Piezocone (CPTu) testing has grown into a very popular method for in-situ soil investigation
in Norway. Traditionally, soil investigation relied on collecting soil samples and
testing these samples in laboratory to determine different soil properties and strength
characteristics. However, this method has limitations, as collecting soil samples can disturb
their natural state, leading to inaccurate results.
The disturbance of soil samples during collection can significantly impact their properties,
depending on soil type, density, overburden pressure, and water content; for example,
a densely packed soil sample may loosen, while a loosely packed sample may become
denser. The delicate structure of the soil is easily disturbed. This soil disturbance can
lead to discrepancies between the behavior of the soil in the field and its behavior in
the laboratory (P. Robertson et al., 1992). Further, laboratory testing on small intact
samples of interbedded or layered soils may not accurately reflect the properties of the
entire soil mass. The difficulty in recreating the exact in-situ conditions in the laboratory
further limits the accuracy of laboratory testing. As a result, engineers often need to
adopt more conservative design approaches, which can increase costs.
To obtain a better understanding of the soil’s properties in its natural state, conducting
in-situ testing is essential. Piezocone (CPTu) testing has emerged as a beneficial
method for this purpose, as it allows for direct measurement and analysis of soil properties
without the need for sample collection. This thesis study is based on cone penetration
tests utilizing a cone with four pore pressure sensors conducted on silt and clay sites. The
objective is to enhance the understanding of silt and quick clay behavior by analyzing
the pore pressure variation along the cone,
During CPTu, three quantities are measured: cone resistance (qc), pore pressure (u),
and sleeve friction (fs). Depending on the position of the pore pressure sensor, the
recorded pore pressure response varies. Pore pressure is commonly measured at the
shoulder location just behind the tip and is denoted as u2. In our study, we measured
pore pressure at four locations:
• The u1 sensor at the cone tip
• The u2 sensor at the shoulder between the cone tip and the friction sleeve
• The u3 and u4 up in the friction sleeve
Moreover, the penetration rate of CPTu significantly influences the measured parameters,
especially in fine-grained soils. The CPTu is commonly performed at 20mmps around
the world. Generally, the undrained response is recorded at u2 location for clay, and
the drained response is recorded for sand during CPTu at a 20mmps penetration rate.
However, pore pressure response at u2 in silt is partially drained during a standard rate
CPTu. Hence, silt is more susceptible to change in penetration rate than clay(Poulsen,
Nielsen, & Ibsen, 2011). That is why we also varied the rate of CPTu using the new
multiple pore pressure sensors cone. The aim was to see the variation in response with
various penetration rates in all four locations along the cone.
It is evident that interpreting CPTu in silt is challenging due to partial drainage and
requires thorough understanding of silt behaviour during cone penetration. Many researchers(
Bihs, 2021; J. T. DeJong & Randolph, 2012; Doan & Lehane, 2018) studied the effect of partial drainage in silt by varying the penetration rate of CPTu and incorporated
the partial drainage effects in interpreting CPTu test results. However, there is a clear
need for more field studies where pore pressures are measured in multiple filter locations.
Currently, these studies are limited. Most of these studies measure pore pressure at only
one location (commonly u2) and multiplied it by a conversion factor to get pore pressure
at other filter location, i.e. u1 or u3 . Recent studies have introduced varied penetration
rates in silt but have not addressed the pore pressure variation along the cone. Moreover,
limited data makes it challenging to interpret silt confidently compared to clay and sand,
highlighting the importance of further research in this area. It is necessary to improve
the accuracy and reliability of soil behaviour assessments. Our study, which aims to
contribute to the database with data including penetration rate effect and pore pressure
variation along the cone, is a new approach to studying silt.
The study was conducted at the NGTS silt site Halsen-Stjørdal, located north of
Trondheim. The site consists of a 200 − 300m thick silt deposit with low plasticity.
Fieldwork was carried out using a combination of multiple pore pressure CPTu and
standard CPTu at 10mmps, 20mmps, and 40mmps penetration rates. The saturation
process was conducted in two steps to account for the loss of saturation and its impact on
the data. Dissipation tests were also performed to understand the development of pore
pressures and drainage conditions. We compared the new data with previously gathered
data. However, previous data only includes u2 pore pressure.
We have also used the multiple pore pressure sensors cone at a quick-clay site (Tiller-
Flotten). The aim is to see if pore-pressure variation along the cone can indicate the
presence of quick clay.
The analysis focused on examining the changes in pore pressure along the cone in
both silt and clay. The results revealed distinct differences in the responses between
the two soil types. Notably, in our silt site, the maximum pore pressure was observed at
sensor u1, while sensors u2, u3, and u4 recorded similar pore pressure values. Additionally,
dissipation tests showed variations in dissipation rates at different sensors. However, we
did not observe significant differences when considering different penetration rates. It is
important to note that further investigations using higher and slower penetration rates
than the penetration rates used in the current study are necessary to uncover more clear
distinctions. The high sensitivity and stress encountered at greater depths significantly
affect the pore pressure response for the quick clay site. However, it is hard to determine
the transition from low-sensitive clay to high-sensitive clay solely based on the pore
pressure change. The pore pressure curves from four sensors seem to converge in the
highly sensitive layer. However, we can not confidently say this convergence is due to the
change in clay’s sensitivity. Further investigation is needed.
The expected pore pressure variation along the cone is observed for our silt and clay
sites. Pore pressure is highest at u1 and decreases as the filter location moves up along the
cone. In Halsen-Stjørdal, this variation is irregular due to stratification, while in Tiller-
Flotten, it is regular and almost constant with depth. Schneider’s classification chart
outperformed Robertson’s for silt and sensitive clay. Although the charts use u2 data,
we found u3 and u4 data better identify silt, sensitive clay, and soil layers. Dissipation
test analysis in our silt site showed that initial pore pressure varies in decreasing order
from the tip to the sleeve. The fastest rate test produced the highest pore pressure and
the slowest rate the lowest, with some exceptions. The t50 value is lowest for u1 and
highest for u4. Pore pressure at u1 drops immediately, while other filters show delayed
responses, with u1 being monotonic and u2 to u4 being dilatory. We could not identify any
clear trend from the rate analysis of the dissipation tests. The graphs showed significant
variation in t50 values and dissipation gradients with penetration rate but without any
distinguishable order.
