Abstract
Takadate loam and Hachinohe pumice, which are two typical volcanic soils distributed widely in the Hachinohe district of Japan, were selected for this investigation. The materials were used in laboratory compaction tests with a rammer under various methods of sample preparation and a given compactive effort. Unconfined compression tests were performed on the compacted soils. The sample preparation methods were categorized into the following four : Method-A (a process which adds moisture and reuses material), Method-B (a process which adds moisture and uses new material), Method-C (uses a drying process and reuses material) and Method-D (uses a drying process and uses new material). Method-D presented the most reasonable results in the laboratory for the real nature in-situ. Different sample preparations produced different curves for the relations of qu, ⋴f, E50 and the moisture content. The relations for qu, ⋴f, E50 and air void ea, however, proved to be unique, irrespective of the method of sample preparation used for both types of soils. Thus, application of the air void in place of the moisture content was recommended as a reasonable parameter for quality control of the rolled fill in this district. It was also clarified that the relation of E50/qu and ⋴f could always normalize the uni-axial characteristics of the two tested materials, irrespective of the method of sample preparation or soil type.
