In the area of reduced-scale model testing, increasing the size of model st
ructures as a means of improving the model performance is often inhibited b
y the capacity limitation of the available seismic testing facilities. This
was the case in a recent experimental programme conducted at NTUA in which
several 1:5.5-scale model frames were tested on an earthquake simulator. I
n order to improve the model reliability, the following measures were taken
: strictly obeying the geometric scaling in almost all aspects; using defor
med bars as model reinforcement; and using stone-aggregate microconcrete fo
r constructing the models, etc. To evaluate the effectiveness of these meas
ures, a parallel column test programme was conducted in which duplicates of
typical columns in the test model frames were prepared, along with those o
f large sizes following similitude laws. The column specimens were tested u
nder reversals of lateral load. Besides the scale effects, general rules re
garding the influences of transverse confinement and axial force on che duc
tility and strength of columns were also confirmed on the small-scale model
columns. (C) 1998 Elsevier Science Ltd. Ail rights reserved.