Journal of South China University of Technology (Natural Science Edition) ›› 2007, Vol. 35 ›› Issue (10): 70-77,104.

• Electronics, Communication & Automation Technology • Previous Articles     Next Articles

Crack Statistic-Based Method for Bridge Performance Assessment

Han Da-pan  Du Jiang   

  1. Dept. of Civil Engineering , South China Univ. of Tech. , Guangzhou 510640 , Guangdong , China
  • Received:2007-03-02 Online:2007-10-25 Published:2007-10-25
  • Contact: Han Da-jian (1940-),female , professor , doctoral supervisor , mainly researches on the construction control and wind-induced vibrations of large-span cable-stayed bridge aswe11 as bridge management E-mail:djhan@scut. edu.cn
  • About author:Han Da-jian (1940-),female , professor , doctoral supervisor , mainly researches on the construction control and wind-induced vibrations of large-span cable-stayed bridge aswe11 as bridge management system.
  • Supported by:

    Guangzhou Scientific and TechnologicalProject (2002Z3-D3031)

Abstract:

This paper proposes a new method for the performance assessment of concrete bridge , in which the capacity or the residual capasity of a bridge is estimated according to the statistic quantities of all the cracks on the bridge surfaces. The finite element numerical analysis instead of "real experiments" is performed to simulate the concrete cracking process , and the simulated experiments are designed to be orthogonal to reveal the
multivariate relationships between the bridge performance and the cracks. In these experiments , the importance of different factors affecting the multiple relationship is determined through a screening analysis , and the rela tionship between the performance and the cracks is illustrated by seuing up a regression model. Experimental results show that the regression model with fewer factors can yield accurate and reliable assesment results.Thus , the proposed method can be used as an effective means for the periodic assessments in bridge management system due to its feasibility , convinience and low cost.

Key words: performance, assessment, crack, statistic, simulated experiment, experiment design