The LATWAK test consists of hitting in the horizontal direction the si
de of a pile on which a horizontal velocity transducer is attached The
blow is delivered with a sledge hammer equipped with a dynamic force
transducer. The force time signal from the hammer (input) and the velo
city time signal from the pile (output) are recorded during the impact
. The experimental mobility curve is obtained as a function of frequen
cy by calculating the modulus of the complex valued ratio of velocity
over force using Discrete Fourier Transforms. Theoretically it is assu
med that the pile is an elastic member with mass and that the soil can
be represented by linear springs and viscous damping. The problem of
the steady state forced vibration of the pile in such a soil is solved
mathematically. It leads to the theoretical mobility curve for the pi
le-soil system. The experimental mobility curve obtained in the LATWAK
test on the pile is matched with the theoretical mobility curve. A sy
stem identification technique is used to match the two curves and to e
xtract the best-fit model parameters, which include the static lateral
stiffness K for the pile-soil assembly. To evaluate the usefulness of
the method, the lateral stiffness K-p predicted by the LATWAK test on
a pile was compared to the lateral stiffness K-m measured in a static
lateral load test on the same pile. A total of 20 pile load tests and
20 LATWAK tests were performed and used to compare K-p and K-m. The r
esults are encouraging.