The first term cN would be zero for a cohesionless material; 

 c = o. Based on Equation B-l, the anchor holding capacity, F™ for a 

 cohesionless soil can be calculated as follows: 



F T = Y^NqAp (B-2) 



where A^ is the anchor fluke area in plan . 



The salvage sand anchor-projectile has a shape that is very 

 difficult to analyze. It cannot be solely represented by either a 

 rectangular, or a circular shape. The present approach is to bound 

 the problem by calculating holding capacity of an anchor based on an 

 equivalent circular area and based upon a continuous strip with 

 comparable width and overall area. The difficulty with this technique 

 lies in arriving at a realistic assumption of the embedment depth 

 at which a particular anchor shape starts behaving as a "deep" anchor. 



For each soil, there is a characteristic relative depth D/B 

 (D/B = ratio of depth of embedment to fluke diameter) beyond which 

 anchor plates start behaving as "deep" anchors and beyond which 

 breakout factors reach constant final values (Vesic, 1969). 

 Experimental data concerning "deep" anchors behavior are available 

 for uniform circular and square anchor plates; however, nothing is 

 available for rectangular sections. 



Preliminary results of studies being conducted at the University 

 of Massachusetts under a contract with NCEL to determine the breakout 

 resistance of circular anchors embedded in saturated sands, indicate 

 that this relative depth, D/B, varies for medium dense sand from 4 to 

 6. This agrees with the results of Baker and Kondner (1966) for dry 

 sand of medium density. Being moderately conservative, all sands 

 are assumed to be of medium density prior to anchor breakout. The 

 sand in the areas where the explosive anchor was evaluated were of 

 medium density (refer to Table 1 in text) . Values of N used in 

 Equation B-l were assumed constant for the circular shape. In 

 addition, for calculations it was assumed that the limiting depth, 

 D/B, for the rectangular shape is D/B = 7. This appeared reasonable 

 after comparing the soil stresses imposed by each shape of anchor. 

 A brief model study to define the behavior of rectangular shapes 

 during pullout is being initiated as part of another program. 



Holding capacity in sand was calculated by first taking Vesic' s 

 results and plotting breakout factor, N q versus relative depth, 

 D/B, Figure B-l,_and extrapolating to D/B = 7 for the rectangular 

 shape. Second, N was plotted versus depth D, in Figure B-2, for the 

 actual width of the sand fluke, B = 2 feet, and for the diameter of 

 a circle with an equivalent area of the sand fluke, B = 6 feet. It 

 appears that for this particular sand fluke, use of both assumptions 

 will yield very nearly the same holding capacities to a depth of 14 feet, 

 since N is directly related to holding capacity. Figure B-3 presents 

 the relationship between static holding capacity, and depth for an 

 ideal sand with the angle of internal friction, <(>, varying from cf> = 30 

 to 40°. 



36 



