Consolidation behaviour of soft soil subjected to on-land and underwater vacuum preloading

Vacuum preloading is a ground improvement technique in which effective stress of soil is increased by reducing pore water pressure. It can be further divided into on-land vacuum preloading and underwater vacuum preloading. To understand the consolidation behaviour of the soil subjected to on-land vacuum preloading and underwater vacuum preloading, theoretical analysis, field investigation, centrifuge modelling and numerical modelling were carried out. Analytical solutions of average reduction of pore water pressure in soft clay subjected to vacuum preloading were developed. Well resistance, smear effect and vacuum loss are considered in combined radial and vertical consolidation. Variation of soil stiffness and permeability, multi-stage loading procedure are considered in an incremental form. Based on constant coefficient of consolidation, it is found that average degree of consolidation is slightly larger in the presence of vacuum loss than that without vacuum loss (one-way drainage). It is also found that radial consolidation is significant only if the drain spacing ratio is less than threshold n value, which is 7-10 for large diameter drain (e.g., D = 0.5 m) or 20-30 for small diameter drain (e.g., D = 0.1 m). Observations obtained from field on-land vacuum preloading and centrifuge model tests on underwater vacuum preloading show that the pore water pressure reduces to a suction line with total vertical stress nearly unchanged. The principle of underwater vacuum preloading is similar to that of on-land vacuum preloading with little difference near the edge of treated zone. The increase in vertical effective stress is equal to reduction of pore water pressure for saturated soil, but the increase in horizontal effective stress varies with locations and generally less than that increase in vertical effective stress. The final coefficient of earth pressure in the treated zone ranged from K0 at the bottom of the treated soil layer to nearly unity near the ground surface at the edge of treated zone. The coefficient of earth pressure also varies across the treated area. Negligible lateral displacement was found at lower depths below ground surface or 2.5H (H thickness of soft soil or depth of vertical drain) away from the edge of treated zone. The water pressure on the membrane plays a positive role on the increase of vertical effective stress. The efficiency (increment of reduction of pore water pressure over increment of water pressure on membrane) is about 0.8 in present centrifuge test with lateral seepage. In field observations, vacuum loss is negligible in a large treated area, whereas vacuum loss is significant in a small treated zone. Penetration depth of a membrane influences the settlement slightly, whereas its influence is reduced in presence of vertical drains. However, penetration depth of a membrane significantly affects lateral displacement. The inward lateral displacement induced by vacuum preloading when the membrane penetrates through the full depth of soil is 1.25 times of that when penetration depth of a membrane into soil is negligible. The ratio of lateral displacement to vertical displacement is 0.55 when the penetration depth of membrane into soil is negligible, and is 0.70 when the membrane penetrates through the full depth of soil. The larger the drain penetration depth, the larger the settlement is developed, unless the vertical drain penetrates beyond the optimum penetration depth. For a given penetration depth of sand drains, settlement decreases with increasing drain spacing from 8D to 1H. Given a penetration depth, the largest settlement is developed with drain spacing 8D when sand drains are used. It is found that the bottom upward seepage (vacuum loss) has a negligible influence on the consolidation when the drain spacing is less than 8D. The optimum penetration ratio is about 0.95 when drain spacing is less than 28D and the permeability of sand drain is larger than four orders of that of clay