Soil investigations in the current project CPT tests

II Standard CPT tests

The previous soil investigations had included weight sounding tests down to 25 metres depth, dilatometer tests down to 18 metres depth, undisturbed sampling to 14 metres depth and pore pressure observations at three levels for about seven months, with intervals between the readings of one to two months. The results had given a good picture of the stratigraphy and soil conditions at the site, and the first tests in the new investigation were two CPT tests performed in order to verify that the soil conditions were uniform over the intended test area. This was readily established. The CPT test penetrated down to 28 metres depth with an ordinary drill rig and the results were very similar and in agreement with the previous results, Fig. 4.2.2.

The results of the CPT tests clearly identify the clayey layer between 4 and 5 metres depth, even if the generated pore pressures remain relatively low and indicate a fairly silty and/or overconsolidated soil. The more fine-grained layers between 9 and 15 metres depth are also clearly identified, as well as the coarser silt/fine sand between 15 and 20 metres depth where only small excess pore pressures are generated and the transition at 20 metres depth, after which no excess pore pressures are measured. The results indicate that the soil is loose to very loose down to about 15 metres depth, where it becomes medium dense.

II Seismic tests and excess pore pressure dissipation tests

Further CPT tests were then performed, one as a seismic cone test in order to measure the initial shear modulus at small strain, G ₀, and one CPT test with stops at every metre of penetration for measurement of the excess pore pressure dissipation with time in order to estimate the drainage conditions in the soil profile.

The results of the seismic cone tests showed initial shear moduli increasing fairly linearly with depth. In view of the almost constant bulk density and thereby fairly constant void ratio of the soil, this is in line with what would be expected from Hardin' s (1978) empirical relation. A direct comparison shows that the measured values are in the same range but in general somewhat less than the empirical values, Fig. 4.2.3.

SGI Report No 54

52

5"" < ~ ,.. ere· C!. 0 ~

...

::, a. 0 ... a. ,.. ~ s·

~-

-?% V, Q, ;;;- u-, w

P E £

l

94Xl27 Rl..o rotio l4 l8 15

l'O'J'lCPT test with measured parameters Site Borlooge DoteLocoloo/bd: Ctwd ..,.roce Ctwd wot.-level: 2.JO mu my. Ecwnent Geo\ech Desigootim BorcptlRelo-.-.:e levo: Ctwd ,afoce PreainJ dq,tlt OJ m Stops ii SOtl'"'f 28,43 >w,dx p-es. """5lfemen\: Yes Stal dq,llt 0.25 m

~

(\>etotor Project nunber:Hyaoslol< P. p-es.: No Slop dq,IJt 2a.J8 m lbservo\o,s; OYed flrn,,-.: lx:<epllcpl Preailed tm\O'id: lry a1.11\ Tip resistonce (Wo) Friction (kPo) Pc<e preSSU'e (kPo) Friction ratio (7.) Pore P.fesslJ'e qT fr u,6u and uRr DPPR0 50 XX)12 ~ 20 0 150 200-XXl 0 XXl 200 300 400 0 2 4 6 8 'O -02 02 0.6 to

j

~~

11

~ "'-

t

10 I ;,, I I I I I 5- 15 1 2 I I I I ~:,.

T ~ ~ 20 I I ::.t, I I I I

-

,!r- \2s

I

L...,J:?"'

I ! I

J- \... 30~-~-~-~-~-~ DesK110lions, see SGI lnforroolioo Fig. 4.2.2 a. Typical results of a CPT test in the test field at Vagverket.

Vl .i,.. Vl ~ ;x, (1) ,::, 0 ;:i z 0 <.n .is.

I

I I

I

^{P fie 'O E _;;-15 0. ., C, 30}

94'027 Rlo 90 '00 X X 15

CPT test evaluated according to SGI Info 15 Site Vogverket, Borlange Page Dote locotioo/levd: Grru-d Sltfc<e Grru-d wolo-level: 2..10 m u my. Eiwr,ent: Geoledl Designation Borcptl ~RefO'er<e levd: Preaii,g deptk 0.Jm Slop$ ii ,ou,oog: 28,43Grru-d "''"" ~otor Hy<rostotic P. pres: No Stq, deptlt 28.38 m tbs<rwtions: OYed nane: borcptlcpt

pros. rreaSltement Yes Slat deptk 015m

~

Project number: Pr,ailed ""t«id: lxy '"'"' Oossificotion Uncroined shear strength (kPo) Friction (Jl9le (0) Relative density (~) Modulus (I.Po) T~ · · · uncorr• • •corr t lo XXX M000 E2SAdditiord dossificotion 0 20 40 60 80 '00 12020 25 30 35 40 45 0 20 40 60 80 '00 0 'O 20 30 40 50 60 70 80 0~ I X IX ~ X X,l,P X HOC ""' F' I if ~ X X 0 XX ~ f<"

'

X X X

't

x• ,~ " g-X X 0 X 'q, : X X

l

11 §~;

... ,.

~ ~' l ~ "x l X{ 20~ ~:;..

...

' ;x""

...

,

...

~ X x-,1;, X ' 119_ l"x ~xbq'o X 0 X "#Oo25' .., xX X X"

:r

g,9

~

XX I 'o ' l<x

= ..,

0 XoP X tIJjSo f;;jSi [Il]a ~Org,P E81F 0esKJ'}Otioos, see SGI Information Fig. 4.2.2 b. Typical results of a CPT test in the test field at Vagverket.

- -

excess pore pressures had dissipated within 5 minutes, Fig. 4.2.4. The only notable exception was the clayey layer between 4 and 5 metres depth, where about 25 % of the generated excess pore pressure remained after 5 minutes. According to the classification proposed by Larsson and Mulabdic' ( 1991) this entails that, for normal loading conditions, the entire soil profile except for the clayey layer can be considered as free-draining. The clayey layer can in the same way be considered as semi-draining bordering on free-draining.

The corresponding times for 50 % pore pressure equalisation and the coefficients for consolidation at horizontal pore water flow, eh, evaluated according to Torstensson (1977) are shown in Fig. 4.2.5. According to Torstensson, the values evaluated theoretically in this way should be empirically corrected to yield representative values for practical use.

Investigations and load tets in silty soils 55

200

0 0

Time, seconds

50 100 150

Coefficient of consolidation, m ²/s

1.00E-06 1.00E-05 1.00E-04

Fig. 4.2.5 Measured time for 50% excess pore pressure dissipation and values of the coefficient of consolidation, eh ,evaluated according to Torstensson

1977.

In most of the pore pressure dissipation tests, the pore pressures were allowed to dissipate fully. An almost continuous profile of the in situ pore pressure was thereby obtained, which was also checked against the readings in the stationary piezometers, Fig. 4.2.6. The pore pressure profile measured at that time showed that the pressure in the bottom layers below about 11 metres depth corresponded to a hydrostatic pressure from a level 1 metre above the ground surface. The hydrostatic pressure head at higher levels started to decline at about 11 metres depth. A maximum rate of decline, corresponding to the lowest permeability, was observed at about 5 metres depth and the free ground water level in the upper layers was found at about 1.5 metres below the ground surface.

Owing to later problems with the ground water, the location of the plate load tests had to be moved about 15 metres and a new CPT test was then performed in order to verify that the uniform soil conditions extended also over that area. The results from all CPT tests are unanimous and yield almost identical pictures of the soil conditions and stratification. The results of the four tests performed with the standard equipment in this field are shown together in Fig. 4.2. 7.

Investigations and load tets in silty soils 57

Pore pressure, kPa

0 50 100 150 200

E

£ 8 - 1 - - - l - - - ~ - - ' - , - - - + - - - 1 - - - ~

c.

_Cl)

C

Fig. 4.2.6 In situ pore pressure distribution with depth in the test field at Vagverket measured in November 1994.

SGI Report No 54

58

Total cone resistance, Qr, kPa

0 5000 10000 15000 20000 25000

o

E :=~;:-+ ----+----+---j

-,.;:

Ji,..

-5 ~--... ""'5c.,_

--.-:"

_.. ^--=-- -...-

-10 ^~

^~

-E ~ ·

.J:.-c.

_Q) ¹⁵

~§:

C

20

25

30

Total cone resistance, qT, kPa

0 2000 4000 6000 8000 10000

0

-

'"---2

~

~~

'EC-4 ^~

~

_{~ ~}

6 41!-:~

E

£

-

^{a. Q) 8}

C

10 12 14 16

Fig. 4.2.7 Results of four CPT tests in the test field at Vagverket in terms of total cone resistance versus depth.

Investigations and load tets in silty soils 59

Pore pressure measurements

The pore pressure situation in the test field was measured by observation of the free ground water level at shallow depths at pre-drilling through the dry crust, by recording the in situ pore pressures in the coarser layers below 20 metres depth in the CPT tests and by measurements in piezometers installed at 3, 5 and 7 metres depth around the area intended for the plate load tests. These pore pressure observations, together with the measurements in the pore pressure dissipation tests at the CPT-soundings, were used to establish the prevailing conditions during the investigation period and in evaluation of the test results, Fig. 4.2.6. The pore pressures were then occasionally read in the piezometers during the following spring and summer in order to check whether there were any significant changes.

A certain increase in pore pressure was observed in May after snow melting and thawing, but for the rest of the summer the measured pore pressures were significantly lower than they had been during the investigation period and more or less corresponded to the situation found in the previous investigation.

Sampling and laboratory tests II Sampling

"Undisturbed" samples were taken with a Swedish standard piston sampler type St II in two bore holes down to 10 metres depth. In order to minimise the disturbance and to obtain a continuous soil profile, the samples were taken at every 0.7 metre of depth. This distance corresponds to the stroke of the sampler and in this way the sampling could be pe1iormed without pushing soil in front of the sampler during installation to the sampling level. The continuous soil profile was obtained by taking care also of the soil in the cutting edge and protecting tubes, which is normally thrown away. The sampling was stopped at 10 metres depth because this was well below any zone that could possibly be affected by the planned plate load tests.

II Classification tests

The samples were classified in the laboratory. In this process, the samples were investigated concerning bulk density, natural water content, fall cone liquid limit and plasticity index. A large number of sedimentation analyses were also per­

formed for determination of the grain size distribution.

The bulk density of the soil varies mainly between 1.9 and 2.0 t/m³ and the natural water content is around 30 %, which corresponds to a fully saturated soil. The liquid limit is close to the natural water content and the plasticity index was found to vary between 6 and 12 %.

SGI Report No 54

60

The visual inspection of the samples showed that the soil mainly consists of silt, but there are thin layers of clay and clayey silt embedded in the silt with a frequency that varies with depth. The upper 3 metres thus contain only a few seams of clayey soil. These seams then become more frequent and between 4 and 5 metres depth they dominate. The frequency then decreases again and between 6 and 9 metres depth the clayey seams are sparse. At 10 m depth they are again somewhat more frequent, Fig. 4.2.8.

Vagverket

Soil classification Water contents, % Bulk density, tfm3 10 20 30 40 1,6 1,7 1,8 1,9 2,0 2,1 2,2 O Top soil _______ _

1 Brown silt

Grey silt with occasional thin layers of clayey silt and clay

Grey layered silt with thin layers of clayey silt and clay

Grey layered silty clay/clayey silt with thin layers of silt

%

E ₍₁₎ ⁵

Grey layered silt with thin Cl layers of clayey silt and clay

Grey silt with occasional thin layers of clayey silt and clay

~riey silt with layers of clayey

0, ·c.,

l::,. natural water content I-plastic limit

8 fall cone liquid limit 0 bulk density

Fig. 4.2.8 Soil profile in the test field at Vagverket determined by laboratory tests and visual inspections in the current project.

Investigations and load tets in silty soils

10

61

According to the Swedish classification system, (Karlsson and Hansbo 1984), which is based on grain size distribution, all the soil in the profile except between 4 and 5 metres depth should be classified as silt. This is also in accordance with the behaviour of the soil in various simple shaking tests used for classification. Since the clay content is less than 10 %, the soil should not even be given the sub­

designation "clayey". Between 4 and 5 metres depth, the average clay content is about 30 % and the soil should accordingly be classified as a silty clay.

This classification is found to be quite different from what would be obtained by using the "Unified Soil Classification System" (ASTM 1992). The latter system uses the consistency limits of the soil to distinguish between clay and silt.

According to the USCS system, the majority of the soil samples would be classified as clay, a few would be classified as clay/silt and only one level as silt.

II Oedometer tests

A large number of oedometer tests were performed on the soil samples, both incrementally loaded tests and tests with constant rate of deformation, CRS tests.

Both types of tests were performed according to Swedish Standard, which in the case of incrementally loaded oedometer tests also conforms to internationally generally accepted procedures.

The incremental tests were performed with automatic registration, which enabled an almost continuous registration of the time-compression curves. Nevertheless, it proved impossible to measure any curvatures that would enable an evaluation of coefficients of consolidation, cv. Instead, within a few seconds after the load application the curves formed the typical straight lines in a compression-log time plot, indicating that there was no hydrodynamic time lag because of delayed excess pore pressure dissipation. Each load step in the tests was applied for 24 hours.

The CRS tests were run at the standard rate of deformation of 15 % per day normally used for clays. It is permitted to increase this rate somewhat in silts, but within limits. This is then done in order to obtain more reliable readings of generated excess pore pressures which enable a better evaluation of the permea­

bility of the soil. In the current case, no significant excess pore pressures were measured and it would therefore have been necessary to increase the rate of deformation by more than one order of magnitude to measure this property, which was not attempted. The only exceptions were samples from the clayey layer between 4 and 5 metres depth, in which small excess pore pressures were measured also at the standard rate. These measurements indicated that the natural vertical permeability of this soil should be about 3-10-⁹ m/s. Several tests were run on

SGI Report No 54

62

specimens from this level and yielded approximately the same value. However, this value must be expected to be too low to represent the behaviour of the bulk volume of the soil at the level. This is partly due to the fact that the most clayey parts were intentionally selected for the oedometer tests in the hope that they had retained their structure and preconsolidation pressure and partly to the horizontal layering of the soil with more permeable layers providing better horizontal drainage.

The interpretation of the separate oedometer curves was difficult. Only in the more clayey layer between 4 and 5 metres depth was it possible to estimate a preconsol­

idation pressure and to clearly obtain the typical shape of the oedometer curve for an undisturbed soil, i.e. an initially stiffer modulus up to a yield stress, then a softening and again a stiffening as the vertical stress increases. The estimated preconsolidation pressures indicated that the soil in this layer is somewhat overconsolidated, with an overconsolidation ratio in the range of 1.5 to 2. The minimum compression modulus after passing the preconsolidation pressure in this layer is around 2.7 MPa.

In the rest of the CRS tests, weaker indications of preconsolidation pressures and constant moduli just after passing these had to suffice or the modulus had to be estimated using the evaluated modulus number M ', (=L\MI L\o-), and the in situ effective vertical stress.

The incrementally loaded oedometer tests were even more difficult to evaluate separately, since no continuous stress-strain relation is obtained but only a number of scattered points along this curve. They can, however, be evaluated in the same way as the CRS tests using the curve pattern from these tests and curve fitting calculations for support.

Altogether, more than 30 oedometer tests were performed and from the combined results there emerged a picture of the compressibility which is well in line with what might be expected with regard to the stratigraphy and the classification of the soil, Fig. 4.2.9. This was possible mainly through the use of CRS tests and also the large number of tests, which permitted evaluations to be made from individually weak indications together forming a consistent pattern. The modulus numbers evaluated from the tests all fall within the range of 30 to 100, which is to be considered as typical for silt, (Janbu 1970).

Investigations and load tets in silty soils 63

Modulus, kPa

Undrained triaxial tests were performed in the SGI laboratory. From the stress paths from these tests, effective friction angles at constant volume, <P'cv' in the range of 35° to 37° were evaluated, with the lower values found in the more clayey parts of the profile. These values are within the typical range for silts (Larsson 1995).

The tests on undisturbed samples showed that the soil had a tendency to dilate at undrained shear, which resulted in relatively high undrained shear strengths, Fig. 4 .2 .10. This tendency has been found to be normal in natural deposits of silt, also when they are relatively loose, (Borgesson 1981).

Samples taken in the test field at Vagverket were also sent to the Norwegian Geotechnical Institute in connection with anotherresearch programme concerning the effect of cyclic loads on the behaviour of silts. In this programme, a number of undrained triaxial compression tests with ordinary monotonic loading were also run and similar values of <P'cv were found in these tests.

SGI Report No 54

64

Undrained shear strength, kPa

0 50 100 150 200

0 I 21

I

4

E 6

£

-

c. ^Q) 8 0

10 Fig. 4.2.1 0 Results of

undrained triaxial tests 12 and field vane

shear tests in 14 , --Field vane shear test

the test field

- Triaxial test at Vagverket. 16

Field vane shear tests

Field vane shear tests were performed in two bore holes according to the recommended standard procedure, (SGF 1993). The average "undrained" shear strength determined in this way ranged from 20 to 60 kPa in the upper 15 metres of the profile, Fig. 4.2.10. The lower value of 20 kPa was measured at 4 metres depth in the more clayey layer, where some relevance for the results may be expected. For most of the profile, the tests can hardly be expected to have been undrained, judging from the results of the pore pressure dissipation tests, and their relevance must be questioned. Both above and below the clayey layers, the vane tests yielded considerably lower undrained shear strengths than the undrained triaxial tests. The reason for the high strength values in triaxial tests is the aforementioned tendency to dilatant behaviour at undrained shear, which accord­

ing to Borgesson (1981) is typical for natural silt sediments in Sweden.

Investigations and load tets in silty soils 65

Dilatometer tests

Two dilatometer tests were performed in addition to the test in the previous investigation. Both tests were performed down to 27 metres depth and the dilatometer could thus be pushed down to approximately the same depth as the CPT probe when using the same drill rig. The results from the tests are very uniform, indicating mainly silt in the profile but with clay layers between 4 and 5 metres depth and 9 and 10 metres depth. Below 15 metres depth, the soil becomes classified as silty sand. The upper 15 metres of the profile are classified as loose or very loose and the soil below as medium dense, Fig. 4.2.11.

The results in the clayey layers yield very low values of the material index / D'

which clearly indicates that the soil has been heavily disturbed at insertion of the dilatometer. This is often the case when the soil is varved with alternating thin layers of silt and clay. Consequently, the evaluated compression moduli at these levels become very low and should be disregarded. In accordance with the SGI recommendations, (Larsson 1990), they should preferably be replaced by moduli from oedometer tests or alternatively, provided that the soil has a certain overcon­

solidation, by empirical relations based on the undrained shear strength and the material index. Apart from in the obviously disturbed layers, the results indicate a small overconsolidation in the profile.

A comparison between the compression moduli evaluated from the dilatometer tests and the oedometer moduli, Fig 4.2.12, shows that in general they are very similar. For the layer between 9 and 10 metres depth, use of the empirical estimation raised the values of the moduli from incredibly low values of about 0.3 MPa to about 3 MPa which is a considerable improvement but not enough to reach the values of about 12 MPa obtained in the oedometer tests. The significance

A comparison between the compression moduli evaluated from the dilatometer tests and the oedometer moduli, Fig 4.2.12, shows that in general they are very similar. For the layer between 9 and 10 metres depth, use of the empirical estimation raised the values of the moduli from incredibly low values of about 0.3 MPa to about 3 MPa which is a considerable improvement but not enough to reach the values of about 12 MPa obtained in the oedometer tests. The significance

In document LARSSON ROLF (Page 54-74)