FEATURE ARTICLE SPECIAL ISSUE : LRFD LRFD Practice for Drilled Shafts Drilled shaft construction at Goethals Bridge, NY-NJ Load and Resistance Factor Design (LRFD) of drilled shafts in 2015 is in a state of transition from past practice by Allowable Stress Design (ASD) to the eventual goal of a reliability-based limit state design by LRFD. LRFD offers the potential for achieving drilled shaft designs that meet quantifiable, consistent and safe levels of reliability. As a profession, we have not yet realized the benefits that are possible by implementation of a robust LRFD procedure. This article provides an overview of where we are and where we need to go to reach that potential. Recent Developments A major sector of the drilled shaft market, transportation, made the leap from ASD to LRFD in 2007 through adoption of the AASHTO LRFD Bridge Design Specifications (see cover story). In 2010, FHWA published its most recent version of Drilled Shafts: Construction Procedures and LRFD Design Methods (GEC No. 10, Brown et al.) in which, for the first time, all design methods are presented exclusively in LRFD format. Both documents provide tables with resistance factors for strength, service and extreme event limit states and for all of the AUTHOR various resistance components. Other market sectors, for example the electrical transmission industry, have incorporated LRFD into design recommendations, but those do not have the same weight as code- driven foundation design methods such as those used in public-sector work. Transmission line foundation design is therefore a mix of ASD and LRFD. In commercial work, such as building construction, drilled shaft geotechnical design is still carried out in predominately ASD format, with LRFD beginning to make inroads as a new generation of engineers educated in LRFD is gradually replacing the previous generation. The eventual transi- tion to all-LRFD design is expected to occur in all sectors, but not without growing pains and not uniformly in all industries that utilize drilled shafts. The Promise of LRFD It is generally accepted (but not without some dissent, kicking and screaming) that LRFD provides a more rational framework for design than ASD. Volumes have been written on this topic, but basically, the reasons are: • ASD, in which the computed available capacity is divided by a global factor of safety to determine the allowable (presumably ‘safe’) load acting on the foundation, accounts for all uncertainty and variability in both loads and resistances through a single lumped parameter, the factor of safety. It is therefore not possible to account for the differences between variability and uncer- tainty associated with resistances and those associated with loads. As a result, ASD does not lend itself to a quantifiable measure of safety in terms of the proba- bility that loads will exceed resistances. • Probability-based limit-state design, of which LRFD is one platform, applies separate partial factors to loads (load factors) and resistances (resistance factors). The load factors can be established using probability theory to account for uncertainty and variability associated only with load, a random variable entirely independent of resistance. Likewise. resistance factors account for uncertainty and variability associated with geotechnical and structural resis- tances, independent of load demand. John P. Turner, Ph.D., P.E., Senior Principal Engineer, Dan Brown and Associates, PC, Professor Emeritus of Civil and Architectural Engineering, University of Wyoming DEEP FOUNDATIONS • MAY/JUNE 2015 • 59
