No. 33. Concrete Engineering for excellence and efficiency - 10th fib Symposium Proceedings - Vol. 1. - 8-10 June 2011 - Prague, Czech Republic

Although the 2010 draft of the fib Model Code is, overall, an excellent document, it has some weaknesses that could and should be corrected. One concerns the material model for creep and shrinkage prediction, which is not only theoretically obsolete but also difficult to justify in view of the “wake-up call” provided by recently collected long-term deflection measurements of 56 bridges.

The second issue is a size-effect formulation for shear strength whose justification is closer to a myth than to sound mechanical reasoning. The third concerns the probability distribution of structural strength for failures occurring at macro-crack initiation. The tail of the distribution that matters, at probability levels on the order of 10⁻⁶, is directly unobservable—much like a black hole. However, indirect evidence from size-effect studies indicates that the distribution transitions from Gaussian to Weibull as structural size increases. This transition may have a significant effect on the safety factor.

Moreover, questionable use of the lognormal distribution, together with another “black hole” in the implicit safety factors embedded in design formulas, makes a meaningful calculation of failure probability impossible. Some possible remedies are suggested.