Soil conditions – variations in plane and profile

The results of the CPT tests showed soil conditions which in most respects were in concordance with earlier descriptions but which in some respects were considerably different. They were also in some respects similar in the two sections but varied greatly between them in other respects.

As in previous observations, the thickness of the upper sand and silt layer was found to vary strongly within the area. From the results of this and previous investigations, it can be estimated that the thickness of this layer has been about 5 metres over the excavated area in Section A. It then decreases with distance from the river, from about 5 metres at the upper crest above the excavation to about 2 metres at a distance of 50 metres behind this. In Section C, the corresponding thickness of the sand and silt layer above the excavated area can be estimated to about 6 metres. However, in this section the thickness of the upper layer then instead increases with distance from the river. It is thus about 8 metres at the upper crest and fully 10 metres at a distance of 35 metres further away.

The thickness of the underlying clay layers was found to vary strongly and mainly be much greater than had previously been assumed. In Section A, it varies from being about 25 metres in the area behind the upper crest and then gradually increasing under the slope to having been about 48 metres down at the river. The thickness of the fine-grained soil layers has decreased because of the erosion and the excavation, but in spite of this there still remains about 35 metres of clay below the excavation and also below the river bottom. The clay is underlain by coarser soil with a thickness that varies from one or two metres to more than 10 metres, which rests on bedrock.

The clay located directly below the upper sand and silt layer contains thin layers and lenses of silt, and varying amounts of organic matter in the form of plant remnants and sand particles. This content decreases gradually with depth and the liquid limit and the water content in the clay increase to maximum values about 15 metres down in the clay layer. From this depth the trend becomes reversed and liquid limit and water content then decrease gradually with depth.

The full thickness of the clay layers has not been established in Section C.

However, the soil profile below the upper sand and silt layer has been found to first consist of clay of approximately the same type as in Section A with an original thickness of about 40 metres. Then follows an about 5 metre thick zone in which the clay gradually becomes more silty, contains thin silt layers, contains gradually thicker sand layers and then in the reverse order returns to clay again. The clay is then first of the same grey sulphide patched type as above. About 7 metres further down it turns into a brown-grey varved silty clay of a type that normally indicates a transition to varved glacial clay in the bottom layers. The zone with coarser soil is slightly inclined with a downward slope towards the river. In spite of the erosion, there are at least 35 metres of “homogeneous” clay on top of it also below the river bottom.

A comparison between the CPT test results indicates that the clay in Section C is of the same character at the same levels in all test points. However, in Section A the results at Point S4 are significantly different from the rest of the results in the section. Below 10 metres depth, the cone resistance is clearly lower and the pore pressure ratio is clearly different and generally higher. This indicates a more normally consolidated, a more silty and/or a more sensitive type of clay. That the clay at this point should have a significantly different stress history than that at Point S3 closer to the crest is unlikely, and the results therefore primarily indicate that something in the composition of the clay and/or its other properties is different.

The results of the routine investigations in the laboratory show that the upper sand and silt layer has a density of about 1.9 t/m³. It consists mainly of sand with an organic content. A few sample tubes with very even-grained material and incomplete saturation and/or large amounts of organic material showed lower values, particularly in Section A. The density in the silty clay directly below is about the same as in the water-saturated sand and silt layer. In Section C, the density then gradually decreases with depth as the silt content in the clay decreases and the water content increases. It reaches a minimum of about 1.65 t/m³ at a level of –5 metres, which is 25 metres below the original ground surface. From this level it gradually increases again to become about 1.85 t/m³ at a level of –45 metres. The density is significantly higher within the zone with infusions and layers of coarser soil where a maximum value of 2.16 t/m³ has been measured, Fig. 19.

The density in Section A in general follows the same trend, but here the differences are considerably greater between the sampling points, Fig. 20. The results indicate that the inclined contour of the firm bottom also has created inclined clay layers.

However, the variation in other parameters such as liquid limit and sensitivity indicates that the composition and properties of the clay also vary somewhat. The values that differ most from the general trends in the latter respect are found at Point S4, which is located 50 metres behind the upper crest.

The uniformity of the soil conditions in Section C is illustrated further when the natural water content and the liquid limit in the various boreholes are plotted versus the sampling level, Fig. 21. The different curves then overlap without any significant difference in values or trends except for a certain normal scatter.

A similar uniformity is not found in Section A. Instead the values vary between the sampling points, Fig. 22. Water contents and liquid limits of the same sizes as those on the same levels in Section C are measured at Point S2 below the excavation and at Point S13 below the river bottom. The values are lower at Point S3 just behind the upper crest and at Point S4 50 metres further away the liquid limit is much lower.

The variation in water content generally follows the variation in liquid limit except for the central parts of the clay layer at Point S4, where the water content is considerably higher than the liquid limit. The measured variations in water contents correspond to the measured variations in density.

Fig. 19. Measured density in soil samples from Section C.

Fig. 20. Measured density in soil samples from Section A.

-50

-50

Fig. 21. Measured water contents and liquid limits in Section C.

a) water content b) liquid limit

-25

Fig. 22. Measured water contents and liquid limits in Section A a) Water content

b) Liquid limit

An alternative and more sensitive way of presenting variations in the general state of the soil is to study its liquidity index. The liquidity index, I_L, is defined as

P

Where w_N= natural water content w_P = plastic limit

w_L = liquid limit

A determination of the liquidity index thus requires that the plastic limit be determined. For simplicity, a quasi liquidity index being the relation w_N / w_L may be used. This has previously been used in Sweden for correlation with sensitivity (Götaälvutredningen 1962), and has here been employed for both sections. The plot of the relation w_N / w_L in Section C shows the same very good correspondence for all points in the section, Fig. 23a. The same type of plot for Section A shows corresponding values for Point S2 below the excavation and Point S13 below the river bottom, Fig. 23b. This indicates a similar structure and “degree of consolidation”

in relation to the composition of the clay, which is reflected in the liquid limit. The values from Point S3 behind the upper crest are similar, but at this point there is a zone in the central part of the clay layer with significantly higher values. At Point S4, 50 metres away from the crest, the values are considerably higher throughout the profile with a maximum at the middle of the clay layer.

A high liquidity index often entails a high sensitivity. When the sensitivity measured in Section A is plotted against the sampling level in the same way as the relation w_N / w_L, the same pattern is obtained with an elevated sensitivity at the middle of the clay layer at Point S3 and a high sensitivity throughout the profile at Point S4, Fig. 24a. It is here very high in the middle of the clay layer. The relation is also illustrated when the relation w_N / w_L in Section A is plotted against the sensitivity, Fig. 24b. In Section C, the variation is too small to illustrate this correlation and here all values fall into the same group as the values for Points 2 and 13 in Section A, i.e. relations between 0.8 and 1.05 and sensitivities between 10 and 40.

The indication of another type of clay at Point 4, which was found in the results of the CPT tests, was thus confirmed by the routine tests in the laboratory.

a)

b)

Fig. 23. The relation water content/liquid limit in the different sampling points in the Torp area.

-25

Fig. 24. Sensitivity of the clay in Section A.

a) Sensitivity versus level

b) Relation w_N/w_L versus sensitivity

1