Acquisition of design soil parameters from In-situ Testing

1 Introduction
Design soil parameters such as compressive strength/stiffness useful for both failure and settlement
analysis can be acquired from in-situ tests. In principle, settlement analyses based on in-situ tests are
suitable for all soil types. However, in practice they are most often used on sandy soils because
obtaining undisturbed samples for laboratory consolidation testing is so difficult. Consequently,
analyses of the allowable bearing capacity and settlement of footing on coarse-grained soils are
often based on empirical methods using test data from field tests. Most of these analyses use results
from the standard penetration test (SPT), the cone penetration test (CPT) or the plate load test
(PLT). Settlements on sands and gravels take place almost immediately.

2 Standard Penetration Test (SPT)

One of the most common in-situ tests is the SPT originally developed in the late 1920s and has been
extensively used in America, the UK and Japan. Because of this long record of experience, the test
is well established in engineering practice. Unfortunately, it is also plagued by many problems that
affect its accuracy and reproducibility and is slowly being used with great caution on larger and
more critical projects. It’s a dynamic test carried out in boreholes during site investigations. The
procedure is specified in BS 1377 and many soil mechanics text books. The procedure is essentially
as follows:
1. Drill a 60 to 200 mm diameter exploratory borehole to the depth of the first test.
2. Insert the SPT split-barrel sampler into the borehole up to the bottom of the hole. The SPT is
connected via steel rods to a 64 kg hammer. The hammer is also used to forced the SPT into
the hole.
3. Using either a rope and cathead arrangement or an automatic tripping mechanism, raise the
hammer a distance of 760 mm and allow it to fall. This energy drives the sampler into the
bottom of the borehole. Repeat this process until the sampler has penetrated a distance of
450 mm.
4. Compute the number of blows, referred to as the standard penetration resistance or N-value,
by summing the blow counts for the last 300 mm of the penetration. The blow count for the
first 150 mm is retained for reference purposes, but not used to compute the N-value because
the bottom of the borehole is likely to be disturbed by the drilling process and may be
covered with loose soil that fell from the sides of the borehole. Note that the N value is the
same regardless of whether the engineer is using imperial or SI units. If 50 blows are
reached before a penetration of 300 mm, no further blows should be applied but the actual
penetration should be recorded.
5. Withdraw the SPT sampler at the conclusion of the test; extract and save the soil sample.
6. Drill the boring to the depth of the next test and repeat steps 2 through 6 as required. Tests
are normally carried out at intervals of between 0.75 m and 1.50 m to a depth at least equal
to the width (B) of the foundation.

The test results are sensitive to various variations, so the N value is not as repeatable as we would
like. The principal variants are as follows: appropriate

1. Method of drilling
2. How well the bottom of the borehole is cleaned before the test
3. Presence or lack of drilling mud
4. Diameter of the drill borehole
5. Crew and condition and efficiency of the SPT equipment including hammer, rope, rope
guide, pulleys, drill rods, anvil, cathead etc.
6. Rate at which the blows are applied.

1
 7. Level of the water table vis-à-vis the test location depth. When the test is carried out below
the water table in fine sands or silty sands, the pore pressure tends to be reduced in the
vicinity of the sampler, resulting in a transient increase in effective stress.
Note: Refer to Soil mechanic textbooks for full details of pictures of the SPT

2.1 Uses of SPT Data

The SPT N value, as well as many other test results, is only an index of soil behaviour. It does not
directly measure any of the conventional engineering properties of soil and is useful only when
appropriate correlations are available. Many such correlations exist, all of which were obtained
empirically. Caution is emphasised especially when using correlations between SPT results and
engineering properties of clays because these functions are especially crude. In general, the SPT
should be used only in settlement analyses of foundations on coarse soils and laboratory
consolidation tests used for clay soils.

2.2 Corrections to the Test Data

We can improve the raw SPT data by applying certain correction factors. It is customary to correct
the N values for various effects such as overburden pressure and energy transfer. A corrected N-
value may be obtained from the following expression suggested by Terzaghi and Peck (1948):
1
N corr .  15  ( N  15) (where Ncorr = N’) (1)
2

2.2.1 Chart correlations for N value

Other chart solutions are also available for correction of N values. Deep tests in a uniform soil
deposit will have higher N values than shallow tests in the same soil. The denser the deposit, the
greater will be the value of N, and hence the higher will be the value of internal angle of friction, .
Some of the correlations are shown in the following Figures:

Figure 1 Correlation between SPT N value, bearing capacity factor N and / value (after Peck,
Hanson and Thornburn 1974)
2
Figure 2 Empirical correlation between SPT N values and Effective Overburden pressure (After
Gibbs and Holtz) [Don’t use a correction value greater than 2].

Figure 3 Terzaghi’s relationship between standard penetration test values and allowable bearing
capacities for footings on sands
3
Settlements should not generally exceed 25mm. The less dense the deposit the greater the
settlements for a given bearing pressure. Hence, there is an approximate relationship between N
values and settlement.

Figure 4 Empirical correlation between SPT N values and Effective Internal angle of friction φ'
(from Schmertmann)

Figure 5 Correction of standard penetration test results for effect of overburden (Thorburn)

4
Relationship between SPT-N and Undrained shear strength (CU)
Based on study of SPT- N data collected from many sites in the United Kingdom together with
triaxial tests performed in insensitive stiff and hard clays. The relationships between SPT-N values
and CU were recommended as:

CU (kPa) = f1×N60. (In this case N60 = N’)

Where N60 signifies the SPT N-values corresponding to 60% efficiency is termed as N60. Seed et al.
(1984) and Robertson et al. (1983) have recommended that the SPT-N value has to be corrected to
an energy level of 60%.
Stroud (1974) stated that the factor f1 is not a constant value but changes with the plasticity index
(PI) of the soil such as:
CU (kPa) = 4.2 N60 for PI>30,
CU (kPa) = 4-5 N60 for 20<PI<30 and
CU (kPa) = 6-7 N60 for PI<20.
Stroud and Butler’s (1975) stated that CU (kPa) = 4.5 N for PI>30% and CU (kPa) = 8N for low
plasticity clay (PI =15%).

Table 1: Comparison of studies with f1 values Figure 6 Correlation between SPT-

𝑁60 values and CU (Kanagaratnam
Balachandran et al 2017)

Table 2: Approximate range of SPT-𝑁60, CU and PL for cohesive glacial tills

5
Table 3: Approximate ranges of CU and corresponding SPT-N (uncorrected) for cohesive soils
(Terzaghi & Peck 1967)

Table 4: Properties of glacial till

3 Cone Penetration Test (CPT)

The equipment and test method for the cone penetration test are described in many soil mechanics
text books as well as BS 1377. Two types of cones are commonly used: the mechanical cone and the
electric cone. Both have two parts, a 35.7 mm diameter cone-shaped tip with a 60 apex angle and a
and a 35.7 mm diameter  133.7 mm long cylindrical sleeve giving an end area of 1000 mm2; when
the tip incorporates a friction sleeve, this has a (cylindrical) area of 1500 mm2. A hydraulic ram
pushes this assembly into the ground at a rate of 20 mm/s and instruments measure the resistance to
penetration, known as cone resistance (qc). qc is the total force acting on the cone divided by the
projected area of the cone (1000 mm2); and the cone side friction is the total frictional force acting
on the friction sleeve divided by its surface area ( as the maximum force recorded during penetration
divided by the end area (1500 mm2).
The CPT defines the soil profile with much greater resolution than does the SPT. Although the CPT
has many advantages over the SPT, there are at least three important disadvantages;
1. No soil sample is recovered, so there is no opportunity to inspect the soils.
2. The test is unreliable or unusable in soils with significant gravel content.
3. Although the cost per metre of penetration is less than that for borings, it is necessary to
mobilise a special rig to perform the CPT.

3.1 Uses of CPT Data

CPT data is better correlated to soil engineering properties such as shear strength than SPT data.
CPT is able to detect fine changes in the stratigraphy. Therefore engineers often use the CPT in the

6
first phase of subsurface investigation, saving the SPT and sampling for the second phase. If a
piezocone is used, the problem of excess pore water pressure in cohesive soils may be overcome.

4 Plate Loading Test or Plate Bearing Test (PBT)

Tests on full sized footings are desirable but expensive. The alternative is to carry out plate load
tests to simulate the load-settlement behaviour of a real footing. The plates are made from steel with
sizes varying from 150-760 mm. Two common plate sizes are used in practice. One is a square plate
of width 300 mm and the other is a circular plate diameter 300 mm. The test is carried out in a pit of
depth at least 1.5 m. Loads are applied in increments of 10% to 20% of the estimated allowable
bearing capacity. Each load increment is held until settlement ceases. The final settlement at the end
of each loading increment is recorded. Loading is continued until the soil fails or settlements are in
excess of 10% of the plate diameter. The maximum load should be at least 1.5 times the estimated
allowable bearing capacity.
For PBT’s to be applicable they must be carried out on soil which is representative of the soil to be
stressed by the foundation. Carry out tests at different depths and locations.

Figure 6 Plate bearing tests on sands

4.1 Problems with the Plate load test
There are several problems associated with the plate load test.
1. The test is reliable only if the sand layer is thick and homogeneous.
2. The depth of sand that is stressed below the plate is significantly lower than the real footing.
A weak soil layer below the plate may not affect the test results because it may be at a depth
at which the stresses imposed on the weak layer by the plate loads may be insignificant.
However, this weak layer can have significant effect on the bearing capacity and settlement
of the real footing.
3. Local conditions such as a pocket of weak soil near the surface of the plate can affect the test
results but these may have no significant effect on the real footing.
4. The correlation between plate load test results and the real footing is problematic. Settlement
in sand depends on the size of the plate. Settlement increases with increases in the plate size.
Bjerrum and Eggestad (1963) found that there is significant scatter in the relationship
between the plate size and settlement for a given applied stress. Bjerrum and Eggestad
(1963) also reported that field evidence indicates that the limit of /p ranges between 3 and
5 rather than a fixed value of 4.
5. Performance of the test is difficult. On excavating sand to make a pit, the soil below the
plate invariably becomes looser and this has considerable influence on the test results. Good
contact must be achieved between the plate and the sand surface but this is often difficult. If the
plate were just above groundwater your results would be affected by negative pore water
pressure.
7