Equation 2 (vane shear testing) reflects this information if the 

 shear strength on the horizontal surface equals the shear strength 

 on the vertical surface. 



? n 3 



T F = 1/2 v D Z H S v + 1/2 ^j- S R (8) 



If S V = S H - S 



t = 1^ <* + lH> S (2) 



2 



2T 



or S = 



F 



,D 2 H (1 + ^) 



Aas (1965) used vanes of different D/H ratios to determine the ratio 

 of the shear strength on the horizontal surface, Sjp to the shear 

 strength on the vertical surface, Sy. Results indicated that the 

 ratio Sjj/Sy varied from 1.1 where the cohesive material was slightly 

 overconsolidated to 1.5 and 2.0 when the material was normally 

 consolidated. In the most severe condition, the assumption that the 

 strengths in both planes are equal could overestimate the strength 

 in the vertical plane by approximately 20%. In the top few meters 

 of sediment, Su should be nearly equal to Sy. 



When the blade is inserted into the soil, the soil around the 

 vane is, to some extent, disturbed and weakened (Gray, 1957). 

 Aas (1965) found that if the vane is left one day after penetration, 

 an increase in strength occurs. The increase was attributed to a 

 dissipation of pore pressures set up during vane penetration. For 

 Aas's tests, the average ratio between the conventional and the delayed 

 vane tests varied from 1.28 to 1.52, the greatest value being found 

 for the vane having the greatest H/D ratio. Unfortunately, delaying 

 each test would consume excessive testing time. 



The failure surface during the vane shear test is assumed to 

 occur along a circular cylindrical surface with a height, H, and a 

 diameter, D. Gibbs (1960) shows that at larger rotational strains 

 failure occurs on such a surface. However, Skempton (1948) suggests 

 that the failure surface need not necessarily be tangential to the 

 vane blades but might be located some distance from them. In this 

 case, the size of the cylinder would be greater than indicated in 

 Equation 2; consequently, the actual vane strength would be less. 



Brand (1967) considers the effect of incomplete stress 

 mobilization. If such conditions exist, the shear stress varies 

 over the cylinder ends. For H = D, the effect may decrease the 

 measured torque by approximately 4%. Brand (1967) believes that 

 such conditions may occur if failure takes place at relatively 

 high strain. 



12 



