The Effect of Laboratory Test Methods on Soil Particle Size Distribution Results

Knowledge of the particle size distribution of soils (including clay content) is required in many geotechnical engineering analyses and designs, including in geoenvironmental engineering. In such cases, the premise is made that that the laboratory test grading results are correct. Engineers have relied on grading results which subsequently turned out to be incorrect with unacceptable design consequences. This research assesses the influence of test methods on the particle size distribution of soils. This was done by sending identical samples of five different soils to nine accredited commercial testing laboratories, in South Africa. The results indicated that the largest variations occurred in the size ranges that are determined using the hydrometer analysis (percentages finer than 0.075 mm, 0.005 mm and 0.002 mm). Furthermore, when considering the clay contents, the range in values from the nine laboratories increased (to a maximum of 23 %) in the case of the sample with the highest average clay content (maximum of 42 %). A similar positive correlation was observed when considering the standard deviations (of the clay contents), which increased (to a maximum of 7.5) with an increase in the average clay content of a soil. Furthermore, in the case of each soil, the values determined by each laboratory for each percentage finer size did not differ significantly from the mean (of that percentage finer size) in 90 % of the tests conducted. In conclusion, the dispersing agent type, concentration and dosage influenced the results of the test methods considered, resulting in an up to 79 % increase in the clay content results.