TECHNICAL paper

Ground improvement for the

foundation of a high-rise building
Luca Mancinelli and Martino Gatti, Rocksoil

Abstract working procedure. of a system of jet-grouted columns for the construction of the columns,
For the construction of a new At the same time, the response of on which the foundation slab of the including the composition of the
office for the Lombardy region the ground, in terms of settlement building in reinforced concrete is stabilizing mixture, the volumes and
administration in Milan, Italy, a under the building load, was set. After structural analyses it was flow rate of the fluid introduced,
161.3m high concrete tower with simulated using FEM models. This clear that the vertical stresses under the number and diameter of the
39 floors and two basement levels, paper describes the development of the slab would have been too high. nozzles, the pressure applied, the
a ground improvement solution the design, from the initial definition In particular, maximum values speed of withdrawal and rotation of
was proposed and implemented. of the geotechnical model, based on of 700kPa and average values of the rods;
This building constitutes the new geotechnical tests, to the best value 500kPa were deduced and judged n verify the results of the treatment
headquarters for the personnel construction parameters of the excessive for the foundation system in terms of the geometry of the
of the Lombardy region local grouted columns; and from FEM without improvement. columns, diameter and continuity,
authority, who were originally in numerical simulation to predict The studies were completed in and the strength of the improved
different locations. settlement to the actual results terms of absolute and differential ground.
The objectives of this intervention measured after construction. vertical settlement; theoretical results The test bed was located in the
were to increase bearing capacity showed, in the worst conditions, area in which the future building was
and reduce settlement under 1. Introduction maximum vertical translation of to be constructed and consequently
the high rise building load. In a This paper looks at the steps taken 40mm and differential vertical the zero level, corresponding to the
geotechnical context consisting of to develop the solution chosen for settlement of 20mm on a horizontal ground level, was assumed to be at
sand and gravel with water and a the ground improvement of the new basis of 25m, which were judged elevation 125.07m.
3m silty layer, with low Young’s office building in Lombardy. unacceptable for the building. The ground improvement
modulus, at a depth of 30m from The vertical structures of the A jet-grouting system was chosen operations were defined on the basis
the surface, jet-grouting technology building consist of a square concrete because the technology for this kind of the field tests and completed
was chosen. After a preliminary core, providing great rigidity from of application is well developed in using finite element numerical
phase of insitu and laboratory tests bottom to top, surrounded by a the area; it was preferred to concrete modeling to predict and estimate
for the definition of the geotechnical system of concrete columns. A injections for timing and cost expected ground settlement. The
model, columns were created 2m thick slab was chosen for the reasons. Before general jet-grouting results of some of the calculations
using different sets of operational foundation, increasing to 3m in the operations began, a test bed was are presented here and compared
parameters and then examined core zone. The solution adopted for designed to achieve the following: with the monitoring data acquired
to establish the most appropriate the ground improvement consisted n identify the operating parameters during construction.

SA1 SA2 SA3 Table 1: Natural characteristics of the gravelly-

122.89m 124.2m 124m sandy levels (A and C)
Fill
Elevation w Gravel Sand Silt Clay
Core
Sandy gravel
bore Sample (m) (%) (%) (%) (%) (%)

A level 1 109.2 14.6 30 45 21 4

2 103.7 9.3 44 40 14 2
3 97.2 9.2 36 42 18 4
4 85.7 15.1 2 70 26 2
B level Silty clay
SA1 5 79.5 11.9 10 60 26 4
1 107.4 6.7 56 28 14 2
2 102.1 10.5 42 43 13 2
3 94.4 14.0 22 57 19 2
Gravelly sand 4 85.6 16.0 9 79 11 1
C level
SA2 5 77.9 14.0 22 65 12 1
1 108.5 9.4 29 41 26 4

Silty sand
2 102.9 10.4 40 40 18 2
3 98.2 12.0 22 53 21 4
4 88.5 12.5 22 51 23 4
Figure 1 – Design stratigraphy SA3 5 80.8 12.7 20 63 15 2

28 ground engineering april 2013

 TECHNICAL paper
Table 2: Natural characteristics of the cohesive level (B)
Core Sample Elevation gnat gdry w wL wP IP Gravel Sand Silt Clay
bore
(m) (kN/m ) 3
(kN/m ) 3
(%) (%) (%) (%) (%) (%) (%) (%)
SA1 A 93.0 -- -- 22.3 36 20 16 0 6 71 23
SA2 A 90.9 18.89 14.31 32.0 59 28 31 0 6 63 31

Table 3: Summary of elastic modulus (values in MPa)

Building Jet columns:
LEVEL Ed EtxCU EtxCD Epress E0 SPT Edin concrete core JL=23m
JL=15m
A 40 140-220 800-950
JL=11m
B 10-20 60-100 70-150 35-45 660-760 SA1 Core bore
C 240-250 900-1,100
SA2
Table 4: Summary of values for cohesion and angle of friction
112.07m
LEVEL Direct shear Triax CU Triax CD SPT SA3
c’ (kPa) f’ (°) c’ (kPa) f’ (°) c’ (kPa) f’ (°) f’ (°)
2D FEM model cross section
A 40-43
B 20-35 22-27 25-45 21-25 20-45 20-26 Building plan shape
113.07m
C 40-41 0 m 20

2. Geotechnical conditions Table 5: Design Figure 2: Plan view of the project

and stratigraphy parameters for cohesion
The soil in the construction area and angle of friction 125.07m Ground level
consists mainly of gravelly-sandy
deposits and in a stratum of silt and LEVEL c’ (kPa) f’ (°) E (MPa)
clay, with very different geotechnical A 0 37 100

10m
Basement volume
characteristics compared to the
surrounding ground. B 20 22 30 115.07m
Examination of the stratigraphy 113.07m
from the boreholes revealed the C 0 37 150 112.07m
following:
n a heterogeneous surface level the water table was assumed to be
consisting of backfill and reworked 106.07m and therefore 19m below
soils with thicknesses ranging from ground level. A level
102.07m (22m)
1.5m to 3.0m;

26m
n a first succession of gravelly- 3. Geomechanical 97.07m
sandy layers with subordinated characterisation of the site
pebbles and silt with elevation from The geomechanical characterisation
125m to 93m defined as LEVEL of the site was performed by means
A (its bottom surface were then at of a series of geological survey and 89.07m B level (3m)
about 32m below ground level); laboratory tests; a summary of the
C level
n a silty-clayey level with an results is given in Figure 1. In terms
average thickness of between 2m of natural characteristics Tables 1
47m
and 3m, ranging from solid to very and 2 summarise the parameters
solid and continuous throughout obtained from the laboratory tests. Figure 3: Cross section of the project (through the minimum width
the area in question, defined as of the slab)
LEVEL B; 3.2 Elastic Modulus
n a final level consisting almost Numerous in situ tests such as SPT, consolidated drained triaxial test and C levels, very common in the
entirely of sandy soils, defined pressure meter and seismic cross- n Epress = Elastic modulus after chosen area, SPT blow counts
as LEVEL C, with more gravelly hole were carried out to determine pressiometric test were considered sufficient to
subordinate levels. This was the design values for the elastic n E0 SPT = Elastic modulus after mechanically describe materials.
discovered below the cohesive moduli, which are very important Berardi e Lancellotta (1992) Table 5 gives design parameters
level, down to the maximum depth for settlement analysis. Laboratory n Edin = Elastic modulus after cross- which were assessed on the
surveyed, set at 50m below ground oedometer and triaxial tests were hole test basis of the data reported; those
level. also carried on gravelly-sandy adopted were statistically based
The surface of the water table in samples. 3.3 Strength parameters on the results of testing and
the area examined was found to be Table 3 gives a summary of the SPT insitu and direct shear and observation, as well as experience
between 105.09 and 105.90m and values obtained in MPa where triaxial laboratory tests were in underground work in the Milan
therefore at a depth of between n Ed = Elastic modulus after seismic performed to determine the design area and similar conditions. For
19.17m and 19.98m below ground test parameters for cohesion c’ and granular soils Young’s modulus
level, with a direction of flow from n EtxCU = Elastic modulus after angle of friction f’. Table 4 gives a design values were mainly based
north-northwest to south-southeast. consolidated undrained triaxial test summary of the values obtained. on SPT tests; for silty-clay the
For design purposes the altitude of n EtxCD = Elastic modulus after For gravelly and sandy soils, A definition base was substantially

ground engineering april 2013 29

TECHNICAL paper
Table 6: Test column data
Column Construction date Column Column head
length elevation (m)
(m)
C1
C2 2007, January 2
C3
35 125
C4
C5 2006, December 15
C6
JB38 2007, February 26 23 112
Figure 4: Photograph of jet- Figure 5: JB38 core, elevation
JA2 2007, March 1 grouted column 99m-94m

constituted by triaxial tests. grouting columns are constituted Table 7; following experiences in are given in Table 8. With the
In both cases a safety factor was by mixing in situ materials and a similar context this technology exception of column C1, which
applied to minimum values. concrete. Formations involved in was considered optimal to obtain broke during excavation, the results
treatment strongly influence then good results. The delivery pressure showed columns with a diameter
4. Description of ground the results in terms of mechanical of the mixture was assumed to be well in excess of the nominal
improvement operations features of improved soil. 45MPa (450bar). The “bi-fluid” design dimension of 1,500mm.
The design of the tower foundations type mixture injected involved the The geometries were regular,
involved ground treatment 5. Description of the test use of cement type 325, with a water especially for the set C4-C5-C6, for
consisting of “bi-fluid” jet-grouting bed to cement ratio of 1 and a cement which the treatment appeared more
columns. This solution was chosen 5.1 Location and geometry dosage of around 6kN/m3. homogeneous, evidence that the
to control building settlement A test bed was employed for the mixing action occurred correctly.
during the construction and service design of the ground improvement 5.3 Summary of the investigations Columns C1-C2-C3 were less
stages. The ground treatments, columns consisting of two stages. carried out regular, probably because of the
columns with a nominal diameter The first stage of tests was performed After an appropriate period of at presence in the ground of larger
of 1,500mm, were performed in a in December and January 2007 least 10 days to allow the cement to chunks of material, which allowed
quincuncial pattern with sides of with the creation of two sets of harden, a number of investigations the injected grout to take specific
3-3.5m and length varying from 11m three columns labelled C1-C2-C3 were carried out to assess the paths. However, these too showed
at the edges of the slab to 23m under and C4-C5-C6, with a length of effectiveness of the treatments: satisfactory amounts of improved
the central core of the building, as 35m running from ground level, n the tops of the columns were ground. Visual examination of the
described in Figures 2 and 3. 125m approximately. Following uncovered for visual examination heads of the columns led to the
The lengths were calculated initial test surveys, the second and measurement of the geometry; selection of columns C5 and C6 as
according to the distribution of the phase investigated the success of n continuous core drilling was the most representative and suitable
loads bearing on the base of the slab the ground treatment in the deepest performed along the central axis of for specific testing; core drilling was
in an attempt to make settlement layers, within and below the clayey the columns, with a description of then performed along their central
uniform and minimise differential formation. the stratigraphy and photographs; axes.
values. In March and April 2007 two n samples were taken, monoaxial
The design parameters employed columns were constructed, labelled compression tests were performed 6.2 Core samples and geophysical
for the dimensions of the jet- JB38 and JA2, running from and exact weight was measured; investigations
grouting operations were as follows: approximately 112m, with a length n seismic surveys were performed. The core drilling on columns C5 and
n Nominal diameter in the gravelly of 23m, enough to completely cross C6, with a diameter of 131mm, was
and sandy formations: 1,500mm LEVEL B. Table 6 summarises the 6 results of the performed with hydraulic rotation;
n Nominal diameter in the clayey programme. investigations performed it showed good performance down
formation: 1,200mm 6.1 Visual examination of the to a depth of approximately 24-26m
n Ultimate compressive strength: 5.2 Operational parameters ground improvement from ground level (el. 101-99m);
8-10MPa adopted Visual and geometrical examination at greater depths the treatment
n Elastic modulus in the gravelly The operational parameters of the ground improvement was had performed poorly and in some
and sandy formations: 8,000MPa adopted for the construction of performed by uncovering the tops of sections ground improvement was
n Elastic modulus in the clayey the columns using a “bi-fluid” jet- C1, C2, C3, C4, C5 and C6 columns virtually non-existent. In detail,
formation 1,500MPa grouting system, where in addition to a depth of approximately 4m column C5 had good ground
Values of elastic modulus appear to the injection of a binary water- from ground level (Figure 4). improvement down to 24m from
to be different for gravelly/sandy concrete mixture air is injected Measurement of the its head, with average RQD values
and clayey formations because jet- with an annular jet, are given in circumferences and the diameters of around 60-70% and some more
fractured zones (RQD of 20-30%).
Table 7: Operational jet-grouting parameters C6 quality was found to be excellent
down to a depth of 25m, with
Columns Number Nozzle Flow Speed of Rate of Withdrawal Specific markedly higher RQD values,
of Ø (mm) rate rotation withdrawal time energy consistently greater than 70-80%.
nozzles (L/min) (rpm) (cm/min) (sec/4.0 cm) (MJ/m) At greater depths the quality was
lower, especially in the cohesive
C1 1 5.00 280 6 20 12.0 57 stratum (LEVEL B). On the basis
C2 and C3 1 5.50 340 7 23 10.5 60 of these results two new test
columns, labelled JB38 and JA2,
C4, C5 and C6 1 4.50 230 5 15 16.0 61 were constructed with five-inch drill
diameters. The new tests gave good
ground improvement results for the

30 ground engineering april 2013

 Table 8: Results of the geometrical examination of Table 9 - Laboratory test results for samples
the tops of the columns C1 - C6 Column Sample Elevation Compressive Specific
Column Column Circumference Direct Diameter (m) strength (MPa) weight
height (m) measurement deduced from (kN/m3)
(m) of diameter (m) circumference (m)
C5 C5-1A 112.6 9.36 23.56
C1 1.70 1.75 0.45 0.55 C5 C5-1B 112.3 11.36 22.75
C2 3.70 6.70 - 2.13 C5 C5-2A 106.9 9.90 23.46
C3 3.55 6.85 - 2.18 C5 C5-2B 106.6 9.22 22.15
C4 2.25 6.90 2.15 2.19 C5 C5-3A 98.8 9.54 23.66
C5 0.80 6.95 2.10 2.21 C5 C5-3B 98.7 9.88 23.36
C6 2.80 6.10 1.85 1.94 C5 C5-4 97.9 8.88 22.35
C6 C6-1A 109.7 8.97 22.45

47m C6 C6-1B 109.3 9.16 22.65

115.07m
C6 C6-2A 103.4 9.68 23.16
A level
(22m) C6 C6-2B 103.3 8.77 22.89
26m
B level C6 C6-3A 94.7 9.39 21.59
(3m)
C6 C6-3B 94.5 8.66 21.89
C6 C6-4A 92.7 4.22 18.77
C level
(30m) C6 C6-4B 92.5 2.85 19.17
170m JB38 JB38-1 98.6 9.00 -
Figure 6: Calculation scheme JB38 JB38-2 92.5 4.26 -

entire length of the columns (Figure columns C5 and C6 gave values JB38 JB38-3 90.4 4.00 -
5), even in the cohesive stratum and constantly higher than 8.5MPa, JA2 JA2-1 109.0 7.00 -
in the granular deposits below. with an average value of 9.4MPa,
In addition to the direct tests on excluding the samples taken from JA2 JA2-2 98.3 9.60 -
the columns, the study included a 97m where the ground treatment
geophysical investigation with cross- had not been successful. JA2 JA2-3 96.3 8.60 -
hole, topographic and mechanical The tests conducted on the JA2 JA2-4 92.6 8.40 -
admittance tests. In the sands and samples from columns JB38 and
gravels under the water table the JA2 also gave values generally
velocity of the shear waves without higher than 7.0MPa, with an average
treatment generally increased value of approximately 8.6MPa, vertically along the lower edge and eccentricity. Figure 8 clearly shows
according to depth and was between excluding the two anomalous results horizontally on the left and right the presence of maximum loads
300m/sec and 400m/sec. Under the of 4.26MPa and 4.0MPa. The latter sides. The discretisation was kept over the largest of the staircase and
same conditions the compression appeared to be associated with local more concentrated in the central lift structures.
waves were between 1,200m/sec test problems like the presence of upper zone where the stresses were In terms of axis, v represents
and 2,300m/sec. In the improved pebbles or uncemented levels which of greater magnitude and interest. vertical displacement, sz vertical
ground the investigations found an affected the scale of the sample. The The study was developed in stress and x the coordinate also
increase in velocity of between two results of the laboratory tests are stages as follows: shown in Figures 2, 3 and 6.
to two and a half times those in summarized in Table 9. n Stage 1 – Geostatic condition Ground improvement were
the untreated ground; even greater n Stage 2 – Implementation of jet- introduced in FEM analyses
increases were found locally in the 7. FEM analysis grouting and rc slab considering materials with better
more permeable strata. Finally, the The study of the foundation on the n Stage 3 – Application of the mechanical features, in particular
tomographic and admittance tests improved ground was conducted design loads increasing elastic modulus and
confirmed the expected diameter using multiple finite elements Design loads are divided into strength, with the values described
reduction in the clayey level. models. Details are given from a dead loads and working loads, in Chapter 4. The system of
two-dimensional section traced where wind contribution plays an columns were transformed into
6.3 Laboratory tests along the axis of the slab along the important role. At Stage 3 Figures 7, homogeneous materials with
The following laboratory tests were shorter side (Figure 6). 8 and 9 show results with only dead average parameters, different in
performed on the core samples This approach was chosen after loads (PP) or with dead and working case of presence of gravelly/sandy
obtained from drilling: deducing that the shape of the loads in worst conditions (SLE). or clayey formations. On the basis
n calculation of the specific weight; building, rectangular in plan, could Settlements at Stage 2 (Figure 7) are of the tests, these parameters were
n compressive strength by means generate a plane stress state at only related to concrete slab and jet- deduced considering the presence of
of a monoaxial compression test, foundation level, especially when grouting weights greater than those treated material (column) inside the
28 days after the jet-grouting was horizontal loads, wind in particular, of the ground. soil matrix, acting proportionally to
performed. were applied. In Figure 8 geostress is always the volumes.
The results of the laboratory tests The calculation mesh employed included in the vertical stress. The This approach doesn’t take into
relating to compressive strength possessed 2,624 triangular elements cases analyzed, SLE and PP, show account any mechanical direct
gave average values higher than and 1,393 nodes. It runs for 170m similar results, although the stresses interaction between columns
8.5MPa. More specifically, the tests on the horizontal plane and for 55m and settlements are more uniform and untreated soil in which these
conducted on the samples from on the vertical, and is anchored in the second case due to the low elements are realized. From

ground engineering april 2013 31

TECHNICAL paper
X (m) X (m)
0 10 20 30 40 50 0 10 20 30 40 50
0 0

σz (MPa)
V (cm)

0.5 0.1

0.2
1
0.3
1.5
0.4
2
0.5

2.5 0.6 el. 110.07m

Stage 2
Stage 3 el. 112.07m
3 0.7
Figure 8: Results of the PP calculation – Stage 3 vertical stresses
Figure 7: Results of the SLE calculation – vertical movements on the earth-slab contact interface (el. 112.07m) and 2m deeper
upper edge of the slab at Stages 2 and 3 (el. 110.07m)

X (m) X (m)

0 10 20 30 40 50 0 10 20 30 40 50
0

V (cm)
0
V (cm)

PP PP
0.2 0.2
SLE SLE
0.4 0.4 M-data

0.6 0.6

0.8 0.8

1.0 1.0

1.2 1.2

1.4 1.4

1.6 1.6

1.8 1.8
Figure 10: Results of the PP and SLE calculations – vertical
Figure 9: Comparison of the results of the PP and SLE calculations movements, upper edge of the slab in Stage 3 without the effects
– vertical movements, upper edge of the slab in Stage 3 without of the movements relating to the previous stages. Comparison with
the effects of the movements relating to the previous stages monitoring data

an engineering point of view this global behavior. These elements form of jet-grouted columns was ground under the footprint of the
aspect was neglected, because demonstrate how ground treatment chosen. tower was modelled using finite
considered of secondary order and had higher performance, in terms of Firstly, computer simulations element software. The results
surely ameliorating in terms of strain control, than what estimated. were created that allowed us to obtained were compared with
strain reduction of the improved soil The cited mechanical direct choose geometry and features those found after monitoring
foundational system. interaction between columns and that would guarantee a sufficient activity had been performed during
untreated soil surely played a role in reduction in building settlement. construction.
8. Monitoring results this better results. Work then began with the The settlement found matched
The construction was monitored construction of a number of test expectations; it is this aspect that
using a complex and co-coordinated 9. Conclusions columns, in order to define optimum represents the real aim of this work
system during the entire course This paper summarises the operating parameters. Continuity, and it clearly confirmed the validity
of the building work. Particular work performed on the design strength, shape and treated soil of the process, from the design stage
attention was paid to the tower, for the ground improvement volumes were verified to optimize to the creation of the columns, and
observed using a horizontal grid of for the foundations of an office the intervention. from displacement prediction to
50 points, 10 of which were on the building in Milan, Italy. The study, From a computational viewpoint, the effective performance of the
perimeter of the central core. Figure developed in stages, began with the stress-strain behaviour of the system.
10 gives the readings for the vertical the characterisation of the site
movement of four points located in (definition of the geotechnical References
the FEM calculation cross section, model) and with some preliminary
and the values given are for the dimensions. Berardi, R and Lancellotta, R (1992). London, UK, 605-627.
completion of the work. The large size of the building, “Deformabilità dei terreni sabbiosi Lade, P V and Overton, D D (1989).”
They therefore take account with a height of approximately 160 da un’analisi retrospettiva del Cementation effects in frictional
of the application of dead loads, m and 39 floors, required ground comportamento di strutture reali”, materials”, Journal of Geotechnical
Rivista Italiana di Geotecnica, Pàtron Engineering, ASCE, 115 (10) 1373–
permanent loads and a live load improvement due to the loads and Editore, Vol. XXVI (4).
component in a theoretically the need to control settlement. 1387.
intermediate condition between the Indeed, preliminary studies showed Covil, C S and Skinner, A E (1994). Mancinelli, L (2008). “Settlement of
“Jet grouting – A review of some of foundation piles in gravel with water”,
simulated PP and SLE cases. a high degree of settlement related the operating parameters that form the
Examination of the results shows to weight and wind action. These Proc. of 2nd British Geotechnical
basis of the jet grouting process”, Proc. Association Int. Conf. on Foundations,
movements of a smaller size than levels were considered unacceptable of the conf. organised by the Institution AJ Brennan and JA Knappett, Dundee,
predicted, with basic confirmation and, after technologic and economic of Civil Engineers, Thomas Telford, UK, 213-222.
of the orders of magnitude and analyses, ground treatment in the

32 ground engineering april 2013