DFI Magazine September/October 2020 - 94

Instrumentation of specimens' rebar cages

shear amplification effect can also occur at
the boundary between soil layers that have
a substantially different stiffness. However,
the effect is most pronounced and most
problematic from a design and construction perspective at soil-rock interfaces.
Since the required amount and spacing
of transverse reinforcements is directly
related to the pile shear demand, an
amplified shear force in the rock socket can
require a substantial increase in transverse
reinforcement - or decrease in spacing -
relative to the rest of the pile. However, it
has long been questioned whether this
predicted amplified shear force near the
rock-socket interface is real or an artifact of
the p-y type analysis. Nationally respected
design standards such as the Caltrans
Bridge Design Practice Manual (Caltrans,
2015) recognize the dilemma by pointing
out the ongoing debate over the validity of
this magnified shear. To date, more
advanced (but computationally intensive,
and therefore less utilized) numerical
methods of analyzing rock-socketed pile
shear demands do not provide a consistent
explanation of the phenomenon.
With support from the DFI Drilled
Shaft Committee, authors Anne Lemnitzer
and Benjamin Turner led a simple, informative experimental program to investigate the behavior of rock-socketed piles
under lateral loading. The program
compared the response of two categories of
piles with different amounts of transverse
reinforcement: piles with adequate transverse reinforcement to resist the predicted
amplified shear force at the soil-rock
interface, and piles with a nominal shear
resistance that was significantly less than
the predicted amplified shear. If the
94 * DEEP FOUNDATIONS * SEPT/OCT 2020

amplified shear predicted by the p-y
method is correct, the second category of
piles should fail in shear well before
developing a plastic hinge (i.e., before
failing in flexure). The research team
hypothesized that both types of piles
would in fact fail in flexure, and that shear
failure in the rock socket would not occur
- despite the p-y method predictions.

Experimental Program
The experimental program was executed at
the Structural Engineering Testing Hall of
the University of California, Irvine. The
university has a large reinforced concrete
(RC) test bed, causally referred to as "the
soil pit" that consists of a strong floor and
surrounding concrete walls that serve as
reaction elements while lateral load is
applied to a pile head. The test bed has a
length of 30 ft (9.1 m), a 20 ft (6.0 m)
width, and a 14 ft (4.3 m) height.
Experimental Setup - To maximize
the use of the soil pit and capitalize on the
tremendous effort required to fill (and
eventually empty) the test bed with soil, a
total of six specimens were installed in the
pit. Three specimens were rock-socketed,

as described below. An additional three
specimens (to be presented in a future
publication) were installed to investigate
response mechanisms of piles with similar
stiffness, but much different bending
behavior and failure mechanisms (comparing an RC pile vs. a hollow steel pile vs. a
grout-filled steel pile).
Pile Design - For design purposes, the
three rock-socketed test piles were initially
analyzed with LPile (Ensoft, 2018), a commonly used program for analysis of laterally
loaded piles. All pile specimens have the
same overall geometry (18 in [45.7 cm]
diameter, 188 in [477.5 cm] length, and a
rock-socket length of 48 in [122 cm]). They
also have the same longitudinal reinforcement (8 #6, with a ratio of 1.41%), and
depth of sand overlaying the rock-socket
(72 in [183 cm]).
The three specimens only vary in
transverse reinforcement details. The
transverse reinforcement for Specimen 1
was designed to satisfy the maximum rocksocket shear predicted using LPile.
Specimen 2 was reinforced with the codeminimum volumetric transverse reinforcement ratio (to satisfy AASHTO LRFD Bridge
Design Specs., 8th Ed.). And the transverse
reinforcement for Specimen 3 was designed
so that the nominal shear resistance is equal
to the maximum applied shear at the pile
head, which results in hoop spacing that
exceeds the maximum permissible spacing
allowed by codes.
For an applied lateral pile head load of
58 kN (13 kips), the LPile analysis
predicted an amplified pile shear force of
463 kN (104 kips) in the rock socket. That
corresponds to eight times the applied
lateral head load and more than double the
nominal shear resistance of Specimen 3.

Specimen 1

Specimen 2

Specimen 3

Designed to satisfy:

Amplified shear

Code minimum

Applied shear

Transverse reinforcement,
bar # @ pitch
Transverse volumetric reinf.
ratio, ρs

Spiral #4 @
114 mm (4.5 in)

Spiral #4 @
152 mm (6 in)

Ties #3 @
305 mm (12 in)

1.27%

0.95%

0.26%

Nominal shear resistance Vn

477 kN (107 kip)

396 kN (89 kip)

222 kN (50 kip)

Predicted failure mode based
on p-y analysis
Predicted pile head load at
failure based on p-y analysis

Flexural failure
58 kN (13 kip)

Shear failure
52 kN (11.6 kip)

Specimen configuration and selected reinforcement

35 kN (7.8 kip)

DFI Magazine September/October 2020

Table of Contents for the Digital Edition of DFI Magazine September/October 2020