8 Feb 04 17:46
In general while the Atterberg limits were used by many to derive simple relationships, the governing factor regarding swelling is related to the soil mineralogy. In my experience despite clays may be classified as very heavy, they do not necessarily exhibit high swell characteristics.
Clays with a high percentage of montmorillonite for example
would have a higer propensity for swelling than on that is composed of predominantly kaolinite and illite minerals.
To quote an example, in one jurisdiction (Guyana) the clays have liquid limits ranging from 60 to 90 and PI's from 30 to 60 The composition of the clay fraction is as follows Kaolinite 20-25 %, illite 30-35%%, chlorite 10-15 % and montmorillonite 0 to 5 %. Hence the importance of understanding the clay mineralogy.
This clay does not exhibit swelling characteristics and this jurisdiction does not undertake swell tests as part of thier evaluation. Based on the Atterberg limits this clay would be potentially high swelling.
Generally, the observations made through the performance of structures indicate wheteher or not swelling clays are present and are of concern. This is often why some jurisdictions decide to use the Atterberg limits as an indicator.
If one is uncertain then the swell test should be undertaken. The free swell test is often sufficient. Jennings et al from South Africa undertook much research in this area. His work is often quoted in publications on swelling soils.
One should review the text "Clay Mineralogy" by Grimm to understand how complex the cay particulate structure is and to appreciate that the geological composition is very important to clay behaviour.
Unfortunately, we do not traditinally undertake x-ray diffraction tests and as such we should only utilize the general correlations unless we have knowledge that the clay soils in a particular area are swelling etc. Even then, I would always undertake a swell test in designing of a remedial measure.
There are a number of technical papers on the subject. I refer you to one that incorporates the empirical approach using the Atterberg Limits and as well as swell tests. This paper is by Diyaljee and Forbes and was published in the Proceedings of the 5th International Conference on Swelling Soils, Adelaide, Australia in 1983. This was based on a case study of heaving of side walk slabs of a shopping centre in Edmonton, Alberta, Canada, which I undertook in 1981.
Please excuse the length