Keywords

1 Introduction

The meteo-oceanographic conditions on the Portuguese coast are suitable for the production of renewable energies in the marine environment (wave and offshore wind). In addition to the energetic potential, the shoreline concentration of the main population groups increase the demand for this energetic resources, being the development of industry and infrastructure facilities, such as modern ports and shipyards with tradition in metal-mechanics, secondary goals, highly supported by the Portuguese state.

In the last decade, the Portuguese offshore has been sought for testing and technological demonstration of devices (Pelamis, Waveroller, Windflot, Demogravi3) for the use of renewable energies. These projects have associated specific needs regarding the geotechnical characterization of the marine soils, as their adequate foundation in the open sea is essential for the feasibility and risk management of the whole project. In this sense, and due to the lack of CPT equipment in Portugal for the offshore characterization of the sedimentary stratigraphy and estimation of geotechnical parameters of the marine soils of the targeted area, the authors selected good quality sampling and alternative laboratory tests to characterize the soils, namely triaxial compression tests. This more traditional approach also allowed the acquisition of real, non-empirical, geotechnical parameters (opposite of those derived from the CPT), which were integrated with geologic information allowing the foundation design for the study area.

This is because, until now, there are no studies that characterize the sedimentary deposits of the Portuguese offshore from the geotechnical point of view. Thus, the present study intends to fulfill the gap in the knowledge of the mechanical characteristics of marine soils and proposes a multidisciplinary alternative to execute the basic geotechnical description of marine sites, necessary to any foundation at the sea bottom. The selected approach is based on the implementation of a geotechnical model, by investigating the mechanical proprieties of the most favorable geological units.

2 The Site

2.1 Location and Background Geology

The site is located 4 km off the village of S. Pedro de Moel, in the northern Portuguese continental shelf, between 30 and 60 m water depths (Fig. 1). It comprises an area of about 15 km2 with a very smooth morphology with no major outcrops. Despite the general gentle slope (about 0.3%), a morphologic structure is recognized at 55 m deep, consisting in a 1 m vertical displacement of the sea bottom. This feature marks the transition of a fine sedimentary deposit (between 30 and 55 m deep) to a coarser one (between 55 and 60 m deep). In the remainder of the studied area no morphological features were recognized.

Fig. 1
figure 1

The studied marine site and vertical sediments sampling

Full size image

Seismic profiles, acquired with a boomer and sub-bottom profiler systems, allowed the definition of a geophysical model of the studied site, based on the description three seismic units (U1, U2 and U3 from the bottom to the top of the seafloor) with different acoustic characteristics (Pombo 2016).

According to Rodrigues (2004), Unit 1 (U1) corresponds to a very deformed and dolomitized carbonate rocks. The basal sedimentary unit (U2), a coarser deposit (sandy gravel sediments), covers the bedrock and is exposed in the deepest area, below 55 m deep. The more recent unit (U3), covering the coarser deposit, was identified in the shallow area, being composed by sandy sediments. According to Pombo (2016), the sandy deposit increases its thickness towards the coast line, from 0.5 to 7 m, while the sandy-gravelly deposit, despite the difficulty in the basal delimitation, seems to be a thicker unit (5–13 m, also increasing eastward).

3 Methods

Once in the laboratory, the corers were submitted to several tests using international normalized methods: namely, grain size distribution (NP EN 933-1:2014 and ISO 13320-1: 2009), particle density, specific weight and water content (ISO/TS 17892-1, 2, 3:2004a, b, c) and triaxial tests.

Triaxial compression tests with CK0D anisotropic consolidation, with seismic wave (P and S) measurements, were carried out on six undisturbed samples. These tests were performed with monotonic increase of vertical stress. For the triaxial tests, adequate preparation of the test specimens according to Part 2 of Eurocode 7 (EN 1997-2: 2007) was required, all of which performed as described below, in compliance with ISO/TS 17892-9: 2004d.

The specimens tested were approximately 70 mm in diameter and 140 mm in height and were previously extruded from the core samplers (Table 1).

Table 1 Identification of specimens analyzed
Full size table

After assembly in the triaxial chamber, ‘water’ was percolated to eliminate the soil air bubbles, followed by saturation with increasing back pressure up to 500 kPa, maintaining the effective confinement of 10 kPa; to ensure the saturation, the Skempton’s B parameter value (>0.95) was controlled. Initial isotropic consolidation was then carried out at an effective stress followed by anisotropic consolidation to the condition of K0 = 0.5 and maintaining this anisotropic condition for 24 h.

Five discharge-charge cycles were carried out for the determination of the modulus of elasticity (Edr), followed by increasing vertical compression, σ1, in drained conditions with volume change variation readings; throughout this final compression stage, the applied axial strain rate to the test specimen remained unchanged. These compression trajectories followed the required condition: that σ3 was constant. At the end of those cycles, the velocities of S and P waves were measured using bender-extender elements.

It should be noted that the percolated and the injected water (for back pressure) inside the specimen during the saturation phase used natural saline deaired water from the studied area.

4 Results

According to the site investigation and testing, the U3 seismic unit, with an average thickness of 3 m, encompasses poorly graded sands, SP, and poorly graded sands with silt, SP-SM (Pombo 2016). The characteristics of the test specimens of these types of sands are summarized in Table 2, and the grain size distribution curves in Fig. 2.

Table 2 Summary of the specimen properties
Full size table
Fig. 2
figure 2

Grain size distribution of the U3 seismic unit soils tested

Full size image

According to the results three specimens (P5, P8 and P9) correspond to poorly graded medium sands (SP) and the other three specimens (P10, P11 and P12) correspond to poorly graded fine sand with silt (SP-SM).

The dry weight of the SP sediments ranges from 17.3 to 17.6 kN/m3, while for the SP-SM sediments it varies between 15.4 and 15.9 kN/m3. Both group of soils have low carbonate contents, around 5%, and their particles densities are around 2.65, which are indicative of quartz particles.

To evaluate the mechanical behavior of these soils, triaxial tests were performed, and the deviatory stress—axial strain curves and the volumetric strain—axial strain of each test specimen are presented in Figs. 3 and 4.

Fig. 3
figure 3

Deviatoric stress—axial strain curves for the six specimens tested

Full size image
Fig. 4
figure 4

Volumetric strain—axial strain for the specimens tested

Full size image

The deviator stress vs axial strain curves (Fig. 3) show that, in general, there is a common tendency to all triaxial tests analyzed: the stress-deformation behavior of the tested marine sediments seems to be initially rigid, apparently linear, to a well-defined yield point, from which the soil undergoes increasing plastic deformations until rupture. The behavior after failure is characterized by a more intense and rapid deviatoric stress drop in the test specimens P5 and P11, and by the existence of a critical state, where the deviatoric stress approaches a stable residual value, for increasing axial deformations. The tested sands present a hardening behavior.

Figure 4 shows the volumetric strain vs axial strain variation for all specimens tested at different consolidation stresses. It is observed that both groups of test specimens present an initial compression followed by a strong expansion until rupture, which coincides with the maximum dilatation and increase in volume. After this point, the expansion gradient decreases, leaning at the end to a constant value.

The behavior described above is typical of medium to dense sands.

Table 3 presents the obtained rupture values, peak deviatoric stress (σ1−σ3)p and residual deviatoric stress (σ1−σ3)cv, for the test specimens and the discharge-charge (pseudo-elastic) modulus, Edr, calculated between the vertices of hysteresis of the five cycles centered on the value of K0 ≅ 0.5.

Table 3 Elastic modulus, peak and constant volume strength of the specimens tested
Full size table

Analyzing the data in Table 3, it can be registered that the modules determined between the vertices of the discharge-charge cycles (Edr) are very variable and are related to the properties of each specimen and the initial consolidation stress. Regarding the deviatoric stresses, it can be concluded that they are also directly related to the type of material of each specimen and to the effective consolidation stresses applied.

Table 4 presents the values obtained in the calculation of acoustic wave propagation velocities for each of the test specimens. These values are used to determine the dynamic deformability of the soil, the Poisson coefficient and the distortion modulus, and can be compared with the values of the same parameters determined in static conditions. However, dynamic modules generally give higher values because deformations involved are very small.

Table 4 S and P waves at end of discharge-charge cycles
Full size table

The results of the effective peak (φ′p) and residual (φ′cv) friction angles, cohesion (c′) and dilation angle (Ψ) of the SP and SP-SM samples tested in this investigation (Table 5) are overall consistent with those published in reference works (Schanz and Vermeer 1996; Lee 2003).

Table 5 Strength parameters of the sands
Full size table

It is also worth mentioning that the angles determined by the graphs q−p′ show an excellent statistical correlation (r > 0.98) and agree with the stipulate in EN 1997-2: 2006.

According to Bolton (1986), the angle of shear strength in the critical state mainly reflects the sediment mineralogy. In the specimens analyzed by this research (Table 5), it is verified that the friction angles for SP and SP-SM soils are 35.2° and 34.5°, respectively, very close to the one reported by Bolton (1986) for specimens composed of quartz particles (33°).

5 Conclusions

This paper presents the results of part of the geotechnical investigation carried out at a small site of the Portuguese continental shelf, mainly characterized by poorly graded sands (SP) and poorly graded sands with silt (SP-SM). The strength parameters of these soils, through triaxial compression tests (CK0D), were calculated to support the foundation design for renewable energy offshore floating structures.

The results of the triaxial tests, performed on the SP and SP-SM soils, presented a typical behavior of medium to dense sands showing, while shearing, a hardening behavior with an initial compaction followed by dilation and increase in volume. The effective friction angles for the two types of sands tested were 35.2° and 34.5°.

The results obtained for both physical and mechanical properties are in accordance with the benchmark values presented in literature for the same type of siliceous sands.

This was the first case study for the Portuguese offshore, where no strength parameters had ever been determined. Further researches should be carried out to complement the presented results.