Simulating Response of Different Lightweight Deflectometer Testing Using Finite Element Modeling

Several state transportation agencies have been moving from a density-based to a stiffness-based approach for quality control/quality assurance (QC/QA) field-testing for determining compaction quality of all types of earthwork including pavement layers. Light-weight deflectometers (LWDs) are being used increasingly by these agencies for rapidly determining geomaterials’ surface moduli. LWDs can be currently procured from a number of manufacturers. American Society for Testing and Materials (ASTM) contains two sets of specifications (ASTM E2835 and E2583) for LWDs. Recent empirical studies have shown that different makes of LWD consistently yield different soil responses, and consequently, different stiffness values. As such, the sources of these device-related differences have to be understood and reflected in a generic compaction specification. The two ASTM types of LWDs were modeled on top of a layered pavement using a finite element (FE) model. Since one of the highest stresses typically experienced in the life of a subgrade layer is during the LWD testing, a nonlinear geomaterial model was incorporated in the model. Despite the fact that the analysis of the LWD data is usually done using elasto-static half-space models, the load is applied dynamically. As such, the model also considers the dynamic nature of the LWD testing. The goal of this paper is not to explain the nature and details of the FE models; rather to describe in a practical manner the implications of selecting a certain type of LWD on the measurements made. It was found that LWDs manufactured by different vendors are not interchangeable. Any specification should be clear about what type and make of LWD should be used in a project and consider target deflection/modulus for acceptance accordingly. In addition to the make of the LWD, the nonlinear model parameters of the geomaterial significantly influence the responses of the LWDs. The depth of influence of the LWD, which is particularly important for testing layered systems, is also dependent on the make of the LWD. The study also confirms that because of the dynamic nature of the LWD loading, the depth of influence of the LWD is between 2 to 4 times the plate diameter, which is deeper than those usually reported based on numerical elasto-static analysis or measurement of the soil pressure in the field.