Plate Anchors in Sand Under Static and Cyclic Loads

This thesis reports on an investigation into the behaviour of circular plate anchors embedded in dry cohesionless soil and subjected to vertical static or cyclic uplift loading. The experimental part of the investigation used model testing techniques, and details of the test rigs, sand, sand placement method and test procedures are given in Chapter 3. A total of thirty static tests were completed in dense (Dr=93.0%) and medium-dense (Dr=59.4%) Leighton-Buzzard sand, using anchor embedment ratios ranging from 2.0 to 15.0. The anchor embedment ratio is the anchor depth (D) divided by the anchor diameter (B). Usually at least two tests were performed at each embedment ratio. The anchors consisted of 6mm thick brass discs, with diameters ranging from 25mm to 100mm. The anchor shaft was a length of smooth brass rod, 6mm in diameter. The static test results established a data base of anchor failure loads for use in setting the load levels in the cyclic tests. The results for dense sand (D/B8, the results were seriously affected by boundary and scale effects, leading to substantial differences in the dimensionless uplift resistance factor, Nu, for the same embedment ratio. The static test results were also used for comparison with the results of a finite element analysis of the anchor uplift problem. The analysis confirmed two characteristics of shallow anchor behaviour : the presence of an elastic wedge of sand above the anchor and the inverted frustum shape of the failure surface in the sand. With respect to ultimate uplift resistance, the finite element analysis predicted failure loads of up to three times the experimental values. The cyclic tests were undertaken principally to investigate cyclic creep, the mechanism whereby the anchor sustains a continuing upward displacement during cyclic loading. An anchor embedment ratio of 4.5 was used in all fourteen cyclic tests. Sinusoidal loading with a frequency of approximately 0.1 Hz was applied to the anchor. The load parameters varied were mean load and amplitude of load, both expressed as a percentage of the static failure load (sfl). Various combinations of mean and amplitude were applied to the anchors during the test series, with some anchors subjected to over 1 million cycles of loading. The cyclic test results show that, with respect to cyclic displacement, the load amplitude is the controlling parameter : the greater the load amplitude, the greater the cyclic displacement, in tests loaded to the same maximum load. In tests with different mean loads but the same load amplitude, the anchor cyclic displacements were similar. A reduction in sand density leads to an increase in cyclic displacement, for anchors subjected to the same relative loading levels. The test results also indicate that the cyclic creep mechanism is affected by attrition of the sand grains in the vicinity of the anchor. For the anchors which failed during cycling, the failure mechanism can be described in terms of the behaviour of simple shear samples of sand subjected to cyclic loading. Design considerations regarding the offshore deployment of plate anchors are discussed