Seismic Cone Penetration Test. Experimental results in

onshore areas.

Alexandra-Ioana Iliescu Jeremy Geron

08.06.2012

Abstract

The use of cone penetration test (CPTu) is an geotechnical onshore site investigation which isoften used as well as other more traditional investigations. To begin with, this article will presenta detailed description of CPTu, SCPTu and shear waves types. The field tests descriptionprovides with information about the soil conditions and draws an idea of the SCPTu set up.For a better understanding of the soil properties, laboratory tests (water content, particle sizedistribution, pycnometer and hydrometer methods) were performed. The CPTu along withSCPTu results provide with reliable information about soil stratigraphy and a lot of data to beinterpreted. Two different methods of finding S-waves are presented along with one method forP-wave in order to analyze and compare with the theoretical assessments.Keywords: seismic, cone penetration test, S-wave, P-wave, soil classification test, shear velocity.

1 Introduction

The cone penetration test (CPTu) is frequentlyused in both onshore and offshore constructionas geotechnical investigation. The seismic conepenetrometer can dramatically reduce the costin time efficiency associated with seismic testing,especially if CPTu is used as part of the regu-lar site investigation program. Comparisons ofonshore seismic cone shear wave velocities withthose measured by both down-hole and cross-hole techniques at sites in Canada, (Rice, 1984),United States, (J.A. Jendrezejczuk) and Bel-gium, (Bouhon, 2010) have already validated theseismic cone technique.This article presents and discusses results

from SCPT preformed in sand and clay in Aal-borg area. The cone bearing, friction sleevestress, cone pore pressure and shear velocitydata can be used to provide a fast and reliabledetermination of soil type and shear strength,according to (P. K. Robertson). The datagiven by the cone can be interpreted to get agood continuous prediction of the soil type andshear strength parameters. When a seismome-ter is integrated into the cone penetration testprocedure, the CPTu becomes SCPTu (SeismicCone Penetration Test). The use of S-Wave ve-locity data in foundation investigations has be-come increasingly popular in recent years, butuse of this valuable and diagnostic study hasbeen delayed because of the difficulty of obtain-ing reliable data, particularly under varying ge-ologic conditions, (Beckstead).To obtain the measurements a rugged velocity

seismometer has been incorporated into the conepenetrometer. Downhole seismic shear wave ve-locity measurements can be made during briefpauses in the CPTu. In (J.A. Jendrezejczuk) theshear wave speed is computed by dividing thedistance between two pairs of receivers by thetime for the signal to travel from one receiver tothe next. There are four types of seismic waveswho propagates through the soil. These canbe divided into two categories. The first cate-gory are the body waves also named flat/volumewaves, their displacement is Longitudinal for theP wave and transversal for the S wave, as seenin Figure 1.The Compression (P) wave are also referred

as irrotational waves which propagate throughsolid and fluids. They propagate at a higher ve-locity than shear waves. The Shear (S) as saidbefore their direction is transversal but the Swaves are also referred as the rotational wavesand are unable to propagate through fluids.

Figure 1: Different types waves. (Rice, 1984)

II

2 Field test

To have an overview of the two different testingsites, Aalborg is located in Jutland which is partof the north of Denmark as seen in Figure 2. Thetests performed in sand were located in the eastpart of Aalborg, the industrial part of the city.On the location of the site a wind turbine bladedeposit will be build. The location for the claysoil is in the center of Aalborg, next to the TrainStation and the Bus Terminal.

Figure 2: Position of the field test, Wikimedia (2006)

Concerning the sand field, this site is situ-ated a few meters from the fjord which meansit is a basin deposit area. The soils compositionis: the top layer (1-4 meters) a fjord deposit ofclay/gyttia, while the lower layers are is mostlycomposed of silty sand. The SCPTu were per-formed until approx. 8 meters depth.

Figure 3: Position of the Boreholes and SCPTu

Nine different SCPTu are executed in a cross-shape positioning. A first line of SCPTu waschosen and the first SCPTu is taken from themiddle of Borehole 126 and 131, aligned un-

til reaching the middle of Borehole 127 and130 (five SCPTu). The distance betweenSCPTu is of ten meters. From the mid-dle SCPTu a perpendicular line of SCPTuis done keeping the same rule of ten me-ters between each other so that another fourSCPTu are to be done, as seen in Figure 3.

Figure 4: Positioningof SCPTu

For the clay soil the siteis situated in the centerof Aalborg. The compo-sition of this in mainlya top layer of sand (2-3)meters of sand and therest is clay. Five dif-ferent SCPTu where ex-ecuted in this location.These SCPTu where real-ized with a 5 meters dis-tance between each otherstarting 5 meters awayfrom the Borehole B4which profile can be seenin Figure 4.Due to a very agglom-

erate area as the centerof the city, different errorscould appear in the resultsof SCPT. These errors canbe generated by the construction site situatednext to the testing site or the presence of thebus terminal and train station. The works fromthe site along with the passing of the trains,buses and cars produce mild vibrations thatcould reach the cone, which is very sensitive onevery interference. In addition, another causefor possible errors in the results could be theappearance of the site. The SCPTu were per-formed on pavement stones and asphalt. Theasphalt in comparison with normal soil, or eventhe pavement stones, absorbs the energy, whichcauses a poor signal for the waves, therefore er-rors in the results. The setup can be seen inFigure 5, where for the P-waves, the plate hasbeen mounted into the ground by drilling intothe asphalt or removing the pavement stones.

Figure 5: Setup for the shear waves

III

2.1 Description of the cone.

The cone used for the tests has standard valuesspecified in the E.U and American standards asseen in Figure 6. The friction sleeve, fs, which isplaced above the conical tip, also has a standarddimension of 150 cm2. A pore pressure trans-ducer is installed to measure the dynamic porepressure during the penetration.The cone pen-etrometer is pushed into the soil at a standardspeed of 20 mm/s.

Figure 6: The standard type of cone.

The memo cone used in this SCPT and theequipment has a certain number of standard val-ues (E.U and American) properties:

• A conical tip;

• A 10cm2 probe with a tip angle of 60◦;

• A 7 channel measuring: point resistance,qc, local friction, fs pore pressure, u , tilt,temperature, electric conductivity, seismic,uniaxial for shear wave measurememnts;

• Depth synchronization;

• Data acquisition system and software;

• Data interpretation, CPT-LOG software.

The signal is transmitted up through the steelof the rods to a microphone on the penetrome-ter. The absence of the cable makes the systemvery easy and time efficient in usage. The mea-surements were used until 8m, engineers choice,

with a penetration rate of 20 mm/s. In certaincases the soil was pre-drilled due to fill. Thecone penetetrometer is advancing through thesoil by being pushed firmly and continuously inthe ground by the truck machine creating me-chanical contact between the seismometer car-rier and the soil. Therefore, allowing good cou-pling and signal response both for clay and sand.Also, the orientation of the cone is controlledby setting up the X and O rings and the depthsynchronizer provides with accurate depth mea-surements.

2.2 Description of machineriesand test setup.

The setup of the machinery starts by positioningthe CPT truck or cabin crawler over the exactposition chosen previously. For the SCPT theseismic sources are a sledge hammer and a pis-ton, which was blown on a steel plate stuck onthe ground. After a 1 meter pre-drill, two plateson both sides of the hole are placed in order toinsure that the left and the right part of the Swave testing are aligned as seen in Figure 7.

Figure 7: Setup for the shear waves

The plates used in this case are ”L” shapedand the bottom of the plate should be equippedwith transversal teeth to improve the contactwith the ground as seen in Figure 8.The distance between the place where the

hammer hits and the SCPTu hole is of 1,4 me-ters. To have a constant and exact same forceon left and right S wave a mounting has beenrealized as seen in Figure 9.Concerning the position for the P waves the

distance for this test is at 1,8 meters from thehole through which the rod string is fed. It hasto be perpendicular to the S wave set up as seenin Figure 10.For the P wave generation is a plate with a

circular part which is pushed into the soil with ahydraulic or pneumatic piston as seen in Figure11. The waves were generated by blowing thepiston on the steel plate from a free fall.

IV

Figure 8: L plate with transversal teeth, Sledge hammerand Penetration Cone Geotech (2012)

Figure 9: Hammer display

Figure 10: SCPT field setup, Geotech (2004)

Figure 11: Introduction of P wave plate into the ground

A specific hammer has been created to havethe comparable wave generated as the one usedfor the S wave setup. It is also designed to gen-erate a constant hit during the hole test. Thehammer is composed a tube of 1,5 meter heightwhich gives a control height on which a chosenweight is dropped off as it can be seen in Figure12. The seismic signal is generated by strikingthe L plate pad with the hammer.

Figure 12: Set-up of the P wave test

2.3 Manipulation

The penetration velocity is measured while therod string is pushed into the soil. Previously, it

V

was defined that the CPT will be performed un-til 8 meters. The Seismic part of the test is per-formed every each meter. When the rod stringhas reached the depth required, the engine of thepenetrometer or drill rig are stopped. This isdone to give the possibility to realize the SCPTtest which is noise sensitive. The interval shearvelocity can be easily checked on field, for qual-ity assessment. As soon as the Seismic part isfinished the CPT can continue the same way un-til the desired depth is reached.

3 Soil Classification Testsfor Sand

With the data obtained both from the bor-ings and the CPTs results, profiles with soiltype stratigraphy, description and correspond-ing depths were plotted using as in Figure 13. Itcan be observed that the first layer is a fill layerfor the first approx. 2 meters, being followed bya a layer of clay and gyttia and after approx 3meters fine sand is reached.

Figure 13: Bore hole 100 and 200 profile with soil typeand depth

3.1 Determination of Water Con-tent

From two of the borings seen in Figure 10, soilsamples were taken in order to perform sev-eral soil classification tests. According to thegeotechnical survey from the boreholes, The firsttest performed was the water content one. Thistest is performed to determine the water (mois-ture) content of soils. The water content is re-quired as a guide to classification of natural soilsand as a control criterion in re-compacted soilsand is measured on samples used for most fieldand laboratory tests according to (DS/S-19000,2004a).The water content, w is defined as the weightloss of the soil in % of the dry weight by dry-ing in an incubator (oven) at a temperature of105 ◦C to a constant weight.

w =Ww

Ws100% (1)

=(W +Bowl)− (Ws +Bowl)

Ws +Bowl100% (2)

where,

Bowl : weight of the bowl [g]W : weight of sample before drying [g]Ws : weight of the dried sample [g]Ww : weight of the water sample [g]

The samples that were tested were taken fromhalf of meter to half of meter until 8 meters werereached. The first two meters of the both bor-ings were covered with topsoil clay. Further-more, gyttia is found for the first boring until3.6 meters, whereas for the second one until 2.8meters. Until 8 meters the presence of fine siltysand was observed.For all the samples taken from both boreholes

Table 1 presents the ranges on which the watercontent varies in the samples. For Borehole 100,the maximum value represents a gyttia samplesituated at 3.5 m depth, whereas the minimumvalue is fine sand found at 8m. For Borehole200, the maximum value represents a clay sam-ple situated at aprox. 2.0 m depth, whereas theminimum value is fine sand found at 4 m.

Table 1: Water content results

Layer Water content, [%]

Borehole Topsoil - clay 23 - 38100 Gyttia 46 - 62

Fine sand 17 - 27

Borehole Topsoil - clay 26 - 55200 Gyttia 54 - 56

Fine sand 20 - 24

VI

3.2 Determination of Particle SizeDistribution

After performing the water content test on everysample, 8 samples were chosen for a sieving anal-ysis, 4 out of each boring from different depthsas seen in the Table 2.

Table 2: Grain size analysis. Borehole sample details

Borehole Sample Depth Type of soilno. no. [m]

100

9 3.6-4 fine silty sand11 4.6-5 fine sand14 6.-6.4 fine sand17 7.6-8 fine sand

200

25 3-3.4 fine sand28 5-5.4 fine sand31 6.-6.4 fine sand34 7.6-8 fine sand

The screenings on each sieve in % of the dryweight of the total of each sample are plottedinto a coordinate system which is function ofthe sieve dimension, as seen in Figure 14. Thescreening percentanges are plotted in the ver-tical axis in an arithmetic scale and the sievedimensions on the horizontal one in a logarithmscale.

Grain Size AnalysisGrain�Size�Analysis100100

80

%s�% 60ngs

nin

een

40cre

40ScS

20

00.01 0.1 1

Sieve mesh sizeSieve�mesh�sizeSample�34 Sample�28 Sample�25 Sample�31

Sample 9 Sample 11 Sample 14 Sample 17Sample�9 Sample�11 Sample�14 Sample�17

Figure 14: Grain size analysis on sand samples

3.3 Determination of ParticleDensity - Pycnometer method

From the results observed from the sievingcurves four different soil samples were taken inorder to perform the pycnometer test in order toobtain the ”relative density” found using Equa-tion (3), (DS/S-19000, 2004b).

Gs =Wv

Wi(3)

where,

Ws : Weight of a given volume soilgrain

[g]

Wi : Weight of the same volume de-ionised water at 4◦C

[g]

The results obtained can be observed in Table3.

Table 3: Relative density results

Sample no Relative density, Gs, [-]

9 2.6614 2.6525 2.6534 2.66

3.4 Determination of Particle SizeDistribution - Hydrometertest

The hydrometer analysis is the process by whichthe weight-related distribution of soil grains af-ter size in the silt fraction (2μm-60μm). Thehydrometer also determines the specific gravity(or density) of the suspension, and this enablesthe percentage of particles of a certain equiva-lent particle diameter to be calculated.According to Section 3.3 the relative density

obtained for Sample 9 is 2.66 as seen in Table3. The grain diameter,d, is found by writingthe Stoke’s law, (DS/S-19000, 2004c) where therelative density is entered instead of the actualdensity so it becomes:

d =

√18η ∗ 100

(Gs − d0)g ∗ 60 ∗√

h

t(4)

The corresponding values of weight percent-ages and grain size are put in the same coordi-nate system as the sieve curve of Sample 9 asseen in Figure 15. The recorded curve part (thebeginning one) is considered as the slurry curve.

Sample 9Sample�9p100100

8080%s�%gs 60ng 60inn

40ee 40recrSc

2020

000 0 1 0 2 0 3 0 4 0 50 0.1 0.2 0.3 0.4 0.5

Sieve mesh sizeSieve�mesh�size

Figure 15: Sample 9 Sieveing curve after HydrometerTest

VII

(a) Sand

(b) Clay

Figure 16: CPT results before and after error filtering for the soil types seen in Figure 13

VIII

4 Interpretation of data

4.1 CPTu Results - Error Filter-ing

The CPTu is considered as a standard methodof assessing soil properties, so measuring errorscan occur during the test. Some of them can bedetected, other might be, wrongly, assumed tobe a real soil property. One of the easiest waysfor detecting the errors is by assessing the rawdata.According to (Andersen, 2011) two criteria

have formed the base for removing the data. Forthe first one, measurements with zero cone tipresistance have been removed as a zero cone re-sistance indicates a cavity in the soil, but forsand deposits is unlikely to happen. The secondone takes into account the errors occurring inconnection with stops of the cone, during thetests. The stops during the penetration testare caused by the need of attaching new rodsor from each meter for performing the seismictests. These errors are due to halts, and it isoften seen as drops in the cone resistance. Therods have a length of 0.95 m and the drops incone resistance and friction sleeve are seen everymeter.For the sand case, when the sand was get-

ting stiffer, around 7 meters, the cone had to behalted more often, as seen in Figure 16. Whenthe penetrometer is halted, the pressure on thecone and the sleeve friction is released. Conepenetration in sands will not occur completelydrained. When the cone stops, small excess porepressure drains away and again builds up duringthe penetration. The peaks marked with therelines represent the halts and they were removedas they are not representative of inherent soilproperties. This is basically based on engineer-ing judgment. Generally, most of the data re-moved was errors produced by halts and the fi-nal results for clay can be seen in Figure 16.

4.2 SCPTu results - verification ofwaves

The seismic signal was generated by the blowsof the hammer for the S-waves and of the pis-ton for the P-waves, only one blow for each typeof wave. The first step of the interpretation ofthe results was to verify if the signal was spread-ing in the right direction along with the depth.The profile sheet is used for viewing the seismicsignal traces at different depths. For sand, inFigure 17 both compressive and shear waves onthe first SCPTu are presented, each graph repre-senting the seismic signal trace loaded file for thecorresponding depth. It can be observed that as

deep the penetration goes the weaker the signalbecomes, which was what it was expected. Thedifference in the depth is of 1 meter. It can beobserved that as deep the penetration goes theweaker the signal becomes.

����������������������������������������������������������

���

�����

�

�

���

�

���

�

��

���

�

��

���

�

���

�

���

�

���

�

(a) P-waves

���������������������������������

���

�����

�

�

���

�

��

���

�

��

���

�

���

�

���

�

���

�

(b) S-waves Left

���������������������������������

���

�����

�

�

���

�

��

���

�

��

���

�

���

�

���

�

���

�

(c) S-waves Right

Figure 17: Shear waves from Sand Sounding

In sand, it was interesting to see if the signalis intercepted correctly by the cone so duringone of the SCPTs performed in sand at a depthof 5 meters 10 successive blows were applied forboth P-waves and S-waves, left and right andcan be seen in Figure 18. Each color representsone blue.

IX

��������������������������������

���

�����

�

���

���

��

����

���

����

���

����

���

����

�

���

��

���

��

���

��

��

�

���

��

��

(a) P-waves

���������������������������������

���

�����

�

���

���

���

��

���

��

���

���

���

�

��

��

��

�

��

�

��

��

��

(b) S- waves Right

���������������������������������

���

�����

�

���

���

���

��

���

��

���

���

���

�

��

��

��

�

��

�

��

��

��

(c) S-waves Left

Figure 18: 10 time blow SCPTu waves in 5 meters depth

For the left shear waves it can be observedthat the signal is the same after multiple blow-ing, but for the right one it can be seen thateach blow’s signal has a different peak at differ-ent places in time. As for the P waves it canbe mentioned that only one of the blows did notfollow the pattern, caused by a possible errorin signal, whereas the rest form a peak signalaround the same time.For clay a different approached was used in

order to verify the accuracy of the signal. Theseismic test was performed both on the cone’sway down into the ground but also while remov-ing it, upwards. In Figure 19 are plotted withred color the down direction waves and with bluethe up direction ones. It can be observed thaton both directions the wave register the samepeak, but when it comes to dissipation the onesthat are measured on the up direction take moretime. This fact can be explained by the void inthe SCPTu hole that prevent the propagationand therefore a proper dissipation of the signalinto the ground.Another problem influencing the accuracy of

the waves was the location of the tests performedin clay. As mentioned before the location, wasnext to a bus terminal and the ground was cov-ered with asphalt which made the placing of theSCPTu plates very difficult. In Figure 20 it canbe observed that for the P-wave problems werenot encountered as the plate was dug into theground.For the S-waves ones it can be observed a dif-

ference in signal distribution from one soundingto another. In Figure 21a and Figure 21b it canbe seen that in the first meters there is not a de-fined peak of signal and a possible cause can bethe asphalt and the topsoil above the clay layerwhich can produce delay in signal, hence a fasterdissipation. For the last meters the signal regis-tered a peak because at a lower level no surfacedisturbances are encountered. Therefore for thelast SCPTus the S-waves plates were placed ona layer of sand added on the asphalt to reducethe effects of the surface disturbances and so thewave distribution can be seen in Figure 21c andFigure 21c .

X

���������������������������������

���

�����

�

��

���

�

���

�

���

�

��

���

�

��

���

�

(a) P-waves

���������������������������������

���

�����

�

������������

�����������

�����������

�������

����������

�������

����������

�

(b) S-waves Left

���������������������������������

���

�����

�

������������

�����������

�����������

�������

����������

�������

����������

(c) S-waves Right

Figure 19: Shear waves from SCPTu in Clay

����������������������������������������������������������

���

�����

�

������������

�����������

�����������

�������

����������

�������

����������

�

Figure 20: P-waves depth profile for Clay

���������������������������������

���

�����

�

���

�

���

�

���

�

��

���

�

��

���

�

���

�

(a) S-waves Left

����������������������������������������������������������

���

�����

�

���

�

���

�

���

�

��

���

�

��

���

�

���

�

(b) S-waves Right

��� ������������������������������������������������������

���

�����

�

������������

�����������

�����������

�������

����������

�������

����������

�

(c) S-waves Left

(d) S-waves Right

Figure 21: Shear waves from Clay SCPTu

XI

4.3 SCPTu results - P-waves

4.3.1 Cross-Correlation Method

For obtaining the velocities generated by theP-waves a relative new method has been used.Using a vertical direction hammer, one blow isapplied on the plate in order to generate thesignal of the wave. In signal processing, cross-correlation is a measure of similarity of twowaveforms as a function of a time-lag appliedto one of them. According to (Schaff and Wald-hauser) in the case of a SCPTu, it is possibleto obtain a correlation function by zero-paddingin the time domain or to be computed by theinverse Fourier transform of the cross spectrum.In this case though, the difference in depth andwave velocity between the two depths is dis-played in Figure22.The P-velocities are displayed in a chart for-

mat depending on depth and type of soil. Theresults for sand can be seen in Figure 23. Thevalue in the fill layer is considered an error inmeasurement as its value is negative, also thegyttia values are relative high whereas the sandones are more or less in the same range of values.

Figure 23: P-velocity for Sand

For clay, the SCPTu results are calculated forboth ”down” and ”up” direction of the test asseen in Figure 24. It can be observed that forthe both directions the velocities follow the samepattern and the values are in the same range.

Figure 24: P-velocity for Clay

4.4 SCPTu results - S-waves

4.4.1 Reverse Polarity Method

The method of determining shear wave velocityfrom seismic CPT data basically involves divid-ing an increment shear wave travel time into anincrement of travel path.The test procedure consists of generating re-

verse polarity shear waves, first by impactingone end of the timber, and then by impactingthe opposite end (left and right). Acceleration-time traces, corresponding to each impact, arerecorded on the computer for subsequent pro-cessing and analysis. In analyzing the data, par-ticular attention is paid to the two records madewith horizontal impacts. The true shear wavesshould reverse polarity, and this characteristic isused as the most important identifying charac-teristic. In some surveys, the shear waves arereadily obvious and this is not difficult. In oth-ers, there may be numerous other arrivals andnoise signals that make identification difficult;hence the need for a clear reversal signature,(J.A. Jendrezejczuk).The first blow (for example, the left side) rep-

resents the seismic record from the SPCTu asseen in Figure 25 part A. To confirm that a realshear wave was obtained, another record is takenby hitting on the other side (in this case, theright) and if the data is correct it should look asFigure 25 part B. The first break is in the oppo-site direction, which confirms that the readingis a shear wave. Many analysts superimpose therecords, for better comparison as shown in Fig-ure 25 part C.

Figure 25: Shear waves reverse polarity when the sourcepolarity is reversed, (Crice).

It has been found that the reverse polarity ofthe source greatly facilitates the identificationof the S-wave and the time for the first cross-over point (shear wave changes sign) is easilyidentified from the polarized waves (forward andreverse) and provides the most repeatable refer-ence arrival time, (R. G. Campanella, 1986).

XII

Figure 22: Cross correlation method for P-waves

Figure 26: Seismic analysis using reverse polarity,(Geotech, 2004)

XIII

4.4.2 SCPTu results - shear velocity

The shear wave velocity is readily computed bydividing the distance between two pairs of re-ceivers by the time for the signal to travel fromthe one receiver to the next. Travel times can becomputed using the start of the S-wave, or anycorresponding prominent feature on the time sig-nals (e.g., zero crossing or peak), as the refer-ence.As an example, using the traces given in Fig-

ure 26, with the start of the S-wave as the refer-ence, the shear wave speed is calculated as fol-lows: we consider X1 = 4m and X2 = 5m asthe reading depths, T1 = 148.14ms and T2 =158.21ms as the tracing amplitudes, ΔXcrt =0.89m as correction factor for the distance bythe depth and from Equation (5) we obtain thefinal result.

Vs =ΔXcrt

ΔT=

0.89

11.07= 80.40

m

s(5)

The results for all the depths in sand soil areplotted in Figure27

Figure 27: Shear Velocities Sand

For clay, the velocities were displayed as acomparison from the ”down” and ”up” directionof the SCPTus as in Figure 28.

Figure 28: Shear Velocities Clay

Another approach is to follow the signal’spath, choosing the strongest signal point as seenin Figure 29, where the traces are placed on the

strongest signal point. The velocities are com-puted using the same routine as shown by Equa-tion (5).

From the Section 4.2 it was observed thatalong the depth the signal follows a path andpattern, so it was considered an interesting ideato see the results if the point of the strongestsignal is followed.For sand and clay the results are seen in Fig-

ure 30 and there can be seen differences in val-ues. The filling layer has negative values andcan be considered as an error in measurement.Unfortunately, the values obtained with the sec-ond method are double in value as the first one.

(a) Sand

(b) Clay

Figure 30: Shear waves from Clay SCPTu obtained usingReverse Polarity Method

4.4.3 Cross-Correlation Method

Cross-correlation calculates the time interval byaligning the signal trains in the time axis, andit utilizes considerably more information in thecollected shear waves than the first arrival andfirst cross-over methods, Liao and Mayne (2006).

Cross-correlation works well if two signals areof the same shape. For both sand and clay theS-waves shapes have been verified in previoussections and in Figure 31 can be shown the signalgenerated by a blow on the left side.

XIV

Figure 29: Seismic Analysis using Reverse Polarity

Figure 31: Cross-correlation method for obtaining S-waves

XV

4.4.4 SCPTu results - shear velocity

The results of the cross-correlation method aredisplayed in a similar matter than the reversepolarity ones. In Figure 32 the sand resultsare displayed and it can be observed that onthe S-left the values are relative small and the2400m/s value is considered to be an error inmeasurement, whereas the S-right ones displaymore realistic values.

(a) S-left

(b) S-right

Figure 32: Shear waves from Sand SCPTu using Cross-Correlation Method

For clay, the ”down” and ”up” SCPTu wasanalyzed and the range of values is similar forboth. Also S-left and S-right signals are similarproviding with the same results for both sidesgenerated wave velocities.

(a) S-left

(b) S-right

Figure 33: Shear waves from Sand SCPTu using Cross-Correlation Method

5 Conclusion

One of the most important conclusion is that thebasic CPTu along with the Seismic test was suc-cessful and provided accurate stratigraphic de-tails for both sand and clay. The seismometeron the cone was able to detect both P-wavesand S-waves signals. However not all the valuesare in the ranges. The seismic test was success-ful, considering that the seismic wave velocitiescould be determined. According to (Andersen),the ranges for sand and gravel, S-waves are be-tween 0 and 200 m/s, P-waves from 500 to 800m/s and for clay, S-waves are between 0 and 250m/s and P-waves from 500 to 700 m/s. Thismeans that the P-waves values are below bothranges. For each site the velocity profiles weredetermined and the values for shear (S-wave)were in general considered to be correct, rangingbetween 0 to 300 m/s for sand (with some neg-ative values on the first meter due to the fill inthe sand) and from 50 to 200 m/s for clay. Asfor the P-waves, they are out of range, varyingbetween 0 to 450 m/s for sand and between 0 to220 m/s for clay.

As highlighted by (R. G. Campanella, 1986),the combination of the seismic downhole methodwith the CPTu provide rapid and reliable means

XVI

of determining stratigraphic, soil properties andvelocity information in one SCPTu. The addi-tion of seismic measurements significantly canimprove the ability of the CPTu, meaning thebearing capacity is obtained from the CPTu andthe soil deformation (elastic parameters) are ob-tained from the SCPTu.It was observed that closer intervals will result

in better resolution of the subsurface layers andmore accurate velocities,but also required moretime to conduct the test. In the end a one meterinterval was a good choice as it could have beenpossible to notice and compare signal traces forboth types of waves.Another important point that was reached is

that the SCPTu downhole survey has advantageslike: faster to carry out (in the same time withthe CPTu), does not require a different hole fromthe CPTu, provides with accurate depth deter-mination during and after the tests and geo-phone orientation is maintained being attachedto the cone. Since modern seismographs havedigital data storage, computers can simplify theprocessing and display of data.Different problems encountered during test-

ing: Since sand was the first time the SCPTutest was realised, it may have caused some er-rors (wrong way of hitting the hammer, not atthe right depth, ect.). This is why it can beobserved that a couple of time we have someShear velocities that are out of range. Thewater level is pretty high since we are 20 me-ters from the fyord and this has not been takenin account. In clay was background frequencynoise caused by the presence of traffic and con-crete and asphalt layer. It was observed thatit seriously affected the signal, at least on thefirst meters, thus the accurate interpretation ofthe both shear and compressive wave signals.Forsome soundings (the last 3) it has been tested toput some soil between the asphalt and the plateto have a better connection. To insure a bet-ter generation of S waves in Clay compared toSand an improvement was brought. One personwas standing on the S plate while the wave wasgenerated to insure that the wave wouldn’t stayin the surface. An other test done was to makethe original SCPTu going downwards but a up-wards test was also brought as it can be seen adifference is shown the closer the test gets to thesurface, this can be explained by the fact thatthere is more air over and under and the con-tact surface is not perfect so it does not allowthe wave to be properly measured by the geo-phone. The final conclusion and one of themost important ones is that, the SPCTus wereperformed for the first time in North of Den-mark and that even though some of the resultsare not accurate, they should be considered as a

starting point to more investigations regardingSCPTu.As a personal statement it would be advised

for a greater experience to go back to the samesand site to add all the evolutions brought in theClay site to observe the difference between thenew values and the previously obtained values.

XVII

References

Andersen. Lars Andersen. LinearElastodynamic Analysis. Aalborg University.

Andersen, Sarah S., 2011.Lauridsen Kristoffer Andersen, Sarah S.Probabilistic site description based on conepenetration test, 2011.

Beckstead. D. O. Beckstead. Example ofDownhole P-Wave and S-Wave UtilizingSignal Enhancement and Polarity Reversal.

Bouhon, 2010. Olivier Bouhon. Apport desmethodes geophysiques dans lareconnaissance liee aux besoins des chantiersde travaux souterrains: TunnelSTAFELTER, 2010.

Crice. Doug Crice. Borehole Shear-WaveSurveys for Engineering Site Investigations.

DS/S-19000, 2004a. DS/S-19000.Geotechnical investigation and testing -Laboratory testing of soil - Part 1:Determination of water content (ISO/TS17892-1:2004). 2004.

DS/S-19000, 2004b. DS/S-19000.Geotechnical investigation and testing -Laboratory testing of soil - Part 3:Determination of particle density -Pycnometer method (ISO/TS 17892-3:2004).2004.

DS/S-19000, 2004c. DS/S-19000.Geotechnical investigation and testing -Laboratory testing of soil - Part 4:Determination of particle size distribution(ISO/TS 17892-4:2004). 2004.

Geotech, 2004. Geotech. SCPT SOUNDING(Seismic Cone Penetration Test). 2004.

Geotech, 2012. Geotech. Seismic CPTadapter. URL: http://www.geotech.eu/index.php/en/seismic-cpt-adapter, 2012.Downloaded: 05/10/2012.

J.A. Jendrezejczuk. M.W. WambsJ.A. Jendrezejczuk. Surface Measurements ofShear Wave Velocity at The 7-GeV APSSite. Argonne National Laboratory ReportLS-129.

Liao and Mayne, 2006. T. Liao and P. W.Mayne. AUTOMATEDPOST-PROCESSING OF SHEAR WAVESIGNALS. Proceedings of the 8th U.S.National Conference on EarthquakeEngineering, Paper No. 460, 1–10, 2006.

P. K. Robertson, R. G. Campanella.D. Gillespie A. Roce P. K. Robertson, R.G. Campanella. Seismic CPT to Measure InSitu Shear Wave Velocity. Journal ofGeotechnical Engineering, V112.

R. G. Campanella, P. K. Robertson,1986. D. Gillespie N. Laing P.J. KurfurstR. G. Campanella, P. K. Robertson. Seismiccone penetration testing in the near offshoreof the MacKenzie Delta. Canadian JournalGeotechnical, V24, pp. 154–160, 1986.

Rice, 1984. Anthony H. Rice. The SeismicCone Penetrometer, 1984.

Schaff and Waldhauser. David P. Schaff andFelix Waldhauser. WaveformCross-Correlation-Based DifferentialTravel-Time Measurements at the NorthernCalifornia Seismic Network. Bulletin of theSeismological Society of America, 95(6).

Wikimedia, 2006. Wikimedia. Aalborg Map.URL: http://commons.wikimedia.org/wiki/File:Map_DK_\unhbox\voidb@x\

setbox\z@\hbox{!}\dimen@\ht\z@

\advance\dimen@-1ex\hboxto\z@{\raise.

67\dimen@\hbox{\char23}\hss}Alborg.

PNG?uselang=fr, 2006. Downloaded:05/10/2012.

XVIII