1

Pitfalls and fallaciesin foundation

engineering designBengt H. Fellenius

Amsterdam

May 27, 2016

www.Fellenius.net

4

The primary base for the design of a piled foundation is the pile capacity as determined in a staticloading test. A routine static loading test provides the load-movement of the pile head...

and the pile capacity?

5

CHIN and DECOURT 235

6

Some people consider the capacity to be the load applied to the pile head*) thatcaused a movement equal to 10 % of the pile head diameter. This is usually claimedto be recommended by Terzaghi. He knew better than that!

Terzaghis's recommendation, and he did give one (Terzaghi 1942), was this: Do not tryto estimate pile capacity unless the pile toe movement is equal to at least 10 % ofthe pile toe diameter. I think you agree that this is a very different recommendation.I have tried to find out where the misquote started and it seems to have been firstput forward by E. DeBeer in the early 1950's and then picked up from there by severalof the "Old Masters".

Terzaghi was referring to pile with diameters of about 300 mm. I have seen themisquoted definition applied to piles with a diameter larger than 1.5 m! At times evento footings! The pile diameter, let alone the pile head diameter, has nothing to do witha pile capacity. It is sad that a few codes and standards have designated this silly andignorant definition as the one to use, notably the EuroCode.

*) EuroCode has now (2016) changed to refer the 10-% movement to the pile toediameter. I do not think this is an improvement.

7

0

2,000

4,000

6,000

8,000

10,000

0 4 8 12 16 20 24 28 32 36

APPL

IED

LO

AD (

kN)

TIME (h)

? ?

Actual Load-Time Schedule of theAraquari Prediction Event

0

2,000

4,000

6,000

8,000

10,000

12,000

0 20 40 60 80 100 120

LOAD

(kN

)

MOVEMENT (mm)

Unplannedunloading

Actual pile-head Load-Movementof the Araquari Prediction Event

Araquari Prediction Event, IFCEE 2015, Brazil. Test results

Actual pile-head Load-Movementof the Araquari Prediction Event

The participants in the prediction event were requested to providethe pile-head load-movement curve from start to 100 mmmovement, i.e., the pile "capacity" of the 1,000 mm diameter pile,as defined by the Brazilian organizers.

The pile was instrumented with strain-gages at several levels.However, the uneven load-holding durations and unequalload increments combined with the (unintended, as it were)unloading/reloading will make the evaluation of the strain-gage records ambiguous and unconvincing.

8

Pile-Head Load-Movements

Araquari Prediction Event — As Predicted and As Compared to Actual Results

Load Distributions

Cnt.

9

I was one of the participants in the prediction event. On receiving the actual test results after the conference, I wroteto all predictors and asked them to tell me what capacity they would asses from the actual load-movement curve.

Note, that assessment is no longer a prediction, but an assessment of fact applying the participants usual method fordetermining capacity as achieved by a static loading test.

Twenty-nine of the participants replied giving me their capacity value.

Araquari Prediction Event, IFCEE 2015, Brazil.Assessments of Capacity

0

2,000

4,000

6,000

8,000

10,000

12,000

0 20 40 60 80 100 120

LOAD

(kN

)

MOVEMENT (mm)

7 values

6 values

3 values

Actual test curve of the Araquari prediction eventwith capacity assessements by 29 participants

Edmonton, Alberta, 2011

Prediction of load-movement and capacity of a 400-mm diameter, 18 mlong, augercast pile constructed in transported and re-depositedglacial till.

0

1,000

2,000

3,000

4,000

0 5 10 15 20 25 30 35 40 45 50

MOVEMENT (mm)

LOA

D (K

N)

10 capacity predictions areat movements > 50 mm

E = 20 GPaE = 35 GPa

7 mm (4 mm + b/120 mm)TESTRESULTS

10

11

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

0 10 20 30 40 50 60

MOVEMENT (mm)

LOA

D (

KN

)

BHF

GeometricMean

260

110

126

COMPILATION OF ALL PREDICTIONS

CapacityPredicted

Compilation of predicted load-movement curves and capacities, TP1

Prediction Event in Bolivia April 2013450 mm diameter, 17.5 m long, bored pile in fine sand

12

is applying a factor of safety or aresistance factor correct, safe, andeconomical, or do we by this approachopen ourselves for litigation anddemands on our liability insurance?

So, when designingfoundations for

capacity,

13

What really do we learn fromunloading/reloading and whatdoes unloading/reloading do tothe gage records?

14

A test on a 2.5 m diameter, 80 m long bored pileDoes unloading/reloading add anything of value to a test?

0

5

10

15

20

25

30

0 25 50 75 100 125 150 175 200

MOVEMENT (mm)

LOA

D (

MN

)Acceptance Criterion

0

5

10

15

20

25

30

0 25 50 75 100 125 150 175 200

MOVEMENT (mm)

LOA

D (

MN

)Acceptance Criterion

Repeat test

0

5

10

15

20

25

30

0 25 50 75 100 125 150 175 200

MOVEMENT (mm)

LOAD

(M

N)Acceptance Criterion Repeat test plotted in

sequence of testing

15

Plotting the repeat test in proper sequence

0

5

10

15

20

25

30

0 25 50 75 100 125 150 175 200

MOVEMENT (mm)

LOAD

(M

N)Acceptance Criterion Repeat test plotted in

sequence of testing

1616

Also the best field work can get messed up if the analysis and

conclusion effort loses sight of the history of the data

The dynamic test (CAPWAP) was performed after the static test.

The redriving (ten blows) forced the pile down additionally about 45 mm.

0

500

1,000

1,500

2,000

0 100 200 300

MOVEMENT (mm)

LOAD

(KN

) STATICTEST

DYNAMICTEST

0

500

1,000

1,500

2,000

0 100 200 300

MOVEMENT (mm)

LOAD

(KN

)DYNAMICTEST

0

500

1,000

1,500

2,000

0 100 200 300

MOVEMENT (mm)

LOAD

(KN

)DYNAMICTEST ina seriesof blows

RepeatedSTATICTEST

1717

NbNqNcr qcu '5.0''

The Pile Toe is Really a Footing:

where ru = ultimate unit resistance of the footingc’ = effective cohesion interceptB = footing widthq’ = overburden effective stress at the foundation level‘ = average effective unit weight of the soil below the foundation

Nc, Nq, N = non-dimensional bearing capacity factors

The Bearing Capacity Formula

Factor of Safety, Fs

Fs = ru/q

1818’

Nq

Min to max Nq ratio is up to≈200 for the same φ’!

The log-scale plot isnecessary to show all curveswith some degree ofresolution.

Nq was determined in tests—model-scale tests

Why is it that nobody hasrealized that somethingmust be wrong with thetheory for the main factor,the Nq, to vary this much?

Let’s compare to the reality?

qt Nqr '

1919

Results of static loading tests on 0.25 m to 0.75 m squarefootings in well graded sand (Data from Ismael, 1985)

0

100

200

300

400

500

600

700

0 10 20 30 40 50

SETTLEMENT (mm)

L O

A D

( K

N )

1.00 m

0.75 m

0.50 m

0.25 m

MOVEMENT

0

200

400

600

800

1,000

1,200

1,400

1,600

1,800

2,000

0 5 10 15 20

MOVEMENT/WIDTH (%)S

T R

E S

S

( KPa

)

1.00 m

0.75 m

0.50 m

0.25 m

Normalized

20

As before, we divide the loadwith the footing area (to getstress) and divide the movementwith the footing width, as follows.

Load-Movement of Four Footings on SandTexas A&M University Experimental SiteJ-L Briaud and R.M. Gibbens 1994,ASCE GSP 41

0

2,000

4,000

6,000

8,000

10,000

12,000

0 50 100 150 200

MOVEMENT ( mm )

L O

A D

(

KN )

3.0 m

1.5 m

3.0 m

1.0 m

2.5 m

0

200

400

600

800

1,000

1,200

1,400

1,600

1,800

2,000

0 5 10 15 20

MOVEMENT / WIDTH (%)

S T

R E

S S

( K

Pa )

Texas A&MSettlement Prediction Seminar

Normalized

21

To repeat, when designingfoundations based on capacityare we not basing our designon an illusion?

22

0

200

400

600

800

0 2 4 6 8 10

PILE HEAD MOVEMENT (mm)

LOA

D P

ER P

ILE

(KN

) Single PileAverage of 4 Piles

Average of 9 Piles

O’Neill et al. (1982)

Group Effect

20 m

Single Pile

9-pile Group

4-pile Group

c/c = 10 b

c/c = 10 b

2323

Shaft Resistance and t-z and q-z functions

Strain-hardening

Elastic-plastic

Strain-softening

The t-z and q-z functions are fundamental to the analysis of pile response

0

20

40

60

80

100

120

140

0 5 10 15 20 25

SHAF

T SH

EAR

(%

of r

ult)

RELATIVE MOVEMENT BETWEEN PILE AND SOIL ELEMENT (mm)

Ratio, θ = 0.200

Hyperbolic (r∞ = 120 %)

Exponential, b = 1.20

Hansen 80 %

rtrg

δtrg Zhang, a = 0.009

0 5 10 15 20 25 30

0 5 10 15 20 25 30

0 5 10 15 20 25 30

0 5 10 15 20 25 30

TOE

HEAD

TOE

"Capacity" ?

24

0

500

1,000

1,500

2,000

0 10 20 30 40 50

LOAD

at P

ILE

HEA

D (

kN)

MOVEMENT (mm)

HEAD

300 mm diameter, 30 m long concrete pileE = 35 GPa; Strain-softening Rs

SHORTENING(Telltale measurement)

0

500

1,000

1,500

2,000

0 10 20 30 40 50

LOAD

at P

ILE

HEA

D (

kN)

MOVEMENT (mm)

HEAD

TOE LOAD vs.Head movement

300 mm diameter, 30 m long concrete pileE = 35 GPa; Strain-softening Rs

SHORTENING

0

500

1,000

1,500

2,000

0 10 20 30 40 50

LOAD

(kN

)

MOVEMENT (mm)

HEAD

SHAFT

300 mm diameter, 30 m long concrete pileE = 35 GPa; ß = 0.30 at δ = 5 mm;

Strain-softening: Zhang with δ = 5 mm and a = 0.0125

SHORTENING

TOE LOAD vs.TOE MOVEMENT

MOVEMENTS AT GAGE LOCATIONS

First shown important by Bramvan Weele back in the 1960s

25

If we want to know the load distribution, wecan measure it. But, what we measure isthe increase of axial load in the pile due tothe load applied to the pile head. Whatabout the axial force in the pile that wasthere before we started the test?

That is, the Residual Force

26

Normalized Applied Load

DEPTH

Load distributions in

static loading tests

on four instrumented

piles in clay

27

B. Load and resistance in DA

for the maximum test load

Example from Gregersen et al., 1973

0

2

4

6

8

10

12

14

16

18

0 50 100 150 200 250 300

LOAD (KN)D

EPTH

(m

)

Pile DA

Pile BC,Tapered

0

2

4

6

8

10

12

14

16

18

0 100 200 300 400 500 600

LOAD (KN)

DEP

TH (

m) True

Residual

True minusResidual

0

2

4

6

8

10

12

14

16

18

0 50 100 150 200 250 300

LOAD (KN)

DEP

TH (

m)

Pile DA

Pile BC,Tapered

0

2

4

6

8

10

12

14

16

18

0 100 200 300 400 500 600

LOAD (KN)

DEP

TH (

m) True

Residual

True minusResidual

A. Distribution of residual force in DA and BC

before start of the loading test

28

FHWA tests on 0.9 m diameter bored pilesOne in sand and one in clay

(Baker et al., 1990 and Briaud et al., 2000)

0

2

4

6

8

10

12

0 10 20 30 40

Cone Stress and SPT N-Index(MPa and bl/0.3 m)

DEPT

H (m

)

SiltySand

Sand

Pile 4

0

2

4

6

8

10

12

0 10 20 30 40

Cone Stress (MPa)

DEPT

H (m

)

Pile 7

N

qc

ClaySilty

Sand Clay

29

ANALYSIS RESULTS: Load-transfer curves

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0 1,000 2,000 3,000 4,000 5,000

LOAD (KN)

DEPT

H (m

)

PILE 4SAND

MeasuredDistribution

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0 1,000 2,000 3,000 4,000 5,000

LOAD (KN)

DEPT

H (m

)

PILE 4SAND

TrueDistribution

ResidualLoad

MeasuredDistribution

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0 1,000 2,000 3,000 4,000 5,000

LOAD (KN)

DEPT

H (m

)PILE 7CLAY

MeasuredDistribution

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0 1,000 2,000 3,000 4,000 5,000

LOAD (KN)

DEPT

H (m

)PILE 7CLAY

TrueDistribution

ResidualLoad

30

0

5

10

15

20

25

30

35

40

45

50

0 500 1,000 1,500 2,000 2,500LOAD (KN)

DEP

TH (

m)

Fellenius et al.(2004)

Static Loading Testat Pend Oreille,Sandpoint, Idaho, for therealignment of US95

406 m diameter, 45 m long,closed-toe pipe pile driven insoft clay

200+ m

Determining True Resistancefrom Measured Resistance (“False Resistance”)

Clay

31Test on a strain-gage instrumented, 406 mm diameter, 45 m long pile driven in soft clay in Sandpoint, Idaho

0

5

10

15

20

25

30

35

40

45

50

-500 0 500 1,000 1,500 2,000

LOAD (KN)

DEP

TH (

m)

ß = 0.60

ß = 0.06

ß = 0.06

"TRUE RES."RESIDUALLOAD

AFTER 1stUNLOADING

AS MEASURED,i.e. "FALSE RES."

A

ß = 0.09

32

Presence of residual force is not just of academic interest

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

0 10 20 30 40 50 60 70 80

LOAD

(kN

)

MOVEMENT (mm)

Toe("False")

Head

Load for 30 mmtoe movement

Shortening

Shaft("False")

OffsetLimit

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

0 10 20 30 40 50 60 70 80

LOAD

(kN

)

MOVEMENT (mm)

Head

Load for 30 mmtoe movementShortening

Shaft

Toe

OffsetLimit

Results from a test on a 15 m long, 600 mm diameter, jacked-inconcrete pile (Fellenius 2014). The manner of testing built-inconsiderable residual force in the pile.

Same pile assumed tested without residual force being present.(The test results were first fitted to a UniPile t-z/q-z analysis,whereafter the analysis was converted to results for an identicalpile and soil, but with no residual force present).

33

From Recent experiences with static pile load testing on real job sites andGeneral Report – Design methods based on static pile load tests.

From Proceedings of ETC3, Symposium on Design of Piles in EuropeLeuven, Belgium, April 2016

Load-distribution from a 508 mm diameterdriven cast-in-place pile installed to 15 m depth.

The authors mention various reasons for the lack of shaft resistance along the lower length, but do not address the main reason, which is thatthe piles have significant residual force, causing the shaft resistance along the upper length to appear too large and that along the lower lengthto appear too small or non-existent.

Figure 10 on Page 74

Load-distribution from a 620 mm diameter, 16.6 m long,screw pile installed to 15 m depth.

Figure 5a on Page 90

34

Proceedings of ETC3Symposium on Design of Piles in EuropeLeuven, Belgium, April 2016

Load-movement curves from a 560mm diameter "displacement screwpile" (CFA pile). Pile length and soiltype not mentioned.Residual toe force can beassumed small.

Load-movement curves from a 508mm diameter driven cast-in-place pile.Pile length and soil type notmentioned. (Not the same pile as thatused to show load distribution).Residual toe force is usuallyconsiderable for driven piles.

Quote from the Proceedings "Thedesign based on the results of pileload tests is based on the results ofthe ultimate or pile capacity at 10% *)

diameter pile base (toe) displacement".

Figure 3 on Page 68

*) Triple mentioning of "base" will not negate the factthat the "10-%" definition of capacity is applied to adisplacement that can be up to 100 % wrong!

Toe and shaftmovements are nowshown after a moderatecorrection for residualforce

Figure 2 on Page 67 Same as Figure 2 on Page 67 withpresumed "true" shaft and toe curves

35

Answer to the questionin the graph:

No, there's alwaysresidual force (axial)in a test pile.

Residual Force

Head-down loading test on a 600 mm diameter,54 m long cylinder pile in Busan, Korea, 2006.(Kim, Chung, and Fellenius, 2011)

3636

Gages were readafter they had beeninstalled in the pile( = “zero” condition)and then 9 dayslater (= green line)after the pile hadbeen concreted andmost of thehydration effecthad developed.

37

Strains measuredduring the followingadditional 209-daywait-period.

0

5

10

15

20

25

30

35

40

45

50

55

60

-300 -200 -100 0 100 200 300 400

STRAIN (µε)

DEPTH (m)

9d

15d

23d

30d

39d

49d

59d

82d

99d

122d

218dDay ofTest

At an E- modulus of 30 GPa,this strain change correspondsto a load change of 3,200 KN

The difficulty associated with wanting to know the pile-toe load-movement response, butonly knowing the pile-head load-movement response, is overcome in the bidirectionaltest, which incorporates one or more sacrificial hydraulic jacks placed at or near the toe(base) of the pile to be tested (be it a driven pile, augercast pile, drilled-shaft pile,precast pile, pipe pile, H-pile, or a barrette). Early bidirectional testing was performedby Gibson and Devenny (1973), Horvath et al. (1983), and Amir (1983). About thesame time, an independent development took place in Brazil (Elisio 1983; 1986), whichled to an industrial production offered commercially by Arcos Egenharia Ltda., Brazil, tothe piling industry. In the 1980s, Dr. Jorj Osterberg also saw the need for and use of atest employing a hydraulic jack arrangement placed at or near the pile toe (Osterberg1989) and established a US corporation called Loadtest Inc. to pursue the bi-directionaltechnique. On Dr. Osterberg's in 1988 learning about the existence and availability ofthe Brazilian device, initially, the US and Brazilian companies collaborated. Somewhatunmerited, outside Brazil, the bidirectional test is now called the “Osterberg Cell test” orthe “O-cell test” (Osterberg 1998). During the about 30 years of commercial application,Loadtest Inc. has developed a practice of strain-gage instrumentation in conjunctionwith the bidirectional test, which has vastly contributed to the knowledge and state-of-the-art of pile response to load.

The bi-directional test

38

The bi-directional testArcos Egenharia

Bidirectional test atRio Negro Ponte

Manaus—Iranduba, Brazil

39

4040

Schematics of the bidirectional test(Meyer and Schade 1995)

-10

-8

-6

-4

-2

0

2

4

6

8

10

0 10 20 30 40 50 60 70 80

MO

VEM

ENT

(mm

)

LOAD (tonne)

Test atBanco Europeu

Elisio (1983)

Pile PC-1June 3, 1981

Upward

Downward

11.0

m2.

0 mE

520 mm

Typical Test Results(Data from Eliso 1983)

4141Inchon, Korea (Fugro Loadtest)

42Sao Paolo, Brazil (Arcos Egenharia)

4343

The bidirectional cell can also be installed in a driven pile. Here in a 600 mm cylinder pile(spun pile) with a 400 mm central void (installed after the driving).

44

Test on a 1,250 mm diameter, 40 m long, bored pile at US82 Bridgeacross Mississippi River installed into dense sand

-80

-60

-40

-20

0

20

40

60

80

100

120

0 2,000 4,000 6,000 8,000 10,000

LOAD (KN)

MO

VEM

ENT

(mm

)

UPPER PLATEUPWARD MVMNT

LOWER PLATEDOWNWARD MVMNT

Weightof

Shaft

ResidualLoad

Shaft

Toe

ResidualForce

Bidirectional load-movement curves(The bidirectional jack is placed at the pile toe)

Bidirectional load-distribution as-measuredand as-“flipped” over to show the equivalenthead-down distribution. (The red curve is thecurve fitted in a UniPile calculation applyingeffective stress analysis with the target beta-coefficients indicated to the right).

Test data from Fugro Loadtest 2002

45

The Equivalent Head-down Load-movement CurveMeasured upward and downward curves

From the upward and downward results, one can producethe equivalent head-down load-movement curve, the curvethat one would have obtained in a routine “Head-DownTest”. The curve needs to be corrected for the increasedpile compression in the head-down test.

Construction of the “Direct Equivalent Curve”

Reference: Appendix to regular reports by Fugro Loadtest Inc.

It also needs to be corrected for the fact that the bidirectionaltest engages the stiffer soil layers first, while the head-downtest engages them last (addressed in the next two slides).

46

The bidirectional load-movement curves(blue lines and square dots) simulatedand fitted using UniPile (red lines).Unit shaft resistance vs. movement as

evaluated from strain-gage recordsand telltale-measured movements.

Fellenius (2015)

Once the simulation of the test is completed,the analysis can produce the equivalent head-down load-movement curves, applying the t-zand q-z functions resulting from the fit to themeasured bidirectional curves.

Cnt. Test on a 1,250 mm diameter, 40 m long, boredpile at US82 Bridge across Mississippi River

-80

-60

-40

-20

0

20

40

60

80

100

1200 2,000 4,000 6,000 8,000 10,000

MO

VEM

ENT

(mm

)

LOAD (kN)

UPWARD

DOWNWARD

Weight ofShaft

ResidualForce

Simulatedcurve

Simulatedcurve

47

0

2,000

4,000

6,000

8,000

10,000

0 25 50 75 100 125

LOAD

(kN

)

MOVEMENT (mm)

MeasuredUpward

( )

Upward TestMeasured andsimulated

As in aHead-downTest

The difference between the measured upward cell movement and the simulated Equivalent Head-down load-movement for the pile length above the bidirectional cell is due to the fact that theupward cell engages the lower soil first, whereas the head-down test engages the upper soilsfirst, which are less stiff than the lower soils. The left graph shows the measured and simulatedupward and downward curves (same as on previous slide) The right graph shows the upward curveas measured and as simulated (in a 1st quarter plot) and compared to an Equivalent Head-down test(shaft resistance only) on the pile using the t-z curves fitted to the measured test.

Shaft resistance only

-80

-60

-40

-20

0

20

40

60

80

100

1200 2,000 4,000 6,000 8,000 10,000

MO

VEM

ENT

(mm

)

LOAD (kN)

UPWARD

DOWNWARD

Weight ofShaft

ResidualForce

Simulatedcurve

Simulatedcurve

Analysis of the results of abidirectional test on a 21m longbored pile in Sao Paolo, Brazil

A bidirectional test was performed on a 500-mm diameter, 21 mlong, bored pile constructed through compact to dense sand bydriving a steel-pipe to full depth, cleaning out the pipe, whilekeeping the pipe filled with betonite slurry, withdrawing the pipe,and, finally, tremie-replacing the slurry with concrete. Thebidirectional cell (BDC) was attached to the reinforcing cageinserted into the fresh concrete. The BDC was placed at 15 mdepth below the ground surface.

compactSAND

CLAY

compactSAND

denseSAND

The sand becomes verydense at about 25 m depth

48

0

5

10

15

20

25

0 5 10 15 20 25

DEP

TH (

m)

Cone Stress, qt (MPa)

0

5

10

15

20

25

0 20 40 60 80 100D

EPTH

(m

)

Sleeve Friction, fs (kPa)

0

5

10

15

20

25

0 50 100 150 200 250

DEP

TH (

m)

Pore Pressure (kPa)

0

5

10

15

20

25

0.0 0.5 1.0 1.5 2.0

DEP

TH (

m)

Friction Ratio, fR (%)

0

5

10

15

20

25

0 10 20 30 40 50

DEPT

H (m

)

N (blows/0.3m)

compactSAND

CLAY

compactSAND

denseSAND

The soil profile determined by CPTU and SPT

0

5

10

15

20

25

0 5 10 15 20 25

DEP

TH (

m)

Cone Stress, qt (MPa)

0

5

10

15

20

25

0 20 40 60 80 100D

EPTH

(m

)

Sleeve Friction, fs (kPa)

0

5

10

15

20

25

0 50 100 150 200 250

DEP

TH (

m)

Pore Pressure (kPa)

0

5

10

15

20

25

0.0 0.5 1.0 1.5 2.0

DEP

TH (

m)

Friction Ratio, fR (%)

0

5

10

15

20

25

0 10 20 30 40 50

DEPT

H (m

)

N (blows/0.3m)

compactSAND

CLAY

compactSAND

denseSAND

49

The final fit of simulated curves to the measured

-50

-40

-30

-20

-10

0

10

20

30

40

50

0 200 400 600 800 1,000 1,200

MO

VEM

ENT

(mm

)

LOAD (kN)

Pile HeadUPWARD

BDCDOWNWARD

15.0

m6.

0 m

50

Simulation performed by effective stress back-calculationwith input of t-z and q-z curves to the UniPile software(see Slide 23)

0

20

40

60

80

100

120

140

0 5 10 15 20 25

SHAF

T SH

EAR

(%

of r

ult)

RELATIVE MOVEMENT BETWEEN PILE AND SOIL ELEMENT (mm)

Ratio

Hyperbolic (ru = 120 %)

Exponential

Hansen 80 %

r1 or ru

δULTZhang

0

5

10

15

20

25

0 200 400 600 800 1,000

DEP

TH (

m)

LOAD (kN)

BDC load

0

5

10

15

20

25

0 10 20 30 40 50

DE

PT

H (

m)

N (blows/0.3m)

compactSAND

CLAY

compactSAND

denseSAND

With waterforce

Lessbuoyantweight

UPWARD

DOWNWARD

The test pile was not instrumented. Had it been, the load distribution of the bidirectionaltest as determined from the gage records, would have served to further detail theevaluation results. Note the below adjustment of the BDC load for the buoyant weight(upward) of the pile and the added water force (downward).

The analysis and results of thesimulation appear to suggestthat the pile is affected by afilter cake along the shaft. Ithas probably also a reducedtoe resistance due to debrishaving collected at the pile toebetween final cleaning and theplacing of the concrete.

51

The final fit establishes the soil response and allows theequivalent head-down loading- test to be calculated

0

500

1,000

1,500

2,000

2,500

0 5 10 15 20 25 30 35 40 45 50

LOAD

(kN

)

MOVEMENT (mm)

EquivalentHead-Down test

HEAD

TOE

Pile head movement for30 mm pile toe

movement

Pile head movement for5 mm pile toe movement

When there is no obvious point on thepile-head load-movement curve, the“capacity” of the pile has to bedetermined by one definition or other—there are dozens of such around. Iprefer to define it as the pile-head loadthat resulted in a 30-mm pile toemovement. As to what safe workingload to assign to a test, it often fits quitewell to the pile head load that resulted ina 5-mm toe movement. In this case, thathappens to be at Q = 750 kN.

The most important aspect for a safedesign is not the “capacity” found fromthe test data, but what the settlement ofthe structure supported by the pile(s)might be. How to calculate thesettlement of a piled foundation is notaddressed here, however.

62

53

-15

-10

-5

0

5

10

15

20

25

0 100 200 300 400 500 600 700 800 900 1,000

MO

VEM

ENT

(mm

)

LOAD (kN)

E

8.5

3.0

If unloaded after 10 min

ß = 0.4 to 2.5 m δ = 5 mm ϴ = 0.24ß = 0.6 to 6.0 m δ = 5 mm ϴ = 0.24ß = 0.9 to cell δ = 5 mm ϴ = 0.24

ß = 0.9 to toe δ = 5 mm ϴ = 0.30rt = 3,000 kPa δ = 30 mm ϴ = 0.90

PCE-02

-30

-25

-20

-15

-10

-5

0

5

10

15

0 100 200 300 400 500 600 700 800 900 1,000M

OVE

MEN

T (m

m)

LOAD (kN)PCE-07

E

7.2

4.3

ß = 0.5 to 2.5 m δ = 5 mm ϴ = 0.30ß = 0.7 to 6.0 m δ = 5 mm ϴ = 0.30ß = 0.8 to cell δ = 5 mm ϴ = 0.30

ß = 0.6 to toe δ = 5 mm C1 = 0.0063rt = 100 kPa δ = 30 mm C1 = 0.0070

A CASE HISTORY Bidirectional tests performed at a site in Brazil on two Omega Piles(Drilled Displacement Piles, DDP, also called Full Displacement Piles, FDP) both with700 mm diameter and embedment 11.5 m. Pile PCE-02 was provided with abidirectional cell level at 7.3 m depth and Pile PCE-07 at 8.5 m depth.

54

0

500

1,000

1,500

2,000

2,500

3,000

0 5 10 15 20 25 30

LOAD

(kN

)

MOVEMENT (mm)

PCE-02

PCE-07

Max downwardmovement

Max upwardmovement

Max upwardmovement

Max downwardmovement

Pile PCE-02Compression

PCE-02, Toe

PCE-07, Toe

PCE-02 and -07Shaft

0

2

4

6

8

10

12

0 400 800 1,200 1,600

DEP

TH (

m)

LOAD (kN)

PCE-07

PCE-02

0

5

10

15

20

25

30

0 400 800 1,200 1,600

MO

VEM

ENT

(mm

)

TOE LOAD (kN)

PCE-02

PCE-07

A

B

Equivalent Head-downLoad-movements

Equivalent Head-downLoad-distributions

After data reduction and processing

A conventional head-down test would nothave provided the reason for the lower

“capacity” of Pile PCE-02

55(Bjerrum et al., 1969)

The ever so scary N.S.F ghost —the drag force

Profile of test site and piles. Heröya site.

56

Distribution of soil stress, excess pore pressure, pile shortening, andload distributions. Heröya site. (Data from Bjerrum et al., 1969).

0

5

10

15

20

25

30

35

0 100 200 300 400

EFFECTIVE STRESSAND PORE PRESSURE (KPa)

DEPT

H (m

)σ'z after full

dissipation ofexcess pore

pressure

σ'zΔu

MarineClay

FILL

Start ofBedrock

Gravel

0

5

10

15

20

25

30

35

0 300 600 900 1,200 1,500

PILE LOAD (KN)

DEPT

H (m

)

Bitumencoated

uncoated

Distributioncalculated from

ß=0.3 times σ'z foractual excess u

Measureddistribution

Notice the distinct ForceEquilibrium, the Neutral Plane

57

Compilation of Norwegian results

0

cm cm

DragForce

58

FHWA Project, Keehi Interchange,Honolulu,Hawaii 1977

59

Results from 3 years of monitoring of a 30-mpile, 50-m pile, and a 30-m bitumen coated pile

3/16 inch = 4 mm

COAT

60

Leung, C.F, Radhakrishnan, R., and Tan Siew Ann (1991) presented a case history on

instrumented 280 mm square precast concrete piles driven in marine clay in Singapore

Note, the distribution of negative skin friction is linear(down to the beginning of the transition zone)indicating the proportionality to the effectiveoverburden stress

TRANSITION

ZONE

61Data from Leung, Radhakrishnan, and Tan (1991)

0

5

10

15

20

25

30

0 200 400 600

LOAD (KN)

DEPT

H (m

)

OldSilt&

ClayFill

MarineClay

WeakShale

Bedrockand

ResidualSoil

Clay

ß = 0.5

Twomonths

after start(57 days)

Two yearslater

(745 days)

Variable loadi.e., Live Load

Live load anddrag forcecannot coexist!

6262

Endo et al. 1969,presented a very

ambitious drag-forcestudy in Japan on four

instrumented steel pilesduring a period of threeyears. The soils consistof silt and clay on sand.The case history is one ofthe few that actually also

measured settlement.

6363

Profile of test site and piles

Closed-toe, Open-toe, Inclined, and Short Pile

(Endo et al., 1969)

64

Load distribution in the three long piles together and settlement of soil andpiles measured March 1967 672 days after start. (Data from Endo et al., 1969).

Neutral plane = Force Equilibrium = Settlement Equilibrium

0

5

10

15

20

25

30

35

40

45

50

0 500 1,000 1,500 2,000 2,500 3,000 3,500

LOAD (KN)

DEPT

H (m

)

Closed-toePiles: Inclinedand Vertical

Open-toePile

SandySilt

Clay

Sand

Silt

( ) ?

0

5

10

15

20

25

30

35

40

45

50

0 50 100 150 200

SETTLEMENT (mm)

Soil

Pile

Closed-toe PilesToePenetration

- NEUTRAL PLANE -

65

Toe forces and toe penetrationsextracted from the graphs of Endo et

al.

0

500

1,000

1,500

2,000

0 10 20 30

TOE PENETRATION (mm)

TOE

LOA

D (

KN

)

From data inpaper

Approximation

66

Inoue, Y., Tamaoki, K., and Ogai, T.,1977. Settlement of building due topile downdrag. Proc. 9th ICSMFE,Tokyo, July 10-15, Vol. 1, pp. 561– 564.

A three-storey building with a foot print of 15 m by 100 m foundedon 500 mm diameter open-toe pipe piles driven through sand and siltyclay to bearing in a sand layer at about 35 m depth. The piles hadmore than adequate capacity to carry the building. Two years afterconstruction, the building was noticed to have settled some 150 mm.Measurements during the following two years showed about 200 mmadditional settlement. The building was demolished at that time.

A Downdrag Case

67Pile Toe Depth

Inoue 1977

68Settlement between piles in Row 6 and Row 10from Sep. 1967 through May 1969 = 150 mm.

Slope 1 : 100(Sep 67 Apr 71)

Inoue 1977

69Gue See Sew, 2011Gue & Partners Sdn Bhd

Singapore Apartments

Looks bad, but, note, thebuilding foundation is OK.No downdrag resultedbecause the neutral plane islocated in the competentnot settling soil. Of course,the pile s have drag forces,but that environmentaleffect is there be thesettlement small or large.Presence of drag force is ofno consequence forfoundations—it is alwayspresent, regardless.

70

Office Building placed on longtoe-bearing piles in Brisbane,Australia. A later built, lightextension was placed on shortwood piles.

Data Courtesy (2007) of Wagstaff Piling Pty. Ltd., Queensland,Australia.

71

View toward the roof

7272

The original piles had a capacityabout three times the applied load,but downdrag got the better ofthem. Luckily, new piles could beinstalled (driven to the sandstone;for some columns to a“four-for-one”ratio).

$ $ $ $

Foundations andunderpinning

Settlement Analysis of Large Pile Groupsby the Equivalent Raft Method

soilpile

soilsoilpilepilecombined AA

EAEAE

The compressibility in thiszone must be of soil and pilecombined

Equivalent Footingplaced at the Locationof the Neutral Plane

2:1 distribution2:1 distribution

G.W.

FILLS, etc.

Settlement of the piled foundation is caused by thecompression of the soil due to increase of effectivestress below the neutral plane from externalload applied to the piles and, for example, from fills,embankments, loads on adjacent foundations, andlowering of groundwater table.

5:1 5:1

For small groups, start byplacing an "Equivalent Raft"at the depth of the Neutral

Plane

N.B., the above approach goes far beyond the 1948 Terzaghi and Pecksuggestion of placing The Equivalent Raft at the lower third depth

73

-20

-15

-10

-5

0

5

10

15

20

-20 -15 -10 -5 0 5 10 15 20

East-West

Nor

ht-S

outh

Pile 11

Pile 16

Pile 7

112 1.0 m diameter, boredpiles installed to 42 m depth

Dense, silty sand to 50+ m depth

Liquid storage tank in Tessaloniki, Greece (Savvaidis 2009)

Settlement

NoSettlement

74

0

10

20

30

40

0 5 10 15 20 25 30 35 40

ACROSS THE DIAMETER (m)

SETT

LEM

ENT

(mm

)

Settlement measured across a tank diameter during a Hydro Test

Curve calculated for a flexible footinglocated at the pile toe level with parametersfitted to the settlement measured at thetank mid-point (calculations were performedwith UniSettle).

75Fellenius and Ochoa 2013; 2016

A quote from a textbook *)

The net effect negative skin friction is that the pile load capacity [sic!] isreduced and pile settlement increases. The allowable load capacity [sic!] isgiven as:

*) Compassion—perhaps misdirected—compels me not to identify the author

negS

negultallow Q

F

QQQ

It could have been worse. Logically,the drag force (Qneg) should have beenincreased by a factor of safety.

But so what! There is so much lackof logic in the approach, anyway.

76

Do not include the drag force when determining the allowable load!

Drag force must neither subtracted fromthe pile capacity nor from the allowableload

HAS INCREASED!

Effect of subtracting the dragforce

If the pile capacity had been reduced with the amountof the drag force before subtracting the drag forcefrom the “net capacity”, so determined, there wouldhave been no room left for working load!

0

5

10

15

20

0 500 1,000 1,500 2,000 2,500

DEP

TH (

m)

LOAD (kN)ALLOWABLELOAD - (Fs = 2.5)

CAPACITY

DRAG FORCE

0

5

10

15

20

0 500 1,000 1,500 2,000 2,500

DEP

TH (

m)

LOAD (kN)ALLOWABLE LOADminus DRAG FORCE

CAPACITY

DRAG FORCE

77

The Euro Code

Unfortunately and regrettably, the recently issued US AASHTO LRFD Specshave adopted the EuroCode approach! A few US State DOTs, e.g., Utah, havewisely rejected the AASHTO Specs and apply the Unified Method.

The European Community has recently completed EuroCode 7, which is supposedto be adopted by all member states. The EuroCode treats the drag force as aload (an “action”) similar to the load from the structure, and requires it to beadded to that load and also that it is subtracted from the pile capacity)!Moreover, the shaft resistance in the soil layer that contributes to the dragforce is disregarded when determining the pile resistance. That is, when acapacity has been determined to, say, 1,000 kN and the drag force is expectedto be, say, 200 kN (an unrealistically low value), the "usable capacity", i.e., theusable unfactored resistance, is a mere 800 kN. This value is then factored. Ifresistance factor is 0.5, the factored resistance is 400 kN. When, as required,the factored drag force is subtracted (applying a drag-force load-factor of,say, 1.5, the amount left to support the factored load from the structure is 100kN!

What “salvages” the economy of some designs must be that the EuroCodeclauses advocate that the designer maintains the faithful approach that “thedrag force cannot really be that large, can it, please?” in determining themagnitude of the drag force. Incredibly, the EuroCode says little on how tocalculate settlement of piled foundations and nothing is stated about downdrag!

78

5.0 m

SOFT CLAY

STIFF SILTY CLAY

11.5 m

FILL

Average unit shaft resistance, rs = 20 KPa

Rs = 94.2 KN; Rs = Qn

Average rs = 50 KPa

Rs = 543 KN

"The settlement due to the fill is sufficient to develop maximum negative skin friction in the soft clay ".

fq*300 + fn*94 ≤ 543/fr

1.35*300 + 1.35*94 ≤ 543/1.0

532 ≤ 543

(Alternative: If fr = 1.1, the lengthin the silty clay becomes 12.4 m)

Q (unfactored) = 300 KN

Eurocode Guide , Example 7.4 (Bored 0.3 m diameter pile)

Rt = 0 KN ?!

CALCULATIONS

The Guide states that the two rs-values are from effective stress calculation. The valuescorrelate to soil unit weights of 18 KN/m3 and 19.6 KN/m3, ß-coefficients of 0.4 in both layerswith groundwater table at ground surface, and a stress of 30 KPa from the fill. Note that theexample presupposes that the analysis is carried out for the long-term conditions.

The Guide states that the neutral plane lies at the interface of the two clay layers. Based on theinformation given in the example, this cannot be correct. But there is a good deal more wrong withthis "design" example.

79

If the settlement is acceptable, there may be room for shortening the pile or increasingthe load. That would raise the location of the neutral plane. Would then the pilesettlement still be acceptable? And, would not some toe resistance develop?

Results of analysis using the given numerical values(ß = 0.4 for both layers and zero toe resistance)

0

5

10

15

20

0 200 400 600 800 1000LOAD (KN)

DEPT

H (m

)

MaximumLoad = 440 KN

Rt = 0?

Fs = 2.1

Qn = 140 KN not 94 KN

0

5

10

15

20

SETTLEMENT (mm)

DEPT

H (m

)

ToeMovement

NeutralPlane

THE KEY QUESTION:is the settlement acceptable?

?

80

Let's assume that a static loading test has been carried out on the example pile and taken well beyond the the usualmaximum movement. (N.B., to assume no toe resistance is too conservative and a small value has therefore beenadded to the example). Toe resistance is a function of toe movement. Let's also assume a toe response per a q-zfunction representative for a stiff clay, and a strain softening shaft resistance. Then, the loading test results in thefollowing pile-head load-movement curves. The curve marked "initial" is from a test assumed carried out during thedesign phase. The curve marked "final" is assumed to have been carried after the consolidation (or most of it) causedby the fill has developed — we could have marked it "forensic", i.e., a test undertaken by the lawyers after the piledfoundation showed to settle excessively.

0

100

200

300

400

500

600

700

800

0 10 20 30 40 50 60

MOVEMENT (mm)

LOAD

(KN

)

Head final

Toe

PileShortening

RULT = 695 KN

Rt = 60 KN

Head initial

Designers do not usually design for the final conditions — it is commendable that they did for this example, but thedesign must also consider the initial conditions. The structure is probably built at the same time as the fill isplaced and before the consolidation process had strengthened the soil. I do not think everyone would becomfortable placing a sustained load of 300 KN on piles that plunge at 500 KN (test at initial condition).

81

0

5

10

15

20

25

30

0 10 20 30 40 50

DEPT

H (m

)

FILL

Silty SAND

SAND

CLAY TILL

LIMESTONE BEDROCK

N

wn

CLAY

m ≈ 100

mr ≈ 600∆σ' ≈ 9 MPa

N (bl/0.3m); qt (MPa); wP, wn, and wL (%);

m ≈ 20

mr ≈ 200∆σ' ≈ 20 KPaw LwP

qt

GW

Example of the Unified Design Approach as Applied to a Refinery Structure

Design for a large refineryexpansion was undertaken at asite reclaimed from a lake inthe 1960s. The natural soilsconsist of sand deposited onnormally consolidated, com-pressible post glacial lacustrineclay followed by silty clay till onlimestone bedrock found atabout 25 m to 30 m depth belowexisting grade. The site will beraised an additional 1.5 m,which will cause long-termsettlement. Piled foundationsare needed for all structures.

Fellenius and Ochoa (2009)

Soil Profile

82

Results of a bidirectional-cell test ona 575-mm diameter test pile, a 26 mdeep bored cylindrical pile. A 1.5 mthick fill will be placed over the siteafter construction. Piles are singleor in small groups.

Results of analysis of test data:Load Distributions

0

5

10

15

20

25

30

0 1,000 2,000 3,000 4,000 5,000

DEP

TH (m

)

LOAD (KN)

Bidirectional-cell Test

Strain-gageValue

Cell

"Flipped"

Long-term (af ter theconsolidation induced bythe Fill is completed).

SiltSand

Clay

Till

-60

-40

-20

0

20

40

60

80

0 500 1,000 1,500 2,000 2,500

MO

VEM

ENT

(mm

)

LOAD (kN)

9 %

UPWARD

DOWNWARD

0

100

200

3000 1,000 2,000 3,000 4,000 5,000

TOE

MO

VEM

ENT

(mm

)

LOAD (KN)

Bidirectional-cell Testq-zextrapolation

83

0

5

10

15

20

25

30

0 2,000 4,000 6,000LOAD (KN)

DEPT

H (m

)O-cell

SiltSand

Clay

Till

Load Distribution ImmediatelyAfter Construction

Qd

Distribution of residual force inthe pile after installation, butbefore load is applied to the pile.

Distribution of load in the pileimmediately after the pile starts tosustain the load from the structure.

Other than at the pile toe, the amount of residualforce and its distribution are not that interesting 84

0

5

10

15

20

25

30

0 2,000 4,000 6,000LOAD (KN)

DEPT

H (m

)

O-cell

SiltSand

Clay

Till

Residual Load DistributionBefore ConstructionResidual Force Distributionbefore Construction

Long-term loaddistribution

The shaft shear is assumed to befully mobilized. However, thetoe resistance value to use is afunction of the toe penetrationdue to downdrag and can only bedetermined from assessing thesoil settlement distribution.

0

5

10

15

20

25

30

0 2,000 4,000 6,000

LOAD (KN)DE

PTH

(m)

O-cell

SiltSand

Clay

Till

Qd

Long-term Load Distribution

Increase oftoe load

85

0

5

10

15

20

25

30

0 2,000 4,000 6,000LOAD (KN)

DEP

TH (m

)

O-cell

SiltSand

Clay

Till

Qd

Pile toe load in the load distribution diagram mustmatch the toe load induced by the toe movement(penetration), which match is achieved by a trial-and-error procedure.

0

5

10

15

20

25

30

0 50 100 150 200

SETTLEMENT (mm)

DEP

TH (m

)

0

5

10

15

20

25

30

0 2,000 4,000 6,000LOAD (KN)

DEP

TH (m

)

O-cell

SiltSand

Clay

Till

0

1,000

2,000

3,000

4,0000 50 100

TOE

LOAD

(KN

)

Qd

Pile toe load in the load distribution diagram mustmatch the toe load induced by the toe movement(penetration), which match is achieved by a trial-and-error procedure. PILE TOE PENETRATION (mm)

Pile Cap Settlement

Soil Settlement

q-z relation

Force and settlement (downdrag) interactive design.The unified pile design for capacity, drag force, settlement, and downdrag

86

0

5

10

15

20

25

30

0 50 100 150 200

SETTLEMENT (mm)

DEP

TH (m

)

0

5

10

15

20

25

30

0 2,000 4,000 6,000LOAD (KN)

DEP

TH (m

)

O-cell

SiltSand

Clay

Till

0

1,000

2,000

3,000

4,0000 50 100

TOE

LOAD

(KN

)

Qd

Pile toe load in the load distribution diagram mustmatch the toe load induced by the toe movement(penetration), which match is achieved by a trial-and-error procedure. PILE TOE PENETRATION (mm)

Pile Cap Settlement

Soil Settlement

q-z relation

Force and settlement (downdrag) interactive design.The unified pile design for capacity, drag force, settlement, and downdrag

87

The Unified Piled Foundation DesignForce and settlement interactive design for

capacity, drag force, settlement, and downdrag

The design is based on three "knowns": The shaft resistance distribution,the toe load-movement response, and the overall settlement distribution.

0

5

10

15

20

25

30

0 50 100 150 200

DEP

TH (m

)

SETTLEMENT (mm)

0

5

10

15

20

25

30

0 2,000 4,000 6,000

DEP

TH (m

)

LOAD (kN)

SiltSand

Clay

Till

01,0002,0003,0004,000

0 50 100LO

AD (

kN)

Qd

PILE TOE PENETRATION (mm)

Pile Cap Settlement

Soil Settlement

q-z relation

0

1,000

2,000

3,000

4,0000 1,000 2,000 3,000 4,000 5,000

LOAD

(kN

)

LOAD (kN)

Qn

≈RULT

88

A recent modern application of a piled pad foundation is the foundations for theRion-Antirion bridge piers (Pecker 2004). Another is the foundations of thepiers supporting the Golden Ears Bridge in Vancouver, BC (Sampaco et al.,Naesgaard et al. 2012), pictured below.

89

Golden Ears Bridge in Vancouver, BC.

90

Piled pad foundation piers supporting the Golden Ears Bridge in Vancouver, BC.

Bored piles:(900 mm; 8 m) to providelateral resistanceandDriven piles:(300 mm; 30 m) piled-padpiles over an about 100 mthick deposit of softcompressible clay

BRIDGE DECK

FOOTING AND PILE CAP

Longslenderpiles

Shortboredpile

Pad

91

92

ShinHo and MyeongJi Housing Project,in the estuary of the Nakdong River, Pusan, Korea

Project Managers: Drs. Song Gyo Chung and SungRyul Kim, Dong-A University, Busan

93

94

95

96

AIR VIEW(Shinho Site)

SITE PLAN (SH Site)

Silty clay

97

CPTU sounding atthe location of theShin-Ho test pile

98

The pile alternative investigated was a 600 mm diametercylinder pile with a 100 mm wall driven closed-toe

The questions to resolve in the design were

1. What is the capacity in the different layers?

2. What is the depth to the force equilibrium/settlement equilibrium, i.e., theneutral plane

3. What will be the maximum load in the pile? Is the structural strengthadequate?

4. What is the settlement of the pile as a function of the location of theneutral plane?

99

The field tests were designed to answer the questions. The foundationdesign was per the Unified Design Method. Total savings were about us$300million over the alternative of steel pipe piles (which included the drag forceas an active load).

100

Thank you for your attention