Lime stabilisation is a ground improvement technique used to improve the engineering properties of cohesive fill materials. During earthworks operations, specialist plant is used to rotovate the clay fill material and intermix lime binder around clay clods. After completion of the lime treatment, the layer is compacted in the usual way. Immediately after mixing, the lime instigate a series of physico-chemical reactions within the clay soil. Where the chemical reactions are favourable and with time after compaction (curing) the material becomes progressively stronger and durable to environmental influences, e.g. inundation by surface or ground water. However, where sulphate is present within the soil, the reactions may change and the ingress of water into the layer can result in the expansive growth of deleterious minerals e.g. ettringite. While sulphate swell issues are relatively rare, when they do occur the degree of expansion can be very high. A high profile sulphate swell failure developed during the construction of the M40, Oxford, UK in 1989. Over the winter period after the lime stabilisation works, a 250mm deep lime treated layer heaved by up to 150mm - destroying the overlying road construction. Since the M40 failure, a substantial amount of effort has been undertaken to better understand the sulphate swell reactions and in this regard the state of scientific knowledge is relatively strong. A fundamental issue for field applications of lime stabilisation is that the vast majority of research has been undertaken on laboratory specimens prepared using methods which do not reflect site practice. Laboratory studies often use oven dried and finely crushed clay, whereas site operations will treat much larger clay clods to result in a more heterogeneous distribution of lime through the compacted soil body. With large clay clods, the chemical reactants must migrate through clods and this may cause the sequence of chemical reactions to change. A further challenge is that laboratory studies are typically undertaken with cure temperatures of 20°C, whereas a typical near surface temperature in the UK is <10°C. This is of particular relevance to sulphate swell failures which are reported to coincide with a reduction in ambient temperature over winter periods. Thus, the direct relevance of laboratory studies to site application was unclear. A series of laboratory experiments using a preparation method which reflects field applications of lime stabilisation was used to investigate the influence of large clay clods on the durability of lime stabilised clay soil. This method was applied to both low and high sulphate clay soils. A fundamental discovery from work on low sulphate clay is that the addition of lime binder to the surface of the clay clods causes a physico-chemical boundary to form. This boundary develops due to the rapid increase to the plastic limit of the clay preventing adjacent clods from joining together during compaction. This causes the engineering properties of each individual clod to develop independent to its neighbours and for each clay clod to be separated by an inter-clod pore space. The strength of each individual clay clod will increase with curing as the added lime dissociates into Ca2+ and OH- and migrates to form C-S-H deep within the clods. Where the material is compacted wet of the optimum water content, this condition improves ion migration and enables development of diffuse cementation deep within clods. The inter-clod porosity remains as a weakness throughout curing especially during specimen soaking, where the pore channels comprise a pathway, accelerating the ingress of soaking water. With low sulphate soil, the soaking water softens the treated material, however, with high TPS soil substantial sulphate swelling may develop. Thus, efforts to minimise this porosity during preparation is important and the use of quicklime with longer mellowing periods can cause the clay clods to develop high strength before compaction. The high strength clods resist compaction and the degree of inter-clod porosity in the compacted mass increases, worsening specimen durability to water ingress. The investigations into high sulphate clays included the development of a Novel Swell Test (NST) to assess volume change. A unique aspect of the NST was that the sulphate swell response of the lime treated material was investigated at site realistic temperatures of 8°C. It was identified that, when compared with standard laboratory test temperatures of 20°C the rate of sulphate swell is substantially higher at the low temperature. The mineralogical testing has permitted the hypothesis that, at 8°C the growth of crystalline ettringite becomes slower and the ettringite precursor, which has a high affinity to imbibe water, remains in this state for much longer. Thus, laboratory swell tests at 20°C may substantially underestimate the degree of swell that may develop in the field. As a pressing need, it is recommended that the industry adapt sulphate swell test methods to appraise the degree of swell at field realistic temperatures i.e. < 10°C. The work also identifies that the primary defence against sulphate swell is to condition the fill so that the risk of post compaction water ingress, via inter-clod porosity, is minimised. The use of GGBS and water addition during extended mellowing periods also reduces the degree of sulphate swell in natural clay soils. This work concludes that working methods for lime stabilisation of medium high plasticity soils of a potentially high sulphate content, should be adapted to encourage diffuse cementation and minimise the degree of (post compaction) inter-clod porosity. Practically this involves the use of hydrated lime and the addition of mixing water throughout extended mellowing periods. Fundamentally, the study recommends that where construction programmes allow, the long term durability of a fill material should be the priority over immediate strength.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:763426 |
| Date | January 2015 |
| Creators | Beetham, Paul |
| Publisher | Loughborough University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | https://dspace.lboro.ac.uk/2134/19583 |
Page generated in 0.0028 seconds