Extensive investigations of free-swell properties were done on 14 South African mudrocks, sampled from various geological formations in different parts of the country. The tests were performed on 50 mm cubes, sawn without the use of water from large rock fragments which had been stored in sealed plastic bags since excavation. Most of the swells were determined over a period of about 24 hours along three cube axes, and expansion with time was monitored on analogue recorders. The cubes from each sample were pretreated in different ways before immersion in water, i.e. some were immersed in water immediately after sawing, some after various times of air-drying (exposure to the atmosphere), and some after oven-drying at 50 and 105 °C. Maximum percentage swells during the second or even third immersions were determined on cubes which did not break down during the first immersion. The samples yielded free swells after oven-drying ranging from 0,1 to 9,1 per cent with the majority of swells below 1,0 per cent. Swells of undried samples from the natural (field) moisture content were much lower. There was little correlation between the amounts of water absorbed during the immersion period and the maximum percentage swells. All the cubes lost moisture during the airdrying periods and swells generally increased with an increased exposure time. Maximum percentage swells after long exposures were similar to swells after oven-drying. The well-stratified samples expanded much more in a direction perpendicular to the bedding than in the direction of the bedding plane when compared with the expansions of the more isotropic or massive samples. It was observed that cubes immersed at their natural moisture content did not disintegrate or slake to the same extent as cubes that were immersed for the first time after they had been exposed to the atmosphere or after they had been dried in an oven. Samples which slaked completely exhibited swells similar to some samples which disintegrated into flakes.
The free swell properties of mudrocks were investigated during a programme of research into classification tests for roadbuilding materials. Olivier (1979) expressed doubts about the possible effects of air-drying (exposure to the atmosphere) and ovendrying at various temperatures on the structure and swell characteristics of the mudrock. It was therefore decided to control strictly all the conditions from the time of sampling unexposed rock and to use various pretreatments before immersion to establish the behaviour and the important factors during a free swell test.
Work of a similar kind could not be found in the literature. Reviews of swelling mechanisms and tests to identify swelling rocks have been given by Lindner (1976) and Einstein and Bishoff (1977). Two samples of a "claystone and a mudstone" possessed higher swell perpendicular to bedding than that parallel to bedding (Murayama and Yagi 1966). Nascimento et al (1968) and Duncan et al (1968) described procedures and apparatus to carry out free swell tests.
