To determine the compaction index K (Equation 4.6), the fraction 1-2 mm was used as
the reference fraction. The content of the 1-2 mm fraction was successively replaced by
aggregates of other fractions (in steps of 10 Vol.-%). Three series of tests were carried
out: with 0.5-1, 2-4 and 4-8 mm fractions. The packing density was determined
according to the standard compaction method. The K-index that fitted the results best
was 3.6 and was lower compared with what De Larrard [1999] recommended for this
compaction level (4.5-4.75). He found a higher compaction index, in spite of lower
initial packing densities of the aggregate fractions. A lower K-index indicates that the
applied compaction process was less effective in compacting the granular skeleton. As
was discussed by Hoy [1998], the shape of a particle also determines the sensitivity
towards a specific compacting method. Appendix B provides an overview on results ofthe study on the K-index. The mean error of the three series for this K-index was 0.46-
0.62%, with a maximum error of 0.81-1.18%.