Theoretical Probabilistic Nonlinear Analysis of Post-Tensioned Bridge Cross-Sections with the Application of Random Fields

Abstract

The paper presents a proprietary procedure for the analysis of normal stress distributions in post-tensioned cross-sections. It has a significant advantage over conventional commonly used approaches based solely on the envelope analysis as it provides stress levels in all components of the cross-section. The procedure was used in a series of probabilistic analyses with the adoption of random fields. These fields represented uncertainties in strain-stress relationship in concrete. The analysis covered several types of cross-sections and several types of random fields. Key observations from the conducted simulations are as follows: (I) the widest ranges of the probable maximum stresses (i.e. the lowest indexes of reliability) were obtained for sections with relatively low heights of the compressive zone. (II) The highest probabilistic sensitivity to the type of random field used was found in tall sections with a relatively large compressive zone. (III) The greatest sensitivity to batch uncertainties was evident in all cross-sections when using squared exponential random fields. (IV) The greatest relative sensitivity to the batch uncertainties in the form of the random field compliant with the guidelines of the Joint Comity of Structural Safety (JCSS) was evident in the analyses of the tallest cross-section corresponding to the incrementally launched bridges.