

When tests are performed at sites where soil strength data are available, 

 checks of strength estimates from penetrometer data will be available. 



The test that was attempted in 1,200 feet of water indicated that 

 the terminal velocity of the penetrometer is about 80 feet per second. 

 This is 20% lower than the design velocity of 100 feet per second. This 

 velocity can probably be increased by smoothing the welds on the penetrometer 

 body and by reducing the volume of the positively buoyant sound source 

 system. However, a terminal velocity in the range of 80 feet per second 

 should give 30 feet of penetration in soft clay which is about the 

 required depth of the fluke of the CEL 20K explosive anchor to obtain 

 its rated capacity in a soft clay. Such a result is satisfactory. 



Because strength data are not available from the sites where the 

 penetrometer has been successfully tested, penetration estimates calcula- 

 ted from penetration formulas could not be made. Therefore, penetration 

 estimates that were made in the development of the penetrometer cannot 

 be verified. However, because the data that were gathered on a mud flat 

 in San Francisco Bay with model penetrometers compared well to the 

 prediction method, this lack of verification is not of concern at this 

 stage in developing the penetrometer. 



While the testing that has been done demonstrates the workability 

 of the instrumentation concept, there is a factor that affects the 

 physical phenomenon of a Doppler shift of a sound frequency. The relation- 

 ship of this factor - the sound velocity v s - to the frequency shift is 

 shown in Equation 1 . The equation shows that the frequency shift for a 

 given penetrometer velocity is a function of the velocity of sound in 

 the medium in which the penetrometer is traveling as well as the velocity 

 of the penetrometer. Therefore, sound velocity changes must be accounted 

 for to prevent data errors. 



The terminal velocity before impact must be calculated. For 

 water, a change of 300 feet per second in the sound velocity (the greatest 

 expected) will cause a 3.1% error in the frequency shift when a midrange 

 value of 4,800 feet per second is assumed. Hence, there will be a 3.1% 

 error in the calculated terminal velocity. However, for most cases, 

 this error can be reduced to less than 1% by using sound -velocity-depth- 

 latitude charts. If the terminal velocity of each penetrometer is found 

 to be nearly the same, this error can be eliminated and the sound velocity 

 of the water calculated from the terminal velocity and the frequency 

 shift using Equation 1 . 



In soil, large changes in sound velocity are possible. Fortunately, 

 in the soft clays and oozes that comprise most seafloors the sound 

 velocity is within ±2% of that of the bottom water. Hence, the assumption 

 that the sound velocity in these soils is the same as the bottom water 

 will result in only small errors. These soft materials can be identified 

 by deep penetrometer penetration. In sands and stiff clays the sound 

 velocity can be 6 to 7% more than that of the bottom water. These soils 

 can be identified by shallow penetration and the sound velocity adjusted 

 accordingly. This adjustment applies only to penetrations greater than 

 the length of the penetrometer because until the penetrometer is fully 



17 



