The mineralogy and fabric of ‘Brickearths’ in Kent, UK and their relationship to engineering behaviour

Abstract

Mineralogical and petrographical investigation of two loessic brickearth profiles from Ospringe and Pegwell Bay in north Kent, UK have differentiated two types of brickearth fabric that can be correlated with different engineering behaviour. Both sequences comprise metastable (collapsing) calcareous brickearth, overlain by non-collapsing ‘non-calcareous’ brickearth. This study has demonstrated that the two types of brickearth are discretely different sedimentary units, with different primary sedimentary characteristics and an erosional junction between the two units. A palaeosol is developed on the calcareous brickearth, and is associated with the formation of rhizolithic calcrete indicating an arid or semi-arid environment. No evidence has been found for decalcification being responsible for the fabric of the upper ‘non-calcareous’ brickearth. Optically-stimulated dates lend further support for the calcareous and ‘non-calcareous’ brickearth horizons being of different age or origins. The calcareous brickearth is metastable in that it undergoes rapid collapse settlement when wetted under applied stresses. It is characterised by an open-packed arrangement of clay-coated, silt-sized quartz particles and pelletised aggregate grains (peds) of compacted silt and clay, supported by an inter-ped matrix of loosely packed, silt/fine-grained sand, in which the grains are held in place by a skeletal framework of illuviated clay. The illuviated clay forms bridges and pillars separating and binding the dispersed component silt/sand grains. There is little direct grain-to-grain contact and the resultant fabric has a very high voids ratio. Any applied load is largely supported by these delicate clay bridge and pillar microfabrics. Collapse of this brickearth fabric can be explained by a sequence of processes involving: (1) dispersion and disruption of the grain-bridging clay on saturation, leading to initial rapid collapse of the loose-packed inter-ped silt/sand; (2) rearrangement and closer stacking of the compact aggregate silt/clay peds; (3) with increasing stress further consolidation may result from deformation and break up of the peds as they collapse into the inter-ped regions. Smectite is a significant component of the clay assemblage and will swell on wetting, further encouraging disruption and breaking of the clay bonds. In contrast, the ‘non-calcareous’ brickearth already possesses a close-packed and interlocking arrangement of silt/sand grains with only limited scope for further consolidation under load. Minor authigenic calcite and dolomite may also form meniscus cements between silt grains. These have either acted as “scaffolds” on which illuviated clay has subsequently been deposited or have encrusted earlier-formed grain-bridging clay. In either case, the carbonate cements may help to reinforce the clay bridge fabrics. However, these carbonate features are a relatively minor feature and not an essential component of the collapsible brickearth fabric. Cryoturbation and micromorphological features indicate that the calcareous brickearth fabric has probably been developed through periglacial freeze–thaw processes. Freezing could have produced the compact silt/clay aggregates and an open porous soil framework containing significant inter-ped void space. Silt and clay were remobilised and translocated deeper into the soil profile by water percolating through the active layer of the sediment profile during thawing cycles, to form the loosed-packed inter-ped silt matrix and grain-bridging meniscus clay fabrics. In contrast, the upper ‘non-calcareous’ brickearth may represent a head or solifluction deposit. Mass movement during solifluction will have destroyed any delicate grain-bridging clay microfabrics that may have been present in this material.