Particle size distribution (PSD) of gypseous soils is important in the soil science community. When gypsum constitutes a major portion of the soil, its removal prior to PSD analysis distorts the results and may lead to textures that do not relate to conditions in the field. In order to understand the true characterization of the soil and the gypsum particles, the entire soil sample should be analyzed. Four different approaches to the BaCl2 method presented in the literature (Hesse, 1976, Matar and Douleimy, 1978, Viellefon, 1979) were used to evaluate the use of BaCl2 solution to reduce the solubility of gypsum by forming a protective coating of BaSO4 around gypsum particles. Results showed that the BaCl2 method was unsatisfactory, as dispersion of clays was not sufficient to allow particle size analysis using the pipette method. A procedure using a laser diffraction particle size analyzer (LPSA) was also evaluated. As gypsum is insoluble in methanol, methanol was selected as a possible solution, but it caused flocculation of clays and could not be used to analyze samples containing silicate clays. Gypsum saturated water containing Na hexametaphosphate was evaluated as a solution. First, 20 non-gypseous samples were analyzed on a sand-free basis using saturated gypsum water with Na hexametaphosphate. Results were used to establish a relationship comparing LPSA results and pipette results. An equation y = 1.37x + 2.03 was established relating LPSA clay percent by volume (x) to the pipette clay percent by weight (y). The equation had a R2 value of 0.84 and was significant at the 1% level. From this equation a comparison of 21 gypseous samples was made, between clay percentages of the pipette method and the LPSA method. Results indicate that LPSA can be used to give a satisfactory particle size distribution of gypseous soils when coupled with sand analysis by sieving.

Particle size distribution (PSD) of gypseous soils is important in the soil science community. When gypsum constitutes a major portion of the soil, its removal prior to PSD analysis distorts the results and may lead to textures that do not relate to conditions in the field. In order to understand the true characterization of the soil and the gypsum particles, the entire soil sample should be analyzed. Four different approaches to the BaCl2 method presented in the literature (Hesse, 1976, Matar and Douleimy, 1978, Viellefon, 1979) were used to evaluate the use of BaCl2 solution to reduce the solubility of gypsum by forming a protective coating of BaSO4 around gypsum particles. Results showed that the BaCl2 method was unsatisfactory, as dispersion of clays was not sufficient to allow particle size analysis using the pipette method. A procedure using a laser diffraction particle size analyzer (LPSA) was also evaluated. As gypsum is insoluble in methanol, methanol was selected as a possible solution, but it caused flocculation of clays and could not be used to analyze samples containing silicate clays. Gypsum saturated water containing Na hexametaphosphate was evaluated as a solution. First, 20 non-gypseous samples were analyzed on a sand-free basis using saturated gypsum water with Na hexametaphosphate. Results were used to establish a relationship comparing LPSA results and pipette results. An equation y = 1.37x + 2.03 was established relating LPSA clay percent by volume (x) to the pipette clay percent by weight (y). The equation had a R2 value of 0.84 and was significant at the 1% level. From this equation a comparison of 21 gypseous samples was made, between clay percentages of the pipette method and the LPSA method. Results indicate that LPSA can be used to give a satisfactory particle size distribution of gypseous soils when coupled with sand analysis by sieving.

Dissolution of soluble substrates such as gypsum presents a major hazard to dams in many parts of the world. This research simulates hypothesised conditions beneath the Mosul Dam, northwest Iraq, where collapse of a karstic system associated with continuous fresh water supply from its massive reservoir water is a recognised problem. The gypsum substrates at Mosul Dam vary in purity and thickness. Experimental work used gypsum rocks and gypseous soils. Gypsum rocks from northern Iraq and similar rocks from Bantycock gypsum mine, UK, were analysed for short-term mechanical response following immersion (5 to 50 weeks) and long-term loading during immersion (maximum 50 weeks). New experimental devices were developed from a conventional oedometer. Cylinder samples (NX, standard diamond drill core size = 54mm diameter, length/diameter ratio equal to 2.5) provided a proxy for massive gypsum strata, while thin samples (NX = 54mm in diameter, 20mm thickness) represented thin layers and lenses. Rectangular bar samples (240 x 40 x 20 mm and 140 x 40 x 20 mm) were tested for short-and long-term mechanical four-point bending behaviour. Samples were permanently submerged at a variety of water pressures, with the influence of groundwater recharge and flow on dissolution simulated by regular changes of water. Stress on each sample was progressively increased to a maximum of 2688 kPa. Small increases in strain were recorded by the end of each test but no failures occurred within 60 days of tests. However, notable failure due to water pressure and axial stress over long time periods of 166 and 238 days occurred. Visible physical changes were observed, notably a decrease in sample mass and volume. Similar change was recorded in ultrasonic velocities. These indicate that gypsum collapse risk beneath dams requires prolonged exposure to dissolution. Gypseous soils from Iraq and similar artificially-prepared soils were also tested. Gypseous soil samples (diameter = 50mm and length = 20mm) and box model strata results showed that gypseous soils are significantly weakened by dissolution over 15 weeks and 50 weeks respectively. Dams built on gypsum substrates are likely to experience ongoing weakening of their foundations, with a progressively increasing risk of failure. This is expected to be enhanced for dams with a large and deep reservoir that induces high ground water pressure.