Single Piles in Lateral Spreads: Field Bending Moment Evaluation

The results of the six centrifuge models of instrumented single pile foundations presented in a companion paper, are used to calibrate two limit equilibrium (LE) methods to evaluate bending response and factor of safety against bending failure of piles in the field subjected to lateral spreading. These six models simulate single reinforced concrete piles in two- and three-layer soil profiles, mostly end bearing but including also one floating pile, with and without a reinforced concrete pile cap, and one model where the liquefiable sand layer was densified locally around the pile to simulate the effect of pile driving. The measured permanent maximum bending moments in the pile, $M_{\max }\text{,}$ invariably occurred at the boundaries between liquefied and nonliquefied soil layers, and in most cases the moments at such boundaries reached their peak $M_{\max }$ and then decreased during shaking. These values of $M_{\max }$ before decrease, which were associated with failure of the soil against the deep foundation, are used to calibrate the two proposed LE engineering methods. For the piles where $M_{\max }$ was controlled by the pressure of the liquefied soil, the measured prototype $M_{\max }$ in the centrifuge tests ranged between about 100 and 200 kN m. It is found that a lateral pressure per unit area of pile or pile cap constant with depth $\left(p_{𝓁}\right)$ of 10.3 kPa, predicts $M_{\max }$ of the single piles tested within 15%. For the cases where $M_{\max }$ was controlled by passive failure of the shallow nonliquefied layer, the prototype $M_{\max }$ measured at the upper and lower boundaries of the liquefied soil in the centrifuge tests ranged between 160 and 305 kN m. The $M_{\max }$ values of 160–270 kN m measured at the upper boundary were reached during the shaking, and then observed to decrease towards the end of shaking. At the lower boundary, the measured $M_{\max }$ of 305 kN m was reached at the end of shaking. Use of passive pressure against the pile of the shallow nonliquefiable soil layer, obtained from the ultimate plateaus $\left(p_{\mathrm{ult}}\right)$ of $p-y$ curves, in conjunction with basic pile kinematic considerations and parameters addressed herein, explains well the development of moments measured in the centrifuge at both the upper and lower boundaries of the liquefied layer. This good accord validates the simplified LE prediction of $M_{\max }$ at the upper boundary. The two proposed simplified engineering LE methods are used to evaluate bending response and distress of end-bearing and floating piles in the Niigata Family Court House building during the 1964 Niigata earthquake, with good agreement between predicted and observed performance.