Identification of Relevant but Stochastic Input Parameters for Fatigue Assessment of Pre-Stressed Concrete Bridges by Monitoring

Abstract

Abstract. Since in the middle of last century when most bridges were erected in Central Eu These joints are in danger of cracking since reinforcements provided according to former code versions at the times of erection do not meet nowadays demands. If the joints crack, concrete releases tensile forces left to be borne by the tendons only. Consequently, stress amplitudes caused by traffic rise significantly along with the fatigue damage induced per load cycle. Damage progress is usually tracked employing S-N-curves and Miner's rule. It is known that uncertainty in the stress amplitude is magnified by the S-N-curve with respect to the number of cycles to failure [1]. To determine stress amplitudes in a tendon one can measure strains in-situ by structural monitoring or predict its size by numerical simulation [2]. Both alternatives highly depend on the quality of input data [3,4]. Hence, the focus has been set to determine the most relevant contributors to precisely predict remaining structural lifetime.Among the scattering input variables assessed by sensitivity analyses are geometrical, material and load parameters, whereas the latter ones turned out to be decisive. Advice to match the findings with structural monitoring measures scheduled for existent bridges is given.

Lifetime predictions of prestressed concrete bridges—Evaluating parameters of relevance using Sobol' indices

“…The initial traffic amount was set to 3.5·10 6 trucks per year. For each time‐step, the relevant temperature gradients ΔT i were accounted for Reference 42, which were idealized as inner constraints in corresponding frequencies 9 . In total, five relevant temperature gradients were combined to five truck‐loads defined in FLM 4.…”

“…They may even change over time. 9 To consider all these aspects, a comprehensive numerical model was developed, which is presented here.…”

“…Since the bearing behavior of prestressed concrete is significantly nonlinear after cracking, the relevant parameters are not known beforehand. They may even change over time 9 . To consider all these aspects, a comprehensive numerical model was developed, which is presented here.…”

“…[2][3][4] Structural monitoring of key parameters [5][6][7] might extend the residual lifetime significantly by reducing the uncertainty of the input parameters of the prognosis. 8 The crucial question concerns the choice of the right parameters to measure, 9 which shall be answered by the results of this paper.…”