A South African overview of gypcrete in road construction

Abstract

Gypcrete occurs in the western arid and semi-arid regions of South Africa and Namibia. These soils exhibit a complex nature and abnormal behaviour due to their gypsum content and as such they have become more prevalent in research. As these soils are finding more use in industry, a keen understanding of their properties and behaviour is required. Powdery and indurated gyprete samples collected from the Northern Cape (Geelvloer) and Western Cape (Rooiberg and R355) Provinces, are subjected to a series of standard test protocols for road construction materials and then compared to similar studies done on gypcrete both in South Africa and abroad, where gypcrete is researched more extensively and used successfully. The samples collected were dried at 40 °C to prevent phase transitions that will affect the properties of the material. The strength of powdery gypcrete is sensitive to density changes; therefore achieving a high density during compaction is imperative, while the strength of more indurated samples remains unaffected by density. The soaking period before the CBR is also deemed unsuitable for gypcrete as four days results in an overestimation of strength. It is apparent that the properties of gypcrete are affected by several factors, including the formation conditions, type of gypsum, and amount of gypsum, particle size distribution, the size of gypsum particles relative to other particles in the soil and the presence of other salts, all of which affects gypcrete differently making it difficult to form clear trends. The samples are also subject to wetting and drying cycles at 40 °C and 60 °C before the CBR values are determined again, to assess the variation in strength due to mineral alteration. An increase is seen for all samples dried at the higher temperature. This leads to the notion that gypcrete possesses self-stabilising properties, where temperatures in hot areas could dehydrate the gypsum, which will then readily re-absorb atmospheric moisture, leading to the formation of cementing between particles and an increase in strength. The testing served as preliminary research to guide further studies into the topic. The natural powdery gypcrete samples, with high gypsum contents and a lot of fine material, and five prepared samples of differing gypsum contents, were subject to falling-head permeability tests using both water and brine. It is understood that particle size distribution contributes to the hydraulic conductivity of soils, where a higher portion fines results in a lower hydraulic conductivity. In the case of gypcrete, the solubility is of significance as well, as it may have long term effects, through leaching and eventual cavity formation. With the intent of evaluating the effect of the aforementioned factors on the hydraulic conductivity of gypcrete in South Africa, the samples used represent differences in particle size distribution, gypsum content and origin. All samples, both natural and prepared, resulted in similar k values, in the order of 7.26x10-6 m/s, for water and the brine, despite the differences in properties. The results show that while the hydraulic conductivity is believed to be influenced by particle size distribution and origin, in the case of gypcrete in South Africa, on a small scale, it remained largely consistent.