Font Size: a A A

Inverse Analysis Of Wind Loads On Super Tall Buildings

Posted on:2017-03-30Degree:DoctorType:Dissertation
Country:ChinaCandidate:M X FangFull Text:PDF
GTID:1362330566453181Subject:Structural engineering
Abstract/Summary:PDF Full Text Request
A large number of super-tall buildings have been built in the southeast coastal areas of China in recent years.Wind load and wind-induced response is the primary control index of security and applicability in designing.The measurement technology of structural dynamic response is advanced,and the measurements of structural acceleration and displacement response are more exact than load measurement,currently.Therefore,the inverse analysis method to identify structural wind load based on the measured wind-induced response become a new approach of indirect measurements gradually.The practical inversion methods in time domain for identifying fluctuating wind load on super-tall structures using several layers' s wind-induced response are studied in detail,which is based on Kalman filter and characteristics of wind-induced vibration of super-tall buildings.The approaches of theoretical analysis,numerical simulation,wind tunnel tests and field measurements are applied.Corresponding mathematical models are established in this dissertation,too.Finally,in order to validate and compare the applicability as well as accuracy of inverse methods proposed in this thesis,three typical super-tall buildings are considered as research objects.The major research contents are as follows:(1)The approximate modal responses of the previous orders are obtained by modal transformation theory and the characteristics of wind-induced response on super-tall buildings in the condition of insufficient measurements.The number of modals controlling structural vibration and the minimal measuring points are determined by Proper orthogonal decomposition(POD).(2)Two inverse methods based on continuous Kalman filter are proposed to identify structural fluctuating wind load(CKF,A-CKF).The dynamic differential equation is transformed into state space equation firstly and the extended form of continuous Kalman filter for estimating wind load is deduced.Then the identification methods of fluctuating wind load and unmeasured wind-induced response are obtained(CKF).Steady state solution of error covariance matrix is raised and then the solving process is simplified.The Augmented-Continuous Kalman Filter method(A-CKF)is acquried through the augmentation to state space equation.Stability of inversion system is determined by PBH rule and the problem of instability is solved by introducing displacement dummy-measurements technology.The L-curve method in regularization theory is introduced into inversion system to identify the most reasonable value of the observation noise intensity.Then the optimal Kalman filter gain matrix is obtained and the anti-noise ability as well as applicability of inversion mathematical modal is improved simultaneously.Transfer function from the accurate load to estimate load in frequency domain is deduced using Laplace transform and the influences of the input response type and the observation noise are analyzed based on numerical simulation.(3)Based on discrete Kalman filter and weighted recursive least squares,three inversion methods are developed(DKF,IRLS,A-DKF).Firstly,the discrete system equations are constructed by the solution of matrix differential equation.Then the extended discrete Kalman filtering method(DKF)to estimate fluctuating wind load and unmeasured wind-induced responses is deduced.An improved recursive least squares approach(IRLS)is developed by combining discrete Kalman filter with RLS.Through the means of amplificating state vector,system equations are reconstructed and then the Augmented-Discrete Kalman Filter(A-DKF)is obtained,the stability of which is determined by the properties of convergent matrix.The covariance matrix equation of estimation error is derived when system reaching steady state.The corresponding steady-state filter gain is given,simplifing the solving process and improving the practicability.(4)The experiment study on wind load inversion is carried in this dissertation.Two wind tunnel tests models of typical super-tall buildings in typhoon area of the southeastern coast of China are designed(Haohe building in Guangzhou,Two International Financial Center in Hong Kong)and the structural wind load under different directions is measured.Then the wind loads from inversion are compared with those determined by wind tunnel tests.The results show that all of the proposed inversion methods can identify structural fluctuating wind load accurately.The time history and power spectra of wind load obtained from inversion matches well with wind tunnel tests.What is more,the precision of the estimated wind load under different levels of noise satisfy the requirements of practical engineering.Then the inversion methods adviced in this paper are proofed to be feasible and accurate.(5)The field measurements and numerical simulation research on wind load inversion were carried in this thesis.Structural wind-induced response is acquired using vibration monitoring system installed in the Taipei 101 Tower during typhoon Matsa.Based on the field measurements date,the structural fluctuating wind load and unmeasured wind-induced responses was identified by inversion.The results show that the time history and power spectra of acceleration from inversion matches well with field measurements.The inversion wind load conforms to the characteristics of actual situation.Wind loads identified from difffrent layers of input responses are similar with each other.Furthermore,large eddy simulation of Taipei 101 Tower is carried in is this research.Then the structural fluctuating wind load and wind-induced responses is identified based on the numerical simulation date by inversion.Then the applicability and reliability of inversion analysis technique proposed in the dissertation have been examined again.The research shows that the inverse analysis technique for identifing fluctuating wind load on super-tall buildings using several layers' s dynamic response is an effective method to acquire the structural wind load.The presented results in this dissertation provide a new approach for wind load evaluation,and should be valuable for analyzing the interaction mechanism of wind load and super-tall structures,the characteristics of wind-induced response and the development of wind-resistant design for super-tall buildings.
Keywords/Search Tags:super-tall building, wind-induced response, wind load, inverse analysis, Kalman filter
PDF Full Text Request
Related items