Treatment of modelling uncertainty of NLFEA in fib Model Code 2020

Abstract

When non‐linear finite element analyses are used in design of new or assessment of existing concrete structures, one should account for the modelling uncertainty before conclusions are drawn based on the results. The present article describes the basis for how this topic is treated in the draft of fib Model Code 2020. There are two components of the modelling uncertainty: (i) within‐model, and (ii) between‐model. The within‐model uncertainty was estimated from a range of series of benchmark analyses gathered from the literature. Each series was assumed analyzed in a consistent manner, termed as the solution strategy. The main result for the within‐model component, is a set of prior parameters for Bayesian updating. In an application setting, the prior parameters should be updated after validating the selected solution strategy with a suite of relevant benchmark analyses, before using the mean and coefficient of variation in the Global Factor Method or calculating the modelling uncertainty factor γRd to be used in the Partial Factors Method. Finally, results from blind‐prediction competitions were studied. With the present data, it cannot be concluded that a‐priori knowledge of the experimental outcome reduces the uncertainty in the prediction. Hence, there is no need for compensating for this effect by a separate partial factor.

Structural reliability of ultra‐high‐performance fiber reinforced concrete beams in shear

“…The effect of errors in such numerical paradigms on the model uncertainty needs to be assessed. Based on a previous study, 50 uncertainty in NLFEA models for less demanding four‐point bending shear tests is characterized by a Lognormal probability density function with a mean value of 1.0 and a CoV of less 0.1. Furthermore, Table 3 presents NLFEA uncertainty with a coefficient of variation of 0.01.…”

Nonlinear analysis procedures for safety assessment of existing RC bridges under traffic loads

“…This is identified by the mean μθ and the coefficient of variation Vθ, which can be estimated considering different informative or non-informative priors following a Bayesian approach [10]. An informative prior for solution strategies for RC structures was proposed in [11] following an analysis of several numerical models from the literature. The procedure develops as follows:…”

Rotational springs for localized crack modeling in structural concrete

“…The mechanical and numerical models described in this chapter are used in two different cases: the experimental validation (Chapter 4) and the cut and cover tunnel case study (Chapter 5). Table 1 summarizes the most relevant parameters of the solution strategy 21 regarding kinematic compatibility and equilibrium, consisting of (i) parameters concerning the global FE solution with iterations over several solution steps until global equilibrium is reached, and (ii) local iterations in the spring elements for each solution‐ and substep. The former comprises the average mesh sizes, the number of substeps (number of load increments in nonlinear FE solution) in each solution step and the acceptable range for , denoted .…”