Comparing laboratory and field samples of lime–cement-improved Norwegian clay

Improvement of the structural integrity of clay-rich soil is crucial to mitigate risks related to ground movements, and representative procedures are needed to quantitatively describe the reproducibility and authenticity of laboratory-mixed (LM) clay model specimens. Here, two standardised industry protocols for preparation of laboratory lime–cement (LC) improved clay samples are compared with field LC-improved clay. Samples from both dry deep-mixed (DDM) clay columns and LM clay, with different binder quantities, were tested for density, water content and unconfined compressive strength, and analysed by X-ray micro-computed tomography. The results showed clear differences in sample density and ‘macro-porosity’ described as entrapped air resulting from field installation and laboratory preparation. For the field samples, the macro-porosity was relatively low, indicating that the air used in the binder injection is successfully evacuated along the drilling rod as sought after during DDM works. For the laboratory samples, prepared with two standard moulding preparation protocols, the macro-porosity was larger and have an influence on the measured strength and stiffness of the improved soil. An inverse relation between the mechanical strength under unconfined compression and porosity was found. In conclusion, the industry-standard laboratory moulding techniques fail to reproduce specimens that are representative for field conditions.