Four tests of the vibratory anchor were conducted. In the first 

 test the anchor embedded 9 feet 6 inches. A short term to breakout 

 test load was applied. Six inches of penetration was lost in "keying" 

 the fluke. The peak anchor line tension measured was 58,000 pounds. 

 One higher tension occurred but the amount was not recorded because 

 it exceeded the scale of the recorder unit before adjustments could be 

 made. However, at the time of occurrence, the anchor fluke was 

 displaced 20 inches and subsequently worked out of the bottom under 

 lower loads. Typical loads and accompanying displacements are shown 

 in Figure 23. 



In the second test, thirty-five minutes of vibration produced only 

 5 feet 6 inches of penetration. Then, a sharp increase in the current 

 to the vibrator indicated that a short had occurred in the electrical 

 circuit. A 9,000 pound cable tension pulled the anchor free. When the 

 anchor was recovered, it was learned that the fluke had not keyed. A 

 shear pin in the tripping mechanism had wedged in the shaft and blocked 

 the tripping slug. 



A refurbished vibrator was used for the third test. Embedment of 

 9 feet 6 inches was attained. After "keying" the fluke depth was 

 8 feet 4 inches. In this test, attempts were made to hold peak cable 

 tensions between 20 and 30 kips so that longer term holding performance 

 would be measured. Efforts were partially successful. The anchor 

 extracted after the breakout test was prolonged over a two-hour time 

 frame. Peaks exceeded 30,000 pounds several times and each peak of this 

 magnitude caused a few inches of displacement. One peak that reached 

 47,000 pounds resulted in a sudden 20-inch displacement. About 50 load 

 peaks between 20,000 to 50,000 pounds were experienced by the anchor. 

 The mean of the peaks was approximately 25,000 pounds. 



In the fourth test, as during the second test, the vibratory unit 

 failed. An embedment of 4 feet was attained. As load was applied, the 

 fluke keyed satisfactorily. Extraction occurred as a peak load of about 

 12,000 pounds was reached. 



A significant aspect of the Operation VIII tests is that good 

 vibrator performance was obtained the first time the vibratory units 

 were used (tests one and three) . When the units were used a second 

 time, (tests two and four) they failed. After the operation, the units 

 were dissassembled and water was found to have penetrated into the 

 eccentric's chamber. 



Optimization of Design 



Background . The need to reliably predict anchor penetration and 

 resistance in a variety of seafloors is paramount if an optimized 

 vibratory anchor system is to be achieved. The basic changes in the 

 first generation vibratory anchor design to achieve optimization are 

 dependent upon the results of an analytical study using basic foundation 

 engineering principles and engineering judgment. 



45 



