Pile Info

TRADITIONAL SITE INVESTIGATION TECHNIQUES
Excerpt from "Swedish National Report", International Symposium on Cone Penetration Testing, CPT´95 Vol.1

The investigation of soil strata by penetration testing has a long tradition in Sweden. As far back as 1914-1922, the Geotechnical Commission of the Swedish State Railways developed and standardized the Swedish Weight Sounding Method. (Note: In the following, the term "sounding" is used in the same sense as "penetration testing"). With the aid of this equipment, in combination with sampling, it became possible to obtain reliable information about the stratification of soil deposits. Over the years, the weight sounding method has become the dominant soil investigation tool in Sweden. Even today, this method is the most commonly used soil investigation technique in Sweden. During the last 20-30 years, the testing equipment has been mechanized. The screw-shaped penetration point, however, has maintained its initial shape as designed by the Geotechnical Commission of the Swedish State Railways. The standard for the Swedish Weight Sounding Method was approved by the Swedish Geotechnical Society in 1974.

With the aid of a mechanical recorder, the total penetration force is automatically registered as a function of the depth. The Geotech Static Penetrometer provides detailed information on the thickness and relative strength of the penetrated soil layers.

Different dynamic probing methods were used very early in Sweden. Around 1940, the Swedish firm Borros designed and developed the Swedish Ram Sounding Method. This method has been commonly used for predicting of the length of end-bearing, driven piles.

Based on extensive investigations by Bergdahl and Dahlberg (1974) the traditional ram sounding method was revised and an improved method and procedure - Method A - was standardised in 1973. The basic elements of the Swedish ram sounding test are described in Figure 5. A free-falling hammer (weight 63.5 kg) strikes a fixed anvil, equipped with a rubber cushion. The height of the fall is 0.5 m In order to reduce the skin friction, the extension rods (32 mm) are rotated two turns every 0.2 m of penetration.

A more sophisticated testing procedure includes measurement of the torque required to turn the extension rods. Based on such measurements, the skin friction can be separated from the total penetration resistance, which in turn enables a more detailed interpretation of the test results.

The Swedish Ram Sounding Method is useful tool for predicting of the length of both end-bearing and friction piles driven in sand and gravel.

Since around 1950, heavy rock drilling machines have been utilized for Soil and Rock Drilling. The main purpose of the Soil and Rock Sounding method is to determine both the depth to bedrock and the quality of the underlying rock. In principle, this type of investigation is carried out with the same equipment as that used for conventional rock drilling. Light, medium or heavy rock drilling machines are used, depending on both the soil/rock conditions and the required maximum depth of investigation.

The time in seconds required for 20 cm of penetration is recorded. This parameter gives an indication of the hardness of the rock or the compactness/composition of the penetrated soil. In accordance with the existing recommendations, the soundings shall extend 3 to 5 m into assumed rock in order to ensure that the borehole does not terminate in a boulder.

The Swedish Geotechnical Society has proposed a standard for the Soil and Rock Sounding method, which includes items such as the outside/inside diameter of the drill rod and casing, size and shape of drill bits, rotation rate, torque and axial force applied to the drill rods, pressure and flow rate of the circulation fluid.

Today the use of in situ methods, such as pressometer and dilatometer are steadily increasing in Sweden.

Parallel to the testing described above, sampling is also carried out by routine in order to obtain additional information about soil types and characteristic soil parameters.

TYPE OF CPT EQUIPMENT USED IN SWEDEN
The Static Penetrometer described in the preceding section may be called a mechanical CPT. Improvement of electronic and computer technologies has facilitated the development of various CPT systems. In Sweden, normal and fairly small drill rigs are used for all types of field investigation. This equipment is relatively compact and can easily reach difficult areas. This has forced the Swedish manufacturers to produce more flexible CPT systems, which can easily be installed on any type of drill rig with capacity to push the probe. Today, there are two manufacturers of CPT equipment in Sweden: ENVI AB and GEOTECH AB. Both produce cordless CPT probes, but using different measuring and recording principles. Almost all CPT tests in Sweden are performed with one of these systems. Other equipment, such as Hogentogler. Van den Berg or GMF, has occasionally been used, mainly for comparison in research projects or measurement of additional parameters.

Equipment
The GEOTECH CPT system is cordless. Instead of a cable, a sound transmitter on top of the CPT converts analogue signals (three or four channels) into a digital coded signal. The signal is transmitted through the rods up to the drill rig or pusher at ground level. A microphone receiving the signal is installed on the feeder of the drill rig. The signal then goes by cable to an interface connected to a PC or some other logger. The system provides full information in real time during the test. Since the hole in the center of the rods is not used for a cable, it is possible to use it for pumping down Bentonite mud and discharging it just above the probe. This reduces the friction along the rods dramatically, and consequently the total force needed to push the probe can be reduced by more than 50%. GEOTECH also manufactures multipurpose mobile drill rigs, from which the CPT is operated.

ENVI MEMOCONE is a CPT system which works with or without a cable. During the test, all data is stored in the internal memory of the probe. The readings are stored once a second for a maximum period of 4.5 hours. On the drill rig, time and depth are recorded. The probe is pushed until refusal, retracted, and the data is transferred into a PC interface or a GEO-PRINTER. The probe can be stopped for a period in order to monitor pore pressure dissipation. This operating principle makes it possible to use CPT at great depths, and the results cannot be affected by disturbance in data transfer. The cone can be operated using the same drill rods as are used for other purposes on the drill rig.

BORRO is the manufacturer of a continuous 20 ton CPT pusher. The rig contains two chucks operating in sequence in such a way that the penetration is uninterrupted. The electronic control panel provides information on rate of penetration, depth, total load in kN and the setting for maximum load cut-off limit. The rig has also been used underwater.

National Codes and Standards
On June 15 1992, the Swedish Geotechnical Society presented a new recommended standard for CPT testing. This standard replaced a standard from 1979. The earlier standard did not include pore pressure measurement and was mainly intended for mechanical CPT testing.

The probe consists of a cylindrical cone with a cross sectional area of 1,000 mm² and an apex angle of 60°. This is pushed vertically into the ground at a constant speed of 20 mm/sec.

The Swedish recommended standard has three different classes depending on factors such as the precision requirements CPT1, CPT2 and CPT3, where class 3 is the highest requirements. The ranges for application of the various test classes are shown in Table 1.

In general, the accuracy, taking into account all possible sources of error (parasitic frictions, errors of measuring devices, eccentricity of the loads on the cone and the sleeve, temperature effects, etc.), shall be better than:

• 2 % of the typical measured value (the average value) for any of the soil layers* in which the results are to be interpreted in terms of classification and soil properties
• 1 % of the measured values for static pore pressures
*) the term "soil layer" in this context refers to separate soil layers or, in case of thick homogeneous layers, depth intervals of one metre

For the different test classes, there are lower limits for the required precision in terms of generally accepted inaccuracies according to Table 1. The precision of a cone must be verified by regular calibration. Calibration procedures have been worked out by Mulabdic et al (1990)

Data format standard
The need for a modem method of interchanging geotechnical information has stimulated the development of a data format standard for geotechnical measurements. This also applies to CPT, and since 1990 there exists a standard recommended by the Swedish Geotechnical Society. Since the 1994 revision, the standard covers the interchange between field systems and office as well as between offices. All geotechnical software used in Sweden today conform with this file standard.

INTERPRETATION OF TEST RESULTS
The most commonly used method of interpretation was worked out by Larsson (1993). Generally, the interpretation of tests results in coarse materials is based on international experience, after some minor modifications. For tests performed in soft fine grained soils, the interpretation relies on semi-empirical data obtained from research carried out in Sweden.

Stratigraphy
To obtain information on soil stratigraphy, the following parameters, measured or derived, are evaluated:

• Total cone resistance, q_T• Total sleeve friction, f_T• Total pore pressure, u

The basic parameters q_T and f_T are obtained after correction of the measured values for pore pressure effects. For interpretation of the results, the following basic parameters are also required:

• Initial in situ pore pressure, u_O
• Initial vertical stress in situ, s_VO (calculated from the density of the soil)

The initial pore pressure may be obtained from supplementary pore pressure measurements made at a number of levels, or by observations of the ground water level in more permeable layers measured during temporary stops in the penetration test.

The initial vertical stress in situ is estimated by using the density of the soil. This estimate can often be by interaction using the classification of soil type and stiffhess obtained from the test results to estimate an approximate density.

Different relations between the basic parameters are used for interpretation of the test results. For a preliminary interpretation, the following parameters are often used:

• Du = u – u₀• Friction ratio, R_f= (f_T/q_T) -100, %
• Differential pore pressure ratio, DPPR = Du/q_T

All the parameters mentioned above (measured, corrected and calculated) are usually plotted against the penetration depth. From this plot, there are excellent possibilities for evaluating the stratigraphy.

Soil classification
The properties of the soil affect the magnitude of the cone resistance, the sleeve friction and the generated pore pressure measured during CPT tests in different ways.

A preliminary classification of the soil can be made using the relations (q_T - s_v0)/ s´_v0 and f_T/(q_T - s_v0). The soil density used for calculating the parameters included in these relations may be obtained from sampling, dilatometer tests or by an iteration process in the interpretation of the CPT test. The soil is usually divided into layers of a given thickness in order to perform the calculation. The principal character of the soil (sand, silt or clay/organic soil) is thereafter estimated by using the diagram presented in Figure 6.

Figure 6. Diagram for evaluation of silt and sand and separation of cohesive soil.

The preliminary soil classification based on CPT tests is thus mainly made on the basis of the relation between the cone resistance, the sleeve friction and the normal in situ stress conditions. In soft clay, the measured sleeve friction is very small and relatively unreliable but in overconsolidated clay, where the cone resistance may be of the same size as for soft, coarser soil, the measured values normally become larger and more reliable. Possible uncertainties in the measurements of sleeve friction normally have a relatively small influence on this division. The main exceptions are highly sensitive clays and/or silty clays. In these soils, the sleeve friction may be very low, at the same time as the measured stiffhess in relation to the overburden pressure places the soil in the region for silt according to the diagram in Figure 6. However, in these soils very high pore pressures are often developed in the tests and a check on whether the factor B_q =Du/(q_T – s_v0) is higher or lower than 0.6 can be used to judge whether the soil should be classified as silt or clay.

In those cases where the soil has been classified as "clay/ organic soil" in the first diagram or after further checks, or if this has been specified at an earlier stage, the classification process passes over to the special classification chart developed for this type of soil. Figure 7. This chart is based on the parameters net cone resistance (q_T – s_v0) and B_q = Du/(q_T - s_v0).

The soil in the groups is also summarily classified with respect to the undrained shear strength as very soft, soft, medium stiff, stiff or very stiff. This division is preliminary because the relation between the net cone resistance and the shear strength among other things depends on the consistency limits of the soil. After a later, more careful evaluation of the undrained shear strength, the subdesignations with respect to the undrained shear strength should be adjusted according to Table 2.

Soil parameters and other data
Correlation between results from the CPT tests and in situ test or qualified laboratory tests have provided relations for evaluation of undrained shear strength in fine-grained soils. The relation between the undrained shear strength and the net cone resistance is sensitive to the liquid limit of the soil, w_L

The undrained shear strength evaluated in this way corresponds directly to the undrained shear strength obtained from corrected field vane and fall cone tests, direct simple shear tests and dilatometer tests.

If values of the liquid limit of the soil are missing, a coarser estimate of the undrained shear strength can be made according to

The accuracy of the evaluation of the undrained shear strength depends mainly on the accuracy achieved in the test. When very accurate estimates are required, the results from the CPT tests should be calibrated locally against field vane tests and preferably also direct simple shear tests.

A preliminary estimate of the preconsolidation pressure in cohesive soils can be made from the net cone resistance:

The relation is sensitive to both the plasticity and the overconsolidation ratio of the soil. The equation is solved by iteration involing the insertion of the estimated in situ effective vertical stress s´_v0 and OCR = s´_c/ s´_v0

The relative density of relatively even-graded sand with a mineral composition of quartz and feldspar may be obtained from the relation between the cone resistance and the effective over burden pressure (Lancelotta 1983). For coarser soils, the relative density can be calculated using the same relation and applying a reduction of 10 - 15 % according to Lunne and Christoffersen (1983).

Also the friction angle in friction material can be evaluated from the relation between the cone resistance and the in situ effective vertical stress. In reality, it is mainly the horizontal stress that governs the relation and, before the evaluation is made, some kind of estimate of the coefficient of earth pressure should be made (Marchetti 1985).

The evaluation of deformation properties is normally not made directly from the CPT test results. In friction soils, some of the well established empirical calculation methods specially produced for the CPT test are used. However, these methods are only suited for calculation of settlements and bearing capacity in normal cases of shallow foundations on sand. For other cases of loading on sand, compression moduli and moduli of elasticity may be estimated according to Robertson and Campanella 1988).

Compression characteristics in cohesive soil should not be evaluated directly from the results of CPT tests, which in this type of soil are performed under almost fully undrained conditions. For evaluation of these properties, undisturbed sampling and oedometer tests are required.

Evaluating software
The use of CPT in Sweden has increased after the Geotechnical Institute developed the CONRAD programme for evaluating of CPT data, (Larsson et al, 1995). The software is interactive and permits use of information from other investigation methods and sampling. The software evaluates different soil parameters according to the methods mentioned above.

USE OF CPT IN GEOTECHNICAL DESIGN

Design concept
Design of shallow and deep foundations in Sweden is carried out using partial safety factors. An analysis shall be performed for the working load (concerning mainly deformations) as well as for the ultimate load (failure conditions).

Calculations are made using the design values obtained from characteristic values via the method of partial safety factors and according to the following relations:

Determination of geotechnical parameters from cone penetration test
Characteristic values of geotechnical parameters shall be based on average (medium) values. In some instances, however, it is permitted to correct these values with respect to the effects of the investigation method, variations in time etc. In friction soils, these characteristic values are usually determined from penetration tests or other in situ tests. Table 4 shows the correlation between different sounding methods and typical geotechnical parameters. It is also more common today to evaluate parameters from the relations described in section 5.3.

Bearing capacity of pile foundations
The design bearing capacity Rd of friction piles in sand is composed of the contribution from shaft resistance Rmd and base resistance Rsd.

The shaft and base resistance can be estimated from cone penetration tests according to the following relations:

a_m Pile shaft capacity factor, c.f. Table 3
q_ci  Cone penetration resistance, average value within the layer
A_mi Pile surface area within the layer
a_s   Pile base capacity factor, c.f. Table 3
q_qs   Cone penetration resistance at pile base, as average value in the depth interval of ± 4 pile iameters measured from the pile base,
A_s Pile area at the base

Table 3. Pile shaft and base capacity factors as and am for driven piles, based on cone penetration tests in friction soils

Cone resistance values higher than 10 MPa may not be used with this design method. Also the method proposed by de Ruiter and Beringen (1979) can be used for designing pile capacity. The base resistance is determined as an average value according to the relation below:

where

I average cone penetration resistance between the level of the pile base and a distance of 0.7 and 4 pile diameters below
II minimum cone penetration resistance between the level of the pile base and a distance of 0.7 and 4 pile diameters below
III average cone penetration resistance between the level of the pile base and 8 pile diameters above.

This method gives similar results, but the maximum design values are limited to 120 kPa for shaft resistance and 15 MPa for base resistance, respectively. When sleeve friction measurements are carried out during the cone penetration test, these values can be used directly for assessment of pile shaft resistance. De Ruiter and Beringen have also proposed limiting values for pile base resistance, which are dependent on soil type.

Generally, the cone penetration test is not used for assessment of pile bearing capacity in cohesive soils.

COMPARISON AND CORRELATION OF CPT WITH OTHER INVESTIGATION METHODS
Using results from a comparative investigation at 14 test sites, a correlation between different site investigation and laboratory methods has been obtained by Bergdahl and Ottosson (1988). The results from this study are summarised for common Swedish sounding methods in Table 4. Both the weight sounding test and dynamic probing type HfA are standardised in Sweden. Further elaboration of the table has been carried out for various purposes, but the contents are almost the same.

In fine silt, the above values should be verified by laboratory tests or other in situ tests (pressure meter or dilatometer). In silty and clayey soils or sedimented soils with some organic content, penetration tests shall be complemented by sampling and compression tests or other in situ tests.

A relation between dynamic probing and SPT has also been established. The blow count for 0.2 m penetration for dynamic probing type HfA is equal to the blow count for 0.3 m penetration for SPT performed in accordance with the European standard.

Table 4. Characteristic values of friction angle and E-modulus for naturally deposited friction soil based on penetration tests.

¹⁾For silt, reduce the value by 3 ; for gravel, increase the value by 2^o .

²⁾ Values correspond to settlements after 10 years. Some investigations suggest that these values may be 50 % lower in silt and 50 % higher in gravel. In overconsolidated soils, the modulus values may be significantly higher. In settlement estimates at loads exceeding 2/3 of the ultimate failure load, the modulus should be reduced by 50 % for higher loads.