1. Prepare and cure each sample. 



2. Place the sample in the test drum and apply a confining pres- 

 sure of 20 psi. 



3. Drive the model pile into the sample. 



4. Perform a pullout test. 



5. Subsample specimens at various depths and distances from the 

 soil-pile interface. 



Two methods of sample preparation, tamping, and raining were used 

 in this investigation. Both methods were based on the procedure used in 

 the previous behavior test to achieve an in-place low and high void ratio 

 density. 



All prepared samples were cured in a 100% humidity room for 3 days 

 before they were placed in the test drum. This curing period was deter- 

 mined on the basis of unconfined compression tests that indicated that 

 3- and 7-day strengths of the samples were about the same. After curing, 

 the test sample was placed on top of a prepared bed of silica sand in 

 the test drum. The space between the test sample and the drum wall was 

 then filled with silica sand compacted in layers with a density similar 

 to the test sample to maintain compliance. The drum was then covered 

 and all pressure lines connected. A hydrostatic confining pressure of 

 20 psi was applied to the sample. 



The pile driving assembly was attached after the sample was in place 

 and the confining pressure applied. Pile driving was done by manually 

 dropping a 15-pound deadweight 12 inches above the pile head. Pile 

 driving continued until the model pile penetrated 12 inches. Pullout 

 tests were performed with the hydraulic pulling system. The pile was 

 connected to the vertical ram of the hydraulic actuator and pulled ver- 

 tically at a speed of about 0.04 inch per second. The pullout test con- 

 tinued until a maximum pullout force was reached. The pullout resistance 

 force and pile displacement were continuously monitored on an X-Y-Y- 

 recorder . 



It was hypothesized that the degree of grain crushing, if any, would 

 be more severe at or near the pile and decrease significantly as the 

 distance away from the pile increased. It was necessary, therefore, to 

 obtain soil specimens at or near the pile wall during this test. Nine 

 soil specimens were taken from each test sample at three horizontal dis- 

 tance intervals of to 1/4, 1 to 2, and 2 to 3 inches from the pile 

 perimeter and at three depth intervals between to 4 , 4 to 8, and 8 to 

 12 inches from the soil surface in each horizontal distance interval. 



Specimens from the cemented samples were analyzed for grain size. 

 In order to do this, the cementing agent and sand grains would have to 

 be separated to assess the changes in grain size distribution. The 

 regular separation method of using physical or mechanical forces was not 

 acceptable since it could break and crush the soft calcareous sand grains. 

 Chemicals could not be used because they could dissolve the calcareous 

 sand grains as well as the cement agent. Instead, the calcareous sand 

 and the cement agent was separated by hand. Extreme care was taken to 



16 



