Civil Engineering July 2022 | Vol 30 No 6

Civil Engineering July 2022 31 γ e = E d E c (7) In the design of structures, γ Ed , M is typi- cally assumed as 1.12 for unfavourable variable actions (fib 2016). EN 1990 (CEN 2002) specifies a range for γ Ed , M between 1.05 and 1.15 which can vary according to a country’s national annex. PARTIAL FACTORS By extrapolating to β = 3.5 for ULS and β = 1.5 for SLS, the following reliability based partial factors are determined. Tables 1, 2 and 3 (page 30) show the reli- ability based partial factors for hogging on two span structures and sagging and shear on single span structures. CONCLUSIONS There are two important observations when the reliability based partial factors are studied: 1. The reliability based partial factors for SLS are smaller than 1.0 2. The reliability based partial factors for ULS are almost insignificant. A β = 1.5 for SLS translates to a return period of 435 years. This is smaller than the 975 years assumed for characteristic loads (see section 1) and hence this leads to reliability based partial factors smaller than 1.0. This indicates that the return period (or quantile) for characteristic loads should be reduced at least to the return period for SLS. Without model uncertainty, this should lead to a PF = 1.0, which is generally accepted as the conven- tion in design codes. The GEV distributions fitted to the daily maxima load effects all tended towards a Weibull (Type 3) EV distribu- tion which is a bounded distribution. Due to this bounded nature, the difference between load effects at a return period of 975 years for characteristic and 5 040 ( β = 3.5) years for ULS is small. It can be said that a PF of close to unity implies very little uncertainty in the load effects. This implies that a much higher level of reliability can be achieved through a negligible increase in the PF. With low uncertainty in the loading indicated by the small reliability based PFs, the assumed FORM sensitivity factor for α E = –0.7 the load should be revis- ited, as it is possible that the resistance contributes more to the uncertainty than assumed. If the characteristic load effects are at the bound of the fitted distribu- tions, then α E ≈ 0.0 for the static load and α R ≈ 1.0, which implies that almost all of the reliability-based uncertainty in the calibration is located in the resistance. This paper highlights two important observations which should be considered in the reliability calibration of design codes. ACKNOWLEDGEMENTS This paper is reproduced from the fib Symposium Proceedings in Lisbon, Portugal (2021) with permission from the International Federation for Structural Concrete ( fib ). New Trends for Eco- Efficiency and Performance, page 1994 – “Reliability calibration of a bridge traffic load model”.  REFERENCES A complete list of references can be ob- tained from the author. Pile Integrity Tester (PIT-Q) • Reveals potential shaft or pile defects such as major cracks, necking, soil inclusions or voids • Can be used to determine unknown pile lengths • Optional PIT-Professional reporting software allows advanced modeling and analysis of shaft and cage alignment • Available in three (cabled or wireless) versions: velocity only, force and velocity, or two velocity channels To learn more about PIT-Q, visit www.pile.com. PIT-S Software