associated with short-term capacities in cohesive soil since suction 

 pressures dissipate almost immediately in sand. Also it is a problem 

 associated primarly with field tests since laboratory tests, such as 

 those presented in Figure 1, are usually performed with anchor bottoms 

 vented to eliminate suction. It is important to recognize that short- 

 term field test results may be unconservatively high, and it would be 

 desirable to be able to predict the extent of these suction forces so 

 that they may be subtracted from the measured holding capacities to 

 yield more reliable results. 



Only limited research has been conducted to investigate the magni- 

 tude of suction forces. Papers published at Duke University (Vesic, 1969, 

 and Ali, 1968) and the University of Massachusetts (Kupferman, 1971) 

 mention the suction effect as important but do not provide information 

 on how to evaluate it. 



The NCEL research on breakout of partially embedded objects is 

 somewhat applicable because suction is thought to be the major contributor 

 in this form of breakout. As a result of this research (NCEL, 1972) 

 it was concluded that the significant parameters in partially embedded 

 object breakout are the soil shear strength, c, and the relative embed- 

 ment depth, D/B. For immediate breakout to occur, rupturing of the soil 

 beneath the object is required, thus the significance of c. It was found 

 that for a relative embedment depth of 1.0, the breakout force was equal 

 to about 7 Ac, where A is the object plan area. It appears that this is 

 a maximum value for the suction force and that further increases with 

 depth will not occur. 



Research of the Hydro-Electric Power Commission of Ontario (Adams 

 and Hayes, 1967) provides additional information on the soil suction 

 problem. Laboratory tests performed with soft (c = 1.5 to 2.0 psi) clay 

 and vented and unvented flukes yielded results which compare favorably 

 to the NCEL results. The data indicate that the suction effect at D/B 

 ratios of approximately 3 and 4.5 is about 7 Ac, which is equivalent to 

 stating that the holding capacity coefficient attributed to suction is 

 ab ou t 7 . 



The NCEL and Ontario Hydro-Electric data can_be used for design 

 p_urposes_. This is done in_Figure 3 in which the N for full suction 

 (N = N + 7) divided by N for no suction, Equation 5, is plott_ed 

 verius tfie relative embedment depth, D/B. The ratio of the two N 's is, 

 if effect, a reduction factor which should be applied to the results of 

 field tests on anchors embedded in soft clay. This may be done by first 

 subtracting the quantity, y, DA, from the measured pullout force, dividing 

 the result by the reduction factor of Figure 3, and then adding y, DA to 

 yield the anticipated short-term holding capacity without suction under 

 static loading conditions. Due to inherent uncertainties involved in 

 predicting the magnitude of the suction effect on a "shallow" anchor, 

 when anchors are to be field tested, it is recommended that these tests 

 be performed on deep anchors (D/B> 5) where shear strength is not a 

 dominant parameter in determining the reduction factor. 



Most seafloor clays would be considered "soft" so this suction factor 



