Deep soil mixing for stabilising deep excavations

Abstract

Excavating tunnels in Scandinavia can be difficult and costly due to deep layers of soft clay. To reduce cost and risk, traditional steel struts can be replaced by deep soil mixing (DSM) struts, as is done at West Link project in Gothenburg. DSM struts are made by mixing columns of lime and cement with the clay in the shaft, usually in rows or a grid pattern. Retaining walls can be designed more lightly while reducing deformations that might affect adjacent buildings. Also, earthworks can be performed more easily. However, the safety of the shaft is dominated by the acquired undrained shear strength of the DSM columns. This strength is hard to guarantee compared to steel struts. DSM strength has a reputation of being ‘poorly understood’ or ‘variable’, so authorities apply restrictions to DSM application as bottom struts. This article comprehensively studies the acquired strength of DSM for excavation support. Undrained shear strengths measured in 315 tests were investigated and two full scale test excavations were back calculated.It has been shown that although the achieved shear strength depends on many different input parameters, when these inputs are taken into account, the acquired shear strength can be predicted. Important inputs are curing time and temperature, soil natural shear strength, confining stress, binder quantity and binder composition. Particularly the curing time has a great impact and improvements are suggested for dealing with measurements made with incomplete curing. It is proven that the acquired DSM shear strength can be much higher than current restrictions of 100 kPa and a strength prediction formula is proposed that takes these inputs into account.