Impact assessment of layered granular materials

Granular materials utilised in the construction of highway foundation layers are currently specified on the basis of index tests. As a consequence, the material acceptability criteria, although developed from many years' experience, do not directly measure a fundamental performance parameter. Once the granular materials are placed and compacted they are rarely checked and as such no assurance can be given to their likely engineering performance in situ. An important performance parameter, the stiffness modulus, describes the ability of the constructed layer(s) to spread the construction (and in-service) vehicle contact pressures and reduce the stresses, and hence strains, transmitted to the lower weaker layers. A significant improvement upon current practice would be to include the specification of 'end product' testing and to include the direct measurement in situ of stiffness modulus to assure performance. A prerequisite of this is suitable site equipment to measure such a parameter, and a sound basis upon which to interpret and utilise such data. Tests do exist that measure stiffness modulus in situ, although in general they measure a 'composite' stiffness, i.e. a single transducer infers the surface strain, under controlled loading, for the construction as a whole and the region affecting the measurement is not precisely known. Currently then, no routine portable device exists for the direct stiffness modulus assessment of the near surface or last layer applied. This would not only provide for consistency of construction, but avoid burying poor or weaker layers. This thesis describes the evaluation of a portable impact test device and research into the behaviour of granular soils subject to rapid transient loads. The requirements for the assessment of pavement granular foundation layers are reviewed, followed by a critical appraisal of current devices that measure the stiffness modulus of material in situ. The prototype impact device, known as ODIN, comprising an accelerometer instrumented swinging hammer, is described. A selection of field data, demonstrating the primary soil influencing factors and correlations with other devices, is presented. Controlled laboratory testing is also described, comprising impact testing with free-falling masses in addition to the ODIN device and for tests on foundations instrumented with pressure cells, that further explains the dynamic behaviour of the material under test. Problems with both hardware and software, associated with high-frequency impact testing are highlighted. In particular, the restraint of the impact mass by the swinging arm mechanical component is observed to lead to a proportion of the impact energy being channelled back into the apparatus during a test. The channelled energy is shown to produce resonance of the apparatus, which in turn leads to problems in interpretation of the accelerometer signal. Numerical methods are then explored and it is demonstrated that the predictions approximated well to the free-falling weights experimental data. Discussion of the research findings concludes with a model for soil behaviour under impact testing, requirements for an improved impact device and the further research work required to realise the potential of such equipment.