This thesis project has examined the moisture dependency of shear strength in the loess soils of Banks Peninsula. These dominant silt materials are generally regarded as having an angle of internal friction between 25 and 30°, and cohesion of zero, when the soil is saturated. However, soil behaviour in terms of slope stability would appear to support a cohesion value higher than zero as vertical cliffs of loess can be seen to stand indefinitely. It is agreed that loess soils upon saturation do show very low shear strengths, but these soils rarely become saturated due to their low permeability (k<10⁻⁶). To address the quantity of water content dependency of shear strength, five field sites have been sampled for unconsolidated undrained triaxial shear strength to determine c and ø at varying water contents in the total stress state, which most closely simulates observed behaviour. The field sites were:

1. Moncks Spur primary airfall in situ loess; 2) Stonehaven Subdivision loess-colluvium;
2. Worsleys Spur primary airfall in situ loess; 4) Duvauchelle primary airfall in situ loess; 5) Whaka Terrace uncompacted loess fill. Nominated water contents for the shear testing program were 6%, 10%, 14% and "as wet as possible". Samples were prepared in the laboratory by immersing the stainless steel sampling tube containing the soil sample in water for up to 2 weeks to a water content maximum, and then drying them back to the nominated water content. Analysis of results of all five field sites tested at the four nominated water contents showed a new cohesion trend, which has not previously been reported in Banks Peninsula Loess. As water contents increase from 6% cohesion increases to a maximum value and then decreases, whilst for angle of internal friction there is a linear decrease over the entire range of water contents, for example 52° at 6% to 13° at 23% for Stonehaven Subdivision loess colluvium. Maximum values for primary airfall (in situ) Port Hills loess fill and colluvium the cohesion maximum was approximately 210 kPa at 10% water content. Friction angle trends were quite similar for all sites. A secondary aim of this thesis was to examine lime stabilising effects on compacted loess fill. Trials were carried out at the Whaka Terrace field site where excavated loess fill from a test pit was mixed with hydrated lime at 2% by weight of the total soil mass, and then compacted back into the same test pit so that the treated soil could be cured under field conditions. Untreated samples were taken from the base of the test pit and treated soils were extracted from the treated compacted layers at intervals of 1 week, 1 month and 2 months after compaction. Samples were prepared at two different water contents, 15% and as "as wet as possible", so that a shear strength dependence on water could also be established. Results suggest that 15% water content treated samples had 30-40 kPa more cohesion and 5° more angle of internal friction than the "as wet as possible" treated samples. For example after 7 days of field curing cohesion was recorded at 30.7 kPa and 4 kPa for water contents of 16.9% and 19.1% respectively, and angle of internal frictions were 30° and 23.4° for the same respective water contents. Shear strength values increased over time with a rapid rate of increase after the first week of curing, and then a slowing down thereafter. A maximum cohesion of 25 kPa and angle of internal friction was recorded for samples cured in the field for 68 days as compared to untreated loess, which had cohesion of 0 kPa and 11.5° for the angle of internal friction. Both were tested at "as wet as possible" water contents. Future work should entail the determination of effective stress parameters c and ø’ by determining matric suctions (negative pore pressures) to determine if the trend seen in this project is something new or an artefact of the analysis used, and more study centred on loess from the Akaroa Harbour region. In conclusion it has been demonstrated in this thesis that Banks Peninsula Loess shear strength has a dependency on water content in terms of total stress parameters c and by way of using the triaxial test apparatus in the unconsolidated undrained condition.