heaving about the footing perimeter. Occasionally a footing would tilt 

 preferentially during the latter loading increments, but still no evi- 

 dence would appear of a distinct failure plane intersecting the soil 

 surface. Footings which had been preconsolidated and then failed in 

 bearing were found to have a "cone" of consolidated soil adhered firmly 

 to their bottom. (See Figure A-2 for condition of soil in these cones.) 

 Key depth variation had no significant effect on performance during 

 vertical loading to failure. 



Lateral Load 



Lateral load was applied to model footings at the center point of 

 the top surface, whereas the resultant force vector resisting this 

 lateral load operated at some elevation below the footing top surface. 

 Thus, the laterally loaded footings were subjected to an overturning 

 force couple during lateral load testing. The footings were free to 

 rotate and/or rise up out of the soil; in most tests the design verti- 

 cal load was maintained during lateral loading to inhibit this tendency 

 to rise up. The exact magnitude of improvement in lateral load capac- 

 ity, resulting from preconsolidation, is not clear from the model test 

 data; however, it can be said that the improvement is probably not 

 greater than two-fold. (See Figure 10.) 



Non-preconsolidated footings failed in the plane of the footing 

 cutting edge. Preconsolidated footings on the Seal Beach silt exhibited 

 a complex failure mode: a non-symmetric wedge of soil adhered to the 

 model base with a shallow, long shear plane on the leading edge, and a 

 steep, short plane of tensile separation on the trailing portion of the 

 wedge. The lateral load failure mode of footings in the Rogers Lake 

 clay was considerably different. The preconsolidated footings failed 

 by rotation about a point about a half-diameter beneath the footing; 

 excavation of one model revealed a spheroid of soil adhered to the 

 model base. The model and soil spheroid rotated within the soil mass 

 like the ball of a ball-and-socket joint. The change in failure mode 

 between the Seal Beach silt and the Rogers Lake clay is attributed to 

 the difference in the soil strength-depth profile: strength in the 

 silt increased rapidly with depth while strength in the clay was much 

 lower and increased less rapidly with depth. 



Skirt depth variation did have a net effect on lateral load capac- 

 ity with the deeper skirts offering much better performance. (See 

 Figure 10, 10 psi curves.) Increased capacity due to increased skirt- 

 depth/footing-diameter ratio derives primarily from the larger passive 

 wedge being pushed in front of the footing and/or from possible in- 

 creased soil strength on the failure planes beneath the footing. The 

 mechanism responsible for the improved performance has not been 

 identified. 



13 



