Interface Based in-Situ Soil Classification

8/10/2019 Interface Based in-Situ Soil Classification

1/8

Interface Based In Situ Soil Classification

Hebeler, Gregory L., Ph.D., P.E.Golder Associates Inc., Atlanta, USA

Frost, J. David, Ph.D., P.E. P. Eng.,School of Civil and Environmental Engineering, Georgia Institute of Technology, USA

Keywords: CPT, MPFA, MFA, interface, soil classification

ABSTRACT: A main objective in all geotechnical site investigations is to determine the type, extent, andproperties of the geologic materials in as much detail as possible within the constraints of the given siteconditions and project budget. Cone penetrometers (CPTs), specifically piezocones (CPTUs), offer a siteinvestigation tool that can effectively identify the behavioral type and extent of tested stratigraphy, and

provide unparalleled profiling ability, with CPTU response times typically sufficiently fast to identify verythin layers (< 5 mm) (Lunne et al., 1997). Accurate geostratification and classification are paramount tosuccessful geotechnical engineering practice, as the soil layering and classification often serve as the basis forall subsequent analyses and calculations. The values of CPT data are functions of a number of fundamentalsoil characteristics, and as such, the variations in measured response can be used to identify both the layeringand properties of tested soils. The interface behavior of soils is known to vary as both a function of soil typeand the contacting interface properties. Most notably soil - continuum interface response is known to be

primarily affected by the angle of internal friction of the soil and the surface roughness of the counterfacematerial. The multi friction attachment devices recently developed at Georgia Tech have the ability to providein situ measurements of interface behavior across the full range of typical surface roughness propertiesencountered in geotechnical engineering and for all soil conditions amenable to penetrometer investigations.This paper looks into analyzing the friction data obtained for the various geologic conditions tested to date, tosee if the use of the Georgia Tech CPT attachment devices: the multi friction attachment (MFA) and multi

piezo friction attachment (MPFA) can provide data allowing for improved in situ soil classification throughthe application of fundamental soil-interface behavior concepts.

1 INTRODUCTION

Two multi friction sleeve penetrometer attachmentshave been developed by researchers at the GeorgiaInstitute of Technology (DeJong and Frost, 2002;Hebeler and Frost, 2006) that allow for the direct insitu characterization of soil-interface behaviorduring conventional CPT soundings. The firstgeneration device, the multi friction attachment

(MFA) developed by Frost and DeJong (Frost andDeJong, 2001; Frost and DeJong, 2005; DeJong andFrost, 2002), was designed to allow for the direct insitu assessment of soil-continuum interface behaviorover a range of surface characteristicssimultaneously within a single sounding. The directmeasurement of interface behavior in situ isachieved through the use of four independentfriction sleeves, connected in series along acylindrical housing equal in diameter to theconventional 15 cm

2CPT device.

The multi piezo friction attachment (MPFA)device, developed by Frost and Hebeler (Hebeler,2005; Hebeler et al., 2005; and Hebeler and Frost,2006) is the second generation of multi frictiondevices developed at Georgia Tech (GT). The MFAand MPFA (MFAs) are designed for use behindconventional 15 cm2 CPT devices to allow forsimultaneous measurement of CPT parameters (e.g.,qc, fs, u2) in conjunction with additional interface

shear measurements, and in the case of the MPFA,the corresponding pore pressure response. The useof a 15 cm

2 geometry, as opposed to the more

conventional 10 cm2 geometry, was chosen to

provide additional internal pass through space toaccommodate the large amount of wiring required inthe MFA devices. Lunne et al. (1997) report that theresponse of corrected parameters from cone

penetrometers ranging from 5 to 15 cm2all provide

nearly equivalent response.The MPFA has been developed and proof tested

to provide for the robust assessment of interface


2/8

performance across a range of soil, counterface,loading, and effective stress conditions. The MPFA

provides the ability to simultaneously measureinterface response across a range of surfaceroughnesses, within a stable shearing regime alongthe penetrometer shaft, and within the framework ofeffective stress theory. The MPFA provides thisassessment through four independent measures ofinterface response (fa1, fa2, fa3, and fa4) and five

independent measures of dynamic pore pressurealong the shaft (ua0, ua1, ua2, ua3, and ua4) of apenetrated probe in addition to, or independent of,conventional CPTU measurements (qt, fs, and u2).The MPFA device consists of four piezo frictionmandrels 260 mm in length, accepting frictionsleeves of 110 mm in length (providing equivalentsurface area to conventional 10 cm2 CPT frictionsleeves) at the front of the mandrel, followed by asolid housing 150 mm in length and 43.7 mm indiameter containing a 12.5 mm diameter button type

piezo element centered 15 mm behind the trailingedge of the friction sleeve, as shown in Figure 1. In

addition to the four piezo friction mandrels, theMPFA is also equipped with a baseline piezomandrel 100 mm in length containing an identical12.5 mm diameter button type piezo elementcentered 15 mm behind the front edge of themandrel. The MFAs overcome commonshortcomings of conventional friction sleeves byaccepting sleeves of specified roughness, providingfour independent measures of interface frictionwithin the stable shearing region along the shaft, andthrough the use of isolated non-subtraction load cells.

2 TEXTURED FRICTION SLEEVES

The knowledge that both surface roughness and

hardness are significant factors in the response of

particulatecontinuum systems guided the

development of the MFAs and led to the use of

replaceable textured friction sleeves of varying

roughness. Wear of the testing materials is

minimized by using hardened steel alloys throughout

the device. The sleeve texture was designed to be

self-cleaning to eliminate the possibility of soil

particles clogging the texture, and changing thesurface properties with depth during penetration. It

was also found through extensive previous research

and proof of concept testing that the texturing should

consist of peak features extending beyond a base

substrate to allow for better soil engagement across

the range of particles sizes and shapes encountered

in situ (DeJong, 2001). Additional considerations

required that the sleeves induce internal shearing of

the soil rather than only sliding along the surface at

higher roughness values, and that the texturing

pattern should be easily machinable. The resultant

texturing pattern consists of an offset diamond

texture (Figure 2) with variations in the height of the

diamonds used to modify the magnitude of surface

roughness.

Figure 1. Multi piezo friction attachment (MPFA) configuredwith conventional CPTU Module. (a) schematic - bracketsindicate sensor offset from tip in meters, (b) design detail, and(c) piezo friction sleeve mandrel design detail.

Figure 2. Schematic of textured friction sleeve pattern

fs

fa1

u2

fa2

fa3

fa4

AttachmentDigitalHousing

Digital Housing

ua1

ua0

ua4

ua3

ua2

qc

Friction Sleeve

Pore Pressure

Tip Load

Dual AxisInclinometer

DigitalBoard

AttachmentDigitalBoards

FrictionSleeve

FrictionSleeve

FrictionSleeve

FrictionSleeve

Mandrel

Mandrel

Mandrel

Mandrel

Mandrel Piezo

Sensor

PiezoSensor

PiezoSensor

PiezoSensor

PiezoSensor

AttachmentSleeveMandrel

ReplaceableAttachmentFrictionSleeve

AttachmentIndividualLoad Cell

AttachmentIndividualPiezo Sensor

(0.67)

(0.81)

(0.88)

(1.07)

(1.14)

(1.33)

(1.40)

(1.59)

(1.66)

fs

fa1

u2

fa2

fa3

fa4

AttachmentDigitalHousing

Digital Housing

ua1

ua0

ua4

ua3

ua2

qc

Friction Sleeve

Pore Pressure

Tip Load

Dual AxisInclinometer

DigitalBoard

AttachmentDigitalBoards

FrictionSleeveFrictionSleeve




MandrelMandrel

MandrelMandrel

MandrelMandrel

MandrelMandrel

MandrelMandrel Piezo

Sensor

PiezoSensor

PiezoSensor

PiezoSensor

PiezoSensor









(0.67)

(0.81)

(0.88)

(1.07)

(1.14)

(1.33)

(1.40)

(1.59)

(1.66)

a b

c

DirectionofPenetration

Passthrough

HW

S

Note:Schematicisaplanarprojectionofthecylindricalsleevesurface.

DirectionofPenetration

Passthrough

HW

S

Note:Schematicisaplanarprojectionofthecylindricalsleevesurface.


3/8

While the base pattern of the diamond texture isheld constant, the height of the diamonds is varied tocreate patterns of varying surface roughness. Thecommon diamond heights used in the preliminarytesting sequence were Rmax = 0.125, 0.25, 0.5, 1.0,and 2.0 mm in addition to smooth sleeves ofroughness equal to conventional CPT frictionsleeves, Ra= 0.50 m. Rmax is the absolute verticaldistance between the highest peak and lowest valley

along a surface profile, and is often sampled overlength intervals on the order of the mean particlediameter. Average roughness,Rais defined as,

=L

a dxzL

R0

1 (1)

where L is the sample length and z is the absoluteheight of the profile from the mean line. The use ofintermittent texturing provides non clogging surfacesthat allow for continuous in situ testing, butsubsequently necessitate increased magnitudes of

Rmax roughness to engage equivalent levels ofshearing achieved with continuously texturedsurfaces (e.g., those found in conventionallaboratory testing and practical interfaceapplications).

While the offset diamond texture creates thedesired combination of particle shearing and slidingto both prevent clogging and to engage the soil mass,a third mechanism exists during the penetration oftextured sleeves. Due to the increased diameter ofthe diamond texture over the remainder of the CPTmodule, a punching shear or bearing capacity typefailure zone exists along the leading edge of thediamond texture. Consequently, the measured sleeve

stress consists of both interface shear resistance andan end bearing component herein termed the annular

penetration force (APF). It has been observed inMFA and MPFA results that the response oftextured sleeves is controlled by a combination oftwo primary mechanisms: interface shearing andsliding along the sleeve length (the isolated interfaceor shear response), and a punching shear or bearingcapacity type failure located at the onset of sleevetexturing (the annular penetration component).

npenetratioannularsheartotalfff += (2)

The isolated interface component of measuredtextured sleeve response is representative ofconditions experienced in situ and in the laboratoryfor continuously textured surfaces, and as such is thequantity applicable to interface design. Results fromDeJong (2001) and Hebeler (2005) have shown thatthe magnitude of APF at every discrete measurementdepth can be calculated by directly scaling thecorresponding CPTU qt response to the appropriateannular area of sleeve texture (Equations 3 and 4).Subtracting the discrete values of the APFcomponent from the total measured response allows

the isolated interface response to be determined ateach measurement depth (Equation 5):

( ) annulustexturet AqRAPF = (3)

sleeve

npenetratioannularA

RAPFf

)(= (4)

sleeve

annulustexture

TtotalshearA

A

qff = (5)

where Atexture annulus = the annular area of texturalasperities, Asleeve = friction sleeve surface area, and

APF(R)= the annular penetration force for a specifictexture height, R. The ability to isolate thecomponents of textured shearing from the totalmeasured values allows the isolated interface

behavior over a wide range of roughnesses to bequantified in situ.

3 IN SITU INTERFACE MEASUREMENTS

An example sounding is presented to highlighttypical results for the MFA and MPFA devices. Theexample sounding shows results from the MFAdevice configured with sleeves of increasingroughness in series. This setup is a typicalconfiguration, and allows the entire interfaceshearing surface roughness relationship of a site to

be characterized in a single sounding. The results ofsuch a sounding are shown in Figure 3, with Rmax=Smooth, 0.125, 0.25, and 0.50 mm for the fourfriction sleeves, respectively. An overlay plot of thefour friction sleeve profiles is shown in therightmost plot of Figure 3 to highlight the effects ofchanging the sleeve surface texture. The soundingshown in Figure 3 was conducted at a predominantlysilica sand site in Vermont USA with thininterbedded silt layers, seen at approximately 6.8and 8.6 meters below ground surface (m-bgs) in theFigure 3 profiles.

A number of interesting behaviors can be noted inFigure 3, most notably the marked increase inmeasured sleeve resistance as a function of increasedtexture height. While the magnitude of sleeve stress

increases with increasing surface texture, it isimportant to note that each sensor trace follows thesame stratigraphic changes with depth, irrespectiveof roughness indicating that a series of interface testscan be successfully conducted within the samesounding. Additionally it is of interest that themagnitude of divergence in the individual sleeveresponses is reduced to almost zero within the finegrained silt seams at 6.8 and 8.6 m-bgs, indicatingthat the critical roughness, or counterface roughnessabove which interface shearing transitions to occurinternally within the soil, is reduced for these layers


4/8


5/8

Isolated interface response can be effectivelymeasured in situ across a wide range of surfaceroughness values and soil conditions.

The range of sleeve textures currentlyimplemented appears to provide sufficientcoverage across both the range of surface

properties utilized in practice and the range ofsoils typically encountered in situ.

fs response was observed to consistently be

slightly higher than measured smooth MFAsleeve responses across a range of geologicconditions; this is believed to be due to largerlateral stresses present in the vicinity of the

penetrating tip. The positions of the MFAs friction sleeves

appear to be within a stable shearing regimeacross all tested soil types for standard steady-state CPT penetration.

The interface behaviors of structured materialsmeasured in situ appear to be controlled byadditional internal and interface interactionmechanisms, which appear to affect both the

magnitude of measured shear resistance and theextent of the influence zone created by the

penetrometer.These results show promise in the ability to

measure geotechnical interface response in situacross a range of geologic conditions. Future workwill attempt to provide a larger database of in situinterface measurements from the MFAs to verifyand substantiate the observed behaviors.

4 INTERFACE BASED CLASSIFICATION

A main objective in all geotechnical siteinvestigations is to determine the type, extent, and

properties of the geologic materials in as much detailas possible within the constraints of the given siteconditions and project budget. Cone penetrometers,specifically piezocones, offer a site investigationtool that can effectively identify the behavioral typeand extent of tested stratigraphy, and provideunparalleled profiling ability, with CPTU responsetimes typically sufficiently fast to identify very thinlayers (< 5 mm) (Lunne et al., 1997). Accurategeostratification and classification are paramount to

successful geotechnical engineering practice, as thesoil layering and classification often serve as the

basis for all subsequent analyses and calculations.The values of CPT data are functions of a number offundamental soil characteristics, and as such, thevariations in measured response can be used toidentify both the layering and properties of testedsoils.

There are currently five methods for determininggeostratification from CPT data: visual examinationof the sensor traces, soil classification charts,

probability methods, variograms, and clustering

analysis. Both the visual and automatedclassification methods are based on experience withCPTU data, and are based on the general changes inCPTU sensor response as a function of soil behavior.It is important to note that classification based oncone penetrometer data does not provide accuratedetermination of particle size and distribution, butrather distinguishes materials based on differences insoil behavior (Douglas and Olson, 1981). All CPT

soil classification methods are based on the generaltrends known to exist for CPT response. The generaltrends being that: sandy soils tend to produce high qt(> 3 MPa), low Rf,and very low u (u = um- u0);clay soils tend to produce low qt(< 2 MPa), highRf,and very high u; and organic soils tend to producevery low qt , very high Rf, and very high u.Sensitive soils tend to produce higher qt, lower Rf ,and higher u; and soils with high OCR tend to

produce higher qtandRfbut lower or even negativeu than similar soils at lower OCR (Lunne et al.,1997). While visual examination of CPT records canoften lead to accurate identification of the main

layers within the tested stratigraphy, subtledifferences in sensor response resulting fromchanges in soil properties (e.g., OCR, PI, , Su, etc.)are often more easily determined through the use ofobjective computer algorithms and classificationcharts.

Interface behavior of soils is known to vary asboth a function of soil type and the contactinginterface properties. Most notably soil - continuuminterface response is known to be primarily affected

by the angle of internal friction of the soil and thesurface roughness of the counterface material. The

multi friction attachment devices have the ability toprovide in situ measurements of interface behavioracross the full range of typical surface roughness

properties encountered in geotechnical engineering,and for all soil conditions amenable to penetrometerinvestigations (Hebeler, 2005). This paper looks intoanalyzing the friction data obtained for the variousgeologic conditions tested with the MFAs to date, tosee if the use of data from the MFAs can provide animproved means for soil classification.

In the search for a parameter to define theinterface response as a function of soil type, several

primary requirements were held paramount:

nondimensionality, normalization with depth and/orstate of stress, maintaining continuity with currentCPT soil classification parameters, and a means toclearly differentiate changes in response as afunction of soil type. A number of different

parameters were investigated to most clearlydifferentiate the behaviors of the tested soil types asa function of the measured in situ interface behavior.In the end, a parameter was chosen that best met thecriteria defined above, and is herein called the MultiFriction Classification Parameter (MFCP):


6/8

1

10

100

1000

0.10 1.00 10.00Normalized Frictio n Ratio: FR= 100*fs/ (qt- vo) [%]

NormalizedConeResistance:

Qt=(qt-

vo

)/

'vo

SRVT - Silica Sand (1-5m) SRVT - Silica Sand (7-8m)LS - Silica Sand (3-5m) LS - V. Loose Silica Sand (9-11m)LPWA - Cem Ca lcareous Sand (1-2m) LPWA - Calcareous Sand (3 -5m)SRVT - Silt Seam A (6.8m) SRVT - Silt Seam B (8.6m)MPSC - Cooper Marl (14-20m) BWDWA - Soft Cl ay (4-8m)

Silt

Seams

Soft

Clay

Dense

Silica

Cooper

Marl

V. Loose

Silica

M. Dense

Silica

Cemented

Calcareous

Calcareous

Sand

12

3

4

5

6

7

8 9

1

10

100

1000

0.10 1.00 10.00Normalized Frictio n Ratio: FR= 100*fs/ (qt- vo) [%]

NormalizedConeResistance:

Qt=(qt-

vo

)/

'vo

SRVT - Silica Sand (1-5m) SRVT - Silica Sand (7-8m)LS - Silica Sand (3-5m) LS - V. Loose Silica Sand (9-11m)LPWA - Cem Ca lcareous Sand (1-2m) LPWA - Calcareous Sand (3 -5m)SRVT - Silt Seam A (6.8m) SRVT - Silt Seam B (8.6m)MPSC - Cooper Marl (14-20m) BWDWA - Soft Cl ay (4-8m)

Silt

Seams

Soft

Clay

Dense

Silica

Cooper

Marl

V. Loose

Silica

M. Dense

Silica

Cemented

Calcareous

Calcareous

Sand

12

3

4

5

6

7

8 9

1

10

100

1000

0.1 1.0 10.0 100.0 1000.0 10000.0

Normalized Multi Frictio n Parameter: MFCP = 2.5*vo / (H0.50 - SM)

Norma

lize

dCone

Res

istance:

Qt=

(qt-

vo

)/

'vo

SRVT - Si li ca San d (1-5m ) SRVT - Si li ca Sand (7-8m)

L S - Si li ca San d (3-5m ) LS - V. L oos e Si li ca San d (9-11m)

LPWA - Cem Calcareous Sand (1-2m) LPWA - Calcareous Sand (3-5m)

SRVT - Si lt Seam A (6.8m ) SRVT - Si lt Seam B (8.6m)

MPSC - Cooper Mar l (14-20m) BWDWA - So ft Clay (4-8m)

SoftClay

Cooper

Marl

Silt

Seams

Dense

Silica

M. Dense

Silica

V. Loose

Silica

Calcareous

Sand

Cemented

Calcareous

1

10

100

1000

0.1 1.0 10.0 100.0 1000.0 10000.0

Normalized Multi Frictio n Parameter: MFCP = 2.5*vo / (H0.50 - SM)

Norma

lize

dCone

Res

istance:

Qt=

(qt-

vo

)/

'vo

SRVT - Si li ca San d (1-5m ) SRVT - Si li ca Sand (7-8m)

L S - Si li ca San d (3-5m ) LS - V. L oos e Si li ca San d (9-11m)

LPWA - Cem Calcareous Sand (1-2m) LPWA - Calcareous Sand (3-5m)

SRVT - Si lt Seam A (6.8m ) SRVT - Si lt Seam B (8.6m)

MPSC - Cooper Mar l (14-20m) BWDWA - So ft Clay (4-8m)

SoftClay

Cooper

Marl

Silt

Seams

Dense

Silica

M. Dense

Silica

V. Loose

Silica

Calcareous

Sand

Cemented

Calcareous

SMH

vMFCP

=

50.0

0*5.2 (6)

where v0= total vertical stress, H0.50= the isolatedinterface stress for a H0.50 textured attachmentsleeve, and SM = the measured sleeve stress for asmooth attachment sleeve.

While CPT and MFA/MPFA measurements aremore directly affected by lateral as opposed to

vertical stress (Salgado et al., 1997), reliableestimates of in situ lateral stress are still notavailable with conventional geotechnical methods.As such, as with other normalized classification

parameters, the total vertical stress is used tonormalize the behavior with changes in stress state.The denominator in the MFCP, the mathematicaldifference of textured and smooth interface response,represents the slope of isolated interface response asa function of surface roughness. Variations ininterface response as a function of surface roughnessare characterized by the difference between the SMand H0.50 isolated interface response. These sleevesare focused towards the sensitive range of lowersurface textures while still providing a significantdifference in response across a wide range of soil

behaviors. Interface response as a function ofsurface roughness is known to deviate as a functionof fundamental soil characteristics as shown inFigure 4. The form of the above expression waschosen so that the response of the MFCPwas seento exhibit similar trends in response the conventionalnormalized CPT friction ratio FR = fs / (qt - v0)],(i.e., values that generally increase from hard sand tosoft clay behaviors). The adjustment factor of 2.5 in

the MFCP expression was chosen so that themagnitude of theMFCPis approximately equivalentto FRfor clean silica sand geologies.

To allow the reader to develop general familiaritywith the new normalized classification parameter,the response from the full range of tested geologiesis presented for both the conventional Qt - FRclassification chart (Robertson 1990, 1991) and thenewly introducedMFCPplotted vs. Qtin place of FRin Figures 5 and 6.

Figure 6 exhibits the expected trend of increasingMFCP with increasing clay-like response. Thecemented calcareous sand exhibits the lowestMFCP,with an average response of MFCP= 0.6, with thestructured silty clay of the Cooper Marl showing thelargest response, with an average value of MFCP=457. The larger range of responses for the MFCPdoes not seem to increase the spread of valuesobserved within each layer. The response for theCooper Marl shows a fair amount of scatter in

MFCP across the layer thickness; however, thisincreased scatter was also seen for the FRparameterand is believed to result from the inherentheterogeneity of this deposit. Using the conventionalRobertson (1990, 1991) classification system, the

Figure 5. Response of selected soil layers across a range ofgeologic conditions within the Qt- FRRobertson (1990, 1991)soil classification framework.

Figure 6. Response of the same soil layers as a function ofnormalized cone resistance (Qt) and the newly developed multifriction classification parameter (MFCP).


7/8

Georgia Institute of Technology - Geosystems Group Multi Fric tion Sleeve CPT Attachment DataTest Site: Timian Yard - South Royalton, VT Oper: JD, GLH, DF MS #2: SM2 MS #5: N/A

Date: Tip Conf: 15cm2 CPT MS #3: 30H.5S3 Pen. Rate (cm/s): 2

Test ID: Z12O0007C MS #1: SM1 MS #4: SM4 Meas Rate (Sa/cm): 1

Notes: Normalized Friction Parameter Comparison

10/12/2000

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

0 5000 10000 15000

Tip Stress (kPa)

Depth(m)

-200 0 200 400

Pore Pressure (kPa)

0 100 200 300

Sleeve Stress (kPa)

0 200 400

SM , H0.50 (kPa)

SM

H0.50

0 5 10

MFCP

0 2 4

FR

2 3 4 5 6 7 8

Rob 1990 Class

Qt - Fr

Qt - Bq

0 5 10

Norm Fric Comp

Fr

MFCP

Figure 7. Response of select CPTU-MFA sensors, FR, Robertson (1990, 1991) soil classification, and the newly developedMFCP for a predominantly silica sand test site with isolated silt seams in Vermont.

Cooper Marl is identified as a silty sand by the Qt-FRbehavior, exhibiting similar response to the veryloose silica sand layer, and as a clay by the Qt - Bq

behavior. The MFCP more clearly differentiatesbetween the Cooper Marl and very loose silica sandlayers, exhibiting the expected shift towards the clayend of the behavior spectrum for the Cooper Marlresponse.

While the classification style plots provide auseful perspective on the data across a range ofcharacteristic soil types, it is also insightful to

observe the continuous behavior of the MFCPwithdepth for the various sites. Figure 7 presents datafrom the a predominantly silica sand site in VermontUSA showing traces of qt; u2;fs; SM;H0.50;H1.00;Robertson 1990-91 Classification; and an overlay ofFRandMFCP in the subplots from left to right. TheRobertson (1990, 1991) classification systemdenotes the stratigraphy at the Vermont USA testsite as sand to sandy silt behavior over the range oftested depths, excepting the thin layers atapproximately 6.8 and 8.5 m, corresponding to thinsilt seams. The conventional normalized frictionratio (FR) and the multi friction classification

parameter (MFCP) both show similar trends inbehavior with depth. The friction parameters bothshow slightly increasing response with depthindicative of a relative increase in soil density withdepth, and all show a marked increase in responsewithin the depth range of the two silt seams. From a

purely visual identification perspective, one couldargue that the MFCP more clearly differentiates

between the sandy and silty behaviors found at thissite as compared to the FR as a result of the largerchanges in response seen within the silt layers.

5 SUMMARY

A parameter has been developed using the availableMFA and MPFA data in an attempt to aid in theclassification and identification of soil type andcharacteristic behaviors. This formulation and

benefits of this parameter, termed the Multi FrictionClassification Parameter (MFCP), are summarizedherein: The MFCP is fundamentally based on the concept

that the differential magnitude between smooth

and moderately textured interface responseprovides a good indication of soil behavior andtype.

The MFCP is mathematically formulated similarto other classification parameters, in that it isdimensionless and is normalized with the totaloverburden stress to reduce the affect ofmeasurement depth on the parameter response.

An adjustment factor is applied within the MFCPformula such that the response within typicalsilica sand geologies was observed to beequivalent to the conventional normalized friction

parameter (FR). The MFCP effectively varies over four orders of

magnitude for the wide range of soil types testedto date, while still exhibiting clustering ofresponse within individual soil layers similar tothe other normalized CPT parameters.

The increased variation in the MFCP across fourorders of magnitude, as compared to the twoorders of magnitude variation representative ofthe FR parameter, leads to more clearly definedchanges in response for a number of the soilconditions tested.


8/8

It is duly noted that the robust formulation of anyin situ based soil classification parameter or systemrequires a large database of data. While the currentdata does cover a fair range of soil behaviors, theauthors readily note the preliminary nature of the

presented parameter and all conclusions made withregard to the response of the presented parameter.

6 ACKNOWLEDGEMENTS

The first author was supported through a NationalDefense Science Graduate Fellowship during hisdoctoral research, and that support is gratefullyacknowledged. The assistance of Vertek team atApplied Research Associates in South Royalton,Vermont for their collaboration in design andfabrication of the multi piezo friction attachment isgratefully acknowledged.

7 REFERENCES

DeJong, J.T. (2001). Investigation of Particulate-Continuum

Interface Mechanisms and Their Assessment Through a

Multi-Friction Sleeve Penetrometer Attachment, Ph.D.

Dissertation, Georgia Institute of Technology, Atlanta,

USA, 360 pp.

DeJong, J.T and Frost, J.D., (2002). A Multi-Friction Sleeve

Attachment for the Cone Penetrometer, ASTM Journal of

Geotechnical Testing, Vol. 25, No. 2, pp. 111-127.

Frost, J.D., and DeJong, J.T. (2001). A New Multi-Friction

Sleeve Attachment, Proceedings, 15th International

Conference on Soil Mechanics and Geotechnical

Engineering, Istanbul, Vol. 1, pp. 91-94.Hebeler, G.L. (2005). Multi Scale Behavior at Geomaterial

Interfaces, Ph.D. Dissertation, Georgia Institute of

Technology, Atlanta, USA, 772 pp.

Hebeler, G.L and Frost, J.D. (2006). A Multi Piezo Friction

Attachment for Penetration Testing, Proceedings of

Geocongress 2006, ASCE, Atlanta, USA.

Hebeler, G.L., Frost, J.D., Schneider, J.A., Lehane, B.M.

(2005). "Cyclic Friction Piezocone Tests for Offshore

Applications", Proceedings of the International Symposium

- Frontiers in Offshore Geotechnics, ISFOG, Perth, AUS.

Lunne, T., Robertson, P.K., & Powell, J.J.M. (1997). Cone

Penetration Testing in Geotechnical Practice, Blackie

Academic & Professional, New York, 312 pp.

Robertson, P. K. (1990). Soil Classification Using the Cone

Penetration Test, Canadian Geotechnical Journal, Vol. 27,

No. 1, pp. 151158.

Salgado, R., Mitchell, J.K., and Jamiolkowski, M. (1997).

Cavity Expansion and Penetration Resistance in Sand,

ASCE Journal of Geotechnical and Geoenvironmental

Engineering, 123, No. 4, pp. 344-354.

Documents

Interface Based in-Situ Soil Classification