INFLUENCE OF THE SCALE EFFECT ON THE MECHANICAL PARAMETERS OF COARSE-GRAINED SOILS

The mechanical parameters of coarse-grained soils are often obtained via indoor or
field tests. In these tests, it is necessary to reduce the particle size of the original graded soils due
to the size limitation of the testing apparatus. Therefore, several scale methods (e.g., the equivalent
substitute method and parallel gradation method) have been proposed to reduce the size of the
original graded soils to the proper testing size. However, the mechanical parameters will be
different if different scale methods are adopted, a phenomenon that has been termed the “scale
effect”. In this research, a group of large-scale oedometer tests were conducted with specimens
that were downsized using different scale methods. The results show that (1) when adopting the
same scale methods, the compression modulus increases with the increase in the nominal
maximum particle size. (2) For the same nominal maximum particle size, samples that are
downsized using the equivalent substitute method have a higher compression modulus and degree
of particle breakage than those adopting the parallel gradation method. (3) Then Duncan-Chang EB
model parameters were back calculated from oedometer tests using an immune genetic
algorithm (IGA). These mechanical parameters were also used in a three-dimensional (3D) finite
element method analysis of the Pankou Rock-fill Dam. The manner in which the scale method
influences rock-fill dam deformation is also discussed.