The procedure has been compared to the results of a series of model 

 dynamic penetration tests performed in a mud flat on San Francisco Bay 

 near Mare Island (Figure 2) . The test models in three sizes were configured 

 similar to the penetrometer; they attained impact velocities ranging 

 from about 70 to 90 feet per second. The figure shows that the penetration 

 prediction method gives reasonable estimates of penetration for long 

 cylindrical objects entering the soil with their axis vertical. 



Terminal velocities were obtained from standard hydrodynamic 

 equations. The drag coefficient used for all the penetrometers in the 

 parametric study was 0.5 as determined by the Naval Surface Weapons 

 Center, White Oak, from model studies in their 100-foot-deep test tank as 

 reported by Waser (1973). This drag coefficient should have been adjusted 

 as the length-to-diameter ratio of the penetrometer was changed. However, 

 for initial results this was deemed unnecessary because it was not 

 expected to vary significantly over the range of penetrometer sizes 

 being studied. Later, when the limits of the size of the penetrometer 

 had been narrowed, drag was separated into a frictional component on the 

 side and a pressure component on the nose and base to obtain more accurate 

 estimates of terminal velocities and resulting penetration. 



The initial calculations were made for three soil profiles: a red 

 clay, a calcareous ooze, and a terrigenous clay. Penetration trends 

 were consistent from soil to soil. Therefore, when the study's limits 

 were narrowed, penetrations were calculated in only one soil, a pelagic 

 clay. 



The parametric study showed that steel-shafted penetrometers were 

 about half as efficient as lead-filled steel pipe penetrometers (a 

 steel-shafted penetrometer equal in size to a lead-filled one would 

 penetrate about half as far) . For this reason lead-filled steel pipe 

 vehicles were chosen for the penetrometer. 



Other data from the initial parametric study were that the optimum 

 vehicle size was between 200 and 500 pounds with a diameter between 2 

 and 4 inches. This size range of penetrometers was then studied in 

 greater detail. In this continued parametric study, friction and 

 pressure drag were considered separately so that improved hydrodynamic 

 drag coefficients could be utilized. Standard pipe sizes were used. A 

 plot of penetrations into a pelagic clay versus pipe diameter is shown 

 in Figure 3 for penetrometers of equal weight and equal length. A 10- 

 pound instrument package of constant volume was attached to each of the 

 hypothetical penetrometers. 



Figure 3 shows that for a given penetrometer weight that penetration 

 will be nearly constant over the range of penetrometer diameters being 

 considered. Therefore, length and other factors govern the design 

 selection. With 35 feet of penetration in a pelagic clay as a target, 

 the weight of the penetrometer should be about 350 pounds. At this 

 stage the instrumentation design had progressed to hardware component 

 selection. A trade-off between vehicle diameter and the diameter of an 

 acoustic projector resulted in the selection of 3.5 inches as the 

 diameter of the penetrometer. Therefore, the penetrometer would be 10 

 feet long, weigh about 365 pounds, and be 3.5 inches in diameter. The 

 vehicle would compose the lower 8 feet of the penetrometer and weigh 355 

 pounds. 



