-F

iN~~--'

FIELD STUDY OF AN ANCHORED SHEET PILE BULKHEAD

by.

1H. Y. Fang

Ladislav Lamboj2

Thomas D. Dismuke3

1Associate Professor of Civil Engineering and Director, GeotechnicalEngineering Division) Fritz Engineering Laboratory, Lehigh University.

2Research Assistant, Fritz Engineering Laborato~y, Lehigh University.

3Consulting Engineer, Bethlehem Steel Corporation, Bethlehem, Pennsyl­vania.

-i

ABSTRACT

This paper presents the field measuring technique and instru­

mentation f~~ measuring the amount of shear transfer mobilized across .

the interlock at the neutral axis of steel arch web sheet piling~ Com-

parisons of tie rod force and positive maximum bending moment by various

existing m~thods together with field measured data are presented.

It is concluded, within the range of applied loads and soil

characteristics, that shear transfer takes place across the interlocks.

Thus, it is bel~eved that the European practice. of assuming that piles

-act as a unit more closely approximates the given field conditions than

the ~erican practice of assuming individual piling action. However,

since American· practice is based on use of sheet piling in all types of

soil, the designer cannot assume that shear transfer always takes place

when interlock crimping or welding is not used. For comparisons of tie

.rod force, it is indicated that there was a trend for the tie rod force.

to increase as the wall height increases. ~owever, the field results are

much smaller than theoretical results. No definite relation could be

found between theoretical and field results for the maximum moment. It

is thought that much of the scatter may be attributed tq composite action

between soil and the piling. Further investigation of the soil-structure

interaction is necessary in order to more clearly understand the phenomenon.

.-1

INTRODUCTION

The section modulus of a structural member is a measure of its

ability to resist bending. It is calculated by dividing the area moment

of inertia of the member about its axis of bending (the neutral axis) by

the distance' from that axis to the outermost fiber of the member cross~

section. The centroidal axis of an individual arch web' sheet piling sec­

tion is located between the axis of the interlocks "and the web, (Fig.·l).

American design practice is to use this centroidal axis as,the neutral

axis for evaluating moment resistance. The location of the centroidal

axis of a wall composed of several interlocked sheet piling sections is

along the line of interlocks (Fig. 1). European design practice is to

use this axis, or an intermediate position, as the neutral axis to eval­

uate the moment resistance. As is evident from Fig. 1, the resistance

of an individual section is about one half the resistance of the composite

group. Consequently, de&igns based on the European method are more eco­

nomical than those based on the American method, when using arch web piling.

A review of literature concerning sheet pile structures were

made by Fang and Dismuke· (1973)." The reviewp indicated that,·

no information related to the amount of shear transfer across the interlock

and the location of the neutral axis of be~ding for a sheet pile wall were

available.

The objective of" this study was· to evaluate the behavior of

interlocked steel sheet piling in an actual field installation. More

specifically) the strain distribution across a sheet pile was measured

-2

in order to experimentally locate the neutral axis of bendi~g of the wall.

This paper presents a detailed description of the measuring technique

and instrumentation used to study shear transfer in a sheet pile wall.. . ~

In addition) the comparisons of tie rod forces and positive maximum

bending moment by various existing methods together with field data are

also presented •. The existing methods include Blum (1931)) free-earth

(Anderson,1956), Tschebotarioff (1951), and Rowe (1952). The results are

summarized in graphical form; discussion of test results and further in-

vestigation needed are also included.

TEST SITE, INSTRUMENTATION AND MEASURING TECHNIQUES

Test Site

The test site was located at Martins Creek, Pennsylvania.

The soil profile was determined from the results of wash borings

and from information supplied by the Pennsylvania Power and Light Company.

Boring logs recorded the surface conditions, the str~ta changes and thick-

ness, the standard penetration values for the soils, and the elevation of

the groundwater table. Figures 2 and 5 show the results of this soil pro-

file investigation. In general, the test site consisted of a thin layer

of sand and silt underlain by a thicker layer of sand, gravel, and boulders.

Instrumentation

The foil-type SR-4 strain gage was chosen as the primary means

-3

of measuring strain in the piling. As strain gages must be protected

during and after pile driving operations, an evaluation of the strain

gage system was undertak~n in the laboratory. Methods for attaching,,~

waterproofing, and protecting the gage were studied. In addition, the

behavior of the gage was observed under simulated field conditions.

In the laborat~ry, a gage was mounted on the outer portion of

a 2" x 2" angle to simulate the actual mounting of a gage on the sheet

piling and driven into soil. A protec~ive epoxy covering was placed

over the gage, but no attempt was made to keep the covering from touching

or adhering to the gage.

The soil was composed of equal amoullts of coarse sand, obtained

directly from the test site, and 3/4" to 1" crushed stone. It was believed

that' the laboratory soil would be more abrasive to the gage, and its covering,

than the soil at the field site. The soil mixture was placed in a 2'

diameter cylinder of 4' height. A drain spout was t~pped into the bottom·

and a manometer was attached so that the level of the water table could be

controlled and measured. The sand and stone mixture was soaked j vibrated,

and allowed to drain. This produced a very compact mixture for the test.

The laboratory testing of the foil-type· strain gage proved that

the gage could be successfully protected against abrasion under controlled

laboratory conditions, and no trouble would arise as long as the epoxy

covering remained unbroken and bonded to the steel. Therefore, care was

-4

taken to properly bond the epoxy to the steel in all subsequent gage

installations •

. After it was shown that the gages could be protected under

controlled laboratory conditions, it was decided to test them in the

field at a nearby construction project.

Two gages were attached to a sheet piling :in the· field and

driven 20' into a loose, silty sand having a high groundwater table.

Both gages were protected by the epoxy covering. One gage was given

additional protection by covering it with a ste~l shoe (Fig. 3) that was

welded to the sheet pile (Fig. 4). During and after driving, both gages

performed satisfactorily, thereby substantiating the results of the la­

boratory' evaluation.

The field testing of the foil-type strain gage showed that the

gage could successfully withstand pile driving forces and the existence

of a high groundwater table. It was decided to protect all gages with

the steel shoe in order to offer prbtection agai~st ~tones that

might be encountered at the test site.

The strain gages were installed on the sheet piling at Fritz

Laboratory prior to delivering the piling to the test site. The piles

were cleaned with high speed grinders to obtain a smooth surface for the

gages. The ribbon wire was laid flat and clamped before the gage epoxy

was applied. After the wires were in place, the gages were attached and

-5

clamped while the epoxy was setting. Each gage was checked after in­

stallation to insure adhesion of the gage to the piling. This was accom­

plished by a "light bulb test" (D.al1y and 'Riley, 1965).

Because of the delicate nature of foil gages, it was necessary

to use low temperature solder to install the wires. Terminal tabs were

used to allow some play in the wires should they be accidentally pulled.

After the wires were installed, gage readings were taken and the p~otective

epoxy covering was applied (Fig. 4).

The strain gages were connected to the 'arch -piles near the in­

terlocks in order to evaluate the shear transfer across the joints. This

information would, in turn, lead to the determination of the location of

the neutral axis of bending for the sheet pile wall. The layout of the

strain gages on the instrumented piling (piling nos. 10, 11, 12, and 13)

is shown in Figs. 6 and 10.

After the gages were installed on the piling, they were "zeroed"

at Fritz Laboratoty with the piling hanging in a vertical position. The

initial readings recorded in the laboratory were checked at random at the

site before driving was started on each pile. The comparison between the

-random checks and the laboratory reading was satisfactory.

Field Test Procedure

The length of the sheet pile wall was 30 feet which was believed.

to be '~ong enough to minimize undesirable end effects. The piling was 30

feet long and they were driven 25 feet below ground level. Consequently,

--6

once the arch piles ~vere in place, approxima~ely 5 feet of pile protruded

from the ground surface. A standard driving rig with a low energy double

action 9B3 steam hammer was used for the driving operations. A guide~

frame was used to insure plumbness of the wall. A transit and a six-foot

level were used to aid in positioning the piles. The wall was anchored

by two tie rods which were held back by H-piles driven vertically 20 feet

into the ground. The tie rods were attached to the wall by means of a

wale welded to the wall at ground level. The wale and tie rods were 10-"

cated at ground level to facilitate. instrumentation :and test .. procedures..

The entire test set-up is illustrated in Figs. 5, 6,.'~7 sand 8.

For all piling, the measured out-of-plumbness during the driving

never exceeded 1 in. in 30 feet in the X-Y plane (Fig. 5). The deviation

from the vertical in the Y-Z plane, however, did increase as the wall was

driven (Fig. 6). This attributed to the tendency 'of sheet piling to close

on themselves during drivJng. The maximum deviation in the Y-Z plane was

5 in. in 30 feet. It should be noted that piles No. 10 and No. 13 met

refusal and could not be driven to the required depth. For this reason

they protrude 15 in. above the other piles (Figs. 6 and 8).

The soil was excavated in front of the wall in four stages.

Each excavation was 5 feet deep as shown in Figs. 5 and 6. The readings

were taken, on all strain gages after the excavation and again one week

later, just prior to the next excavation sequence. Excavation time for·

each stage was one day. This sequence of events was repeated until the

excavation reached the 20 ft. level. The plan view excavation was approx-

imately 20' x 35' as shown in Figs. 7 and 9. In addition, wall deflections

-7

(measured by transit) were recorded for several wall locations. An

initial load of 2000 Ibs. was applied to each of the tie rods. De­

tailed field data is summarized and given by Brewer and Fang (1968).

DISCUSSION OF RESULTS

Figure 11 shows the pertinent date for shear transfer in

graphical form. The distribution of vertical strain across the sheet

pile interlocks is shown for all gage levels, and at all stages of ex­

cavation.

Although there is considerable scatter in the data and a num-·

ber of the gages failed completely (see Fig. 10), the graphs suggest that

shear stress transfer across the interlocks between piles does occur. This

is apparent because the vertical strain distribution across the interlocks

may be reasonably approximated by a single continuous straight line atall stages of ~xcavation at which there is sufficient data.· If there was

only partial or no shear stress transfer across the interlocks, the ver­

tical strain distribution across the piles would be shown by two discon­

tinuous straight lines.

Figure '11 shows the vertical strain distribution interlock of

joint 1, piles 10 and 11. Other data on piles 11, 12, and 13 are given

by Brewer and Fang (1968, 1970).

It is pf interest to note from Fig. 11 that the location of

the neutral axis of bending for the sheet pile wall, given by the inter-

-8

section of the vertical strain curve with the line of zero strain, does

not 'al~ays lie within the pile cross-section. For discussion purposes,

consider the behavior of piling Nos. 10 and 11 at gage level Gl (Fig. 11).t

It can be seen that just prior to excavation, the neutral axis lies com-

pletely outside the piling cross-section toward the fill side of the wall.

This could indicate that the piling is in tension' due· to bend~ng induced

during driving, and the compressive bending stresses are carried by the

soil behind the piles. Such b,ehavior may be considered comp~site action,

with the wall and-the soil, acting as a unit, however, there is no further

substantiation of this.

Comparisons of tie rod forces versus wall height is shown in

Fig. 12. There is a trend that tie rod force increases as wall height

increases. However, the field results are much smaller than the theoret~

ical results and are less sensitive. This may be the result of the soil

strain conditions. It should be noted that the walls of the excavation.did not have to be shored to stand vertically for the full depth of the

excavation.

Comparison of the positive maximum bending moment is, carried ,out

in a graphical form in Figs. 13 and 14. Bending moment computations are

based on strain records. Sheet piles are considered as single acting units.

Numerical values of strain were obtained from reading of strain gages which

were located near interlocks. It is shown that there is no definite relation

between theoretical and experimental results. Possibly, much of the scatter

was attributable to composite action between soil and piling.

-9

Other comparisons include theoretical comparisons of tie rod

forces, embedment depth, and maximum moment with various existing methods

and internal friction angles are given by Lamboj and Fang (1970) togethert

with computation procedures.

SUMMARY AND CONCLUSIONS

The analysis of field study of an anchored sheet piling bulk-

head can be summarized as follows and conclusions presented considered

applicable within the limitations of the assumptions used:

1. Strain gage instrumentation installed on sheet piling prior

to driving may be successfully protected against damage during driving.

2. Within the range of applied loads and soil conditions en-

countered in this investigation, the available data suggests that shear

transfer takes place across the interlocks of arch web steel sheet piles.

3. Due to scatter in moment values evaluated from field measure-

menta it is believed that composite action between soil and the piling

occurs, however) further investigation of the soil-structure' interaction

is necessary in order to more clearly understand this phenomenon.

ACKNOWLEDGEMENTS

This paper is a part of the general investigation on the

tlBehavior of Soil-Pile Systems" currently being carried out at Fritz

Engineering Laboratory, Lehigh University. The investigation is sponsored

and financed by the American Iron and Steel Institute

-10

The authors express their thanks to Mr. R. J. Carle; Beth­

lehem Steel Corporation and Mr. H. A. Lindahl, U. S. Steel Corporation

for their assisatnce in every phase of the investigation.

Thanks are also extended to Mr. C. Ee Brewer who conducted

the field test and Dr. T. J. Hirst who reviewed the manuscript.

-11

. N.A.

li.!~_._

U. S. PRACTICE

EUROPEAN PRACTICE

N.A.

I

Fig. 1 Location of Neutral Axis (U.-8. and European ,Practibe)

-Elev.235 i-Etev.235''<//1''"../j Y///',

---225 t225'

GWT v ~\\\\1--" 2]5' GWT v 2151

---"205' - 205 1

i95' 1951

185' ISS'

175' 175'J70 1

170'I-P~ 300' I-N

Section A-A .....: I

Plan View of Boring Locations

'Legend

~ Silt.a Sand

~ Sand, Gravel~a Boulders

I Soft Lime Stone

~~ Sand a Gravei

IBLimestone a- - t Soft Limestone

-Elev.235t

--225'

" 2151

--2051

112Section c-c

GV/T v

tIlj

J-N

~ of Bulkhead

1-0

cr>8

I

',r-Q

,A:=I

f II c ,Al41-P ~'#2

L--s

Data for Holes #l-N, '-0" ~-p, J.. Q Supplied by the

Pennsylvania Povler and laight Co.

Fig. 2 Soil Profiles and Location Diagrami

}-JN

Fig. 3 Steel Protective Shoe

Fig. 4 A Steel Sheet Pile being Equippedwith strain Rosettes

-14

Surface Temp.Gage

H Pile201

Long~

-0C\J

Anchored Weier

30·

Lx

Sheet Piling Wall~JII~rain· Gage ~Loading DevicesDatum

StrainRosettes

Soil Profile Determined During. Excavation

Coarse Sand a Boulders

Coarse SandSand a SiltCoarse Sand

Sand a Silt

""",,,,\).

.. " ,.o .. ,: 0 '

, '" ..~~~~m~~~l

5

25

& ... \) ".:.-

o ... -... ; : tl· 0··.. '\).

~'.>~~J--=-=- Ground Water Table \!ll...~ ..

:;::?'-·0"......... 0

LU··: .... ,·, ~ 10 • 0 .

30 · : ~::

,...:;:0...CD

°20

.....(1)

If 15

~~=:;:==:::::~=b£======-; 11vJt,'fir~;;~w'lj=i-"/TiiII~q,'litl~=/,W'j1Elev. 235 ~ ~ --l

Ground Surface~ r "i"'/lI,;.m"IJ/::>/II"I/i f tTie Rod~ AnchorageDrs" <'i 1?" I.""" ."" -" "­,,, ""'" "" ",'-

" ",,, "" "­'" ", ",,, "IOr-R>;\~·~.:7Ji

})Hf.~::~

Fig. 5 Cross-Section of Sheet Pile Wall If-1Vi

•.~~~:~~~~~ ---- ~dllit1K£:....:_ Qt:::t.::::e::&:f.>QZ~i.~::..:..:..:..;;~~_.;.~~~.'::...-

\.

-16

Excavation.

I

GroundSurface7Elev. / 235 t

- - //I;:;tll/;;,! ;E:.l1/G II~

oDODD

o

4/=10 ifF II :il:12 ~i=13

Wales

I» 23 PDA27 , Arch Web Sheet Piling, Nominal Width 16" Each

4 Sheets With

I Gages I Anch Vveb Sheet--.J~-I~-\.-J--3-- ,Piling 301 L<?ng

J"2

lO

10-

0 0DO 00

-00+

0 0-- - DO DO

0 0 -C\J

00 00-r<)

J-I J-2 J-3

o Locations Of Strain Rosettes (2 rosettes at each location)

Fig. 6 Front of Sheet Pile Wall Showing Location of StrainGages

-17

30·1~-----{---------1--------~

Loading Devices

(-rLlrnbucl{les)

---Left Tie Rod

~ Right Tie Rod

H- Piles

Fig. 7 Plan View of Sheet pile Wall

-18

Fig. 8 Photograph of,'- Field -Te.st . in P'rogressTL~ Rods, W~~e, ~and Strain IndicatingEquipment shown.

-19

· Fig. 9 Photogra~h of Excavation at the 10' Level

Wale

Ground ~Level

Level Gr

Level G2 -

Level G3

Level G4

~··T----l I --

I I-- --~- -~-------- -- --- .--. --........ .......... - ......-.-. ...... -...--.---..

~-~- --j- C\J N - to "~ C\l - to ~

:t ~ ~ ~ 3: 3: ~ :: 3: ~~ ~

0 0 0 0 0 0 0 0~

a 0a:: 0:: n:: 0:: 0::: 0: 0:: a:: 0: cr:

1971 ~4 34 19 43 48 24 f9 43 4B 24 1971 I' /1 f' ?I l' "?1 f' -(1 l\

20 2f 22 23 2322 2( 20 4445 46 47 2322 21 20 4445 46 47 23 22 21 20

13 18 18 13 . 37 42 18 13 37 42 f8 1371 1'\ 74 f' 71 f' C\J ~ t' Jt; 1" 7j f'

14 15 lSl7 - i7'6 15 14 3839 4041 17 16 15 14 38 5940 41 oro 17 16 15 14-:> -:> -,

.... ~ l-

I z ~ z6 0 0J ? ~

~ ~~

7"/1 12 12 ~7 31 36 12 7 31 36 12 7K 7i ?1 f' ~ ~8 ?I K Jl7f ~

8 9 10 'f 11 10 9 8 3233 34 35 II 10 9 32 33 34 35 "10 9 8

17} 1"6 6 f'1 25 30 6 I 125 30 6 15'7I ?l f' 5'1 f\\2 ~ ft 71 f22 :5 4 5 4 3 2 26 27 28 29 4 3 ~6 27 28 29 5 4

LC)

\0

to

Col

(\J

f'()

I

o.1oJ"{)

'''F~

Pile #'0

0:,"~~

Pile if/: I"'

~~~~

~~~~~

Fig ~ 10

"-=-"~

~"*­~

Pile #12"

~~"J--

T..tocaticl1 (yf Strain Gages

4-~~

• Denotes Gages ThatFailed To Balance

Pile *13

~~

,~"*'

I~o

oo--n-

oo

PILE II

~~.10 \~~l~rGAGEG~f)JEL~

II

-20 ·15 -10·5 a 5 10 15 20-20·15 ·10 .. 5 0 5 10 15 20-20 ·15-10 -5 0 5 10 15 20

STRA IN JL IN.lIN.

COMPRESS10N- +TENSION

I I I

I I I

'" I I

-20 -15 ·10 -5 a 5 10 15 20-20-15 -10-5 0 5 10 15 20I

II

GAGE LEVELG4

DEPTH OF EXCAVATfCN- DRIVEN 5 ft. 10ft. 15ft. ZQft.

Pl-otted Strains - Strain Gagli! Reading - Laboratory Hung Zero (Vertical -Gage Only)

Fj.g. 11 Vertical Strain Distribution Interlock (Joint 1, Piles. 10 and 11)

IN~

,........

t. ~1II3eo

+->4-i

~.'3

110pcf'-'" ym~

E-t .1' ~~

J::~"'O<IJCJ T .- 2.H '(K/FT.)

4,... '.J..."i §:"

0 <.¢'~

~

'"d0~

OJ ~d Mea.u",menl'"rt :c...¢- __ ---o-----o

E-1

0 1O 20

H (FT.l

Wall Height above Dredge Line, H. ft.

-22

Fig. 12 Comparisons of Tie Rod Force for ~ = 38°

-20

40

- 23

~--_. Free Earth50 --- ----- Rowe's Method

.--. --:... - Anderson's Method------ -.-a Blum's Method·•. ~ ..•.......... Tschebotarioffls Method

~ Field Measurement.s

30

-10

i~ comparison of Theoretical Moment Curves withField Measurements for a wall 20 ft. high

Fig.

M( KIP-FI)

••

•

-24

Free EarthRowe's Method

Anderson's MethodBlum's MethodTschebotarioff's MethodField Measurements

....... ......

10

-to

-20

Figo '14 Comp~rison' of Theoreti9al Moment Curves withField Measurements for a' wall ·15 ft. high

M(~< IP-FT:) Ol;,·~--0---------'"1b---/-+-J-15---

2.....a-10--;;liiir.....

.,;';:r/' H (FT.)

{o

I

REFERENCES

Anderson, P. (1965). Substructure Analysis and Design, RonaldPress Company, New York.

Blum, H. (1931). Einspannungsverhaltnisse Bei Bohlwerken, WilhelmErnst and Sohn, Berlin.

Brewer, C. E. and H. Y. Fang (1968). Field Study of Shear Transferin Steel Sheet Pile Bulkhead, Fritz Engineering LaboratoryReport No. 342.2, Lehigh University.

Dally, J. W. and W. F. Riley (1965). Experimental Stress Analysis,McGra"tv Hill, p. 406.

Fang, H. Y. and C. E. Brewer (1968). Field Study of Sheet-PileBulkhead, ASCE Annual Meeting on Structural Engineering,Pittsburgh, Pennsylvania, September.

Fang, H. Y. and T. D. Dismuke (1973). Bibliography on Steel SheetPile Structures, Envo Publishing Co.

Lamboj, L. and H. Y. Fang (1970). Comparison of Ma~imum Moment,Tie Rod Force and Embedment Depth of Anchored Sheet Pile,Fritz Engineering Laboratory Report No. 365.1, Lehigh Uni­versity.

Rowe, P. W. (1952). Anchored Sheet Pile Walls, Proe. Institutionof Civil Engineers, Vol. 1, No~ 1, London, January, pp. 27-70.

Tschebotarioff, G. P. (1951). Soil Mechanics, Foundations andEarth Structures, McGraw-Hill Book Company, N. Y. Chapters10 and 16.

-25