sleeves in the platform base and while the inserted piles were being hammered into the sea bed. Max- 

 imum tip displacement amplitudes (cross flow) of 3.2 to 3.8 m (10.5 to 12.5 ft) from equilibrium were 

 predicted for currents as low as 0.6 m/s (0.31 kt) at the platform site. These large-scale motions were 

 expected to create difficulties while "stabbing" the piles into the sleeves, and they could also increase 

 the risk of buckling and fatigue failures during the pile driving operations. 



Experiments were conducted with model piles in three laboratories (40), for both the pile lower- 

 ing and the pile driving operations. Uniform and nonuniform (shear) flows were modelled in the 

 experiments. The shear parameter that characterizes a nonuniform flow is defined as 



I^REF Az 



where V is the magnitude of the incident flow velocity and z (in this case) is the depth of the water. 

 The reference value Vref 'S usually taken as the velocity magnitude at half the distance along the cable 

 or structure, although the maximum value from a velocity profile is sometimes used. For the small- 

 scale experiments reported by Fischer et al (40) the shear parameter was /3 = 0.01 which matched the 

 actual Cognac site value at depths between 100 m (330 ft) and 250 m (820 ft). 



The results from some typical model-scale experiments are plotted in Fig. 2.24. The tests were 

 conducted with a 1:168 scale model of the large marine piles of diameter Z) = 2.1 m (7 ft). Both the 

 full-scale and the model piles had specific gravities of 1.5. It is clear from the results in Fig. 2.24 that a 

 shear flow with /3 = 0.01 to 0.015 had virtually no mitigating effect on the peak vortex-excited displace- 

 ment amplitudes in the cross flow direction. The data plotted in the figure correspond to a free- 

 cantilever flexible beam with no tip mass at the free end. This configuration matched closely the 

 "stabbed" pile before an underwater hammer was attached for driving it into the sea bed. The structural 

 damping of the PVC model in Fig. 2.24 was Cj = 0.063 and for a similar stainless steel model the 

 damping was C,^ ~ 0.015; the two flexible cylinders experienced tip displacement amplitudes of 2 K = 

 3D and 4D, respectively. These damping ratios and displacement amplitudes fall well toward the left- 

 hand portion of Fig. 2.2 where hydrodynamic effects are dominant. 



24 



