Please use this identifier to cite or link to this item:
|Title:||Fatigue reliability assessment of steel bridges instrumented with structural health monitoring system|
|Keywords:||Iron and steel bridges|
Railroad bridges -- Reliability
Steel, Structural -- Fatigue.
Hong Kong Polytechnic University -- Dissertations
|Publisher:||The Hong Kong Polytechnic University|
|Abstract:||Fatigue is among the most critical forms of damage potentially occurring in steel structures, and therefore it is essential to assess the fatigue damage status as well as the remaining fatigue life of steel bridges. As a paradigm of integrating structural health monitoring (SHM) data into practical structural condition assessment, the continuously measured dynamic strain data from a long-term SHM system are significantly useful in assessing the fatigue life and reliability of steel bridges. The research presented in this dissertation has been devoted to investigating fatigue life assessment and reliability analysis of welded connections of steel bridges using long-term monitoring data from the permanently installed bridge health monitoring system (BHMS). A monitoring-based stress-life fatigue evaluation method is firstly developed for deterministic fatigue life assessment of steel bridges by use of long-term measured dynamic strain data from an on-line SHM system. The proposed method takes the benefits of resemblance and consistence in daily stress spectra obtained under normal traffic (highway and railway) and wind conditions, and accounts for predominant factors which affects the prediction of fatigue life, such as the diversity of daily strain data subjected to diurnal and seasonal traffic variations and the typhoon effect. With the continuously measured dynamic strain data from the instrumented Tsing Ma Bridge (TMB), the proposed method is exemplified to evaluate the fatigue life of the failure-critical welded details of the bridge. A combined method of finite mixture distributions and hybrid mixture parameter estimation is proposed to formulate the multi-modal probability density function (PDF) of the stress spectrum acquired from the monitoring data. This method is capable of modeling complex probability distributions and enables the statistical modeling of random variables with multi-modal behavior where a simple parametric model fails to adequately represent the characteristics of observations. The rainflow counting algorithm is utilized to obtain the rainflow stress matrix including the information such as stress range, mean stress, and number of cycles. The method of finite mixture distributions is applied to formulate the PDF of the stress range, while the joint PDF of the stress range and the mean stress is derived by means of a mixture of multivariate Weibull-normal distributions. The formulated PDF fits the measurement data fairly well and thus provides a crucial basis for fatigue reliability evaluation.|
The determination of stress concentration factor (SCF) and its stochastic characteristics for a typical welded bridge T-joint is perceived by conducting full-scale model experiments of a railway beam section of the TMB. The strain data at the pre-allocated measuring points are acquired and the hot spot strain at the weld toe is determined by a linear regression method. The SCF is then calculated as a ratio between the hot spot strain and the nominal strain which is obtained directly from the strain gauge. To fully account for the effect of predominant factors on the scatter of SCF, the experiments have been carried out under different moving load conditions. The statistical properties and probability distribution of SCF are achieved, which reveals that the SCF for the welded steel bridge T-joint conforms to a normal distribution. Two reliability-based fatigue life assessment methods are proposed. In the first method, the stress cycle of specific stress range is treated as a random variable. The statistics obtained for all concerned stress ranges is combined with the S-N relationship along with the Miner's damage cumulative rule to conduct a probabilistic fatigue life assessment. In the second method, a fatigue reliability model integrated with the probability distribution of hot spot stress range is formulated based on the S-N relationship. The joint PDF of the hot spot stress range is constructed from the obtained probability distributions of the stress range and the SCF. The two methods have been applied for probabilistic fatigue life assessment of critical welded details at the TMB by use of the long-term monitoring data, and the mean value and the standard deviation of the fatigue life as well as the failure probability and reliability index versus fatigue life are achieved. In summary, the research addressed in this dissertation chiefly contributes to the development of a systematic framework for S-N approach-based fatigue reliability assessment of welded joints of steel bridges making use of long-term monitoring data. Aiming to narrow the gap currently existing between SHM technologies and structural condition assessment practices, the approach developed in this PhD study is capable of performing fatigue life and reliability evaluation with taking into account uncertainty and randomness inherent in the nature of fatigue phenomenon and measurement data. Following this approach, reliable fatigue condition assessment can be achieved for instrumented steel bridges and rational strategies on bridge inspection and maintenance can be executed in accordance with the correlativity between reliability indices and predefined inspection and/or maintenance actions.
|Description:||xxvi, 227 leaves : ill. ; 30 cm.|
PolyU Library Call No.: [THS] LG51 .H577P CSE 2010 Ye
|Rights:||All rights reserved.|
|Appears in Collections:||Thesis|
Show full item record
Files in This Item:
|b2393069x_link.htm||For PolyU Users||162 B||HTML||View/Open|
|b2393069x_ir.pdf||For All Users (Non-printable)||4.1 MB||Adobe PDF||View/Open|
Checked on Dec 4, 2016
Checked on Dec 4, 2016
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.