Probabilistic investigation of transmission length formulations considering model uncertainties

Abstract

Transmission length is a key parameter for detailing and assessing shear and anchorage resistance of pre-tensioned concrete members, and according to the main design codes, semi-empirical models are adopted to estimate its value. In these approaches, model uncertainty may affect the results and often plays an important role in achieving the desired reliability. The purpose of this study is to investigate the current code design formulations for the transmission length described in the upcoming Eurocode2:2020 and fib Model Code 2020, and to propose new coefficients to consider the model uncertainties. To this end, a database with 951 experimental measures of transmission length was collected from the existing literature and properly filtered to obtain a homogeneous sample of 561 experimental observations on which model formulations were applied. The uncertainties of the two models were estimated and suitable probability models were derived. Results show that the prestress force release methodology strongly affects the uncertainties, both in terms of mean value and dispersion of the prediction. Finally, new probabilistic coefficients were proposed to adjust the formulations and ensure the achievement of the target reliability length to be adopted in each design situation, both at the serviceability and ultimate limit states.