SHAFT GROUTING EFFICIENCY INVESTIGATED BY BIDIRECTIONAL LOADING TEST

OF BARRETTE PILE AT VINHOMES GOLDEN RIVER PROJECT IN HO CHI MINH CITY,

VIET NAM
Lan V. H. Bach(1); Hai M. Nguyen(2)
1
University of Architecture Ho Chi Minh city, 196 Pasteur str., District 3, Ho Chi Minh city, Vietnam,
2
Department of Geotechnical Engineering, Geotech Engineering and Testing, Houston, USA
Email: lan.bachvuhoang@uah.edu.vn ; haitdmu@gmail.com

ABSTRACT
To evaluate the shaft grouting efficiency, a bidirectional static loading test programme was performed on a shaft-grouted barrette pile at
Vinhomes Golden River project in Ho Chi Minh City, Vietnam in February, 2016. The soil profile consisted of fill sand to 6,1 m depth on loose to
medium dense silty sand to 35,2 m depth deposited on stiff clay to 63,2 m depth and underlain by very dense sand. The cross-section area of the
barrette pile was 2,800 mm by 800 mm. The pile was excavated to 69 m depth using grab-bucket excavation techniques with bentonite slurry and
guide wall advanced ahead of the hole. The cell assembly was placed 10,5 m above the pile toe level and the reinforcing cage was attached with
diametrically opposed vibrating wire strain-gages at three levels below and five levels above the cell level. Shaft grouting was carried out along a 10m
length above the pile toe. The static loading test was done 22 days after placed concrete. The analysis of strain-gage records reveals that the average
unit grouted shaft resistances on the nominal pile diameters were larger than the resistance along the non-grouted lengths about 30%. The pile toe
stress-movement responses to toe stiffness were soft with no tendency toward an ultimate value.
KEYWORDS: Bidirectional load test; Shaft grouting; Barrette pile; Analysis of strain-gage record
1. INTRODUCTION AND BACKROUND
In recent years, the population and economy of Ho Chi Minh City in Vietnam have increased rapidly, so the development of high-rise buildings
is a necessary factor to solve the problem of housing. However, the soils at the city consist of thick deposit of alternating alluvial soil layers of organic
soft clay up to hundred meters. For the drilled shaft foundation design for the high-rise buildings in this area, the toe resistances are normally ignored
by virtue of cleaning difficulty of the soft soils accumulated at the bottom of hole during drilling. Moreover, negative skin friction within the surficial
soft clay soil will introduce the additional settlement, downdrag, for the drilled shaft foundation, which requires the shafts to be significantly
lengthened to meet the required limits for settlements. Therefore, the deep foundations are usually constructed by the barrette or bored piles having
length from 70 m to 100 m to ensure load bearing and settlement caused by negative friction. But, this solution has resulted in a significant cost for
foundations of high-rise buildings. In order to save construction costs and time, shaft grouting technique has been developed to improve pile bearing
capacity and shorten pile length. The mechanisms proposed to account for the enhancement of shaft resistance following shaft grouting have been
summarized by Stocker [1] and Troughton & Stocker [2] as follows:
- The concrete on the perimeter pile is cracked and pushed out against the surrounding soil with the grout bracing the pile against the soil.
- The increased lateral pressure causes a local increase in soil density in the interface zone of the pile bore which has become softened or loosened
by the pile construction process.
- In granular soils, cementation of the soil particles in the interface zone may occur due to the infiltration of grout into the pores of the soil.
- Voids, fissures, cavities may be filled with grout providing an improved contact between the pile and the soil.
Field studies related to improvement of shaft resistance using post grouting technique were reported by Littlechild et al. [4]; Suthan et al. [5];
Nguyen and Fellenius [6]... In cases where the shaft grouting
technique has been deployed, it has proven very effective,
multiplying the shear resistance of pile by between 1,5 and 3
times. However, successful level of each project depends on
many factors such as the characteristics of soil, the grouting
pressures, the grouting mix ratio, the arrangement of grouting
tubes, the perforation along the installed grouting tubes, etc. In
addition, data on these effects is scarce and there were no local
codes of practice and the guidance for design and construction
of shaft grouted barrettes available to advise engineers how best
way to use shaft grouting technique.
The paper presents the test results of the drilled shaft
constructed for the Front project, which consists of three
apartment buildings from 36 to 50 storey in Vinhomes Golden
River complex in Ho Chi Minh City, Vietnam. The test barrette
was named TN6, which has a rectangular cross-section area of
2,800 mm by 800 mm and 69 m depth. The pile had a
bidirectional cell assembly placed about 10,5 m above the pile
toe and the reinforcing cages were instrumented with three pairs
of diametrically opposed vibrating wire strain-gages. Shaft
grouting was carried out on barrette pile after completion of
concrete placement about 10 m lengths from 1 m above the pile Figure 1. Ho Chi Minh City area with the meandering Saigon River and the
bottom. The purpose of the bidirectional loading test was to location of the Vinhomes Golden River Project
determine the pile capacity and the effective of post grouting.

-1-
2. SOIL CONDITIONS
The project is located along Saigon River and next to Thi Nghe canal the soil consists of thick deposit of alternating alluvial soil layers of
organic soft clay, compact silty sand with some gravels, and medium dense to dense silty sand, underlain by dense to very dense silty sand. Figure 1
shows a map over the area of the Vinhomes Golden River complex and Saigon River, which establishes the upper soil layer of the Ho Chi Minh City
and the site.
The site investigation included 10 boreholes (HK), 5 field vane shear tests (FVST) and 1 observation water level (SP) at locations shown in
Figure 2. The soil profile at the borehole #HK 12, which located closest to the test barrette, consisted of fill sand to about 6,1 m depth, on compacted
loose sand to 15m depth, next by the medium dense sand layer to 35,2 m depth, followed by very stiff to hard clay about 63,2 m depth and underlain
by dense to very dense sand to at least 70 m depth (see Figure 4).

Figure 2. Layout of boreholes and test barrette TN6 Figure 3. Water content; Friction angle and cohesion from direct shear
tests; N-indice at the borehole #HK12

The average wet density of the clay and

the sand are about 1950 kg/m3 and 2100 kg/m3
respectively. The pore pressure distribution was
hydrostatic and corresponded to a ground-water
table at 1,2 m depth below the ground surface.
The subsoil conditions, which located up to 35
m depth, are very sensitive character of poorly
graded uniform silty sand with low to medium
SPT N values. Below 63 m depth, the N-indices
are about 50 blows/ 0,3m, which showed the
conditions to be very dense. Figure 3 shows the
distribution of water content, consistency limits;
friction angle () and cohesion (c) from direct
shear tests; SPT N-indices of the borehole #12.
The soil profile between the borehole #12 and
borehole #10 at the site is shown on Figure 4.
The coefficient of permeability medium
dense sand layer (from -15 m to -35,2 m) is
k=3,76.10-4 m/s, which was determined by the
field permeability tests.

Figure 4. Soil profile between the borehole #HK12 and #HK10

3. CONSTRUCTION OF THE TEST PILE
February 5th, 2016, Bachy Soletanche Vietnam completed the construction of the test barrette. The instrument pile was constructed using rope
grab excavation techniques with bentonite slurry. Before lowering the reinforcing cage and placing concrete, the shaft was cleaned by clamshell grab
during recycling of the bentonite slurry. Thereafter, reinforcing cages with the O-cell boxes (see Figure 5), which was attached at 10,5m above pile
toe, were lowered into the borehole and temporarily held by millet bars on the operation floor.

-2-
 Figure 5. Installation of reinforcing cage of the test barrette TN6

The concrete was poured through two 300 mm O.D. tremie pipes to the bottom of the barrette pile, displacing the bentonite slurry until the
concrete reached the elevation +2,09 m. The concrete grade of barrette pile is B40, and the unconfined compressive concrete strength was determined
22 days after casting to 56,9 MPa. Representatives of Bachy Viet nam and Fugro Singapore supervised the construction of the test barrette.

-3-
 Figure 6 illustrates the orientation of the bidirectional cells, reinforcement bars, telltales, strain-gages, grouting tubes… over the cross-section
of the test pile. The barrette was supplied with a reinforcing cage of forty-four 32-mm bars; resulting in a steel reinforcement area of 353,9cm2 and a
reinforcement ratio of 1,58 % for the 2,24 m2 nominal pile cross section.
Figure 7 shows the locations of the vibrating wire strain-gages (SG) attached to the reinforcing cages (three levels below and level 5th through
level 8th above the bidirectional cell level) of the test pile. Each gage level contained three diametrically opposed pairs: Gages A and D; Gages B and
E; Gages C and F, which cover equal area of the barrette cross section (Fig.6).

Figure 6. Instrumentation Orientation Layout of test pile TN6

Figure 8. Tubes à-Manchette (TAM) pipes arrangement Figure 7. Details of grouting and instrumentations

The procedure of shaft grouting barrette pile

The system adopted for shaft grouting consisted 8 numbers of 49 mm diameter mild steel Tubes à-Manchette (TAM) pipes (Note: 2 numbers of
grout tubes for spare), with manchettes spaced at 1,0 m intervals along the pipes. The TAM pipes were fixed, using normal tie wire, to the outside of
the reinforcement cage and within the zone of the concrete cover. The TAM pipes extended the full depth of the shaft grout zone and were generally
equi-spaced around the perimeter of the piles and the barrettes (see Figure 8). The shaft grout zone is to be measured from 0,5 m above the manchette
to 0,5 m below the bottom manchette.
Water Cracking: The TAM pipes for shaft grouting, were flushed 24 hours after concreting to ensure there were no blockages in the pipes.
Grouting Methodology: Five days after placing the concrete, the shaft grouting was implemented by first cracking the pile concrete cover by
pumping high pressure water through the grout tubes.
A water/cement grout mix used, with a retardant and plasticiser, Cement : 105 kg
to maintain the fluidity of the grout. During the grouting process the Water : 66,6 liters
viscosity and density were checked for compliance with the Bentonite : 0,6 kg
specification and were found to be consistently achievable. The grout Daracem 100 : 400 ml
mix per 100 liters of grout is as follow: Bentocryl 86 : 150 ml

Actual volume of grout per sleeve was 42 liter. The total grout volumes for barrette TN6 was about 2,52 m3, equal to 35 liter/m2. Grouting
pressure of all TAM was recorded range from 8,6 MPa to 25,1 MPa, which were smaller than the limiting pressure set at 40 MPa. All grout volume,
grouting pressure were controlled by SINNUS system, which was developed by Soletanche Bachy Group.
-4-
4. ANALYSIS OF TEST RESULTS
a. Load versus Movement
On February 27th 2016, 22 days after concreting barrette TN6 was performed by Fugro Singapore Pte. Ltd.. The loads applied by the
bidirectional cell act in two opposing directions, mobilizing resistance above and below the cell. The loading test for pile was carried out in two
loading cycles. Figure 9 shows the load versus Time schedule of the bidirectional assemblies, which were placed at 10,5 m above the pile toe.
The Cycle 1 loading for pile was performed by four increments ranging about 5,78 MN to a maximum load of 17,12 MN. Each of the first
three load increments of Cycle 1 was held constant during one hour and the level 4th was held for 6 hours and then the test pile was unloaded phase in
four steps.
In Cycle 2, the barrette was first reloaded to the same 17,12 MN in two increments, after that the loading continued in four increments ranging
from 3,61 MN to 4,12 MN until a maximum load of 32,06 MN. In this cycle, the test pile was unloaded in six steps. Almost unload levels were
maintained for 30 minutes. Only level 1st the load was held for 24 hours, because the pile toe suddenly had more downward movement in the phase. It
should be noted that the records shown in Figure 10 are the result of a successful placing of the bidirectional cell as the upper and lower lengths of the
pile (above and below the bidirectional cell) and the lower length of the pile moved appreciably by the applied load.

Figure 9. Time versus Bidirectional Cell Load Figure.10. Time versus Displacement of the test pile

Theoretically, the cell does not impose an upward movement until its expansion force exceeds the buoyant weight of the pile above the cell
plus the residual load, if any, acting at the cell level. The pile buoyant weight above the cell was 1,9 MN. Figure10 and 11 describe the measured load
- upward and downward movement curves of the TN6 bidirectional test in the Cycle 1 and 2. Loads are those measured and are not adjusted for pile
weight and water pressure at the cell level.
In Cycle 2, the maximum downward cell movements was 130,5 mm, the pile toe movement was 126,7 mm, whereas the upward cell
movement was 12,5 mm, and the pile head movement was 2,4 mm. The movements of the cell are more than the pile head or pile toe, because the pile
has axial “elastic” shortening under axial loaded.

Figure 11. Measured load-movement curves of test pile TN6

b. Strain Gage Measurements
Strain measurements in pile must always be in pairs placed diametrically opposed. If so, the average strain will offset any bending effect to be
representative for true axial strain. The records of both gauges in the pair are needed. If one gage of the pair becomes unreliable, the data of remain
gage of the pair should be discarded. It is, therefore, advisable to schedule three gage pairs to important levels in the test pile. Having three gage pairs,
-5-
which were placed four sides of the barrette cross section (Figure 6), will also improve the representativeness of the measurements because some
variation of stress from one side of the pile to the other is unavoidable.

Figure 12. Load-strain measured at gage level 5th of the barrette test
Figure 12 shows the TN6 load versus strain recorded at Gage Levels 5 in Cycles 1 and 2. In the loading phases, five of the gage records (A, B,
C, E and E) are similar, whereas the fourth (D) significantly deviates from the others. Possibly, the gage SG5-D didn’t work properly, so that the
records of the gage pair (SG5-A and SG5-D) were discarded in average strain values of level 5.
The scatter of strain gage records not only appeared in level 5, but also found at the level 6 and 8 in both cycles. The divergence of strain
values were recorded by SG6-B, SG6-E in level 6 and by SG8-A, SG8-D in level 8. Therefore, the records of these gage pairs were discarded in the
average strain values of level 6 and 8, respectively. The other gage levels of the barrette: SG1, SG2, SG3, SG4 and SG7 (not shown here) appeared to
functioning adequately and the averages of all gages were considered representative for the axial strain measured at the respective levels.

Figure 13. Bidirectional cell load versus measured strain for TN6 in Cycles 1 and 2
Figure 13 combines the average strains records for the two load cycles applied to the test barrette. As is usually the case, the load-strain
relations are slightly curved, which is due to the influence of shaft resistance and, but to a minor degree, to the fact that concrete modulus reduces
with increasing strain. Short portions of the curves will always appear rather straight, however.
It is easy to recognize that, in the begin of the cycle 2 the average strains records of all gage levels are not equal zero (right diagram in Fig.13),
because these strain gages were consisted of residual strain from the unload phases of cycle 1. Moreover, the measured strains of SG1 increased
rapidly in the end of cycle 2, so the load - strain relation of SG1 was suddenly strongly curved, that cause of that cause of a smaller stiffness of the
pile portion. It is likely that the concrete of pile toe was not good quality or the pile was narrowed at the gage level SG1 resulting in a smaller cross
section, because it was affected by presence of slurry filter cake in the bottom of the drilled shaft.
c. Axial Stiffness of the test Barrette
The upward shaft movements measured in the test were small. At pile head, the maximum relative movements between the pile shaft and the
soil were no more than about 2,5mm. This movement is not likely sufficient to have mobilized the shaft resistance fully. The downward movements at
pile toe and base of O-cell are 126,7mm and 130,5mm, respectively which are larger than the minimum movement necessary to fully mobilize the

-6-
shaft resistance along the pile portion below the cell level. The full shaft resistance mobilization is reached where the slope (stiffness) of the load-
strain curves at SG2; SG3; SG4 and SG5 (closest to the cell assembly) become approximately linear.
Axial tangent stiffness (EA) of the drilled shafts is an important parameter for evaluating the shaft resistances measured from the installed
strain gages. Normally, to convert the measured strain values to load, it is necessary to know the modulus of the drilled shaft material; however, the
cross-section area of drilled shafts is variable significantly versus the drilled shaft depth. Therefore, the combination of shaft material modulus and
cross-section area of shaft at the gage levels, which is known as axial stiffness of shaft (EA), is important to evaluate the shaft shear resistances from
the measured strain gage data. Determining the modulus to use is simple in theory, but complex in practice. Therefore, for this case, the authors have
preferred to rely on the linear portions of load-strain relations and convert the measured strains using constant stiffness EA to convert the measured
strain to force by equation:
Q
  E  Q  EA (1)
A
Where:
: Stress in the pile section
Q: load in the pile section
A: cross-section area of the pile
E: Young modulus (elastic) of the pile material
 : Measured strain
The force at a strain gage level is equal to the load applied by the
bidirectional cell minus the shaft resistance, which engaged between the
bidirectional cell and the strain gage level. Once the relative movement
between the pile shaft and the adjacent soil is sufficient to fully mobilize
the shaft between the bidirectional cell and the strain gage level (and
assuming the continued shaft shear is plastic), the additional load applied
is transferred in its entirety to the section where the strain gages are.
Therefore, beyond this load, the continued load-strain is linear, as
expressed in Equation 2:
Q  EA    EA  Q  (2)
Where:
Q  Qn 1  Qn : Change of load from one load increment to the next
Figure. 14. Barrette TN6 Stiffness Plot
  n 1  n : Change of strain from one load increment to the next
Using the records of the test piles in Cycle 2, Figure 14 shows the incremental stiffness (Q/) against the strain () as an approximately
straight line with an ordinate intercept equal to the pile stiffness (EA), for low strain. In fact that, the increment stiffness - strain relations converge to
a sloping line, which indicates a pile modulus reducing with increasing strain. A slight slope is indicated, that is the diminishing stiffness with
increasing strain, but the range of imposed strain is not large enough to ensure this assessment, according to Fllenius: “Ideally, the maximum strain
should have been at least 600 με for the final stiffness to be established accurately and also to establish the strain dependency of the stiffness”. The
procedure is known as the “Tangent modulus” or "Incremental Stiffness” (Fellenius [7]), the method assumes that for load increments applied after
the shaft resistance at the studied gage level has been fully mobilized, the continued incremental stiffness values will plot along a slightly sloping line.
Note that in Figure 14, only some strain gage levels closest to the bidirectional cell, were used in the determining the stiffness. The
differentiations involved in the process are sensitive to the precision of the measured strains and applied loads in relation to the actual value of strain
resulting from the applied increment of load. The maximum strains, which were appeared in SG2, SG3, SG4 and SG5, about 260 με to 370 με
induced by the applied loads. The actual stiffness of test barrette is reasonably about EA =70 GN, which corresponds to a Young modulus of about
E= 31,25 GPa on the nominal cross section.
d. Shaft Stiffness from Elastic Shortening Measurements
The elastic shortenings had obtained from difference of the movements measured at the cell levels and at the top and bottom of the tested
barrette. To confirm the shaft stiffness estimated from the strain gage measurements, it is necessary to employ the theoretical elastic compression
equation (3) to calculate the shaft stiffness from the measured elastic shortenings as follows:
QL TABLE 1: Back- Calculation of Elastic Shortening from the Estimated Shaft Stiffness
Se  (3)
EA Maximum Shaft Shaft Elastic Shortening of Elastic Shortening of
Where: Load test Length Stiffness shaft below cell level shaft above cell level
(MN) (m) EA (MN) (mm) (mm)
Se : Elastic shortening
Below Above Below Above
Q : Applied load cell cell cell cell Calculated Measured Calculated Measured
L : Length of shaft level level level level
EA: Stiffness of shaft 33,3 30,0 10,5 58,5 70 000 4,9 3,8 25,1 10,1

The elastic compression in this case refers to the shaft deflection that would occur if 100 percent of the applied load was transferred to the toe
and head of the shaft. The results of elastic compression are given in Table 1. Note that, the maximum load above the cell level is the respective cell
loads minus the buoyant pile weight above the cell level and the load below cell level is the respective cell load plus the load resulting from the water
pressure at the cell level. No consideration of residual load was included.
Logically, if absence of the positive residual loads in the drilled shafts before testing, the measured elastic shortenings must be smaller than the
theoretical calculated elastic compression from equation (3) due to presence of the soil shear resistances along the constructed shafts. The calculated
-7-
results presented in the above table show that the calculated elastic shortenings for the stiffness estimated from the strain gage measurements are less
than the measured elastic shortenings. Therefore, the shaft stiffness estimated from the strain gage measurements is reasonable for the drilled shaft
resistance analysis.
e. Load distribution along pile shafts
The derived EA = 70 GN pile stiffness was used to convert the average of strain measured at each gage level for each applied load. The
selected stiffness is the reasonable stiffness to use for the back calculation of the load distribution along of the pile. A larger stiffness would have
resulted in load magnitudes at SG3 and SG4 too close to the applied cell loads or even larger than the cell loads.
Figure 15 shows the load distributions of the barrette TN6 in Cycle 1 and 2. The right diagram of the Fig.15 also shows the equivalent head-
down load-distribution curves of the tested shafts in the end of Cycle 2 (dashed lines). It is noted that the buoyant pile weights above the cell levels
were not subtracted and the residual loads were not included. The water pressures at the cell levels were not measured and were not considered.
In Figure 15, it is easy to recognize that load distributions from the gage level SG 2 through the gage level SG1 are not reasonable. Hence, the
load distributions of this pile portion will become more reasonable if extending the load-line between SG3 and SG2 to reach level of SG1 (dotted
line). So, gage data of SG1 have not been used in the further analysis.

Figure 15. Load distributions of the Barrette in Cycle 1 and 2

f. Unit Shaft Friction versus Movement

Figure 16. Unit shaft resistance versus Average zone movements

The pile movements were measured at pile head and the pile toe by the records of the corresponding telltale. Moreover, the upward top of O-
cell and the downward base of O-cell were determined by records of the O-cell expansion and the upper compression telltales, which were located at
elevation -56,3 m closed of the O-cell top. The movements at gave level SG4 (above the cell) and SG3 (below the cell) were calculated as the upward
and downward movement measured at the cell level minus the elastic shortening of segments GS4 - Cell and Cell - GS3, respectively. The
movements at another gage levels were calculated with respect to the movements at SG4 and SG3, respectively, and so on for the next level.

-8-
 The difference in load between the gage levels is the shaft resistance between the pile and the soil. The left diagram of Fig. 16 shows the unit
shaft resistances versus upward movement for the gage levels above the cell levels and the right diagram represent the unit shaft resistances versus
downward movements for the gage levels below the cell levels. The curves were obtained by dividing the shaft resistance by the surface area between
the gage levels and then plotting against average movements at the pile portion between the considered gage levels.
For test pile, the unit shaft resistances below cell were mobilized fully at a movement of about 10 mm to 12 mm; thereafter shaft resistance
became plastic. In contract, the upward movements at SG5; SG6; SG7 and SG8, were rather small. This movement is not likely sufficient to have
mobilized the shaft resistance fully, so the unit shaft friction of this pile portions (SG5 to SG8) increased throughout the test. Only, the pile portion
between the cell and SG4 (3,5 m above the cell) the upward movement was larger than 10 mm and the shaft resistance was fully mobilized .
TABLE 2. Movements at gage levels and back-calculated beta-coefficients.

Beta- Effective Unit shaft

Gage Depth Movement
coefficient stress resistance Note On the other hand, according to Bjerrum Burland
level (m) (mm) (  'z - kPa)
() (rs - kPa) (1973), the unit shaft resistance can be determined by the
Head 0.0 2,4 formula (4) through the coefficient  and the effective
- - -
overburden stress of the soil at each level (method ).
GL8 15,0 2,8 0.18 222 40.2
rs   'z    rs  'z (4)
GL7 27,0 4,0 0.14 342 48.8
Where:
GL6 29,0 5,8 0.17 442 76.6
 : Bjerrum-Burland coefficient (or beta-coefficient)
GL5 37,0 7,9 0.24 522 125.8
 'z : Effective overburden stress
GL4 55,0 10,3 0.46 580 267.1
Table 2 includes the effective overburden stress,
 12,5
O-Cell 58,5 - 597 - the unit shaft resistance and movements at the gage levels.
130,5 The beta-coefficients () were back calculated by the
GL3 61,0 129,7 0.54 637 345.4 grouted formula (4). The unit shaft resistances (rs), which are
GL2 64,0 128,6 0.52 667 348.9 grouted represented in the Table 2, are data at the ultimate
GL1 67,0 127,1 - 702 - grouted resistance values in Cycle 2. According to Fellenius [8]
the beta-coefficients (  ) usually have a value between 0,2
Toe 69,0 126,7 - 722 -
to 0,5.
f. Toe resistance versus movement
The toe resistance was determined as the load of gave level 2 (SG2) of the test barrette, because the deepest gage (SG1), which was located very
close to the pile toe, is not unsuitable with the chosen stiffness of barrette. The toe resistance is calculated by the load of SG2 subtracted by the shaft
resistance between the SG2 to pile toe. In there, the unit shaft resistance of this portion gets approximately equal to the corresponding value of the
length between SG2 to SG3. Figure 17 shows the toe resistance versus the telltale-measured barrette base movement with solid line.
The curve was quite fitted to a q-z function curve shown with
dashed lines (Figure 17), which is determined by the ratio function
(Gwizdala; Fellenius [8]) according to Equation 5:

  trg 
R trg
  (5)
R n  n 
Where:
R trg : Reference or target resitance
R n : Any resistamce
 trg : Movement mobilized at R trg
 n : Movement mobilized at R n
 : An exponent; 0    1
Using ratio of the load to the corresponding movement at any
point as the R trg trg , and all other load-movement pairs are
determined the θ-exponent in Equation 5. As mentioned below, the
θ-exponent equal to 0,45 was found to provide the toe load –
movement curve that fitted the measured response for the pile toe.
According to Gwizdala (1996) the q-z curves for toe resistance Figure 17. Toe load versus telltale-measured pile-toe movement
usually have an θ-exponent between 0,4 to 1,0. and q-z curve for  - exponent of 0,45

6. DISCUSS
The barrette was designed to use in the site, which is located along Saigon River. It length is about 69m to ensure the pile base located in the
very dense sand layer (having SPT index N=50), which can be limiting the settlement of pile foundation. In order to proportion between the pile
capacity from soil properties and load bearing capacity of the pile section, the grouted length was calculated 10 m, which is placed in the hard clay
layer and dense sand strata, from 1 m above the pile bottom.

-9-
 In practice, designers often choose pile portions, which are placed in the sand layers to apply the shaft grouting technique, because they think
that in sand strata the efficiency of shaft grouting better than that in cohesive soils. In fully saturated clay layer, grouting pressure will increase the
pore water pressure increase rapidly when the grout injection begins, making the effective stress of soil can reach to zero and this liquefaction allows
the grout to form a homogeneous layer between soil and pile. Due to the permeability coefficient of clay is small, the excess pore water pressure will
not totally dissipate for a small time and therefore the effect of shaft grouting in clay does not reach the peak in the short term. In contrast, an
infiltration of the grout into sand strata may quickly improve their strength local interface. However in the test, the gage data of level SG1 was
discard, so we cannot evaluate the unit shaft resistance of the grouted pile portion located in the dense sand layer (from gave level SG2 to SG1).
At the maximum O-cell load of 32,06 MN, the downward movements at pile toe and base of O-cell are 126,7 mm and 130,5 mm respectively,
which are larger than the minimum movement necessary to fully mobilize the shaft resistance along the pile portion below the cell level. The unit
shaft resistance of two segments Cell to GL3 and GL3 to GL2 ranging from 345 to 360 kPa are relatively reasonable for the grouted shafts placed in
the hard clay. Whereas, the upward movements at pile head and top of O-cell are 2,5 mm and 12,4 mm respectively, as a result the unit shaft
resistances of sand layer from the gage level SG1 through GL6 (the left diagrams of Figure 16) are very low, only about 40 and 76 kPa. Only the pile
segment between gave level SG4 to O-cell reach the ultimate shaft resistance, thus the maximum measured unit shear resistance of about 267 kPa. It
has shown an improvement in the friction of grouted pile portions. The unit shaft resistance of the pile lengths from SG4 to O-cell (non-grouted) is
smaller 30% than that of grouted pile portion between O-cell to SG3, which are located in the hard clay layer. So, the effectiveness of the grouted
shaft estimated from the strain gage measurements is about 1,3 for clay layer.
The pile toe resistance – movement showed no tendency toward an ultimate value. The toe load-movement curve fitted the q-z function curve
with -exponent about 0,45. The maximum toe resistance of test pile is 9,8 MN, and accounting of 15,2% of the maximum applied load in the end of
Cycle 2. The pile toe response was very soft and not representative for a pile in very dense sand. Moreover in the procedure of shaft grouting, the
grouted pressures of all TAM were recorded range from 8,6 MPa to 25,1 MPa, which were so smaller than the limiting pressure set at 40 MPa. It
shows that the soil around the barrette test is considerably disturbed and loosened by the pile construction process.
7. SUMMARY
The bidirectional-cell test on the grouted shaft was reported. The analysis of the load-movements curves and the strain gage measurements
were also performed and presented. The following conclusions can be drawn from the present study.
The bidirectional test reported in the paper was unbalanced, that is the test showed a good response of the lower pile length (good downward
response) but did not engage the upper length fully. Because, the maximum upward movement of the pile head was 2,4 mm, which means that the
shaft resistance along the upper pile length was not quite fully mobilized. The barrette TN6 established that the allowable bearing capacity of pile
satisfied the specified value, which was designed equal 30 MN.
The strain gage levels installed close to the cell level are useful for evaluating the shaft stiffness, but not reliable for evaluating all strain gage
record. The evaluated axial stiffness of the barrette test is about 70 GN, which corresponds to a Young modulus of 31,25 GPa on the nominal cross
section.
The shaft grouting technique has been used on the barrette to enhance the pile capacity. The unit shaft resistance of the clay layer was increased
about 130 percent after grouting. The average unit shaft resistance of the 10-m grouted shaft is about 1,5 times greater than that of the non-grouted
shaft. The pile toe resistance was very soft and indicated that debris might have been left at the bottom of the pile shaft after cleaning the slurry and
the tremie concreting.
The shaft grouting techniques for drilled piles applied in Vietnam can be one of the best solutions for high-rise buildings and infrastructural
projects especially for projects could not find below the bedrock layer to embed into. To broaden the technique, Vietnam needs to issue local codes of
practice and the design and construction guidelines for estimating the capacity of deep grouted piles.
ACKNOWLEDGEMENTS
The authors wish to thank Mr. Bui Truong Nghia - Vice director of Bachy Soletanche Vietnam for permission to use the project data.
REFERENCES
[1]. Stocker. M., 1983. “The influence of post-grouting on the load-bearing capacity of bored piles”. Eight European conferences on Soil
Mechanics and Foundation Engineering. Helsinki, pp.167-170
[2]. Troughton, Stocker. M., 1996. “Base and shaft grouted piles”. Proceedings of Institution of Civil Engineering, Geotechnical Engineering
119, pp.186-192
[3]. Gouvenot.D. and Gabaix.J., 1975. “A new foundation technique using piles sealed by cement grout under high pressure”. Proceedings
Offshore Technology Conference, Texas, pp 645-655.
[4]. Littlechild, B., Plumbridge, G., Hill, S., and Lee, S. (2000) “Shaft Grouting of Deep Foundations in Hong Kong”. New Technological and
Design Developments in Deep Foundations: pp. 33-45.
[5]. Suthan P., Sherif E., Walt V., and Gray N., 2010. “Large scale laboratory testing of low mobility compaction grouts for drilled shafts tips”.
Geotechnical Testing Journal, 33(5) 13.
[6]. Nguyen, H.M. and Fellenius, B.H., 2015. “Bidirectional cell tests on non-grouted and grouted large-diameter bored piles”. Journal of Geo-
Engineering Sciences, IOS Press, 2(3-4), pp.105-117
[7]. Fellenius, B.H., 1989. “Tangent modulus of piles determined from strain data”. ASCE, Geotechnical Engineering Division, the 1989
Foundation Congress, F.H. Kulhawy, Editor, Vol. 1, pp. 500-510.
[8]. Fellenius, B.H., 2017, “Basics of foundation design”, a text book. Revised Electronic Ed., [www.Fellenius.net].
[9]. Fugro Singapore Pte. Ltd, 2016. “Reports on Barrette Bidirectional Testing”, Vinhomes Bason, HCMC. Vietnam
[10]. Portcoast Consultant Corporation, 2016. “Report on Soil Investigation Work – Project: Vinhomes Bason, HCMC. Vietnam”.

- 10 -