An influential American specification for deep foundations is being rewritten in a project supported by the USA Nat iona l Cooperat ive Highway Research Program under the Transportation Research Board of the National Academy of Science.
The project aims to rewrite AASHTO (American Association of State Highway and Transportation Officials) Deep Foundation Specifications for 2001.
These specifications are traditionally observed on all federally aided projects and generally viewed as a national code of US Highway practice, hence influencing the construction of all the deep foundations of highway bridges throughout the US.
The project team, headed by the author, is divided into three major groups broadly dealing with static analyses (University of Florida), probabilistic approaches and structural analyses (University of Maryland) and dynamic analyses (University of Massachusetts at Lowell).
The AASHTO specifications are based on load and resistance factor design (LRFD) principles. In LRFD, partial safety factors are separately applied to the resistance and load components. A factored (reduced) strength (capacity) of a pile is larger than a linear combination of factored (magnified) load effects.
In this format, strength is reduced and load effects increased, by multiplying the corresponding characteristic (nominal) values with factors, which are called strength (resistance) and load factors respectively. The intent of the LRFD method is to separate uncertainties in loading from uncertainties in resistance and to assure a prescribed margin of safety.
The AASHTO specifications (as well as other foundation codes using LRFD) were developed with insufficient data. Judgement combined with back-calculated factors was commonly used in resistance factors evaluation. The main challenges of the project are to compile large, high quality databases and a framework for a procedure and data management to enable LRFD parameter evaluation from probabilistic analysis of data and future updates.
These challenges include two requirements: organisation of the factors following the design-construction-quality control sequence (ie independence in resistance factors according to the chronological stage and the evaluation procedure) and to overcome the generic difficulties of applying the LRFD methodology to geotechnical applications, for instance incorporation of indirect variability (site or parameters interpretation), judgement (eg previous experience), type and level of quality control (in particular small and high strain dynamic testing) as well as other similar factors.
A large database (PD/LT2000) containing information of 389 dynamic measurements and static load tests to failure on 210 driven piles is the backbone of the dynamic methods' performance evaluation. The keynote lecture and paper provide a background for design methodologies and the LRFD. Database PD/LT2000 is presented and analysed. The state of practice and the selected dynamic methods are described, followed by an initial evaluation of the signal matching technique and examination of the parameters that control the accuracy of the dynamic predictions.
Analyses indicated the importance of certain mechanisms associated with the pile penetration (especially soil inertia) and their dynamic simulations. The performance of the dynamic methods is then provided, categorised according to the controlling parameters. The results are used for the development of probabilistic based resistance factors utilising FORM (First-Order Reliability Method), compatible with the parameters employed for the superstructure design. The obtained resistance factors are to be recommended for the new AASHTO specifications.
Figures 1 and 2 present examples of the obtained analyses in the form of histograms and frequency distributions for the performance of all the dynamic analyses utilising wave matching (CAPWAP) and the simplified energy approach, respectively.
Rational codes based on solid data and a probabilistic approach allowing for a prescribed risk are a major development that would bring immediate change in separating fact from fiction and in the long run, a substantially enhance quality control techniques such as the dynamic methods.
Samuel Paikowsky, Geotechnical Engineering Research Laboratory, Department of Civil and Environmental Engineering, University of Massachusetts, USA.
Paikowsky SG and Stenersen KL, The performance of the dynamic methods, their controlling parameters and deep foundation specifications. To be published in Stress Wave 2000.