Vertical Load Test on a Drilled Shaft Installed in Glacially Derived Sediments

"There are several methods for estimating the axial capacity of drilled shafts installed in coarse-grained soils. Most methods estimate the ultimate capacity for side friction and tip resistance. Design capacity is then calculated by applying a resistance factor to the ultimate capacity. In reality, however, ultimate resistance may not be mobilized except after a significant displacement that is not structurally tolerated in either compression or tension modes.A full- scale load test is the best way to obtain the load settlement curve for a drilled shaft, especially in soil formations where accurate estimation of soil frictional properties from a field investigation may be difficult. This paper presents the results of a static, axial, compressive bi-directional load test, using the Osterberg method (O-cell test). The test was designed so that the shaft above the O-cell would move up and the shaft below it would move down when loaded. At the start of the test the concrete around the O-cell is fractured generally on a plane near the bottom of the cell.The test was designed to validate the design capacity of drilled shafts embedded into the Vashon recessional outwash (Qvro), which is a glacially derived sediment consisting of thick deposits of cobbly sand and gravel relatively free of silt and clay found in the Seattle, WA area. The test shaft consisted of a 2.5- feet (0.75 m) diameter drilled shaft. The test shaft was installed to a depth of about 47 feet (14.3 m) into this formation. A maximum load of 437 kips (1.95 MN) was applied during the test, in each direction. The maximum load was attained at upward and downward displacements of less than 0.1 inch (2.5 mm), respectively.Although the load test did not yield ultimate values, the results are compared to the design assumptions using methods adopted in the FHWA and AASHTO guidelines, which verified the adequacy of the design assumptions. We conclude that more emphasis is needed in the prediction of the service loads and corresponding mobilized displacement that may lead to a more cost effective design. It is also clear that more testing to higher loads could lead to more efficient designs on specific projects and generally as data is added to the local knowledge base."