Sieving Error from Dry-Separating Silt-Sand-Gravel Soils

The dry-separation method is an alternative to the wet-preparation in the current European Standard for the determination of particle size distributions by the sieving of soils. Due to the risk of error, dry-separation is cautioned against in the standard; however, there is no additional guidance as to when it is unsuitable nor for the magnitude of error that it may introduce. This study investigates the dry-separation method as an alternative by comparing with the conventional method of Wet-preparation in terms of particle-size distributions of eight cohesionless sand-gravel soils with varying amounts of nonplastic fines. The findings indicate a gradually increasing sieving error for fractions at minus 0.5 mm with the amount of fines in the soil, and depending on the fines content of the soil, dry-separation introduced errors upwards of 45% in silt-sand-gravel soils. An empirical best-fit formula is proposed for the estimation of the error using the dry-preparation method on this type of soil. Furthermore, to avoid sieving errors, the results suggest that the dry-separation method should not be used for silt-sand-gravel soils exceeding 2% silt size fractions.