear characteristics and the input peak acceleration, and the equivalent shear strain is formed by considering the whole shear strain time history instead of the maximum value. The EQL-H method can better estimate site response and on different site classes under wide range of shaking intensity, the mean relative error of the predicted PGA by it is within ±20%. While on site classes D and E under strong shaking intensity, the method errors of the existing EQL and NL methods cannot be neglected, and on average, the predicted PGA by them significantly lower than the actual values. The proposed EQL-H method can be used not only for one-dimensional problem, but also for multi-dimensional site seismic response analyses, and furthermore, may provide new clues for solving strong nonlinear dynamic problems in other fields.
A holistic equivalent linear method for site response analysis
Abstract In order to solve the problem of site response analysis under strong nonlinearity, this paper proposes a concept of a Holistic Equivalent Shear Strain (HESS) and establishes a Holistic Equivalent Linear (EQL-H) method. The key point of the HESS is that the seismic wave components that affect the nonlinear response should be selected to join the iterative calculation of the equivalent linearization. A simple formulation for calculating the HESS is constructed by the parameters related to the nonlinear characteristics of soils themselves, the intensity of the seismic excitation and all the contributing waves in the shear strain time history. The difference between the EQL-H method and the traditional Equivalent Linear (EQL) method such as SHAKE2000 lies in the equivalent shear strain solution and the traditional EQL method basically is a special solution of the EQL-H method under the weak nonlinear condition. Comparison of the EQL method, EQL-H method and the existing fully nonlinear (NL) method is made by using 1468 actual downhole records on the level sites, especially on site classes D and E. The comparison indicates that the EQL-H method show quite a noticeable improvement for site classes D and E and slight improvement for site class C than the traditional EQL method and the existing NL method. The mean relative error of the predicted PGA by the EQL-H method on different site classes under ground shaking intensities less than 1.0 g is within ±20%.
Highlights A new equivalent shear strain formula which can better estimate site response in cases of strongly nonlinear is presented. To construct the new equivalent shear strain formula, the principle of only including contributors to nonlinear reactions is proposed and can be achieved by setting an effective shear strain threshold. The threshold is determined by the soil nonlinear characteristics and the input peak acceleration, and the equivalent shear strain is formed by considering the whole shear strain time history instead of the maximum value. The EQL-H method can better estimate site response and on different site classes under wide range of shaking intensity, the mean relative error of the predicted PGA by it is within ±20%. While on site classes D and E under strong shaking intensity, the method errors of the existing EQL and NL methods cannot be neglected, and on average, the predicted PGA by them significantly lower than the actual values. The proposed EQL-H method can be used not only for one-dimensional problem, but also for multi-dimensional site seismic response analyses, and furthermore, may provide new clues for solving strong nonlinear dynamic problems in other fields.
A holistic equivalent linear method for site response analysis
Sun, Rui (Autor:in) / Yuan, Xiaoming (Autor:in)
21.10.2020
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Newly Developed MATLAB Based Code for Equivalent Linear Site Response Analysis
Springer Verlag | 2017
|Newly Developed MATLAB Based Code for Equivalent Linear Site Response Analysis
British Library Online Contents | 2017
|Newly Developed MATLAB Based Code for Equivalent Linear Site Response Analysis
Online Contents | 2017
|Propagation of Uncertainty in Equivalent Linear Site Response Analyses
British Library Conference Proceedings | 2018
|