Evaluation of SCPT-surveys as method for accessing dynamic modulus

(1)

Evaluation of SCPT-surveys as method for accessing dynamic modulus

Joakim Granskär

Civil Engineering, master's level 2018

Luleå University of Technology

Department of Civil, Environmental and Natural Resources Engineering

(2)

(3)

PREFACE

This master thesis is the final part of the Civil Engineering program with a master towards Soil and Rock Engineering at Luleå University of Technology. The work done with this thesis corresponds to 30 credits, which is approximately the same as 20 weeks fulltime work. This project has been conducted at and for Sweco Civil.

With this master thesis a five years long chapter of amazing experiences and amusing times are coming to an end. With wisdom and friends that will last for a lifetime gained here at Luleå University of Technology there is no doubt that a bright future lies ahead, which leaves me looking forward to future chapters of life.

First I would like to thank my supervisor at Sweco Civil, Per Nilsson. Without your vast geophysical knowledge and guidance this thesis would not have been completed. At Luleå University of Technology I would like to thank my supervisor and examiner Jan Laue for his patience and rewarding discussions about dynamic properties of soil. Furthermore I would like to thank Luke Chapman at Sweco Civil, who is responsible for the geotechnical part of the East Link project, for his help guiding me through this huge project. I also have gratitude for Geotech, which have helped me by providing software and support to interpret the data.

Lastly I would like to thank my colleagues at the geotechnical department at Sweco Civil in Luleå for their support and geotechnical knowledge, with a special thanks to Jens who shares my interest in geophysical surveys.

Luleå, May 2018

Joakim Granskär

(4)

ABSTRACT

The purpose of this master thesis is to evaluate the results from the completed SCPT (Seismic Cone Penetration Test) surveys with respect to the methods ability to estimate dynamic young´s modulus. This has been done by comparing the result from SCPT to other seismic methods and dynamic parameters converted from “static” geotechnical surveys.

SCPT is a relatively new method which has not been used by Sweco or in Sweden to any greater extent and is therefore of interest to be looked in to. By adding two geophones to an ordinary CPT (Cone Penetration Test) equipment as well as including strike plates, hammer and logger system one can log dynamic parameters in addition to the regular parameters logged by a CPT system. This is done by stopping the otherwise continuous CPT survey every meter and striking the strike plates, which are placed under the tracks of the drilling rig.

Seismic waves will then travel through the soil down to the geophones and be used to calculate dynamic parameters.

For this master thesis a total of 47 geotechnical and geophysical surveys have been

considered. The location of these 47 are approximately 7 km north of Norrköping and have been conducted between the years 2016-2017 for the East Link Project. The geotechnical surveys are composed of weight-sounding, ram-sounding and CPT while the geophysical ones are SCPT, refraction-survey and MASW (Multichannel Analysis of Surface Waves).

Static elastic parameters have been calculated using the geotechnical survey results according to the Swedish transportation administrations standardized methods. These have then been converted to dynamic parameters with the help of different relationships. When these have been converted they can be compared to SCPT results and other seismic survey methods which also use wave velocity, Poisson’s ratio and density to calculate dynamic elastic parameters.

Based on the results from this thesis it can be concluded that the seismic test add-on to a standard CPTu survey is a good method for accessing dynamic modules and it gives extended information of the soil stratigraphy from one single survey point.

After the studies and analyses conducted for this thesis it can be concluded that the proposed conversion between static and dynamic young’s modulus using the Alpan curve gives slightly higher values than the ones derived directly from shear waves using SCPT. The other

conversion using cone tip resistance to shear wave velocity gives on the other hand slightly lower values than the ones measured with the SCPT.

The analysis also indicate similar trends in results between the different seismic methods.

These results does however also shows that the used assumptions are somewhat general for accurate comparisons between the methods.

(5)

SAMMANFATTNING

Syftet med detta examensarbete är att utvärdera resultaten från undersökningar utförda med SCPT (Seismic Cone Penetration Test) baserat på metodens förmåga att uppskatta dynamiska elasticitetsmoduler. Detta är gjort genom att jämföra SCPT resultaten mot andra seismiska metoder samt dynamiska parametrar konverterade ifrån ”statiska” geotekniska

undersökningar.

SCPT är en relativt ny metod som inte tidigare använts av Sweco eller i Sverige i någon större utsträckning och är därav intressant att kolla närmare på. Genom att addera 2st geofoner till en vanlig CPT (Cone Penetration Test) utrustning samt med slagplattor, hammare och logger- system så kan dynamiska parametrar loggas utöver de vanliga parametrarna som CPT loggar.

Detta görs genom att stanna den annars kontinuerliga CPT undersökningen varje meter och slå på slagplattorna som är placerade under banden på borriggen. Seismiska vågor kommer då färdas genom jorden ner till geofonerna och kan utifrån detta framräknas till dynamiska parametrar.

För detta exjobb så har sammanlagt 47 geotekniska och geofysiska mätningar beaktats.

Platsen för där de 47 mätningar är genomförda är ca 7 km norr om Norrköping. Mätningarna har blivit genomförda 2016-2017 och med anknytning till Ostlänken projektet. De

geotekniska undersökningarna innefattar viktsondering, hejarsondering och CPT medan de geofysiska består av SCPT, refraktionsseismik och ytvågsseissmik.

Statiska elastiska parametrar har räknats fram utifrån dem geotekniska undersökningarna enligt trafikverkets standardiserade metoder. Dessa har därefter blivit konverterade till dynamiska parametrar med hjälp olika samband. Då parametrarna har blivit konverterade så kan de jämföras med SCPT och andra seismiska metoder som använder sig av våghastigheten, tvärkontraktionstalet och densiteten för att räkna fram dynamiska elastiska parametrar.

Baserat på resultaten från detta exjobb så kan det konstateras att det seismiska tillägget till en standard CPT undersökning är en bra metod för att bedöma dynamiska moduler och ger utökad information om jordens egenskaper och lagerföljd från en enstaka undersöknings punkt.

Efter resultat och analyser utförda för detta exjobb så kan det konstateras att den förslagna konverteringen mellan statisk och dynamisk elasticitetsmodul med hjälp Alpan kurvan ger något högre värden än dem härleda från direkt från skjuvågor från SCPT. Den andra

konverteringen som använder qt till VS ger å andra sidan något lägre värden än dem uppmätta med SCPT.

Analysen indikerar likartade värden för de seismiska metoderna. Resultatet visar också att de antagna värdena är något för generella för att ge en exakt jämförelse mellan metoderna.

(6)

LEGEND

Abbreviations

SCPT Seismic Cone Penetration Test CPT Cone Penetration Test

RST Ram-sounding

WST Weight-sounding

SGI Swedish Geotechnical Institute

MASW Multichannel Analyses of Surface Waves CRS Constant Rate of Strain

GWT Groundwater table

P-wave Primary/compression wave S-wave Secondary/shear wave R-wave Rayleigh wave

Cl Clay

Si Silt

Sa Sand

RTK-GPS Real Time Kinematic-Global Positioning System

Roman upper case

ES Static Young’s modulus [MPa]

ED Dynamic Young’s modulus [MPa]

Rf Friction ratio [%]

U Pore Pressure [kPa]

G Shear modulus [MPa]

K Bulk modulus [MPa]

M P-wave modulus [MPa]

(9)

LEGEND

VP Primary/compression wave speed [m/s]

VS Secondary/shear wave speed [m/s]

VR Rayleigh wave speed [m/s]

Roman lower case

i Angle of incoming wave [º]

r Angle of refracted wave [º]

qt Cone tip resistance [MPa]

fs Sleeve friction resistance [kPa]

g Gravity constant [m/s²]

Greek lower case

ν Poisson’s ratio

ρ Soil mass density [kg/m³]

ε Strain

σ Stress [MPa]

τ Shear stress [MPa]

τfu Undrained shear strength [MPa]

γ Specific weight [kN/m³] α Change in angle [º]

(10)

(11)

INTRODUCTION

1 INTRODUCTION

1.1 Background

The economic growth of the lower central and east region of Sweden has increased the need of high capacity infrastructure. Today’s railway network is overburdened, which leads to an increased use of cars and trucks for transportation in the region. This leads to traffic jams and other disturbances, which consequently can lead to environmental problems as well as

hindering the development of nearby communities and societies. (Trafikverket, 2017). The solution presented by the Swedish Transport Administration to this demand is The East Link Project.

The East Link Project is a high-speed railway project between Järna (close to Södertälje/

Stockholm) and Linköping, with a planned construction start in 2017. This 150 km stretch is the first part of the new generation railway core network. In the future the network will also reach Jönköping where it splits towards both Gothenburg and Malmö. The completion of the new railway will unlock great possibilities for commuting and freight transportation by increasing the capacity and shortening travel times. (Trafikverket, 2017). An overview of the planned expansion of the new generation railways can be seen in Figure 1.

Figure 1 Overview of the planned expansion including The East Link Project. (Trafikverket, 2017)

Railway construction like this present new and interesting challenges from a geotechnical point of view. The trains on this new railway will induce heavy and instantaneous loads on the ground. This will put a lot of oscillating strain, similar to those of a small earthquake, to the soil, which can also lead to an accumulation of deformation. These vibratory conditions create the need to investigate how the soil will react under these conditions. Consequences of these

(12)

nearby surroundings or even trigger landslides according to the Swedish Geotechnical Institute (SGI) (Möller, Larsson, Bengtsson, & Moritz, 2000).

These consequences are not allowed to happen and will have to be countered with different ground reinforcements, ground improvement methods or by simply replacing the soil. One of the crucial properties of the soil for these kind of loads are the dynamic young’s modulus (ED) under small strains. Dynamic young’s modulus is the ratio between stress (σ) and strain(ε) under vibratory condition and will have a large impact on where and how much of the ground needs to be improved, reinforced etc. (Möller, Larsson, Bengtsson, & Moritz, 2000).

Sweco is responsible for designing and planning one of the sections nearby Norrköping in the East Link Project. To correctly design and plan this new generation of railway a variety of different surveys had to be done to get an accurate estimation of the soil conditions. Surface wave surveys, seismic refraction surveys and SCPT are some of the methods used in this project to investigate the dynamic modulus of a soil. These methods measures shear wave velocities (VS) and compression wave velocities (VP) in soils, which for small strains are directly linked to the dynamic modulus. (Möller, Larsson, Bengtsson, & Moritz, 2000).

SCPT is a rather new and unique method used for investigating dynamic modulus and has not been tested immensely in Sweden. Because of the national inexperience with this method Sweco is interested to further look in to the results from this survey as a method for

investigating the dynamic modulus. This is to gain experience about the method but also to see how its compares to other ways of investigating dynamic modulus’s such as empirical correlations based on geotechnical investigations or other seismic methods.

If it can be concluded that the results of the SCPT-surveys are correct and reliable then it can be used as a standard method along with the other survey methods that a geotechnical drilling rig can perform. This would make the drilling rig a good overall solution for this kind of surveys because of its ability to both perform the seismic survey as well as the highly recommended calibration drilling with soil/rock probing and other methods.

(13)

INTRODUCTION

1.2 Aim

This purpose of this master thesis is to evaluate the results from some of the performed SCPT- surveys at the East Link Project with respect to its reliability as a method for assessing

dynamic modulus compared to other methods.

Another aim is to evaluate the accuracy of the used relationships for converting static young´s modulus to dynamic young´s modulus. This evaluation will be done by comparing seismic wave velocity acquired dynamic young’s modulus against dynamic young´s modulus correlated from calculated static young´s modules.

1.3 Limitations

One of the limitations for this master thesis is that the collected data only comes from the Malmölandet part of the East Link Project and may thereby not correspond to the vast majority of conditions that soil can possess and thereby limit the validation of SCPT.

The surveys conducted and used for this thesis are placed with respect to Swecos project of surveying and designing as first priority. This means that the placement of different survey points and methods will not always be optimal for this thesis aim to compare and confirm results between them.

Another limitation is that no laboratory tests have been used or conducted on the soil

regarding dynamic modulus, which decreases the ability to confirm the acquired SCPT survey values. Since no laboratory test have been taken into account for this thesis the densities and Poisson’s ratio have been derived empirically.

In order to keep the thesis within a reasonable time limit a categorization of strain rates has been done. This categorization consists of assuming small strains belonging within the dynamic range and medium to large strain belonging to the static range. This will therefor make the method and analysis of this thesis more accurate for linear-elastic conditions but will not necessarily be very accurate for other soil mechanic models. In reality the boundary between these two categories are more transient and this thesis is ignoring the middle part of the range. In combination with this categorization only one general correlation between static- and dynamic young’s modulus have been used and evaluated. One correlation between cone tip pressure from cone penetration test and shear wave velocity is also used and evaluated in this thesis.

(14)

2 THEORY

Dynamic young’s modulus can be acquired in a series of different ways. Either from seismic surveys such as SCPT, MASW and refraction wave analyses, or from correlating it from geotechnical investigations, which are based on the static young’s modulus (ES). Lastly, both static and dynamic young’s modulus can also be calculated based on different laboratory tests.

With this thesis subject there is a need to explain much of the underlying theory, hence this chapter. This chapter will first explain young’s modulus then later move on to some general theory about seismic waves and describing the different survey methods that can be used to acquire both static and dynamic young’s modulus with a focus on the methods used for this project.

2.1 Young’s modulus

2.1.1 Static young’s modulus

Young’s modulus (E) is described according to Hooke’s law in isotropic materials as the slope of the materials stress-strain curve. Hooke’s law is defined as:

𝐸 = 𝜎

𝜀 (1)

Stress (σ) is for these magnitudes measured in mega pascals (MPa) and strain (ε) is

dimensionless, giving E the unit of MPa. To further clarify the concept of stress and strain;

stress is the ratio of applied force to an area while strain is the deformation caused. Strain is expressed as the ratio between the change in length or volume to its original length or volume.

(Tavakoli, 2016).

As seen in Equation (1), the relationship that creates this slope can be represented as the materials resistance against elastic deformation. This is however only valid during the elastic part of the soils deformation i.e. before the material reaches its peak/yield strength. Young’s modulus is also called the elastic modulus hence its notation E. (Larsson, 2008).

A common stress-strain curve for soil with young’s modulus included can be seen in Figure 2.

(15)

THEORY

Figure 2 General stress-strain curve for soil

Young’s modulus is however often calculated using a series of other elastic parameters as can be seen in Equation (2) and (3), where K is the Bulk modulus, G is the Shear modulus and ν is Poisson’s ratio.

𝐸 = 3𝐾(1 − 2𝜈) (2)

𝐸 = 2𝐺(1 + 𝜈) (3)

The Bulk modulus (K) is measured in MPa and can be described as a materials resistance to uniaxial compression. See Figure 3. (Tavakoli, 2016).

Figure 3 Conceptual figure of uniaxial compression (Tavakoli, 2016)

The shear modulus (G) is also measured in MPa and can be described as a materials resistance to shearing, hence its rigidity. (Larsson, 2008). See Figure 4 where τ is shear stress and α is the change in angle.

(16)

Figure 4 Conceptual figure of shearing (Larsson, 2008)

Poisson’s ratio (ν) is the ratio between horizontal and vertical strain. A simplified explanation of the ratio is measuring how much thinner a specimen gets when pulling on its ends in comparison with its original width and length as can be seen in Figure 5.

Figure 5 Conceptual description of Poisson’s ratio. (Luna & Jadi, 2000)

This is a ratio and hence it is dimensionless. The value for Poisson’s ratio is difficult to measure in-situ and is often estimated for soils between 0,2-0,4 based empirically on different soil types. When calculated, Poisson’s ratio is often derived from equations such as Equation (2) or (3) (Luna & Jadi, 2000).

Young’s-, bulk-, shear modulus and Poisson’s ratio are important parameters when investigating soil conditions and behaviour. With knowledge about these parameters, a geotechnical engineer can assess different needs of ground reinforcements or how much ground improvements needed for a road, building foundation, etc.

When generally calculating young’s modulus one refers to the slope according to Figure 2.

However this young’s modulus applies for medium to large strains and is referred to as the static young’s modulus. As it can be seen in the same figure, the stress-strain curve is not linear for different stages of stress. For some constructions such as railways and wind power plants the young’s modulus for small strains are of more interest due to its vibratory loads consisting of small stresses, which causes repeatedly small strains. The young’s modulus for small strains are often referred to as the dynamic young’s modulus. (Luna & Jadi, 2000).

Small strain levels are considered to be less than 10^-3 percent deformation while medium to large strain levels are considered to be in the range 10^-2to 5 percent. This of course varies with different soils. (Luna & Jadi, 2000).

(17)

THEORY

2.1.2 Dynamic young´s modulus

The dynamic young’s modulus (ED) is calculated according to the same relationship with Hooke’s law as its static counterpart. The steeper slope at the beginning of a stress- strain curve will grant a higher value for the dynamic young’s modulus due to the higher gradient as can be seen in Figure 6.

Figure 6 General stress-strain curve for soil with dynamic modulus

The reason why young’s modulus decreases with increasing deformation is highly relatable to the behaviour of the shear modulus for different stresses and deformations. Shear modulus has similar static-dynamic behaviour as young’s modulus and we know from Equation (3) that young’s modulus is heavily dependent on the shear modulus as it can be seen in Figure 7.

One extra thing worth mentioning about the dynamic range of a stress-strain curve is that the short duration of the vibratory loads which creates small strains to the soil can be considered as undrained conditions. (Wayne, 2000).

Figure 7 Static and dynamic shear modulus (Luna & Jadi, 2000)

However most in-situ geotechnical survey methods such as Ram-Sounding Test and Cone Penetration Test measures when the soil reaches failure (medium to large strains) and are able to evaluate young’s modulus and shear modulus after this hence acquiring the static young’s modulus and static shear modulus. When acquiring static parameters from these methods but the dynamic is sought after, a correlation or conversion needs to be done.

(18)

The border between static- and dynamic properties are determined by the strain rate and is not as simple as it has been defined in the 1.3 Limitations chapter, but with that limitation there is correlations which can be easily used between static and dynamic young’s modulus. The correlation used for this thesis is the one proposed by I. Alpan (Wichtmann & Triantafyllidis, 2009).

Generally the difference between correlations are how they define which levels of strains that are considered dynamic and which ones that are static. Some correlations are only valid for different soils or different survey methods. (Wichtmann & Triantafyllidis, 2009). Which correlations that have been used for this thesis and how they are used will be further described in 3.3 Geotechnical correlations.

However there are also methods, which are able to measure parameters during the small strain phase. The standard geotechnical surveys cause medium to large strains as mentioned earlier and thereby measure parameters within that range. Within the range of small strains consists vibratory deformations. Therefore, methods of inducing vibrations and measuring how the soil reacts have proven to be an excellent way to conduct measurements within the dynamic range of a soil. (Wayne, 2000).

Seismic behaviour is strongly relatable to soils elastic parameters. When using seismic surveys the wave propagation speed within the soil is measured. The compression/primary wave (VP) and the secondary/shear wave (VS) is of the most interest when evaluation elastic parameters as can be seen in Equation (4) and (5). (Tavakoli & Rasmussen, 2016).

𝑉_𝑃 = √𝐾 +4𝐺

𝜌 3 = √𝑀 𝜌

(4)

𝑉_𝑆 = √𝐺 𝜌

(5)

In Equation (4) a new modulus is presented, the P-wave modulus (M). This modulus

relationship exists to make the primary wave velocity more usable. The standard way of using it often generates two unknown variables: (K, G) while the equation with M generate none.

The P-wave modulus can still be converted to G or ED or other sought parameters which makes it more favourable. (Wichtmann & Triantafyllidis, 2009).

(19)

THEORY

The density (ρ) of the soil is measured in kg/m³and can either be measured from core samples or acquired empirically with rather good accuracy. The elastic wave properties of soil or rock are highly dependent on its density as seen in Equation (4) and (5). Density is in turn

depended on a variety of different aspects such as degree of compaction, porosity, water content and of course the composition of the material. In general both the density and velocity of the seismic waves will increase with increasing depth. A correlation between density and P-wave velocity can be seen in Figure 8.

Figure 8 Correlation between P-wave velocity and density (Tavakoli & Rasmussen, 2016)

However a closer look at Equation (4) and (5) reveals that the density is the denominator and thereby should the wave velocity decrease with increasing density. The answer to why it is the opposite is that the shear (G) and bulk (K) modulus also increase with density and with a faster rate. (Tavakoli & Rasmussen, 2016).

Further theory about seismic waves and seismic survey methods will be mentioned in the chapter 2.2 Seismic waves and for each respective methods chapter

.

(20)

2.2 Seismic waves

To further understand the seismic methods used for this thesis an underlying knowledge about how seismic waves propagate in soil and their different behaviours are needed. As mentioned in 2.1.2 Dynamic young´s modulus the seismic behaviour of a soil is strongly relatable to the soils elastic parameters. There are different types of waves and also different parameters that affect these waves and their velocities such as damping, water content and interference to name a few. Waves and parameters will be described in the chapters below, 2.2.1 General wave theory and 2.2.1 Wave types.

2.2.1 General wave theory

Seismic waves are a gathering name of the vibration and acoustic energies that either are transmitted into the ground or caused from within the ground itself. These waves can have a large energy level span from less than the vibrations caused by the wind blowing across the soil up to the seismic energies released close to a continental ripping earthquake that can be measured with ease around the globe.

The later example will cause large deformation and will not correspond well to the soils behaviour during the elastic phase when the deformations are small. The elastic phase is of interest for seismic surveys since it will reveal the soils elastic parameters i.e. how the particles move with respect to each other within the soil when subjected to different forces.

(Möller, Larsson, Bengtsson, & Moritz, 2000)

Seismic waves propagate from a source through the soil according to a pattern, which can be described with the concept of wave rays and wave fronts. The wave rays can be thought of as arrows dashing out in all directions from the epicentre in a spherical 3-D pattern. Wave rays indicate the direction that the wave front is propagating and is perpendicular to the front.

Wave front can be defined as points of equal value similar to the ripples on a pond. (Tavakoli, 2016). A conceptual sketch of wave rays and wave fronts can be seen in Figure 9.

Figure 9 Propagation of seismic waves described with wave rays and wave fronts (Tavakoli, 2016)

The perfect spherical propagation is however only valid for purely isotropic materials. In soils there is often different layering and obstacles which will disturb and alter this pattern. One altering factor is the damping factor. Some of the wave energy will always convert into heat

(21)

THEORY

or form micro movements in the soil material and the water within it when the energy is partially absorbed. This will make the wave lose some of its energy as it propagates. Higher frequency waves are more easily damped by the soil than lower frequencies, which will penetrate deeper and further. Another altering factor is interference, a phenomenon which can occur when there is a series of waves interacting with each other. If the difference in

frequencies is small between the waves they could cancel each other or create a wave with increased amplitude. Interference does not influence the wave velocity but by altering the amplitude it can make it difficult to identify the waves. (Möller, Larsson, Bengtsson, &

Moritz, 2000).

The presence of rocks or other obstacles will of course also alter how the wave react and bounces but probably the most important phenomena of acoustic energies within layered soil and rock is the concept of reflection and refraction. Reflection and refraction occurs when a wave hits the interface between two layers with different elastic parameters, which will alter the wave velocity. A reflected wave occur when an incoming wave hits an interface between two layers. Some of the energy will then reflect back upwards at the same angle creating a reflected wave as seen in Figure 10. Wave energy will tend to take the easiest way i.e. in the medium with the elastic parameters where it can maintain the highest velocity. If the lower layer of the interface has elastic parameters, which allow the wave to travel at higher velocity refraction will occur. The refracted wave will propagate at a different angle according to Snell’s Law, which can be seen in Equation (6) and Figure 10 where V1 is the wave velocity for the top layer, V2 is the wave velocity for the bottom layer, i is the angle of the incoming wave and r is the angle of refracted wave. (Tavakoli, Seismic method, 2016).

𝑆𝑖𝑛 𝑖 𝑆𝑖𝑛 𝑟 = 𝑉₁

𝑉₂ (6)

Figure 10 Refraction and reflection of an incoming wave. (Tavakoli & Rasmussen, 2016)

Depending on V1, V2 and the incoming waves angle a critical refraction could happen. This is when the angle r is 90° meaning that the refracted wave will travel along the interface

(22)

within the higher velocity layer. As the wave loses energy some of the energy will refract back towards the surface with the same angle as the incoming waves angle. (Tavakoli &

Rasmussen, 2016). This phenomenon is used in Refraction surveys, which will be mention more in 2.5 Seismic refraction survey.

2.2.1 Wave types

The text above has described some general notations on how seismic waves behave but regardless of the behaviour it exists different types of waves. Seismic waves can be split into two mayor groups namely body waves and surface waves.

Body waves are waves which can form in homogenous materials. These body waves can then be further divided into compression waves, (also called primary waves or P-waves) and shear waves, (also called secondary waves or S-waves). (Tavakoli & Rasmussen, 2016).

The other mayor group of seismic waves are surface waves. Surface waves are created when there are one or more free surfaces available. Surface waves can be divided into a series of different waves but the most common ones are Rayleigh waves and Love waves. However Love waves will not be mentioned any further in this thesis due to its irrelevance to any of the used survey methods. (Tavakoli & Rasmussen, 2016).

Primary waves (P-waves) have the fastest wave velocity of all the seismic waves hence its name. The P-waves moves forward according to a pattern composed of compressions and dilations of particles as can be seen in Figure 11 where the white arrow symbolize the propagation of the wave and the red arrows indicate the movement of the particles within the medium. This pattern is quite similar to the movement which a worm travels. (Möller, Larsson, Bengtsson, & Moritz, 2000).

Figure 11 Movement behaviour of compression waves (Dahlin, Larsson, Leroux, Svensson, & Wisén, 2001)

The velocity of the P-wave is governed by how compressible a material is. A material which is more difficult to compress will result in greater P-wave velocities, which also can be seen with Equation (4). Some common velocities of the P-wave for different materials can be seen in Figure 12.

(23)

THEORY

Figure 12 P-wave velocities for some different materials (Dahlin, Larsson, Leroux, Svensson, & Wisén, 2001)

As it can be seen in Figure 12 the velocity is higher if the material is below the ground water table (GWT) than above. This phenomenon is depending on the porosity of the soil and caused by the fact that the P-wave velocity in water is higher than in air. This leads to soil with water filled pores having a higher velocity then the same type of soil with pores filled with air or gas. (Tavakoli & Rasmussen, 2016).

Secondary waves (S-waves) generally are the waves that arrive first after the P-wave and thereby have the second fastest wave velocity of all the seismic waves hence its name. The S- waves moves forward according to a transversal pattern where the particles move

perpendicular to the orientation of the wave propagation. This can be seen in Figure 13 where once again the white arrow symbolizes the propagation of the wave and the red arrows

indicate the movement of the particles within the medium. This pattern is instead quite similar to the movement which a snake travels. (Dahlin, Larsson, Leroux, Svensson, & Wisén, 2001).

Figure 13 Movement behaviour of shear waves (Dahlin, Larsson, Leroux, Svensson, & Wisén, 2001)

The velocity of the S-wave is governed by how prone a material is to shearing. A material which is more resistant to shearing will generate higher velocities. S-wave velocity can be acquired using Equation (5) and is better to use then Equation (4) when dynamic parameters are sought because it is not dependent on water content. (Tavakoli, Seismic method, 2016).

Some common velocities of the S-wave for different materials can be seen in Figure 14.

(24)

Figure 14 S-wave velocities for some different materials (Dahlin, Larsson, Leroux, Svensson, & Wisén, 2001)

In Figure 14 it can be seen that the S-wave velocity is within a much lower range then general P-wave velocities. Another interesting phenomenon which also can be seen is that the velocity is the same for a material below and above the GWT. The reason for this is that the material movement within a shear wave is dependent of its shear modulus as seen in Equation (5).

Waters shear modulus is zero which will result in that S-waves can’t propagate through water.

Rayleigh wave, shortened R-wave is a type of surface wave and is measured with MASW (Multi channel Analysis of Surface Waves) surveys. R-waves moves according to an elliptical backwards rolling pattern quiet similar to the way an ocean wave moves. A principal sketch of R-wave movement can be seen in Figure 15.

Figure 15 Movement behaviour of Rayleigh waves (Dahlin, Larsson, Leroux, Svensson, & Wisén, 2001)

Much of the destruction from earthquakes are caused by these waves which contain a lot more energy than P- and S-waves. R-waves are however dispersive, meaning that the amplitude of it decreases exponentially with depth. Unlike body waves this wave contains a series of different frequencies, which will affect its velocity. (Dahlin, Larsson, Leroux, Svensson, &

Wisén, 2001). Generally an approximate velocity of the Rayleigh wave is calculated with Equation (7) according to (Tavakoli & Rasmussen, 2016)

𝑉_𝑅 = 0,9 ∗ √𝐺 𝜌

(7)

(25)

THEORY

2.2.2 Seismic survey equipment

There are many ways to measure seismic waves and many different survey methods. The basic principal does however stay the same. Firstly one need something that creates a

recognisable signal, which will travel through the medium of interest towards a receiver that can register this same signal. To get a bigger area surveyed the amount of receivers are often quantified and then there is also the need for cables or other equipment, which connects and collects all the data registered from the receivers. All this information then needs to be gathered and converted with respect to time, amplitude and other parameters of interest.

Lastly a computer with appropriate software for the task is needed for interpretation and presentation of the data. (Milsom & Eriksen, 2011).

The source of energy for the survey i.e. the emitter of signal can be many different types of sources depending on what the purpose of the survey is. The standard source of energy for the surveys done in connection with this thesis are explosives and hammer with strike plate.

Another example of ground signal source could be a falling weight or a vibrotruck. Surveys can also be conducted within and between boreholes, then a down-hole hammer or down-hole geophone is needed. Offshore surveys for gas or oil often use airguns, sparkers or boomers.

Different sources have different amplitudes and frequencies which will decide what source is the best for the task. One thing that all sources have in common is that they are connected with a trigger and the seismograph. This trigger will open the channels for a predetermined limited time allowing the geophones to send information through the channel cable to a seismograph. (Tavakoli, 2016).

The seismograph is the centrepiece equipment of a seismic survey. It synchronizes the arrival time of incoming information from the geophone with the triggered source and digitalizes it all thus enables for further calculations and interpretations. (Milsom & Eriksen, 2011)

Geophones are one of the most crucial pieces of equipment but is basically an electromagnet, which is composed of a coil wrapped around a magnet. The magnet is suspended with a spring and will move according to the changes in ground movements thus generating small electrical currents, which the seismograph can interpret as seismic waves. Geophones can also have different frequency intervals and orientations depending on the task. (Milsom & Eriksen, 2011) A sketch of a vertically moving coil geophone can be seen in Figure 16.

(26)

2.3 Geotechnical survey methods

Mankind has for a long time needed to survey the ground for constructional reasons. In Sweden these surveys are dominated by geotechnical surveys. In southern and central Europe, geophysical surveys are also common. The reason for this is both traditional but also based on the geological differences caused by the effects of the ice age up here in the northern

hemisphere. Geotechnical surveys could however be considered to be the standard way of investigation. Every geotechnical method consist of penetration of the ground, either by pushing, hammering or screwing down in different ways all while measuring how the soil counter reacts this external force. These methods measure when the soil reaches failure and even past, which means that it is in the domain of medium to large strains.This chapter will further explain the geotechnical survey methods used for this thesis.

2.3.1 Cone Penetration Test

Cone penetration test (CPT) is a geotechnical method which got increasingly popular in the 1980s due to the progressions in electronics (SGF:s Fältkommité, 2013). The reason for this is that CPT logs and derives a series of different parameters: Cone tip resistance (qt ), Sleeve friction resistance (fs), Friction ratio (Rf), depth, tilt and temperature. Pore pressure (U) is also logged in CPTu, which is considered to be the standard way of cone penetration test today.

(SGF:s Fältkommité, 2013). The method for measuring with CPT in its standard way is pushing it down into the soil at a rate of 20mm/s until sought depth or until a maximum cone tip resistance (qt)-value has been achieved.

The equipment is composed of a 60° angled conical tip as seen in Figure 17. The tip is filled with a liquid and used to measure how much pressure is needed to push the rod and test equipment down. Placed above the tip is the friction sleeve which measures the friction, often in the form of cohesive forces caused to the side to the rod. Filters to measure pore pressure can be placed on the tip, between the tip and sleeve or after the sleeve and is often composed of either a thin slot or porous stone and are also filled with a liquid. Lastly there are electrical sensors for data acquisitions from the other components. (Iliescu & Geron, 2012).

Figure 17 Conical tip (10cm²) for CPTU with porous filter (SGF:s Fältkommité, 2013)

(27)

THEORY

CPT is a good method for accessing the layering of friction and cohesion soils, which are free of boulders and rocks. Since the equipment works by applying pressure from the soil to the liquid within the equipment, pushing it down to a rock could overload the equipment and render the measurement useless. The measurements from CPT can also be used in

combination with empirical relationships to calculate deformation and strength properties for the soil such as undrained shear strength (τfu), friction angle and from those two the static young’s modulus (ES). This will be further explained in 3.2.2 CPT calculations.

2.3.2 Ram-Sounding Test

Ram-sounding test (RST) is an old Swedish developed survey method originating from the 1930s. RST can be used to evaluate deformation and strength parameters of the soil in combination with empirical relationships and also how deep point bearing piles needs to be placed. A big advantage with RST is that it is a heavier and more rigid method then CPT and weight-sounding, which means that it can be used in moraines and other firm soils where the other methods cannot be driven down. The equipment used for this method is a standard geotechnical survey drilling rig with a free falling weight device and a conical cylindrical probe tip as can be seen in Figure 18.

Figure 18 Ram-sounding conducted by Sweco

This method is performed by lifting a weight of 63,5 kg 50 cm up then letting it fall freely and land upon the probe rod, which will transfer the force down into the soil through the conical cylindrical probe tip. How many blows per 20 cm penetration are then measured and the survey is ended if it reaches 200 blows/20 cm. (SGF:s Fältkommité, 2013).

(28)

2.3.3 Weight-Sounding Test

Weight-sounding test (WST) is the oldest Swedish standardized survey method dating back to 1917. This method can be used in loose to semi firm soils without rocks and boulders and is mainly used to determine soil layer stratigraphy and relative firmness of the soil. The equipment used for this method is a standard geotechnical survey drilling rig equipped with probe rods with a WST-tip. The WST-tip can be seen in Figure 19.

Figure 19 WST-tip (SGF:s Fältkommité, 2013).

This method is performed by pushing the rod with tip into the ground with a pressure up to 1 kN (100 kg) while measuring achieved penetration. When 1 kN is reached a rotation is initiated and the number of half revolutions are counted per 20 cm penetration. (SGF:s Fältkommité, 2013).

(29)

THEORY

2.4 Seismic Cone Penetration Test

The main subject of this thesis is the Seismic Cone Penetration Test (SCPT). SCPT is a hybrid test combining seismic downhole technique with a CPT probe. This is done by the adding of two or more geophones to an otherwise complete CPT-setup. Additional equipment needed except the CPT with geophones, is a seismograph, sledgehammer, strike plates and cable system, which can be seen in Figure 20 as a schematic sketch of the system.

(Holmsgaard, Ibsen, & Nielsen, 2016).

Figure 20 Schematic sketch for a SCPT setup (Holmsgaard, Ibsen, & Nielsen, 2016)

The L-shaped left and right strike plate is favourably placed under the tracks of the drilling rig with its weight on them to ensure a good contact with the ground. The sledgehammer can either be used by hand or with an arrangement as in Figure 20. The hammer is connected with a trigger, which goes off when it hits the strike plate sending a signal to the seismograph to open the channel to the geophone. Lastly there exists a P-wave strike plate not shown in Figure 20 but can be seen on the left side of Figure 22.

SCPT measures VP and VS in addition to the parameters acquired from a standard CPT

mentioned in 2.3.1 Cone Penetration Test. This method however faces the same limitations as CPT when used in soils containing rocks and boulders. Seismic surveys can however be used in predrilled holes in the same fashion as downhole and cross-hole seismic surveys without the CPT being active, a good contact between the soil and geophone must however be insured. The drilling rig pushing the probe down often insures a good mechanical contact between probe and soil in “new” holes and thereby allowing a good signal response from the source of energy. (Iliescu & Geron, 2012).

The method for SCPT surveys are similar to standard CPT with a depression rate of 20mm/s until sought depth or max qt value but the difference is that the survey is stopped every meter for conducting the seismic part of the survey. While the probe is stopped a pore pressure dissipation test could also be conducted to measure what the pore pressure is at that depth level without the increased pressure caused by the probe moving.

(30)

The seismic survey is composed of hitting the strike plates, which will initiate seismic waves that will travel into the ground and reach the geophones at their current depth. Hitting the L- shaped strike plates will cause primarily dominant shear waves while hitting the P-wave strike plate will cause primarily dominant compression waves. How it looks in the field can be seen in Figure 21 and Figure 22 which are pictures from the actual surveys conducted by Sweco.

Figure 21 S-wave initiation

Figure 22 P-wave initiation

As it can be seen in Figure 21 and Figure 22, the sledgehammer was operated by hand. This will generate good results if the blows are consistent. A hammer arrangement will perform the same blow every time and does not get tired, which makes it favourable especially for test sites with many survey points. For each level of depth and survey one strike per plate should suffice for the seismic survey but to confirm repeatability and to insure a good signal to noise ratio several strikes can be made. The reasons for it being two L-shaped strike plates are to gain the ability to create left and right polarized shear-waves. Since the compression wave is faster than the shear-wave, first pick of the shear-wave is not always straightforward. With the left and right polarized shear-waves you will get two waves that if done correctly looks like mirrored copies of each other that can then be used to confirm the first arrival of the shear- wave, which will be used to calculate its velocity. (Holmsgaard, Ibsen, & Nielsen, 2016).

(31)

THEORY

2.5 Seismic refraction survey

Seismic refraction surveys is a purely seismic survey method, which is based on the fact that seismic waves travel at different velocities depending on different elastic parameters of the medium it travels within. It also take advantage of the behaviour that waves follow when they hit the interface between two layers with different elastic parameters as it is mentioned in 2.2.1 General wave theory. In refraction surveys the P-wave is often used due to the fact that it is the fastest wave and thereby easiest to identify. Shear-waves can also be measured with the same method but is less common.

The general concept of this method is to measure the time it takes for the P-wave to travel through different layers of soil towards a geophone compared to the time it takes for the direct P-wave traveling along the surface to reach the same geophone. This will result in known VP

(P-wave velocity) for different depths and layers. Those different VP can then be used to determine the border between soil and rock, depth and thickness of different layers, material types using an empirical table like the one in Figure 12 , zones of fracture and depth to groundwater table. With known P-wave velocity, elastic parameters can also be calculated together with Equation (4) but with some uncertainties due to P-wave velocity being dependent on water content etc. (Möller, Larsson, Bengtsson, & Moritz, 2000).

The equipment used for this survey method consists of a seismograph for collecting data from the geophones via the signal cable. The number of geophones used are limited by how strong the source of signal is. How much depth penetration a survey has is also dependent on the number, and especially the length, of the geophone lines; longer line gives a deeper survey.

An initiation of signal is also needed, which is often consisting of a sledgehammer with trigger and strike plate or explosives. These components are parts of almost every type of seismic survey but used in different fashions. These are also explained in 2.2.2 Seismic survey equipment. (Milsom & Eriksen, 2011). The setup in the field looks almost the same for

MASW and Refraction surveys and in Figure 25 a picture from the actual surveys in Malmölandet can be seen.

Seismic refraction surveys are conducted in field by planting the geophones into the ground in a straight line with a fixed distance between them. These then need to be attached to the signal cable, which runs to the seismograph. The geophones also need to have a known position, which can be acquired with for example a RTK-GPS (Real Time Kinematic-Global

Positioning System) system. A signal is initiated in the form of an explosion or sledgehammer strike at several different places along the line that will send seismic waves throughout the soil, which can be registered by the geophones. (Dahlin, Larsson, Leroux, Svensson, &

Wisén, 2001). A sketch of the course can be seen on the left side of Figure 23.

(32)

Figure 23 shows how the wave propagates through the surface (red arrow) which is also called the direct wave and will reach each geophone at a certain time. At the same time the wave will reach a layer with different elastic parameters, which will allow a faster velocity (green arrow). After a while the wave within the deeper layer will refract up towards the surface and reach the geophones faster than the direct wave. By observing first arrival for the P-wave for each geophone a time-distance graph similar to the one on the right side of Figure 23 can be achieved which can be used to calculate depth and VP. (Tavakoli, 2016).

Figure 23(Left) Sketch of refraction survey course. (Right) corresponding time-distance graph. (Tavakoli, 2016).

Refraction surveys are an advantageous method to get a general picture of the geology when supplemented with geotechnical surveys for calibration. The method does however suffer from some disadvantages. In Figure 23 a flat stratigraphy with two different soil layers are shown, if there are steeply angled surfaces or layers underneath a flat one the refraction analysis will be much more difficult to calculate due to how the waves will refract and travel.

The biggest disadvantaged with the method is that it demands soils with increasing elastic parameters towards the depth. By looking at Figure 10 and Equation (6) one can see that if a lower VP layer is underneath a higher one, then angle r will be less than the incoming angle i.

This will make the wave go further down into the deeper layer instead of refracting along it and later go up towards the geophone line. Practically this means that lower velocity layers in- between faster layers can’t be detected with this method. (Milsom & Eriksen, 2011).

S-waves can be used in the same way with refraction if dynamic modules are primarily sought. This will eliminate the uncertainties regarding how the water content influence the wave velocity but is also increasingly more difficult to conduct. The reason is the increased difficulty of picking out first arrival of S-waves. S-wave refraction is done by using

geophones, which are oriented alongside the ground instead of into the ground; this will make the geophones more susceptible for S-waves rather than P-waves. The initiation of signal can also been done similar to how it is done with SCPT, which will generate more S-wave dominant seismicity. (Luna & Jadi, 2000).

(33)

THEORY

2.6 Multi-Channel Analyses of Surface Waves

MASW is the youngest of methods investigated in this thesis with modern MASW dating back to the early 2000s. MASW is another purely seismic survey method, which is very similar to refraction survey. The main difference is that the Rayleigh-waves (R-wave) velocity is the one of interest. This method takes advantage of the fact that the R-wave is strongly dependent on the S-wave in layered soil according to Equation (7). R-waves are a type of surface wave which is mentioned in 2.2.1 Wave types. These waves are dispersive ,which means that the R-waves are composed of a frequency interval where the velocity in different layers are depending on the frequency. (Dahlin, Larsson, Leroux, Svensson, & Wisén, 2001).

When an R-wave is initiated the wave will propagate throughout the soil with different frequencies that will have different velocities depending on depth, which can be seen in Figure 24. As mention in 2.2.1 General wave theory this is due to the fact that higher frequencies are more easily damped than lower frequencies, which will penetrate deeper.

Body waves, like P- and S-waves, rarely show this behaviour, which separates this method from SCPT and Refraction surveys. However the equipment and field setup for MASW is the same as for refraction analyses with the geophones, seismograph, signal cable and hammer + strike plate. This will, similar to refraction surveys, also create a 2D velocity profile by initiating a signal at different places along the geophone line. (SGF:s Fältkommité, 2013).

Figure 24 Surface wave with different velocities depending on frequency (Park Seismic LLC, 2017)

There are two ways to conduct MASW surveys. One way is active, which is the way used for this project and is the one described above similar to refraction surveys. The other one is the passive method, that gathers data for a longer time and uses the ambient sound as a signal.

How MASW is conducted in the field is similar to how refraction surveys are conducted but afterwards there is a need for an advanced dispersion and inversion analysis to acquire S- wave velocities. How the S-wave is calculated from MASW is described further in 3.6 MASW analyses.

To gather data in the field you place geophones in a straight line with equal spacing between them in the same fashion mention in 2.5 Seismic refraction survey. In Figure 25 a geophone and signal cable from the actual survey can be seen. The number of geophones are often 24 but can be less or more. If more, then you often need an extra seismograph since they have a limit of how many geophones they can receive data from. A signal needs to be created that will travel through ground.

(34)

One big difference between refraction survey and MASW is that a single shot/strike at one end of the geophone line is often sufficient for MASW surveys. A heavier source of impact will generate lower frequencies, which will increase penetration depth. (Ólafsdóttir, 2014).

MASW surveys can be used to investigate the same things as a refraction survey or SCPT but with some advantages. It does not have the limitation of needing increasing velocity with depth such as the refraction survey and also with a better resolution. One of the largest advantages with MASW is that a sledgehammer and strike plate is often sufficient for a good signal to noise ratio while refraction surveys often require explosives. This makes it ideal for seismic surveys in urban areas. The reason for this is that the amplitude of the surface wave is larger than for P- or S-wave. (Ólafsdóttir, 2014).

Figure 25 Geophone and signal cable at Malmölandet

(35)

METHOD

3 METHOD

This chapter describes the test sites, positioning and type of the different survey conducted and used for this thesis. How the analysis of the results are calculated and evaluated for each survey method is also explained within each respective subchapter.

3.1 Site investigations

Located approximately 7 km north of Norrköping is the Stavsjö-Loddby area, which is also called Malmölandet. This area consist of mostly farmlands with mainly clay/silt soils and some moraine hills but in the future it will be the location of the East link projects high speed railway. According to the projects length measurement the area of interest picked for this thesis is located between 114km+150 and 119km+600 whereas 114km+150 is closer to Norrköping. Within this length, 4 separate sites have been defined with respect to different methods and distances between them. Each of these sites have a SCPT in the center and is including surrounding surveys in a radius of approximately 50 m around it. The sites and an overview of Malmölandet can be seen in Figure 26. The Refraction survey is located along the length measurement and will cross each of the sites while the MASW only crosses Site 3 and 4. Within these 4 test sites there are in total 47 surveys conducted between 2016 and 2017 which will be used for this thesis.

(36)

3.2 Calculations from geotechnical surveys

3.2.1 Empirical values

Density is an important property needed when calculating both static and dynamic modulus, especially for the seismic related equations where density is the denominator. The Swedish Transport Administration, which sets many of the standards of how geotechnical calculations should be done in Sweden, have created an empirically based table of specific weight (kN/m³) with the abbreviation γ for some crushed and natural materials as seen in Table 1.

(Trafikverket,TK, 2014). This table will be used in combination with dividing with the gravity constant g = 9,81 m/s² to acquire the density when the density is not known from laboratory testing etc.

Table 1 Characteristic values of the heaviness for some crushed and natural materials (Trafikverket,TK, 2014).

(37)

METHOD

Poisson´s ratio is a parameter often taken from empirically based table values, which can be seen in Table 2. The table used for this thesis is provided by (Sefindia, 2017) and corresponds well with the values used by the geotechnical engineers at Sweco in Luleå.

Table 2 Typical values of Poisson´s ratio (Sefindia, 2017).

Type of soil 𝜈

Clay (saturated) 0,4 – 0,5 Clay (unsaturated) 0,1 – 0,3 Sandy clay 0,2 – 0,3

Silt 0,3 – 0,35

Sand (dense) 0,2 – 0,4 Sand (course) 0,15

Sand (fine) 0,25

Rock 0,1 – 0,4

Loess 0,1 – 0,3

3.2.2 CPT calculations

Static young’s modulus is calculated from CPT surveys using a program called CONRAD 3.1 provided by SGI. This program evaluates parameters from measured values and mathematical relationships based on Swedish natural soils. One of the empirically based mathematical relations can be seen in Figure 27 here below. This method of evaluating CPT-surveys is the recommended method according to Swedish Transport Administration and used by

geotechnical engineers throughout Sweden. (Trafikverket,TR, 2014).

Figure 27 Young´s modulus from CPT-probing in sand (Trafikverket,TR, 2014).

(38)

3.2.3 RST calculations

The Swedish transport administration also have recommended methods for evaluating RST and WST, which will be used in combination with Microsoft Excel program. The empirical relationship for RST proposed by the Swedish transport administration can be seen in Figure 28. The RST survey results have been viewed and presented with the Novapoint Geosuite program provided by Trimble.

Figure 28 Young’s modulus from RST-sounding in sand (Trafikverket,TR, 2014).

3.2.4 WST calculations

The empirical relationship for WST proposed by the Swedish transport administration can be seen in Figure 29 which is also used in combination with Windows Excel. Like the RST survey results, WST survey results have also been viewed and presented with the Novapoint Geosuite program provided by Trimble.

Figure 29 Young’s modulus from WST-sounding (Trafikverket,TR, 2014).

(39)

METHOD

3.3 Geotechnical correlations

Static young’s modulus is calculated according to 3.2 Calculations from geotechnical surveys and dynamic young´s modulus can be calculated using wave velocities, density and Poisson’s ratio, which will be described within each of the geophysical seismic survey methods.

However, to be able to use the geotechnical surveys as means of acquiring dynamic

parameters and thereby comparing the results from SCPT, a correlation is needed. The general correlation between ES and ED used for this thesis is the one that is proposed by Alpan, which can be seen in Figure 30 below.

Figure 30 Correlation between ES and ED according to Alpan (Wichtmann & Triantafyllidis, 2009).

Using a program called Graphgrabber, Equation (8) could be derived for the Alpan curve from Figure 30.

SCPT is an ideal method for comparing static and dynamic modulus’s since it is a hybrid method, which can be used to measure and calculate them both. The static modulus can be derived from the CPT part of the survey while the dynamic part comes from the wave

velocities acquired through the geophones. Studies by (Holmsgaard, Ibsen, & Nielsen, 2016) came up with a best fit equation for sandy silts with clay stripes, which will be used for this thesis and can be seen below in Equation (9) with qt given in MPa.

𝐸_𝑑𝑦𝑛/𝐸_{𝑠𝑡𝑎𝑡} = 27,799 ∗ 𝐸_{𝑠𝑡𝑎𝑡}^−0,48 (8)

𝑉_𝑆 = 99.45 ∗ 𝑞_𝑡^0,428 (9)

(40)

3.4 SCPT analyses

The seismic adapter of the SCPT needs to have two geophones or more to be able to receive accurate wave velocities. This is because one can identify similar arrival patterns hence being able to pick first arrival more easily with two geophones. Wave velocity can then be

calculated by dividing the distance between the geophones with the difference in arrival time between them. (Iliescu & Geron, 2012).

This makes the SCPT a semi-point survey, which will generate wave velocities with

resolution depending on the distance between the geophones and also the number of test done with respect to change in depth. The data analysed can then be presented as for example seen in Figure 31.

Figure 31 SCPT-example

In Figure 31 one can see that the distance between the geophones are one meter and the test is conducted every meter. The program used to analyse the data is SeismicLog, which is

provided by Geotech who also provides the SCPT equipment.

P- and S-wave velocity calculations are based on when they arrive but picking this first arrival of waves are not as straight forward as it sounds. Noise can influence picks and especially the S-wave is more difficult to pick since it is not the first wave to arrive to the geophones.

Initiating the signal several times and thereby stacking it will lead to better identification of the signal and a better signal/noise ratio. This is done by calculating the mean signature of the wave and removing of anomalies. (Holmsgaard, Ibsen, & Nielsen, 2016).

(41)

METHOD

The SCPT surveys conducted at the east link projected faced some problems with

measurements of the P-wave, which lead to the results of those not being used for this thesis.

The S-wave however was calculated using the method mention in 2.4 Seismic Cone

Penetration Test using left and right polarized shear-waves. When put together, as in Figure 32 and mirrored, one can identify the first arrival and then calculate the wave velocity. This method of determining first arrival is called reverse polarity (Iliescu & Geron, 2012).

Figure 32 Mirrored left and right polarized shear waves (Iliescu & Geron, 2012)

With known VS from the survey, density and Poisson’s ratio acquired either from laboratory test or based on empirical values from 3.2.1 Empirical values the dynamic young’s modulus can be calculated. This can be done by using Equation (5) and Equation (3) resulting in Equation (10) shown here below.

𝐸_𝐷 = 2(𝑉_𝑆²∗ 𝜌)(1 + 𝜈) (10)

(42)

3.5 Refraction analyses

The general method of refraction survey is described in 2.2.1 General wave theory and 2.5 Seismic refraction survey. To analyse the received data a program called SeisImager/2D was used, which is provided by the company Geometrics that also provides parts of the equipment.

SeisImager/2D quantifies the data from several geophones and several shot points, which can then be used to create a 2D velocity profile.

The P-wave is picked by identifying the first arrival of the signal reaching the geophone. In Figure 33 the noise, first arrival and damping can be seen. Picking first arrival is often relatively simple if there is a homogenous ambient noise environment. Conducting a survey near a loud construction site while the signal is acquired from a hammer and strike plate, the signal might be masked by noise and thereby making it difficult to accurately pick first arrival. (Milsom & Eriksen, 2011)

Figure 33 P-wave first arrival

An equation was provided by (Wichtmann & Triantafyllidis, 2009) that can be used in

combination with Equation (4) to calculate dynamic young’s modulus. The result is Equation (11) which needs VP, ν and ρ as input values.

𝐸_𝐷 =𝑉_𝑃² ∗ 𝜌(1 + 𝜈)(1 − 2𝜈) 1 − 𝜈

(11)

Equation (11) is however only valid above the groundwater table. The reason for this is the increased measured velocity caused by the P-wave going through the water rather than the soil grains. This phenomenon can be seen in Figure 12 aswell. A solution for this is provided by (Studer, Laue, & Koller, 2007) who suggest a reduction of VP based on soil material type.

For the rather fine soils observed at Malmölandet a reduction of factor 2 was chosen, meaning that the measured VP are reduced to 0,1 of its original velocity. If the P-wave and S-wave are both known for an area then Equation (10) and Equation (11) can be combined to back calculate Poisson’s ratio to see if the empirically based assumption is correct.

𝜈 = 𝑉_𝑃²− 2𝑉_𝑆² 2(𝑉_𝑃²− 𝑉_𝑆²)

(12)

Evaluation of SCPT-surveys as method for accessing dynamic modulus