Soil Disturbance . Soil disturbance affects both laboratory and 

 in-situ vane shear test results. However, disturbance effects are 

 considered to be more severe in laboratory testing. Laboratory vane 

 strength data are affected by sampling disturbance and also by disturbance 

 caused during vane insertion, while soil disturbance during in-situ vane 

 testing is limited to disturbance caused by vane insertion. The degree 

 of soil disturbance caused by vane insertion is related to the area of 

 the vane divided by the area of a cylinder generated when the vane is 

 rotated about its axis. To minimize soil disturbance due to vane 

 insertion, all laboratory and in-situ vanes have area ratios less than 

 the recommended maximum value of 0.15 (Brand, 1967). 



It is probable that the greatest errors in laboratory testing are 

 generated by soil disturbance which originates from poor sediment sampling 

 techniques. In the field of seafloor soil mechanics, it has been 

 difficult to obtain good quality sediment samples, primarily because 

 all of the samplers were designed for geological investigations with 

 little emphasis on preserving the sediment engineering properties. 



Friction on the Torque Rod . The in-situ vane shear assembly was 

 designed to minimize the effect of friction or adhesion on the torque 

 rod. The torque rod diameter is small in relation to the vane width and 

 the torque rod length between the vanes and transducer is small. However, 

 the torque rod lengths are such that the transducer does not interfere 

 with the failure surfaces generated by vane rotation. The torque 

 produced by friction or adhesion on the torque rod is estimated at less 

 than 2 percent of the total torque. Therefore, the actual vane shear 

 strengths may be 2 percent less than those measured with the NCEL test 

 device. A similar error is expected in the laboratory vane shear test 

 results because a portion of the torque rod is immersed in the soil 

 during shear. 



Since laboratory and in-situ test results are both somewhat equally 

 affected by friction on the torque rod, the discrepancies in the strength 

 data are probably not due to torque rod friction. Also, it is most likely 

 that data scatter is sufficient to mask a 2 percent error in strength data. 



Correlation of Vane and Cone Results 



A comparison of the cone penetrometer and vane shear test results 

 is presented in Figure 23 for Site 1 and Site 2; the ratio of average 

 unit cone load to average vane shear strength versus penetration depth 

 is plotted. The data in Figures 17 through 19 were reduced to average 

 unit cone load and average vane shear strength at 10-inch penetration 

 intervals. Similar experimental data have been presented in the 

 literature. Skempton (1951) found that the immediate bearing resistance 

 of deep circular or square footings (the ratio of depth of footing 

 embedment to footing width is greater than 6) on clay was approximately 

 nine times the vane shear strength. However, actual bearing capacity 

 failures in the field have shown this ratio may be as low as five (Lambe 

 and Whitman, 1969), and empirical static cone penetrometer results of 

 Begemann (1963) demonstrate that this ratio may be as high as 13.4. 

 For similar tests on granular soils, higher ratios are expected. 



15 



