Vorbelastung und Erdruhedruck eines Kreidetons

The widening of the Mittelland Canal, which runs east-west and is located slightly north of Hanover, F.R. Germany, has been in progress for some years. This involves several design and construction Operations such as steepening of slopes, use of bulk heads, longer and wider bridges, new docks with greater capacity, etc.The soils in the vicinity of Hanover are of the upper cretaceous period and are very stiff to hard. The region has been through several glacial periods such as Elster, Saale, Weichsel (DIETZ 1973).The index properties of the investigated clay are as following: Water content decreased frofti 29,5% at 1,5 m to 19%. Below this it ranged between 17,5% and 19,5% down to the bottom of the boring at a depth of about 20 m below ground surface. Liquid limit ranged between 47 – 64% with slight increasing tendency with depth. Plastic limit was invariable around 19%. Clay fraction content (particles smaller than 2 microns) were between 30 and 55%. Carbonate contents ranged between 42 and 67%. So the clay should better be called marl. Independent of the carbonate contents the soil may be classified as a clay of medium to high plasticity.The mineralogical study using X-ray diffraction technique indicated no diagenetic bonds. A slight evidence of weathering was observed only in one specimen, 2,8 m deep. Major clay components were montmorillonite (16 – 23%), illite (9 – 16%), kaolinite (4 – 5%) and chlorite (2 – 4%).The soil was found to exhibit different degrees of swelling, in vertical and horizontal direction, upon contact with water. In order to fully understand its behaviour the following Saturation, load and deformation conditions were studied for undisturbed specimens.1.The test specimen had no access to free water during the entire test.2.The test specimen was immersed in water at a stress equal to or greater than the effective over-burden stress.3.The specimen was immersed in water with the very first load increment, but swelling was completely restrained.4.The specimen was permitted to swell under a very low stress of 5,2 kN/m2 while immersed in waterRemolded specimens Consolidated from a slurry in a large oedometer were immersed in water with the very first load increment of 10,4 kN/m2. In an attempt to arrest swelling two additional load increments were applied in less than one hour.From comparison of fig. 1 to 5 the following can be concluded: all tests showed lower precompression stresses than can be deduced from geological estimations. Horizontal tested samples (H) showed higher precompression stresses than vertical (V) tested samples. H-samples also showed greater amounts of swelling when immersed. From these findings it is concluded that the more a sample was able to swell in one direction, the more it will be unable to recall its past maximum consolidation stress. When Hand V-samples were tested in one dimensional compression after füll swelling under a very low load, no anisotropy and no marked differences in precompression stress were to determine.From tests with laboratory prepared samples the swelling properties showed to be quite normal. The precompression stress estimated from Standard methods was in good agreement with the real value. So the different behaviour of the undisturbed samples is due to their different history.For engineering works the knowledge of the coefficient of earth pressure at rest is needed very often. For overconsolidated soils normally this coefficient is related to the overconsolidation ratio as defined from oedometer tests. This ratio is open to some doubt, so, in spite of several research works with laboratory prepared samples, for practical purpose it is not easy to establish a reliable earth pressure coefficient K for overconsolidated soils.From a literature review it follows, that only self-boring instruments, WROTH and HUGHES (1973), or push-in spade-shaped instruments, TEDD and CHARLES (1981), seem to be able to measure this quantity in a direct way. Since often it is not possible to have additional field tests, the derivation of a simple expression for estimating K from laboratory tests has been considered to be of interest.The determination of K follows the way proposed by SKEMPTON (1961) and the calculations are based on the principles of the critical State concept (ROSCOE et al. 1958). Therefore the overconsolidation ratio was taken according to COOLING and SKEMPTON (1942).The unconsolidated, undrained shearing resistance, normalized by the equivalent normal stress σ_e’ proved to be rather constant and coincided quite well with the value given by SKEMPTON (1957) according to the plasticity index of this soil, fig. 8. The normalized cohesion intercept from the results of a comprehensive series of Consolidated drained and undrained triaxial tests on V- and H-samples was found by dividing p and q by σ_e’ based on the actual void ratio at the point of maximum deviator stress. Form curve fitting with linear regression from 27 data points the true angle of internal friction σ_e’ was found, fig. 9.By comparing the undrained normalized and the drained normalized shear strength an expression for the coefficient of earth pressure at rest was derived.The value of Ag has a pronounced influence on K, but information on A_s are very small. In this research work no further information on A_s was available, so K was calculated for three different values of A_s (fig. 10).The results of the calculation are in rough agreement with the measurements of TEDD and CHARLES (1981) when using the smaller values for A_s.