STP1198: A Pavement Evaluation Procedure Incorporating Material Non-Linearity

    de Almeida, JR
    Research assistant, University of Nottingham, University Park, Nottingham

    Brown, SF
    Professor, University of Nottingham, University Park, Nottingham

    Thom, NH
    Lecturer, University of Nottingham, University park, Nottingham

    Pages: 15    Published: Jan 1994


    Abstract

    Two new computer programs, LEAD and FEAD, have recently been developed at Nottingham for the back-calculation of elastic stiffnesses of pavement layers from deflections measured with the Falling Weight Deflectometer (FWD). These programs represent a significant improvement on the program PADAL, previously developed at Nottingham. A revised algorithm for the iterative back-calculation process, based on the Gauss-Newton method, has been successfully implemented. Appropriate stress-strain relationships for the granular layers and the subgrade were built into the programs, enabling the non-linear behaviour generally exhibited by these materials to be evaluated. The influence of overburden, pore pressures and “locked-in” horizontal stresses on the in situ stiffnesses is also considered. Additionally, a rigid bottom may be included in the analysis. The main difference between the two codes is in the sub-routines adopted for the computation of deflections. LEAD makes use of layered elastic analysis, while FEAD uses a finite element approach which, although more accurate for non-linear problems, is more time consuming. Data from FWD surveys have provided a means of validating the new software. Examples are presented, including comparisons with laboratory test data. The results suggest that the stress-dependent nature of the lower layers can be of significance for the overall behaviour of the pavement.

    Keywords:

    elastic stiffnesses, deflections, non-linear models, back-calculation algorithm, FWD data, validation


    Paper ID: STP18150S

    Committee/Subcommittee: D04.39

    DOI: 10.1520/STP18150S


    CrossRef ASTM International is a member of CrossRef.