Comparative study on using static and dynamic finite element models to develop FWD measurement on flexible pavement structures. (10th July 2018)

• Record Type: Journal Article
• Authors: Hamim, Asmah; Yusoff, Nur Izzi Md.; Ceylan, Halil; Rosyidi, Sri Atmaja P.; El-Shafie, Ahmed
• Is Part Of: Construction & building materials. Volume 176(2018)
• Page Start: 583
• Page End: 592
• Publication Date: 2018-07-10
• Subjects: Finite element -- Falling weight deflectometer -- Flexible pavement; Building materials -- Periodicals; 624.18
• DOI: 10.1016/j.conbuildmat.2018.05.082
• Languages: English
• ISSNs: 0950-0618

Abstract

Highlights: A FE model was developed using static and dynamic analyses in ANSYS program. Transient dynamic analysis is best method for simulating FWD test by using FEM. A 5000 × 5000 mm model geometry is sufficient for developing an FE model. Abstract: The deflection basin obtained through backcalculation analysis is compared with the measured deflection basin to determine the moduli of each pavement layer. Most computer programs use multi-layered elastic theory (MET) to perform backcalculation for determining deflection basin. Other structural analysis techniques, such as finite element method (FEM) and finite difference method (FDM), can be used to model flexible pavement structures when conducting FWD tests. Unlike FEM, MET analysis does not take into account nonlinear materials and dynamic loading. This study aims to develop a better finite element (FE) model by using the static and dynamic analyses in the ANSYS computer program. A comparative study was conducted by using varying sizes of model geometry and different types of elements and sizes to determine how they affect the developed FE model. The results of the analyses show that transient dynamic analysis is the best method for simulating FWD test. The percentage of errors between FE deflection basin and measured deflection basin range between 0.94 and 5.01%. Model geometry of 5000 × 5000 mm is sufficient to produce a good deflection basin which approximates the measured deflection. To ensure the accuracy of the developed model, the information on material properties must be valid. Additionally, finer and higher order elements should be used close to the loading region, for instance four or eight-node quadrilateral element (CAX4 or CAX8) with quadratic interpolation function.