The consecutive application of the original El Centro earthquake
record with varying amplitude scales enabled the assessment of the
effects of increasing peak ground acceleration on the tested GRS
wall configurations. In general, measured accelerations increased
from the bottom to the top on the wall face and from the back to the
front on top of the wall, as reported in the literature. Accelerations
measured by the uppermost accelerometer on the face and the front
accelerometer on the top were the highest.
Amplification factors were calculated by dividing the peak acceleration
measured during shaking at the instrumented location on
the wall face by the peak shaking table acceleration. The peak values
measured were not necessarily concurrent. Fig. 3 shows the change
in amplification factors for configuration 1. Fig. 3 illustrates that
constant amplification factors may be assumed for the lower parts of
the wall face. The behavior became nonlinear closer to the top of the
wall; amplification factors decreased with increasing shaking table
acceleration. This decrease in amplification factors at the top were
observed for the configurations 1–3 and for Part 1 of the test on
configuration 4. The highest amplification factors observed were in
the range of 1.5–2 for these tests.
Extremely large amplifications were measured by the uppermost
accelerometer on the face and the front accelerometer on top in Part 2
of the test on configuration 4, where the input acceleration record
was doubled, as explained previously. These extreme measurements
implied high-frequency vibrations of the top facing blocks because
of the loosening of the connection between the top two rows of
facing blocks, which were fixed in the first part of the test.
Similarly, very high accelerations were recorded at the top of the
wall face in the tests on configurations 5 and 6, indicating highfrequency
vibrations of the top blocks that were not fixed to the
more stable lower blocks. Figs. 4 and 5 show the change in
amplification factors with increasing peak shaking table acceleration
for the tests on configurations 4 (Part 2) and 5, excluding the top part
undergoing high-frequency vibrations. Constant amplification
factors may be assumed up to the top of the wall, but amplification
factors are more variable near the top.
With the top block layers left unfixed, the increase in reinforcement
length from 85 cm (configuration 6) to 145 cm (con-
figuration 7) prevented the high-frequency vibrations of the top
facing blocks. Variation of amplification factors with changing peak
shaking table acceleration for the test on configuration 7 is shown in
Fig. 6. For this configuration and for configuration 8, which was the
top rows’ fixed version of configuration 7, the amplification factor at
the top of the wall face increased with increasing peak shaking table
acceleration.