IC4_I

Introduction and Research Background
Cone Penetration Testing (CPT) has become a widely used in situ testing tool in soil engineering and science. Besides of its cost effectiveness when compared to sediment coring and other geotechnical soil investigation methods, CPTs are increasingly employed for many reasons: By pushing a cone e.g. from beneath a heavy truck at a controlled and constant rate into the soil, this device provides continuous readings in vertical soundings without extracting soil samples. Electronic load cells and multiple sensors allow the simultaneous measurements of the parameters at the cone head and along the sleeve: 1. cone resistance qc; 2. sleeve friction fs; 3. pore pressure u. The cone resistance, obtained by measurement of the total force over the conical tip area, expresses the resistance of sediments to penetration (Robertson and Campanella, 1988). The sleeve friction fs is determined as the load over a standard 150 cm2 sleeve and is indicative for the adhesive strength of the material. Sleeve fiction measurements are usually used for the determination of the friction ratio which relates, directly to the soil behavior type and approximately to the sedimentological soil type. Penetration pore water pressures u are monitored using an internal transducer connected to a porous filter element located on the exterior of the probe (Schneider et al., 2001). Filter element position is variable in different instruments, but is mostly located either mid-face on the cone (u1) or immediately behind the tip at the shoulder (u2), behind the sleeve (u3) is rarely used but valuable for certain questions (Lunne and Robertson, 1997).

In saturated soils the absolute pore pressure increases with depths. Pore pressure values are additionally of importance for correcting qc and fs for the unequal end-area effect resulting from the inner geometry of the cone. The correction is negligible in coarse-grained soils like sands, where qc is large relative to the pore water pressure u2, but crucial in soft fine-grained soils like clays and silts, where qc is low relative to the high induced water pressure around the cone due to the undrained penetration process (Lunne and Robertson, 1997).

For this study Geotechnical Offshore Seabed Tool (GOST), a recently by the Marine Engineering Geology working group at the MARUM Institute developed, was to be utilized for soil investigation. This tool provides CPT data along with pore pressure measurements, called CPTu or piezocone following all required national (DIN ISO 22476-1, Entwurf Oktober 2009) and international standards (ASTM D5778 -07, ISSMGE, 1999). It utilizes a hydrostatically compensated cone of 5 cm2 projected cone area which allows the very precise and high resolution measurements of soil strength parameters. Compared to penetrometers with a standard 10 cm2 projected cone surface, smaller cones react more sensitive to rapid changes in thin-layered soils hence increasing vertical resolution and providing more detailed profiles. A hydraulic system is used to push the penetrometer with a penetration speed of 2 ± 0.5 cm/s into the soil while near continuous readings at two second intervals are taken. An integrated gravity sensor monitors the verticality of penetrometer rod to warn against excessive bending of the push rod. GOST is designed to operate from the seafloor. When operating in the so called ”dangling mode” GOST is slightly lifted and the ship moves slowly to the next penetration site without lifting GOST out of the water on deck, making multiple operations highly time and cost efficient.

The newly developed GOST is universal to the effect that it enables the determination of a number of soil properties in different environments. Cone penetration testing can be done over the entire soil spectrum ranging from sand to clay sized particles, and is therefore useful in soil science, as well as geotechnical engineering and design. Providing repeatable, nearly continuous, high quality, real time results, GOST can enormously improve strata delineation and enhance stratigraphic resolution. This is of particular importance since it allows for a better identification of thin-layered, for engineering purposes potentially problematic, material. This can e.g. be the existence of soft, eroded till, or sensitive pyroclastic material in harbor basins, which can pose problems for port construction and infrastructure. With its provision for vibration GOST can also apply in situ cyclic stresses, provide cyclic soil properties, and evaluate the response of harbor foundations subjected to vibrations. Given its modular design GOST can not only resolve this thin layered potentially liquefiable or otherwise hazardous material in the marine, but also in terrestrial environment, additionally enabling the reconstruction of dynamic changes in the harbor seabed.

It is evident that GOST can be used for a great variety of research questions and engineering problems in highly differing environmental settings, and that its use can considerably support our understanding in soil science.
In the following, background information, research questions, and objectives for three major manuscripts are presented:

1. Correlating sea- and landward CPT profiles: A methodological approach
Although the primary focus of GOST is the offshore site investigation it also works onshore, for instance mounted on a truck. This allows penetration testing on the land, as well as seaward side of the quay where retaining walls are planned, while keeping costs and efforts much lower than core taking below the water column, with no need for an expensive ship or platform. This study therefore intends to compare GOST CPT profiles taken from the land and seaside and deduce similarities and differences in the stratigraphy in this highly variable environment at the rim of rivers and estuaries. Where soil layers coincide again and how well they correlate with depths is thereby of main interest and will help evaluate the currently utilized approach. Dynamic process like e.g. erosion in the seabed and deposition can additionally be studied through this comparison.
Key questions: How well do land-and seaside CPTs correlate? At which depth do profiles correlate in a seaside erosive/depositional environment? What can be said about the upper young soil layers in the harbor? Can dynamical processes be reconstructed?

2. Analysis of sensitive layers of volcanic origin in the Port of Tauranga
Following a large storm during May 2005 a number of landslides occurred across the Tauranga basin region (Bay of Plenty, NZ), most notable in the suburb of Otomaetai where sensitive layers which are associated with volcanic ash materials that have subsequently weathered to soils with high silt and clay contents occur. In clays with high sensitivities, i.e. clays where the ratio between the unconfined compressive strength of an undisturbed specimen and the strength of the same specimen at the same water content but in a remolded state is great. The degree of sensitivity is different for different clays, and it may also be different for the same clay at different water contents. Most clays have sensitivities ranging between 2 and 4. Sensitive material can reach values between 4 and 8, and extrasensitive clays between 8 and 16 respectively while clays with even higher sensitivities are known as quick clays. Evidence of sensitive soils and hence potentially dynamically weakening material in the Tauranga harbor basin exists since construction of a bridge in the Stella Passage where vibrations due to pile driving caused the pile to subside for several meters. For future mitigation of slope stability hazards/liquefaction and best practice engineering, it is therefore of crucial importance to determine the presence and extent of potentially liquefiable sediments. This study shall make use of the exceptional resolution of piezocones for detailed soil stratigraphy and detection of thin seams and lenses. Main target is to investigate and if existent map these sensitive layers along the Waimapu estuary and Tauranga harbor basin and to quantify their volume.
Key questions: to solve are thereby: Can we use GOST to identify similar sensitive materials in the marine environment? What characteristics do they show on CPT testing and is this enough to distinguish them from other similar materials in the profile? If so, what is their distribution, both vertically (thickness) and laterally? Is the weathering sequence and clay mineralogy similar to the terrestrial weathering sequence (i.e. is halloysite still a key mineral)? Do they have similar geomechanical properties to the on-land sequences? What are the implications of these materials for harbor infrastructure, dredging and port development?

3. Comparison of static versus dynamic CPT testing
It is a matter of common experience that vibrations due to pile driving, traffic, blasting, wind loading, or the operation of machinery usually increase the density of a sand causing its surface to subside (Terzaghi et al., 1996). In order to evaluate e.g. the response of foundations subjected to vibrations and the manner of vibrations and its transmission through the ground, the dynamic properties of soils must be determined. When e.g. evaluating seismic hazards we would ideally obtain an undisturbed soil specimen, apply the same stress conditions that exist in the field, and then subject the soil specimen to the anticipated earthquake-induced cyclic shear stress and analyze the resulting soil behavior. The method to be used for such an approach is freezing the subsurface and subsequently drilling it in a frozen state. This is only an option in coarse grained sediment without too much content of fines. However, in engineering practice, this approach is far too expensive and sophisticated to be followed. The within this study utilized GOST piezocone has a provision for vibration which e.g. allows the detection of liquefaction resistance by applying dynamic stresses in situ. In this part of the study a comparison will be made between static and dynamic GOST CPT data during field work in the North Sea. Possibly static/dynamic GOST measurements will also be obtained on the Swedish Geotechnical Institute (SGI) test site.
Key questions will thereby be: How do soils respond to different stimuli, e.g. different frequencies? Why do these materials respond this way? What differences can be seen between static and dynamic data sets? How do different variables react to application of different loads? How do in situ cyclic in situ and lab tests differ?