Chapter 2. SOIL INVESTIGATION METHODS < Previous | Table of contents | Next > 2.1 Standard Penetration Test (SPT) The SPT is a well-established and unsophisticated method, which was developed in the United States around 1925. It has since undergone refinements with respect to equipment and testing procedure. The testing procedure varies in different parts of the world. Therefore, standardisation of SPT was essential in order to facilitate the comparison of results from different investigations. The equipment is simple, relatively inexpensive and rugged. Another advantage is that representative but disturbed soil samples are obtained. The reliability of the method and the accuracy of the result depend largely on the experience and care of the engineer on site. A split-barrel sampler is driven from the bottom of a pre-bored hole into the soil by means of a 63.5 kg hammer, dropped freely from a height of 0.76 m. The diameter of the pre-bored hole varies normally between 60 and 200 mm. If the hole does not stay open by itself, casing or drilling mud should be used. The sampler is first driven to a depth of 15 cm below the bottom of the pre-bored hole, then the number of blows required to drive the sampler another 30 cm into the soil, the so called N30 count, is recorded. The rods used for driving the sampler should have sufficient stiffness. Normally, when sampling is carried out to depths greater than around 15 m, 54 mm rods are used. The quality of test results depends on several factors, such as actual energy delivered to the head of the drill rod, the dynamic properties (impedance) of the drill rod, the method of drilling and borehole stabilisation. The actually delivered energy can vary between 50 - 80% of the theoretical free-fall energy. Therefore, correction factors for rod energy (60 %) are commonly used, Seed and De Alba (1986). The SPT can be difficult to perform in loose sands and silts below the ground water level (typical for land reclamation projects), as the borehole can collapse and disturb the soil to be tested. The following factors can affect the test results: nature of the drilling fluid in the borehole, diameter of the borehole, the configuration of the sampling spoon and the frequency of delivery of the hammer blows. Therefore, it should be noted that drilling and stabilisation of the borehole must be carried out with care. The measured N-value (blows/0.3 m) is the so-called standard penetration resistance of the soil. The penetration resistance is influenced by the stress conditions at the depth of the test. Peck et al. (1974) proposed, based on settlement observations of footings, the following relationship for correction of confinement pressure. The measured N-value is to be multiplied by a correction factor CN to obtain a reference value, N1, corresponding to an effective overburden stress of 1 t/ft2 (approximately 107 kPa), N1 = N . CN (1) where CN is a stress correction factor and p' is the effective vertical overburden pressure. CN = 0.77 . log10 (20/p') (2) Seed (1976) proposed a similar correction factor for the assessment of liquefaction problems in loose saturated sands. This relationship was developed for earthquake problems and is based on extensive laboratory tests on mainly loose to medium dense sands, CN = 1- 1.25. log10 (s 0'/s 1') (3) where s 0' is the effective overburden pressure (in t/ft2) and s 1' is the reference stress (1 t/ft2). The correction of SPT results with respect to the effective overburden pressure is of importance for the evaluation of compaction results. Therefore, consideration should be given to this aspect when compaction criteria are to be based on N-values. Unfortunately, this fact is not always appreciated. The resistance (N30) has been correlated with the relative density of granular soils. Sand and gravel can be classified as shown in Table 1, Broms (1986). Relative Density Standard Penetration Resistance  (N30, blows/0.3 m) Loose £ 10 Medium Dense 10 - 30 Dense ³ 30 Table 1. Classification of sand and gravel after Broms (1986) The Standard Penetration Test is mainly used to estimate the relative stiffness and strength (bearing capacity) of soils. Deformation characteristics of granular soils can be estimated from empirical correlations, Peck et al. (1974). It is also possible to get some indications from SPT of the shear strength in cohesive soils. The SPT used frequently for the evaluation of the liquefaction potential of water-saturated, loose sands and silts in seismic areas, Seed and De Alba (1986).

2.2 Cone Penetration Tests (CPT)

The CPT was invented and developed in Europe but has gained increasing importance in other parts of the world, especially in connection with soil compaction projects. Different types of mechanical and electric cone penetrometers exist but the electric cone is most widely used. A steel rod with a conical tip (apex angle of 60° and a diameter of 35.7 mm) is pushed at a rate of 2 cm/s into the soil. The steel rod has the same diameter as the cone. The penetration resistance at the tip and along a section of the shaft (friction sleeve) is measured. The friction sleeve is located immediately above the cone and has a surface area of 150 cm2. The electric CPT is provided with transducers to record the cone resistance and the local friction sleeve.

A CPT probe, equipped with a porewater pressure sensor is called CPTU. It is important to assure complete saturation of the filter ring of the porewater (piezo) element. Otherwise, the response of the piezo-transducer, which registers the variation of pore water pressure during penetration, will be slow and may give erroneous results. The CPTU offers the possibility to determine hydraulic soil properties (such as hydraulic conductivity - permeability) but is most widely used for identification of soil type and soil stratification. The CPT can also be equipped with other types of sensors, for example vibration sensors (accelerometer or geophone) for determination of vibration acceleration or velocity. The "seismic cone" is not yet used on a routine basis but has, because of the relative simplicity of the test, potential for wider application especially on soil compaction projects.

The CPT is standardised and the measurements are less operator-dependent than the SPT, thus giving more reproduceable results. The recent geotechnical literature contains comprehensive information about different types of cone penetration tests, detailed descriptions of the test procedures and data evaluation/interpretation, Lunne et al. (1998). Therefore, only some aspects of importance for soil compaction projects will be discussed below.

The CPT measures the cone resistance q_c and the sleeve friction f_s from which the friction ratio, FR can be determined. FR is the ratio between the local sleeve friction and the cone resistance, expressed in percent (f_s/q_c). In spite of the limited accuracy of sleeve friction measurements, the valuable information, which can be obtained in connection with compaction projects, has not yet been fully appreciated. As will be discussed below, the sleeve friction measurement reflects the variation of lateral earth pressure in the ground, and can be used to investigate the effect of soil compaction on the state of stress, as will be discussed later. Cone and sleeve friction measurements are also strongly affected by the effective overburden pressure. It is necessary to take this effect into account, similar to the SPT. A correction factor CM for the cone resistance was proposed by Massarsch (1994),

CM = (100 / s m')0.5

(4)

where s m' is the mean effective stress. It should be noted that for SPT correction the overburden pressure is used, which does not take into consideration the effect of lateral earth pressure. The mean effective stress can be determined from

s m' = s v' (1+2 Ko) / 3

(5)

where s v' is the vertical effective stress and K_o is the coefficient of lateral earth pressure at rest. Figure 3 compares the different correction factors for SPT and CPT, respectively. It should be noted that for Massarsch uses the mean effective stress while the SPT correction factor is based on the vertical effective stress.

Figure 3. Stress correction factors C_N (SPT) and C_M (CPT) with K_O = 0,57, cf. equation (1, 2 and 3)

The measured cone penetration resistance, qc can be corrected for the effect of the mean effective stress sm' at any given depth

qco = qc CM = qc (100 / s m')0.5

(6)

where qco is the normalised cone penetration value. When considering the difference between the vertical and the mean effective stress, there is good agreement between the different correction factors for the SPT and CPT, respectively. However, as will be shown later, soil compaction can significantly increase the horizontal stress and this effect should be taken into consideration when evaluating the densification effect. It is recommended to limit the correction factor C_M to a value of 2,5. It is suggested that the corrected cone resistance qco be used for specification of compaction criteria, as this will assure more homogeneous soil layers and avoid unnecessary overcompaction close to the ground surface.

One important objective of the CPT investigations in connection with soil compaction is to obtain information concerning soil stratification and variation in soil properties both in horizontal and vertical direction. The friction ratio is often used as an indicator of soil type (grain size) and can provide valuable information when evaluating alternative compaction methods.

Measurement of the excess pore water pressure with the CPTU can detect layers and seams of fine-grained material (silt and clay). It is also possible to obtain more detailed data information concerning soil permeability and thus soil stratification.

2.3 Comparison between SPT and CPT

In areas like North America, many geotechnical engineers have developed considerable design experience, based on local correlation between the SPT and performance of foundations. Therefore, engineers may feel more comfortable by converting CPT data to equivalent SPT N-values and then comparing these with their SPT-based design methods. A considerable number of studies have been carried out in the past to quantify the relationship between SPT N-value and CPT cone bearing resistance (q_c), Robertson and Campanella (1986). The author has compiled data from various case histories, which are compared in fig. 4 with the data published by Robertson and Campanella (1986).

In spite of some scatter between individual test points and different soil types, there exists a clear correlation between the q_c /N ratio and the mean particle size.

2.4 Pressuremeter Test (PMT) The PMT was invented by the Menard in 1962 - 1963 in France, where this test is widely used. National standards exist and geotechnical design is based almost exclusively on this type of test. Over the years the PMT has been further developed in France, the United Kingdom and Japan, and has found increasing acceptance in several countries. However, the PMT is still a specialist tool, which requires experience in test performance and data interpretation. The standard pressuremeter is either inserted into a pre-bored hole or directly jacked or driven into the ground. A slotted tube protects the measuring cell, which consists of a cylindrical rubber membrane. In order to reduce the influence of soil disturbance during probe insertion, the self-boring pressuremeter was developed. This type of pressuremeter is, however, limited to fine-grained soils, while the standard pressuremeter can be used in most soil types. A detailed description of the PMT is beyond the scope of this paper. However, guidelines for data evaluation and interpretation as well as design recommendations were published by Baguelin et al. (1986). The PMT is an intermittent test and can thus not provide a continuous profile. The test is comparatively time-consuming and thus expensive. Figure 4. Variation of qc /N ratio with mean grain size (note that the cone resistance in MPa) From the PMT, a stress-deformation curve (applied pressure vs. volumetric strain) can be obtained in situ. From this curve, a deformation modulus and a value of the limit strength can be obtained. Also the "at rest" lateral earth pressure can be estimated, which is of considerable interest for soil compaction project. Few correlations exist between the PMT and other in situ tests. Because of the necessity to drill borehole, the quality of the test results may be suspect in loose sand below the ground water table.

2.5 Dilatometer Test (DMT) The dilatometer test (DMT), which is a simple and reliable in-situ testing tool, was developed in Italy and later introduced in Europe and North America, Marchetti (1980). The dilatometer consists of a flat, 15 mm thick and 95 mm wide blade and has a length of 220 mm. A flexible, stainless steel membrane, 60 mm in diameter, is located on one face of the blade. Inside the steel membrane there is a pressure chamber and a distance gauge for measurement of the movements of the membrane when the pressure inside is changed. The probe is pushed into the soil with the aid of hollow sounding rods and does not require drilling of a hole. When the membrane is inflated, the pressure required to just lift the membrane off the sensing device (p0) and to cause 1.10 mm deflection (p1) are recorded. As the pressure is released and the membrane returns to its initial lift-off position, another reading can be taken. The pressure values p0 and p1 can be used to define three index parameters. Marchetti (1980) calls these parameters the material index (ID), the horizontal stress index (KD) and the dilatometer modulus (ED), respectively. ID = ( p1 - p0 ) / (p0 - u0 ) (7) where uo = pore water pressure at rest (not excess pore pressure). The ID value varies from about 0.6 to 1.8 for silt and is about 1.8 for sand. The DMT is especially suited for monitoring of compaction projects as it can be used to assess the deformation characteristics of soils. From these index values, empirical relationships have been developed to determine geotechnical parameters. For instance, assuming that the soil behaves elastically, the dilatometer modulus can be deduced from the relation: ED = 48.1 ( p1 - p0 ) (8) The dilatometer modulus is commonly used to assess the compression (oedometer) modulus M of sand, silt and clayey silt. Experience has shown that the following relation obtains a good estimation of M M = 1.1 Rm ED (9) where Rm varies depending of the soil. Schmertman (1986) has suggested design procedures for settlement estimates based on the DMT.

2.6 Seismic tests Conventional seismic tests, such as wave refraction measurements, have been used in the past primarily for soil and rock layer identification. However, during the past decade, several new seismic in situ tests have been developed and applied successfully on a variety of soil compaction projects, Massarsch and Westerberg (1995). The "seismic" cone penetrometer, which has been briefly discussed above, incorporates a small rugged velocity sensor in an electronic penetrometer. Woods (1986) has published a detailed description of different seismic field testing methods. The most common seismic test for compaction control is the down-hole test. A vibration sensor is installed in a borehole, or by pushing the sensor into the ground (cf. seismic CPT). A polarised shear (and/or compression) wave is generated at the ground surface and the time required for the wave to travel across the soil layers to a receiver is measured. Different methods of signal interpretation can be used to determine the first arrival time of the signal. From the known distance the wave propagation velocity (shear wave or compression wave) can be calculated. Down-hole tests are relatively easy to perform, as only one sensor must be installed in the ground. The down-hole test is suitable for compaction control as it measures the average properties of a relatively large soil volume, compared to penetration tests. Signal interpretation is basically simple (determination of first arrival time at the two sensor locations), but more complex evaluation concepts (e.g. signal cross-correlation) are used. An important advantage of the shear wave velocity is that the ground water level does not affect the measurements. In the case of a cross-hole test, two sensors are installed in the ground and the wave is generated in a borehole at the same level. The distance between the vibration source and the sensors must be determined accurately in order to obtain sufficient accuracy. The distance between the sensors is typically 3 - 6 m. The cross-hole test is more cumbersome to perform than the down-hole test and is mainly used for research purposes. Another, increasingly popular seismic testing method is the Spectral Analysis of Surface Wave Technique, SASW, Woods (1986). It uses a seismic source (impact or vibration generator) at the ground surface and at least two vibration transducers at the ground surface. The vertical transducers record the propagation of surface (Rayleigh) waves. By analysing the phase information for each frequency contained in the wave train, the Rayleigh and shear wave velocity, can be determined. The evaluation of SASW measurements is relatively complex and requires specially developed computer software. However, user-friendly hardware and software has been developed which simplify the application of SASW also for compaction control. SASW measurements can determine wave velocity profiles to depth exceeding 20 m, which is sufficient for most foundation projects. The main advantage of SASW is that large soil volume can be investigated relatively rapidly. From the calculated shear wave velocity Cs, the shear modulus G can be calculated from the following relationship G = Cs2 . r (10) where r is the bulk density of the soil mass. As the strain level of the propagating shear wave is low (< 10-4%) and the elastic wave velocity is measured. It should be noted that the "dynamic" (small-strain) shear modulus decreases with increasing strain level and can thus not be directly converted into a "static" (large-strain) modulus value. A modulus reduction factor can be used to estimate the static modulus Gstat from the dynamic modulus, Gmax. Approximate values of the modulus reduction factor R can be obtained from Table 2. Semi-empirical correlations can be used to estimate the equivalent static shear modulus (secant modulus), Gstat Gstat = Gmax . R (11) where R is a reduction factor, which takes into account the strain-softening effect of soils (at strains of approximately 0,1 % strain). Table 2. Modulus reduction factor R for determination of the static (secant) modulus from dynamic tests, Massarsch (1984) Soil Type Reduction Factor, R Gravel 0,20 Sandy Gravel 0,19 Loose Sand 0,18 Medium Dense Sand 0,15 Dense Sand 0,12 It is interesting to note that the static modulus in sand is only about 10 to 20 % of the dynamic modulus. The Young's modulus, Estat and the constrained (oedometer) modulus Mstat can be readily calculated from the shear modulus Gstat (Poisson's ratio of n: 0,3) Estat = 2,6 . Gstat (12) Mstat = 3,5 . Gstat (13) From the above relationships it is obvious that the Young's modulus and the constrained modulus are significantly larger than the shear modulus. This aspect must be taken into account when evaluating the results of dynamic soil tests. The damping (attenuation) characteristics of the soil can also be determined using seismic techniques, Woods (1986). However, the practical application of these measurements for compaction projects is not yet well understood.