Missouri University of Science and Technology

Scholars' Mine

International Conference on Case Histories in (1984) - First International Conference on Case

Geotechnical Engineering Histories in Geotechnical Engineering

08 May 1984, 8:00 am - 10:00 am

Performance of a 30 m Deep Instrumented Diaphragm Wall

M. M. Soares
Federal University of Rio de Janeiro, Brazil

Follow this and additional works at: https://scholarsmine.mst.edu/icchge

Part of the Geotechnical Engineering Commons

Recommended Citation
Soares, M. M., "Performance of a 30 m Deep Instrumented Diaphragm Wall" (1984). International
Conference on Case Histories in Geotechnical Engineering. 50.
https://scholarsmine.mst.edu/icchge/1icchge/1icchge-theme2/50

This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

This Article - Conference proceedings is brought to you for free and open access by Scholars' Mine. It has been
accepted for inclusion in International Conference on Case Histories in Geotechnical Engineering by an authorized
administrator of Scholars' Mine. This work is protected by U. S. Copyright Law. Unauthorized use including
reproduction for redistribution requires the permission of the copyright holder. For more information, please
contact scholarsmine@mst.edu.
Performance of a 30m Deep Instrumented Diaphragm Wall
M. M. Soares
Associate Professor, Federal University of Rio de Janeiro, Brazil

SYNOPSIS This paper summarises the performance of the second instrumented research section of the
Rio de Janeiro underground. The first aim of this programme was to establish a simple procedure
to design diaphragm walls in soft clay soils. The model of a beam on elastic supports was chosen
due to its simplicity and usefulness in the prediction of the wall behaviour. The concrete wall
was built using the slurry trench technique and was instrumented to measure the concrete strain, thP.
load on the reinforcement bars, the horizontal displacement and the horizontal earth pressure.
It was measured the loads in the two level of struts. The geotechnical profile consists of a soft
layer between two layers of sand. The surrounding soil received instruments to measure the pore
water pressure, the horizontal and vertical displacements at different locations. This paper
presents the comparison between predicted and measured values during the second stage of excavation.

SPT PARTICLE SIZE

INTRODUCTION BOREHOLE A533 w 5u lkNtnr:) llkN/m1} OlSTRIBUTION (•t.J
30 SO 70 0 30 60 100 0 .05 0.1 15 17.5 20 0 ltD BO
The Civil Engineering Department of the
Coordination of the Post-graduation Programmes
in Engineering (COPPE) of the Federal University
of Rio de Janeiro (UFRJ) instrumented and
observed the behaviour of the Rio de Janeiro ...-4 A A

underground construction in nine places. The

instrumentation was undertaken in the majority 8:: A A

of the places to reduce risks of damage to the

surrounding buildings. Two test sections were
.
6
..,.
~ "'AA
instrumented to furnish data to establish a
,._L,
simple procedure for designing diaphragm walls
in clay soils, as well as to improve other ,L..
methods of design and develop instrumentation
techniques. This paper summarizes some of the
field measurements made at the second
instrumented section and compares the predicted
and measured performance during the second .., UU TRIAXIAL TESTS
-CLAY
stage of the excavation works. More details 0 SILT
about the performance of this test section can ~SAND

be found in the DSc thesis of the author

(Soares, 1981). Fig. 1. SOIL PROPERTIES SUMMARY FIRST PROGRAM OF
INVESTIGATION

SOIL CONDITIONS The 120 em thick, 30m deep concrete wall

supported at two strut levels was designed by
Knowledge of the soil properties was obtained applying the standard design procedures used for
in two distinct programmes. The first by the Rio de Janeiro Underground Company. These
investigation programme was carried out to procedures were based on the free earth support
design the wall and the second to give and fixed earth support methods. At the same
additional information about the soil parameters time it was decided to instrument a typical
thus permitting more realistic analysis of the section to give more data to establish an
behaviour of the instrumented test section. improved procedure applicable to the design of
The first programme consisted of borings with concrete walls in soft clay soils.
standard penetration tests every 0.5 m and
laboratory index tests on disturbed samples. The second programme of soil investigation with
Conventional compressibility and shear strength two phases was started at the same time as the
laboratory tests were carried out on 50 and installation of the instruments and excavation
100 rom shelby samples. Figure 1 shows some in the area. The first phase of this programme
results from this programme. The most important consisted of:
layer for the design is the clay layer 15m
deep with SPT about 2 and cone point resistance • boreholes for installation of the
60 kN/m 2 • On that occasion the Designer instruments and soil sampling with some
decided to adopt for safety and undrained shear standard penetration tests
strength equal to 30 kN/m 2 •

405
 TABLE I
• conventional laboratory tests on
undisturbed soil samples (shelby and Average undrained shear
block) strength ( kN/m 2 )
• standard laboratory tests on concrete Routine High quality
samples obtained from the wall Depth Layer contractors block samples Su ( kN/m 2
• in situ tests - deepsoundering, vane and
(m) UU Triaxial UU Triaxial Adopted
pressuremeter Tests Tests
7-13 Sandy Clay 45 75 80
Figure 2 shows data obtained from this first
phase of the second programme. The distribution 13-19 Silty Clay 20 - 40
of the undrained shear strength adopted in 19-22 Sandy Clay 45 - 90
calculations in this paper is given in Table 1.
The values of the SPT's in the different borings 22-25 Silty Clay 40 - 80
and at different depths were valuable in
establishing this adopted undrained shear The second phase consisted of special laboratoJ
strength distribution. It was impossible to tests on undisturbed samples to establish the
improve the quality of the site investigation stress-strain behaviour of the soils on differE
because the work was performed by a Private stress paths. These tests are being carried 01
Contractor not under our control. The adopted at our laboratory COPPE-UFRJ. Initial results
values presented on Table 1 are believed to showed that the adopted undrained shear streng·
represent the most likely undrained shear represents satisfactorily this soil property.
strength profile. The relation between the initial modulus of
deformability in undrained extension tests and
the undrained shear strength does not vary roue:
with OCR and for practical purposes a value
between 200 and 300 can be utilised for this
relationship. More information about the soil
STANDARD PENETRATION SIMPLIFIED CLASSIFICATION TESTS parameters can be found on Lins (1980) and
TEST ISPTJ DISTRIBUTION
OF THE Correa ( 1 9 81 ) .
10 20 30 LAYERS 0 20 40 60 80 100%

·~ 0
SAND .......,,_., ....
FIELD INSTRUMENTATION

The instrumented area was chosen in a place

sufficiently distant from buildings to make th
ol ,,
10
!0 SANDY
CLAY
analyses of the data easier.

SILTY .I
CLAY '. ...,:• .. ; EXTERNAL FACE INTERNAL FACE
[0
9 I. LEVEL

20
)"~ SANDY
CLAY
lo
- " . ..
ii 3
t~~~ij~l~ ~~~~~ij~~ OF STRUTS
=-
2. LEVEL

( SILTY
" l ~

F 0 l
OF STRUTS
=-- 0 l
25
- CLAY

COARSE ........ .......... •"'·'·"

~10
c, o,
'
SAND '0
30 i ' 0

''I

~15
SPT AVERAGE - % CLAY } PARTICLE
SPT BOREHOLE SL-01, 0 %SILT SIZE
=
~
v
% SAND
LIO.UJD LIMIT
PLASTICITY LIMIT
NA !URAL WATER CONTENT
OJSTRIBUTION

:g I
:X:
>-
I•

rzo
I'
~
0

Fig. 2. DEFINITION OF THE SIMPLIFIED DISTRIBUTION OF . 25

THE LAYERS ID

30
l:'"lllllrli f lilllllllll~llll
CAGE A CAGE B CAGE A CAGE B
KEY

Do - 9 TOTAL EARTH PRESSURE CELL MDS 72

10 BAR EXTENSOMETERS MOS 24

'- 37 CONCRETE EXTENSOMETERS MOS 53a

III - NULL INDICA TORS

Fig. 3. PANEL INSTRUMENTATION

406
 Qre 3 shows the instrumented diaphragm wall CONSTRUCTION AND MEASUREMENTS DURING THE SECOND
el with the transducers to measure the STAGE OF EXCAVATION
crete strain at 16 sections, reinforcement
loads at 5 sections and total horizontal The 1.2 m thick and 30.0 m deep concrete
th pressure at 9 positions. The installation diaphragm wall was built by the slurry trench
cedures for these instruments has been technique excavated in panels 7.5 m long. The
cribed elsewhere (Soares, 1983). bentonite slurr¥ used has a specific gravity
about 10.5 kN/m • Each panel had a preassembled
ure 4 shows the soil instrumentation steel reinforcing cage weighing 350 kN.
centrated near the instrumented panel. Pore
ssures were measured near the wall inside and The excavation between the concrete walls was
side the excavation by pneumatic piezometers. done in three stages; excavation started on
December 8th, and took 85 days to reach the
final level 13 m deep. Figure 5 portrays the
progress of the construction. The walls were
supported at two levels by steel beams 1.80 m
apart and not preloaded.

- - - - 14.80 - - -
~- , .,- ~RIGI~_L GL
""""'lA :1st. STRUT LEVEL IIn O.Q 'db
FINE f2==:===:=;:;::=t:"=:=::,_o?1" ·•.
·. SAND
2nd. STRUT LEVEL ~=3.6 ;?: I

FIRST LEVEL OF STRUTS 12.JOml

. SANDY
CLAY
:§ -------~------------------
:z SECOND LEVEL OF STRUTS i5,90ml

:~
;'--. ·. ·; ~-. ·.-·- ;-··
~
<
> -----------------
SILTY
<
• CLAY 5
... "
. ·-~· -··
.
·- -
:5
....
::t:
10

12
0..
~
14

120~~~~
100
80
KEY

L~~~L ~~~ :!! ~ L~~~ll I

2nd. 251 • 251
:
--o-- 2nd. I
.,;~\~~
1000
M AA

7SO
kNM I

soo

60 LEVEL 252. 25L _.....,._LEVEll 20(! 160 100 I

~30

40
KEY RESIDUAL SOIL
20
• • 12 PIEZOMETERS ~
• · 17 SETTLEMENT GAUGES
o • 6 RING MAGNET HEAVE GAUGES "SPIDER" 100
o 2 4 6m
SCALE lml -~"~- 80
J. · INCLINOMETER TUBES
60
40
Fig. 4. SOIL INSTRUMENTATION
20

-20
-40
-60
-80
AND 2nd. STAGE
-100
1e vibrating wire piezometer was installed 5 10 15 20 25 30 10 15 20 25 30 10 15 20 25 30
1side the excavation near its final level. JANUARY I FEBRUARY I MARCH I APRIL
1ven inclinometer tubes were installed, six in
1e soil behind the wall at different distances Fig. 5. LOADS AND BENDING MOMENTS MEASURED ON THE
>art and one in the centre of the wall. The STRUT 5.
>rizontal movements of the tubes were measured
.th digitilt inclinometers models 50301 and
1306 manufactured by SINCO. The settlements at
,fferent positions were measured by means of
1rvey control and the bottom heave on the
:ntre of the excavation at different depths was
:asured with ring magnet gauges. Strut loads Figure 5 also shows the loads and the bending
:re monitored by four vibrating wire moments measured in one of the seven instrumented
(tensometers distributed in two sections. struts. There was very good agreement between
the measured loads at the two ends of the struts.
The maximum average and minimum loads at the
end of excavation are 560, 420, 340 and 110, 90,
60 kN/m for the first and second bracing level
respectively.

407
Figure 6 shows the concrete strain measured at Empirical and semi empirical methods such as t
the end of the second stage of excavation apparent earth pressure, free earth support an
(before installation of the second level of fixed earth support, were used to predict the
struts). From these measurements, the bending behaviour of the instrumented section. The
moments in each section of the wall were apparent earth pressure method as described by
calculated assuming a linear distribution of Terzaghi and Peck (1967) only estimates the
the strain at each section and a value for the maximum loads on the struts. The generalisati
Young's modulus of the concrete of 30 x 10 6 kN/m 2• of this method as suggested by some authors do
This value of elasticity modulus was slightly not give a reasonable prediction of the bendin
higher than the average (28 x 10 6 kN/m 2 ) obtained moments in the wall. The use of the free and
in the static tests on concrete samples obtained fixed-earth-support methods does not give good
from the wall. The value of this modulus from results either. It was suggested (Soares, 198
dynamic tests showed a good homogeneity of the that these methods should not be applied to th
concrete with the depth of the wall and the design of diaphragm walls in clay soils.
average value was 3 6 x 10 6 kN/m 2 •
The finite element method was not adopted, at
this stage of the analysis, due to the
difficulties of simulating excavation and in
most field cases the necessary data is not
available. The author believes that
sophisticated analytical techniques may be no
more appropriate than simpler solutions when
EXTENSION COHPRESSION
these are used with engineering judgement.
G.L 100 I
r---T---~--~roFo--~zo~o--~3oo 0 From its first use i t was seen that the model
a beam on elastic supports (Winkler hypothesis
could be useful in the prediction of the wall
behaviour. Many suggestions were obtained frc
Fages et al (1971a, 1971b and 1973), Dalerci a
Torrigiani (1976) and Miyoshy (1977) and all
available methods were used to estimate the
horizontal coefficient of subgrade reaction.
The structural pattern of the wall was a beam
supported by the struts and by the soil below
the excavation level. These supports behave
like elastic springs. The constants of the
strut springs were determined from the measurE
loads and horizontal displacements. It is
important to remember that the effective
rigidity of the struts depends on the
.tiiAAst ~:. construction conditions. For preloaded strut~
SAIIJ ,. this rigidity can be a tenth or less of the
'· structural rigidity. For struts that have not
ICEY been preloaded the effective rigidity can be
IIITEIINIAL FACE much smaller even than this.
o EX1ERHAt. FA!I
• 11JU. llliCATORS

LOAD ON THE STRUTS BENDING MOMENT HORIZONTAL

Fig. 6. CONCRETE STRAIN DURING THE SECOND STAGE lkN/ml lkN.m/m.l DISPLACEMENT lmml
OF EXCAVATION.

15 s
COMPARISON BETWEEN PREDICTED AND MEASURED VALUES
20
~6
- INTERNAL PRESSURE
The importance of establishing a simple design DISTRIBUTION
procedure was the aim at the first stage of the 25
analysis of the instrumentation data. The
simple procedure must be able to predict the 30
load~ on the struts, the distribution of the ka ::0~~ :.~~TAL
~nd~ng moment and the horizontal displacement ICa SPRING COHSTAHT ikN/M)
o t e wall. The settlement of the surroundin
ground, could be estimated from the horizontal g Fig. 7. COMPARISON BETWEEN MEASURED VALUES WITH THOSE
movements using an existing method (for example PREDICTED BY MIYOSHY METHOD.
Caspe, 1966; Juca, 1981). This simple
procedure could then be incorporated into the
design approach suggested by Rio de Janeiro
Underground System (RJUS).

408
che suggestion of Miyoshy (1977) when applied
:o the Botafogo wall conditions gives the
ralues shown on Figure 7. These calculated LOAD ON THE STRUTS BENDING MOMENT HORIZONTAL
ralues overpredict the measured bending lkN/ml ikN.m/m.l DISPLACEMENT (mml

noments, the horizontal displacements and strut 400 200

Loads. Miyoshy (1977) considers the loading on

f\~-·,
:he external face to be the active pressure up
:o the level of the excavation and from this
Joint up to a length equal the width of the
=xcavation a linear decrease of the loading
10
(Figure 7). It was decided to reduce the load LAYER 2
suggested by Miyoshy (1977), after verifying k=BOOO
:hat the soil modulus did not have a substantial 15
LAYER 3
influence on the results (Figure 7) . k:6000
INTERNAL PRESSURE 20
?igure 8 shows the predicted values for a LAYER 4
k:9000
f- DISTRIBUTION
loading equal to the active pressure up to the LAYER 5 KEY
level of the excavation. This condition gives k=SOOO 25
~ reasonable prediction of all measured values. LAYER 6
k:10000
rhese results show that there was a substantial 30
restriction to the change of the stresses in 20000 0
k= lkN/m'l
the soil inside the excavation. 30000 lSI

2soooo a

LOAD ON THE STRUTS BENDING MOHENT HORIZONTAL

lkN/m J (kNm/ml DISPLACEMENT (mml Fig. 9. INFLUENCE OF THE EFFECTIVE RIGIDITY OF THE
STRUTS ON THE VALUES PREDICTED WITH ACTIVE
SOIL PRESSURE ABOVE THE EXCAVATION LEVELS.

10

LOAD ON THE STRUTS BENDING MOMENT HORIZONTAL

IS lkN/ml lkN/ml DISPLACEMENT lmml
HYPOTHESES
1,00 zoo
20

25
LAYER 6
k•\0000
30
k• MODULUS OF HORJZONTAL REACTION lkN/m31
LAYER 2

Fig. 8. COMPARISON BETWEEN MEASURED VALUES WITH THOSE LAYER 3

PREDICTED WITH ACTIVE SOIL PRESSURE ABOVE THE
. 20
EXCAVATION LEVEL. LAYER 4

LAYER S 25

LAYER 6 k: 2000 SPT S

lO
k: 5000 SPl ~
Figure 9 shows the influence of the magnitude
of the effective rigidity of the struts. The
increase of this value increases the maximum
bending moment above the level of the excavation
and decreases the moments in the embedded Fig. 10. INFLUENCE OF THE SOIL MODULUS ON THE PREDICTED
length of the wall. VALUES

Figure 10 portrays the influence of the soil's

modulus. The best value of this modulus was
obtained when using the following relationship:
CONCLUSIONS
k (kN/m 3 ) = 100 to200 Su (kN/m 2 )
The measurement of bending moment, strut loads
It was concluded that this value of soil's and horizontal earth pressures at the second
modulus corresponds to the undrained modulus of instrumented section at the Rio de Janeiro
deformability from extension tests. Considering Underground made it possible to develop a simple
a safe relation between SPT and Su this design procedure for diaphragm walls in a
horizontal soil modulus can be estimated by: stratified soil profile with soft clay soils.
The model of a beam on elastic supports has been
k = 1000 to 1500 SPT (kN/m 3 ) used and the horizontal subgrade reaction
modulus hasjbeen related to standard laboratory
and field data.

409
 Symposium on Field Measurements in
ACKNOWLEDGEMENTS Geomechanics, zurich, September.
The author wishes to express his appreciation Terzaghi, K. and Peck, R.B. (1967)
to the Rio de Janeiro Underground Company for Soil Mechanics in Engineering Practice,
permission to publish the results presented in Wiley.
this paper. To carry out the research
programme and to obtain the mass of summarised
field data required the cooperation and efforts
of all staff of the Soil Mechanics Area from
Civil Engineering Department of COPPE/UFRJ.
The author would particularly like to
acknowledge the cooperation from Prof. P. Carim,
Eng. J.B. Ribas, Prof. W. Lacerda and Prof. D.
Velloso from COPPE/UFRJ and Dr. J.A. Charles
from Building Research Establishment.
Special appreciation is expressed to the
Brazilian Research Council CNPq for its
financial support during the postdoctoral
programme of the writer and to Building
Research Establishment BRS and the Norwegian
Geotechnical Institute NGI, for receiving him
into this programme.

REFERENCES
Caspe, M.S. (1966)
Surface Settlement Adjacent to Braced Open
Cuts. ASCE, Journal of Soil Mechanics and
Foundation Division, Vol. 92, SM4, pp 51-59.
Correa, M.H.C. (1981)
Compressibility Behaviour of Botafogo Soft
Clay on Conventional and Hydraulic Oedometer
Tests. Master of Science Thesis COPPE/UFRJ
(in portuguese).
Dalerci, G and Torrigiani, M. (1976)
Anchored Bulkheads Embedded into Soil with
Non-Linear Elastic Response. Proceedings of
the Sixth European Conference on Soil
Mechanics and Foundation Engineering, Wien,
Vol. 1.1, pp. 115-118.
Fages, R. et Bouyat, C. (1971a)
Calcul de Rideaux de Parois Moulees ou de
Palplanches. Travaux October.
Fages, R. et Buy at, C. ( 1971b)
Calcul de Rideaux de Parois Moulees ou de
Palplanches. Travaux, December, pp. 49-50.
Fages, R. et Gallet, M. (1973)
Calculations for Sheet Piled or Cast in situ
Diaphragm Walls. Civil Engineering and
Public Works Review 68, December,
pp. 1067-1071.
Juca, J.F.T. (1981)
Settlements on Buildings due to Excavations.
Master of Science Thesis, COPPE/UFRJ
(in portuguese).
Lins, A.H.P. (1980)
Extension and Compression Triaxial Tests on
Botafogo Soft Clay. Master of Science Thesis
COPPE/UFRJ (in portuguese).
Miyoshy, M. (1977)
Mechanical Behaviour of Temporary Braced
Walls. Proceedings of the Ninth International
Conference on Soil Mechanics and Foundation
Engineering, Tokyo, Vol. 1, pp. 655-658.
Soares, M.M. (1981)
~ Simple P~ocedure to Design Diaphragm Walls
~n Clay so~ls. Doctor of Science Thesis
COPPE/UFRJ (in portuguese).
Soares, M.M. (1983)
The Instrumentation of a Diaphragm Wall for
the Excavation for the Rio de Janeiro
Underground. Proceedings International

410