Bayesian-based prediction and real-time updating of axial deformation in high-rise buildings during construction

Abstract

Estimating the axial deformation of concrete vertical members in a structure is challenging due to the shrinkage and creep characteristics of concrete together with the uncertainties involved in the construction process. To address the resulting inaccuracies, this paper presents a novel method that employs Bayesian theory to enhance the predicted results by incorporating measured strain from vertical members and equations for time-varying properties of concrete. The following five uncertain parameters are considered: environmental humidity, concrete strength, load transfer, concrete shrinkage calculation errors, and creep equation errors. The proposed method was experimentally evaluated on a twin tower high-rise building, with results showing that even with limited measured strain information, the uncertainty of the predicted results was significantly reduced and the predicted values were aligned closely with the measured values. The study also analyzed the impact of vertical deformation on the elevation deviation of adjacent wall-column members and the connected parts of the two towers.