The influence of bulk shape factors on settling velocities of natural sand-sized sedimentary suites

Settling rates of natural sand-size particles are influenced, to some extent, by their shapes and this may be an important factor in using settling rates to estimate grain size. In order to gauge the sensitivity of this influence, two natural sand populations from the Mesozoic Nubian Sandstones of Southern Israel, with a high probability of being similar in their bulk shape characteristics, were examined in ¼φ sieved fractions for their shape characteristics and settling rates. Fine surface features (roundness and surface roughness) were evaluated using Fourier shape analytical methods. Significant differences in bulk shape were detected and their influence on settling rates was measured empirically in a settling tube. The most marked differences were in the coarse grain sizes and, to a lesser extent, in the intermediate sizes. Sampling of raw settling data at closely-spaced time intervals yielded high-resolution grain size frequency plots which were usually polymodal in nature. Subtle shape contrasts, which are an important influence on settling rates, are thus an important consideration when working at this level of sensitivity. Natural sand populations which have followed a more varied provenance or process pathway could be expected to have even greater contrasts in settling rates than the samples analysed here. Thus it is recommended that the bulk shape factor should be taken into account in order to minimize errors in the conversion of settling times to grain size. An easy method, outlined in this paper, is through the establishment of an empirically derived calibration curve for each individual suite of sand undergoing analysis. Sieved ¼φ samples, derived from a split of the total composite sample undergoing analysis, forms the basis of the calibration and hence a correction factor converting sieve diameters to true diameters must be applied. In this research, nominal section diameters were obtained optically through an image analyser.