buildings

Article
A Numerical Study of Reinforcement Structure in Shaft
Construction Using Vertical Shaft Sinking Machine (VSM)
Chunjing Ma 1, *, Haijian Hong 2 , Long Yu 3 , Keqi Liu 4 and Jiadong Huang 2

1 Guangzhou Municipal Engineering Group Co., Ltd., Guangzhou 510060, China

2 Guangzhou Municipal Construction Group Co., Ltd., Guangzhou 510060, China
3 Shanghai Tunnel Engineering Rail Transit Design and Research Institute, Shanghai 200235, China
4 School of Resources and Civil Engineering, Northeastern University, Shenyang 110819, China
* Correspondence: machunjinglw@163.com

Abstract: Using the Vertical Shaft Sinking Machine (VSM) for shaft construction is an emerging
modern technology, which is also currently one of the most advanced techniques in the field of
shaft sinking. Current research on VSM technology primarily focuses on mechanical and technical
issues, neglecting the impact of the construction on the surrounding soil and the structure itself. This
oversight leaves structural design lacking a reliable foundation. Additionally, there is insufficient
information on the role of reinforcement design during construction in soft soils. These engineering
challenges have hindered the widespread implementation of this new technology. Therefore, a series
of numerical models were used to analyse the mechanical behaviour of shaft and soil during or
after the sinking process, with the aim of addressing these gaps by investigating the influence of
shafts constructed using VSM technology, and providing a scientific basis for reinforcement design in
soft soil. The case study shows that increasing the soil-cement strength does not have a significant
effect on the overall deformation of the soil surrounding the shaft, but leads to a notable reduction
in the plastic zone volume, subsequently enhancing the overall stability of the neighbouring soil.
The ring bottom beam design effectively reduces convergence deformation by about 50%, while also
improving the horizontal internal force distribution near the cutting edge. However, this approach
significantly escalates local vertical bending moments, necessitating thorough consideration in the
design stage.
Citation: Ma, C.; Hong, H.; Yu, L.; Liu,
K.; Huang, J. A Numerical Study of Keywords: vertical shaft sinking machine; underground parking garage; structural design; assembly
Reinforcement Structure in Shaft sinkhole; numerical simulation
Construction Using Vertical Shaft
Sinking Machine (VSM). Buildings
2024, 14, 2402. https://doi.org/
10.3390/buildings14082402
1. Introduction
Academic Editor: Eric M. Lui The current parking problems in major cities around the world are primarily mani-
fested in a huge gap in urban parking spaces, insufficient total quantity, prominent struc-
Received: 4 July 2024
Revised: 30 July 2024
tural contradictions in urban parking facilitiess, and backward urban parking management.
Accepted: 1 August 2024
In order to meet the growing demand for parking lots, there is a need for more intensive
Published: 3 August 2024 land use, as well as more effective parking options and methods for utilising urban space.
As a solution, the shaft sinking technology for building underground parking garages has
the characteristics of low ground interference, low construction site area requirements, and
minimal overall cost [1]. Therefore, the mechanized construction for rapid shaft sinking
Copyright: © 2024 by the authors. has gained increased attention [2,3].
Licensee MDPI, Basel, Switzerland. For the construction of underground shafts, the general method involves on-site
This article is an open access article pouring of concrete to make the shaft wall, and uses excavators and grabbing machines for
distributed under the terms and excavation and slag removal, and sinking is achieved by overcoming the friction between
conditions of the Creative Commons the shaft wall and the soil layer through the weight of the structure. This technology
Attribution (CC BY) license (https://
is suitable for various cross-sectional forms and is relatively mature. However, slow
creativecommons.org/licenses/by/
construction efficiency, poor stability, susceptibility to sudden sinking, stop sinking, etc., as
4.0/).

Buildings 2024, 14, 2402. https://doi.org/10.3390/buildings14082402 https://www.mdpi.com/journal/buildings

Buildings 2024, 14, 2402 2 of 25

well as a significant impact on the surrounding environment are among its shortcomings [4].
The pneumatic caisson method has a relatively high degree of mechanization, but it cannot
fundamentally solve the problems of difficult sinking and slow sinking speed; especially
when the cutting edge encounters isolated hard rocks, forcibly increasing the downward
pressure can easily cause the attitude of the shaft to tilt and sink suddenly [5].
With advancements in engineering technology and increasing practical engineering
demands, shaft sinking has been developing towards larger cross-sections and greater
depths. The shortcomings of traditional self-sinking shafts have become increasingly
evident, leading to a sharp decline in their use since the 1990s. They have been replaced
by more rational methods such as the Space System (SS) caisson method [6], the press-in
caisson method, and the Super Open Caisson System (SOCS) method [7]. In the SS method,
the outer contour line of the cutting edge extends approximately 15–20 cm beyond the
boundary of the shaft wall. This design creates a gap between the shaft’s outer wall and the
surrounding soil during sinking. The space gravel is filled in the gap in order to reduce the
peripheral face frictional resistance without compromising stability of the natural ground.
Unlike the self-sinking method, a new shaft sinking method that utilises ground anchor
reaction devices to press the shaft into the ground has emerged, known as the press-in
caisson method [8]. This method requires the addition of a ground anchor reaction press-in
device, and employs through-hole jacks to apply downward pressure on the top of the
shaft at the surface, addressing the issues of difficult and slow sinking. Additionally, by
adjusting the downward pressure at different positions, the shaft’s attitude can be adjusted
and corrected, making it suitable for construction projects with stringent environmental
control requirements. Based on the press-in method, a newer sinking technique is the Super
Open Caisson System (SOCS), which was developed in Japan [9]. It uses an underwater
excavator mounted on the inner shaft wall to dig the soil below the footing to the centre
of the shaft, and then discharges the sludge by means of a grapple or a slurry pump,
thus reducing the resistance to sinking of the cutting edge, and then the shaft is pressed
into the ground by the ground anchor counter-force device. The SOCS method is well-
suited for construction projects in urban areas with high-speed construction requirements
and stringent environmental controls. Engineers have also improved the underwater
excavator, developing a new type of excavator suitable for non-standard cross-section
shafts, significantly broadening the application scope of the SOCS method.
Compared to these techniques, shaft sinking using the Vertical Shaft Sinking Ma-
chine (VSM) is a more advanced sinking method proposed using Herrenknecht AG in
Germany [10,11]. The excavation machinery is has become intelligent and digitised, fea-
turing a cutting drum to efficiently slice through the soil [12,13]. The crushed soil is then
mixed with water and pumped out, thereby enhancing the overall excavation process. The
shaft wall is prefabricated by concrete lining segments, and over excavation is carried out
around the wall, and the gap is filled with bentonite slurry to reduce resistance. A low-
ering unit is used to control the wall movement, which prevents sudden sinking of the
shaft. The VSM method incorporates the shield tunnelling lining construction technique
to create a new type of prefabricated shaft wall [14]. Regarding friction reduction, both
the VSM and SS methods are similar, employing the injection of bentonite slurry around
the exterior of the shaft wall to reduce resistance [15]. The key distinction between the
VSM method and traditional sinking methods lies in the sinking process: with the VSM
method, the shaft wall is supported by a sinking unit, preventing it from sinking during
underwater excavation. This enables precise and active control of the sinking process by
adjusting the suspension force. The method has the advantages such as deep excavation
depth, small construction disturbance, fast construction speed, minimal site usage, high
economic benefits, and broad geological applicability. Combining the characteristics of
prefabricated construction, intelligent construction, and environmentally friendly construc-
tion processes, the market is developing rapidly. At present, this method has been used
in many countries and regions for practical engineering applications, and has achieved
favourable outcomes [12,16–18]. Some cases are shown in Table 1. Although there are
Buildings 2024, 14, 2402 3 of 25

many cases of using VSM technology to construct starting shafts or ventilation shafts, the
successful construction of a parking garage in Nanjing, China, and the engineering case
presented in this article fully demonstrate the feasibility of this technology in parking
garage construction.

Table 1. Examples of VSM application.

Location Diameter/(m) Depth/(m) Geological Condition Functionality

Neapel, Italy 4.5 50 Clay, silt, sand, gravel ventilation
Loamy coarse-grained clay, sandy clay,
Girona, Spain 6–10 20 metro
marl, driftstone
Dortmund, Germany 9 23 Sand, silt, marl tube jacking tunnel launch
St. Petersburg, Russia 8.4 85 Clay, sand, driftstone sewerage
Moscow, Russia 6.4 30–73 Sand, lime, clay ventilation
Seattle, USA 9.8 45 Clay, silt, sand, gravel shield tunnel launch
Hawaii, USA 10.8 36.2 Coral, basalt harbor
Sand, colluvium, clayey sandy soil,
Kuwaiti 8.8 15–27 tube jacking tunnel launch
limestone
Jeddah, Saudi 11 45 Soft soil, clayey sandy soil, sand, gravel sewerage
Singapore 11.2 57 Limestone, sandstone, mudstone sewerage
Nanjing, China 12.8 65 Clay, silt, sandstone underground parking
Shanghai, China (Case study) 22.6 50.5 Mucky clay, silty clay, medium silty sand underground parking

In recent years, based on the design concepts, system composition, and functional
configuration of VSM, several companies, particularly Chinese enterprises, have under-
taken modifications and upgrades to the VSM and optimised the related construction
methods [19]. China Railway Construction Heavy Industry Corporation (CRCHI) devel-
oped three different product series of VSM based on excavation diameter, with the largest
being the 23-meter level which was employed in the studied project in 2023. Additionally,
China Railway Engineering Equipment Group (CREG) developed the dry mucking shaft
sinking machine, CJM. in contract, CITIC Heavy Industries has developed the multi-arm
shaft sinking machine, JSB [20,21]. The difference between the CJM and the aforementioned
equipment lies in the utilisation of a free-profile cutting head, enabling excavation to be
prformed in both dry and underwater situations, and a mechanical grab for muck disposal,
while the JSB uses a toothed roller cutter combined with a rotating and revolving cutting
head for excavation, with a muck disposal method similar to that used by the VSM. In
addition, Shanghai Tunnel Engineering Co., Ltd. in Shanghai, China developed the APM,
which features a more flexible cutting arm, and the shaft wall is made to sink by actively
pressing down. These advancements reflect the ongoing research and development in
VSM technology. Neye et al. [3] described the role of VSM in mechanical shaft sinking
technology and presented the associated equipment components and their performance.
Schmäh et al. [22,23] described different projects that have been constructed using the VSM
method, demonstrating the potential applications of VSM and highlighting the different
requirements that should be met by the VSM technique. Rahimi et al. [24] highlighted
geotechnical considerations when VSM is combined with conventional drilling and blasting
methods for excavation. Zhou et al. [17] took the Zhuyuan Bailonggang Sewerage Project
in Shanghai as an example; through a comparative analysis of on-site measurements and
numerical simulations, it was concluded that the maximum ground settlement caused by
the VSM was 15.2 mm, and the maximum horizontal displacement was 3.74 mm. The
influence of the shaft excavation on the ground settlement was about 30 m from the centre
of the shaft.
Previous research on vertical shafts has mainly favoured theory, numerical analysis,
monitoring or experiments regarding the general method [25–33]. However, there are
currently few systematic presentations of new shaft technology based on actual projects [16].
Although there are a number of practical cases of shaft sinking using VSM, there is still a
lack of research on the mechanical responses it induces and the effectiveness of the support
structures, and even fewer studies on the detailed simulation based on this technique for
Buildings 2024, 14, 2402 4 of 25

large-diameter shaft sinking in soft ground. Similarly to conventional methods in terms

of load-bearing, the VSM method determines its specific load-bearing characteristics and
structural response as shown in Table 2.

Table 2. Comparison of load-bearing characteristics of different construction methods.

General
Load Type SS Press-in SOCS CJM VSM
Method
Lateral earth
Lateral earth
Lateral earth Lateral earth and water
and water
and water Lateral earth and water pressure,
Sinking pressure,
Lateral earth pressure, and water pressure, slurry pressure,
process: slurry pressure,
Lateral load and water lateral gravel pressure internal water internal water
equipment
pressure pressure pressure pressure, strut
load
shoe load
drained
Lateral earth and water pressure
condition:
Lateral friction,
Rolling friction Lateral friction,
Lateral friction, cutting edge
resistance of cutting edge
cutting edge Lateral friction, cutting edge resistance,
Sinking resistance gravel, cutting resistance,
resistance, resistance, buoyancy buoyancy,
edge resistance, suspension
buoyancy suspension
buoyancy force
force
Self-weight
Self-weight
Sinking load Self-weight Self-weight (+press-in
(+press-in force)
force)
Vertical load after completion Buoyancy, self-weight, lateral friction, uplift pile resistance (if any)

This paper will take an underground parking garage project in Shanghai as a case
study, and introduce the application of the VSM technology and some types of support
structures adopted in this project for large-diameter shaft sinking in soft ground. In addi-
tion, the changes in ground deformation and mechanical responses induced by excavation
are computed, and the analysis will provide references and suggestions for future simi-
lar projects.

2. The Application of VSM and Structural Design in the Case Study

This section will briefly introduce the composition of equipment and construction
process with VSM, focusing on the specific execution process in the underground parking
garage project in Shanghai, as well as the structural design for reinforcement for the shaft
construction, including the concrete ring foundation, strata reinforcement, ring bottom
beam and anti-float piles in the water-rich soft soil.

2.1. Project Overview

The shaft being constructed in Shanghai features a large diameter and incorporates
VSM technology. The project comprises two shafts, with the first one already completed
as of July 2023. Each shaft measures 23.02 m in diameter, with a final excavation depth
of approximately 50.5 m and a base plate depth of 44 m. The underground space spans
836 m2 , accommodating a 19-layer parking garage with a total of 304 parking spaces.
The shaft wall traverses various strata, including miscellaneous fill soil (2.7 m thick),
clayey silt (3.3 m), muddy clay (10.2 m), silty clay (26.3 m), silty clay interspersed with
sandy silt (6.6 m), and a 7.5 m interlayer between silty clay and clayey silt. Predominantly,
the area beneath the shaft bottom consists of silt and fine sand, indicating that all excavated
soil layers were soft.

2.2. Components of VSM

The VSM system comprises several key units: the cutting unit, the lifting unit, the
lowering unit, and the separation unit. Figure 1 illustrates the setup for the case study.
 sandy silt (6.6 m), and a 7.5 m interlayer between silty clay and clayey silt. Predominantly,
the area beneath the shaft bottom consists of silt and fine sand, indicating that all exca-
vated soil layers were soft.

Buildings 2024, 14, 2402 2.2. Components of VSM 5 of 25

The VSM system comprises several key units: the cutting unit, the lifting unit, the
lowering unit, and the separation unit. Figure 1 illustrates the setup for the case study.
Excavationisisexecuted
Excavation executedthroughthroughaacutting
cuttingdrum
drummounted
mountedon onthe
thetelescopic
telescopicarm armof ofthe
themain
main
machine,which
machine, whichmills
millsand andbreaks
breaksthe thesoil.
soil. If
If necessary,
necessary,over-excavation
over-excavationcan canalso
alsobe becarried
carried
out.This
out. Thismethod
methodisisideally
ideallysuited
suitedforforsoft
softsoil
soil oror soft
soft rock
rock environments,
environments,but butititis
is important
important
toconsider
to considerincreased
increasedtool toolconsumption
consumptiondepending
dependingon onthethesoil’s
soil’scompressive
compressivestrengthstrengthand and
abrasiveness[23].
abrasiveness [23].
The VSM
The VSM used
used in in this
this case,
case, developed
developed by by CRCHI,
CRCHI, has has anan overall
overall height
height of of approxi-
approxi-
mately
mately 1010 m mand
andcancanexcavate
excavatevertical
vertical shafts
shafts withwith aa diameter
diameter ranging
ranging 18 18 mmto to23 23m.m.The
The
maximum
maximumexcavation
excavationdepth depthcan canreach
reachup upto to6060m, m,with
withan anexcavation
excavationrate rateof
ofup upto to44m/d.
m/d.
The
The main
main machine has has aapower
poweroutput
outputofof2400 2400 kW. kW. It suitable
It is is suitable for for
softsoft
soil soil
and and soft
soft rock
rock
stratastrata
(with(with a compressive
a compressive strength
strength of less
of less thanthan 80 MPa)
80 MPa) andand cancan adapt
adapt to to a variety
a variety of
of complex
complex geological
geological conditions
conditions withwith different
different working
working modes,
modes, makingmaking it effective
it effective for
for both
both dryhigh-water-pressure
dry and and high-water-pressure conditions.
conditions. Due to Due to its
its large large diameter,
diameter, it is equipped
it is equipped with two
with two cutting arms operating simultaneously, which enhances
cutting arms operating simultaneously, which enhances construction efficiency compared construction efficiency
compared to traditional
to traditional methods. The methods.
cuttingThe cutting
process canprocess can be conducted
be conducted underwater, underwater,
eliminating elim-
the
inating
need tothelowerneedthetowater
lowertable.
the water table. This
This approach approach
facilitates facilitates
balancing balancing
internal and internal
external
and
waterexternal water
pressure, pressure,
ensures ensures
structural structural
stability, stability,
controls soil controls
deformation soil deformation
and is environmen- and is
environmentally
tally friendly. friendly.
For
For this project, the
this project, thesystem
systememploys
employsfive five lifting
lifting units
units located
located directly
directly above above five
five sup-
support arms to raise and lower the mainframe. Additionally, twelve
port arms to raise and lower the mainframe. Additionally, twelve lowering units are used lowering units are
used to suspend the shaft structure, and the overall setup includes
to suspend the shaft structure, and the overall setup includes two winch towers, as de- two winch towers, as
depicted in Figure
picted in Figure 1. 1.

Buildings 2024, 14, x FOR PEER REVIEW 6 of 27

VSM lifting Machinery installation

Figure1.1. VSM
Figure VSM set-up
set-up for
for the
thestudied
studiedproject.
project.

2.3. Structural Design for Reinforcement

Given the large excavation diameter of the vertical shaft and the soft strata, reinforce-
ment measures are crucial to manage soil and structural deformation, thereby bolstering
safety and structural integrity throughout construction.
1. Concrete ring foundation
The project’s foundation has been meticulously devised to withstand heavy machinery
and various stresses encountered while sinking shafts. To significantly increase the bearing
capacity of the delicate soil, a hybrid foundation approach has been implemented (Figure 2).
 ment measures are crucial to manage soil and structural deformation, thereby bolstering
safety and structural integrity throughout construction.
1. Concrete ring foundation
Buildings 2024, 14, 2402 The project’s foundation has been meticulously devised to withstand heavy machin- 6 of 25
ery and various stresses encountered while sinking shafts. To significantly increase the
bearing capacity of the delicate soil, a hybrid foundation approach has been implemented
(Figure
A 2). Aring
concrete concrete ring foundation,
foundation, poured
poured from from the
the surface surface
down to 3down to 3 m
m depth, depth,
offers offers
a robust
a robust base for the core components of the VSM, ensuring the precise descent
base for the core components of the VSM, ensuring the precise descent of the shaft. Before of the
shaft. Before erecting the ring foundation, triple-axis mixing piles reinforce
erecting the ring foundation, triple-axis mixing piles reinforce the soil beneath it. Thisthe soil be-
neath it. This mitigates
reinforcement reinforcement mitigates
uneven uneven
settlement risks settlement risks during
during construction and construction and
elevates the soil’s
elevates the soil’s
load-bearing load-bearing
capacity. The rangecapacity. The range of soil
of soil reinforcement reinforcement
extends extends
is 3 m radially andishas3m a
radially
depth ofand
21.5hasm. a depth of 21.5 m.
2. Shaft
Shaft wall
wall and
and ring
ring bottom
bottom beambeam
Due to the soft strata and extensive excavation diameter, the shaft wall segments in
this project feature a relatively thick design, determined as 800 mm through engineering
analogies and design calculations considering the construction loads and and various
various actions.
actions.
Given the
the lengthy
lengthy circumference,
circumference,each eachring
ringconsists
consistsofof1010concrete
concrete segments,
segments, each
each2m 2m in
height
in andand
height weighing
weighing 27.427.4
tons. Using
tons. Usingprecision-processed
precision-processed steelsteel
cutting edges,
cutting thesethese
edges, seg-
ments areare
segments assembled
assembled into rings,
into rings,filled
filledwith
withself-compacting
self-compactingconcrete,
concrete,and and then
then stacked
vertically up to 4.5 m from the opening. The The uppermost
uppermost section
section of of 4.5
4.5 mm employs
employs aa cast-
cast-
in-place shaft wall. To mitigate resistance and facilitate sinking, the gap between the shaft
wall and soil layer is filled with bentonite slurry. slurry.
To enhance
enhance the rigidity
rigidity of thethe shaft
shaft wall
wall near
near the bottom,
bottom, a reinforced ring bottom
beam is incorporated,
incorporated, as asdepicted
depictedininFigureFigure2a.
2a.Within
Withinthethethree
three prefabricated
prefabricated rings,
rings, an an
80
80
cmcm thick
thick reinforced
reinforced concrete
concrete sidewall
sidewall was poured,
was poured, andcentral
and the the central ring is
ring beam beam
linkedis linked
to this
to this sidewall
sidewall via connecting
via connecting pillars.
pillars. The The coreofbenefits
core benefits of this
this design design are
are reducing thereducing
thicknessthe of
thickness of the base concrete, providing structural reinforcement to the
the base concrete, providing structural reinforcement to the lower shaft wall (exposed to in- lower shaft wall
(exposed to increased
creased soil pressure), soil pressure),
reducing wall reducing walland
deformation, deformation,
enhancingand the enhancing
stability of the
the stability
connec-
of the connection between the plain concrete for bottom
tion between the plain concrete for bottom sealing and the shaft wall. sealing and the shaft wall.

Lifting Unit Concrete Ring

Energy Winch Foundation
Tower

Lowering
Unit

Shaft
Wall

Ring Bottom
Beam
Pile

(a) (b)

Figure 2. Structural design (a) side view; (b) top view.

3. Piles
The comprehensive anti-float design comprises three key components: the shaft wall,
the top ring foundation, and the pile foundation. At the base of the ring foundation and
surrounding the vertical shaft, 20 piles with a 1.2 m diameter, approximately 70 m length,
and with 3.9 m spacing are strategically arranged, as depicted in Figure 2b. The pile tops are
securely embedded within the top concrete foundation, with a embedment depth of 1.5 m.
Buildings 2024, 14, 2402 7 of 25

These piles serve the roles in bearing a portion of the vertical load, resisting lateral soil
pressures and counteracting buoyancy forces. Once the shaft is sealed and water is pumped
out, the structural self-weight helps mitigate buoyancy, while the remaining buoyancy will
be counteracted through the combined action of the top ring foundation, the shaft wall and
the piles.

3. Methods
3.1. Model Development
Numerical modelling is conducted considering the geological conditions, incorpo-
rating the model geometry with the specific construction characteristics and structural
attributes. For this purpose, the Lagrangian finite difference software FLAC3D 6.0 is em-
ployed to conduct the numerical analysis.
To minimize boundary effects, the model width is set at 120 m, approximately six times
the
Buildings 2024, 14, x FOR PEER REVIEW inner radius of the shaft. Given the shaft’s structural features and pile arrangement, a
9 of 27
symmetric model is adopted for simulation. The domain height is 104 m, twice the final
sinking depth, as depicted in Figure 3a,b.

(a) (b) (c)

Figure 3.
Figure 3. Geometry and reinforcement
Geometry and reinforcement structures
structures (a)
(a) side
side view;
view; (b)
(b) top
top view;
view; (c)
(c) support
support structures.
structures.

3.2. Mesh and Boundary

The shaft’s depth is 51 m, with an inner wall diameter of Φ21 m. Each
Conditions
total sinking
concrete
The ring
model is 2ism high andof
composed 0.8111,688
m thick. The
solid ring foundation
hexahedral elements.measures 22.6 m near
The elements in inner
the
diameter,
excavation3 m in width,
area and theand 3 mwall
shaft in height. The lower
are relatively zonewith
dense, reinforced
length,by triple-axial
width mixing
and height of
piles
aboutmaintains
1 m. The the
mesh same
sizethickness
increasesas atthe ringoffoundation,
a rate 1.1 times outsideextending
this 14 m in
area. height.
Piles Both
are simu-
the
latedring foundation
using structural and the reinforced
elements, zonepile
with each arecomprising
impermeable. 34 elements. The modelling of
structural elements is not restricted by the specificwhile
The ground surface elevation is set at 0 m, positionthe ofexcavated
the solidzone’s bottom
elements, which is
positioned at − 52 m. The piles, with a diameter of 1.2 m, are located
allows for easy group pile analysis. The interaction between the pile and the solid ele- between depths of
− 2 m and
ments −70 m. through
is achieved In this model,
coupled 11springs.
piles areThese
arranged,
coupled spaced 3.7 are
springs m apart, withand
nonlinear the slid-
pile
centre positioned 1.6 m away from the shaft wall’s outer edge. To simplify
able connectors that can transfer forces and moments between the pile nodes and the solid the complexity of
the naturally curved excavation surface, it is modelled as a plane in
elements. The tangential springs function similarly to the friction mechanism of grouted this study. Furthermore,
throughout
anchors. Thethe sinking
normal process,
springs arethe
usedvertical distance
to simulate thebetween
effect ofthe bottom
normal surface
loads, and and the
the for-
cutting edge remains fixed at 1 m.
mation of gaps between the pile nodes and the solid elements can be simulated, as well as
Before commencing
the squeezing effect of the the sinking process,
surrounding soil onthethe
initial
piles.support structure comprises a 1 m
tall cutting edge, a 9 m high shaft wall, and
A fixed boundary condition is imposed at the bottom, a ring bottom and beam (Figure 3c).
horizontal A sidewall
displacements
1are
m constrained
thick and 6 m forhigh and vertical
all four five pillars connect the
boundaries. ringthe
Since bottom
shaft beam to the shaft
wall remains in a wall,
sus-
with the sidewall’s bottom aligned with the top of the cutting
pended state throughout the construction process by means of wire rope traction, the edge. The initial support
sink-
ing of the shaft wall is precisely controlled by regulating the cutting edge’s velocity, with
a cycle sinking depth of 2 m.

3.3. Soil–Structure Interaction

In practice, a 10 cm gap between the shaft wall and the soil is filled with bentonite or
Buildings 2024, 14, 2402 8 of 25

structure and excavation face commence at an elevation of −6 m, marking the beginning of

the first step in a total of 24 sequential excavation phases. These steps gradually lower both
structures until they jointly reach the final depth of −51 m. Each sinking cycle involves
a depth of 2 m, necessitating the installation of a new concrete ring atop the shaft wall to
maintain a minimum of 2 rings (4 m) above the ground surface. Once the shaft wall reaches
its final depth, it aligns with the ground surface. The thickness of the plain concrete for
sealing the bottom is 8 m, and the final depth is 52 m.

3.2. Mesh and Boundary Conditions

The model is composed of 111,688 solid hexahedral elements. The elements near the
excavation area and the shaft wall are relatively dense, with length, width and height
of about 1 m. The mesh size increases at a rate of 1.1 times outside this area. Piles are
simulated using structural elements, with each pile comprising 34 elements. The modelling
of structural elements is not restricted by the specific position of the solid elements, which
allows for easy group pile analysis. The interaction between the pile and the solid elements
is achieved through coupled springs. These coupled springs are nonlinear and slidable
connectors that can transfer forces and moments between the pile nodes and the solid
elements. The tangential springs function similarly to the friction mechanism of grouted
anchors. The normal springs are used to simulate the effect of normal loads, and the
formation of gaps between the pile nodes and the solid elements can be simulated, as well
as the squeezing effect of the surrounding soil on the piles.
A fixed boundary condition is imposed at the bottom, and horizontal displacements
are constrained for all four vertical boundaries. Since the shaft wall remains in a suspended
state throughout the construction process by means of wire rope traction, the sinking of
the shaft wall is precisely controlled by regulating the cutting edge’s velocity, with a cycle
sinking depth of 2 m.

3.3. Soil–Structure Interaction

In practice, a 10 cm gap between the shaft wall and the soil is filled with bentonite
or cement slurry to adjust the frictional resistance accordingly. In this model, an inter-
face element is utilised to simulate the lubrication effect between the shaft wall and the
surrounding soil. The constitutive model is defined by a linear Coulomb shear-strength
criterion that limits the shear force acting at an interface node. By default, pore pressure is
used in the interface effective stress calculation. Normal forces are calculated through the
nodes and distributed on the target surface, while shear forces on the opposite surface are
connected to the nodes. These forces are distributed to the nodes of each surface through
weighted averaging. The nodal stress is assumed to be uniformly distributed over the
representative area. Each node is associated with contact surface parameters. The interface
stiffness and shear strength are initially set to low values, which allows for the simulation
of force transmission and sliding effects.
The normal and shear stiffness of the interface are set to 4.1 MPa and 2.1 MPa, respec-
tively. The cohesion is 2 kPa, and the internal friction angle is 5◦ .

3.4. Structural Parameters

1. Concrete structures
The load applied to the top ring foundation is determined by subtracting the buoyancy
of the entire structure from the combined gravity of the equipment units, the main machine,
the shaft wall, and the ring bottom beam. This load is then uniformly distributed across the
corresponding load areas shown in Figure 3b.
After calculations, the results indicate that the stress imposed on the load area of the
lifting unit is 28.3 kPa. The load area of the lowering unit experiences a load of 552 kPa,
with an incremental increase by 48.7 kPa for each sinking step. Additionally, the load area
of the power winch tower bears a load of 178 kPa.
Buildings 2024, 14, 2402 9 of 25

Steel sheet piles surround the outer edge of the ring foundation, standing at a height
of 12 m and modelled using shell elements. Considering the presence of water in the shaft,
a normal gradient pressure related to the elevation is applied vertically on the inner surface
of the shaft wall. Similarly, normal pressure is applied to the surface of the ring support
beam and the excavation face.
The reinforced zone is modelled using the Mohr–Coulomb model, a widely recog-
nized elastic-plastic model for simulating soil behaviour. The soil strata in the simulation
primarily comprise 18 m of mucky soil, followed by 33 m of silty clay, and 8 m of clayey
silt, culminating in a fine sand layer at the bottom. To accurately represent the structural
components, the shaft wall, ring foundation, and ring support beam are simulated with
the elastic model, while the pile foundation is modelled using a combination of structural
and pile elements. The selected key parameters for these components are given in Table 3,
based on previous studies [34,35]. A noteworthy aspect is the consideration of buoyant
density of the concrete located below the water table. This buoyant density is calculated
by subtracting the density of water (1000 kg/m3 ) from the density of concrete, effectively
accounting for the buoyancy effect in the simulation.

Table 3. Parameters of structures.

Structure Shaft Wall Ring Foundation Reinforced Zone Ring Support Beam Plain Concrete
Bulk modulus, K/[MPa] 1.92 × 104 1.8 ×104 314 1.8 × 104 1.75 × 104
Shear modulus, G/[MPa] 1.44 × 104 1.35 × 104 235 1.35 × 104 1.3 × 104
Cohesion, c/[kPa] - - 177 - -
Friction angle, φ/[◦ ] - - 28 - -
Density, ρ/[kg/m3 ] 2500 2500 1900 2500 2350

The parameters of the pile elements are derived from a circular section with a diameter
of 1.2 m, and the normal stiffness is calculated according to the following equation [36]:

K + 4G/3
kn = (1)
∆zmin

where ∆zmin [m] is the minimum width of the element in the vicinity of the pile.
The shear stiffness ks [kPa/m] is taken as follows:

k s = 1000 × (21.8 + 0.438σn ) (2)

where σn [kPa] is the normal stress from the surrounding soil layer. The cohesion and the
friction angle of the pile–soil interface are taken as 0.8 times that of the surrounding soil.
2. Soils
• Incremental elastic law
The soil structure is appropriately simplified using the HS model, i.e., plastic harden-
ing soil model, to better describe the nonlinear and stress-dependent deformation behaviour
of soft or hard soils prior to failure [37]. It is formulated within the framework of hard-
ening plasticity, allowing the removal of the main drawbacks of the original non-linear
elastic model formulation, and is well-established for soil–structure interaction problems,
excavations, and settlements analysis, among many other applications [38].
The HS mode adopts hypo-elasticity for the description of elastic behaviour,

∆p = −K∆ϵve (3)

∆sij = 2G∆ϵij (4)

Buildings 2024, 14, 2402 10 of 25

where K and G are the elastic bulk and shear moduli, which can be derived from the elastic
unloading-reloading Young’s modulus, Eur :

Eur
K= (5)
3(1 − 2v)

Eur
G= (6)
2(1 + v )
where the Young’s modulus is a stress-dependent parameter:
re f
Eur = Eur Z m (7)

where
ccotφ − σ3
Z= > f cut (8)
ccotφ + pre f
re f
and Eur is the reference unloading-reloading stiffness modulus at the reference pressure
pre f . The current unloading-reloading stiffness modulus Eur depends on the maximum
(minimum compressive) principal stress, σ3 , the cohesion, c, and the ultimate friction angle,
φ, as well as the power, m. For clays, m is usually close to 1. For sands, it is usually between
0.4 and 0.9. The default cut-off factor fcut is 0.1.
The HS model also employs an additional stiffness measure, E50 , which defines the
initial slope of the hyperbolic stress–strain curve. Parameter E50 obeys the following
power law:
re f
E50 = E50 Z m (9)
re f
where E50 is a material parameter, which could be estimated from a set of triaxial compres-
sion tests with various cell stresses.
• Shear yield criterion
The shear yield function determining the onset and development of shear hardening
is defined as
Eur qa q Eur γ p
fs = · −q− =0 (10)
Ei (q a − q) 2
 
where γ p is a shear hardening parameter (one of the internal variables), Ei = 2E50 / 2 − R f ,
deviatoric stress q = σ3 − σ1 , and q a is given as
qf 2sinφ
qa = = (ccotφ − σ3 ) (11)
Rf R f (1 − sinφ)

where the failure ratio R f = q f /q a has a value smaller than 1 (typically, R f = 0.9 is used),
and q f is the ultimate deviatoric stress, which is consistent with the Mohr–Coulomb failure
criterion (Figure 4). For a standard drained triaxial test, the connection between the axial
(vertical compressional) strain, ε1 , and deviatoric stress, q, is graphically represented in
Figure 4 with the cut-off qf .
• Volumetric cap criterion
The volumetric (cap) yield function is defined as

∼2
q
fv = + p2 − p2c = 0 (12)
α2
∼
q = [σ1 + (δ − 1)σ1 − δσ3 ] (13)
δ = (1 + sinφ)/(1 − sinφ) (14)
Buildings 2024, 14, 2402 11 of 25

where α is a constant based on a virtual (numerical) oedometer test. The evolution of the
hardening parameter, pc , is given by the relation:
 m
ccotφ + pc
∆pc = kK p ∆γv (15)
ccotφ + pre f

where γv is the volumetric hardening parameter, k is a correction factor with a typical value
re f re f
of 0 < k ≤ 1, K p = K1 K2 /(K1 −K2 ), K1 = Eur /[3(1 − 2ν)], and K2 = Eoed (1 + 2Knc )/3.
re f
Parameter Knc denotes the normal consolidation coefficient, and Eoed stands for the tangent
oedometer stiffness at the reference pressure pre f .
Buildings 2024, 14, x FOR PEER REVIEW 12 of 27
• Tensile yield criterion
The model checks for the tension failure condition. Tension failure and potential
functions
where the are
failure ratio 𝑅 = 𝑞 /𝑞 has avvalue smaller than 1 (typically, 𝑅 = 0.9 is used),
t
f = −
and 𝑞 is the ultimate deviatoric stress, which is consistent with the Mohr–Coulomb fail-
σ3 σ (16)
ure criterion
where (Figure
σt is the 4).limit.
tension For aBy standard
default,drained triaxial
σt is zero, test,user
and the thecanconnection
provide between the
a value with
axial (vertical
an upper limitcompressional) modelεdoes
of c/tanφ. Thestrain, 1, andnot
deviatoric
considerstress,
tensionq, is graphically represented
hardening.
in Figure 4 with
The key the cut-off
parameters forqf.the soils in the case study are given in Table 4.

Hyperbolic Asymptotic Line

qa
Hyperbolic Line
qf
Mohr-Coulomb Failure Line
E50

qf /2
Eur

ε1
Figure
Figure 4.
4. Hyperbolic
Hyperbolicstress–strain
stress–strain relation
relation in
in primary
primary shear
shear loading.
loading.

4. Key parameters
•TableVolumetric of the soils.
cap criterion
SoilsThe volumetric (cap) yield function is defined
Mucky Clay asSilty Clay Clayey Silt Fine Sand
Reference stress, pref /[kPa] 𝑞100 100 100 100
𝑓 = 0.3+ 𝑝 − 𝑝 = 0.3 0 (12)
Poisson’s ratio, ν/[MPa] 𝛼 0.3 0.25
Cohesion, c/[kPa] 8 24 11.8 4
Friction angle, φ/[◦ ] 𝑞 = 𝜎 + (𝛿 − 1)𝜎 − 𝛿𝜎
11 21 28 34 (13)
Dilatancy angle, Ψ/[◦ ] 0 0 0 4
re f
Secant Young’s modulus at reference stress, E50 /[MPa] 8.3 18.9 13.8 41.1
𝛿 = (1 + sin𝜑)/(1 − sin𝜑) (14)
Unloading-reloading secant Young’s modulus at reference stress,
37.8 132 97.5 164.4
re
Eur
f
where
/[MPa] 𝛼 is a constant based on a virtual (numerical) oedometer test. The evolution of the
Oedometric tan gent modulushardening parameter,
at reference stress, Eoed𝑝 /[MPa]
re f , is given by the relation:
18.4 22.5 16.8 41.1
Exponent in the power law, m 0.95
𝑐cot𝜑 + 𝑝 0.9 0.7 0.5
Normal consolidation coefficient, Knc ∆𝑝 = 𝑘𝐾 0.81 0.64∆𝛾 0.53 0.44 (15)
Failure ratio, Rf 𝑐cot𝜑 + 𝑝
0.9 0.9 0.9 0.9
where 𝛾 is the volumetric hardening parameter, 𝑘 is a correction factor with a typical
value of 0 < 𝑘 ≤ 1, 𝐾 = 𝐾 𝐾 /(𝐾 −𝐾 ) , 𝐾 = 𝐸 / 3(1 − 2𝜈) , and 𝐾 = 𝐸 (1 +
2𝐾 )/3. Parameter 𝐾 denotes the normal consolidation coefficient, and 𝐸 stands
for the tangent oedometer stiffness at the reference pressure 𝑝 .
• Tensile yield criterion
The model checks for the tension failure condition. Tension failure and potential
functions are
Buildings 2024, 14, 2402 12 of 25

The hydraulic computation uses the isotropic fluid model, and the hydraulic heads
inside and outside the shaft are at the same depth; therefore, seepage is therefore not
considered in this case.

3.5. Model Validation

In this study, the following sign conventions for displacement, axial force, and bending
moment are adopted: for displacement, in the absence of a specified coordinate axis
direction, positive values represent horizontal movement towards the shaft centre. Vertical
displacements with positive values indicate upward movement. When the displacement
moves towards the interior (or inside the shaft), it is considered positive. Axial forces and
stresses are deemed positive if they tend to compress the member at the section being
analysed. Bending moments that deflect the cross-section towards the shaft centre are
positive. Negative values represent movement in the opposite direction.
To validate the model, several monitoring points were established on-site along three
distinct directions (L1, L2, and L3) around the shaft’s perimeter (see Figure 5a). These points
correspond to locations in the numerical model where we collect data. The horizontal
Buildings 2024, 14, x FOR PEER REVIEW 14 of 27
displacements of the monitoring piles and soil layers are monitored vertically, as illustrated
in Figure 5b. Upon reaching the final excavation depth, the monitored results are compared
with numerical results. In the L1 and L3 directions, a relatively minor difference can be
Despite
observed the observed
between discrepancies
the numerical values andbetween the monitoring
the measured data, withand numerical
a maximum results,
deviation
the overall similarity in trends suggests that the difference is acceptable. Given this
of 1.09 mm. In the L2 direction, the maximum difference is 1.76 mm, which is considered overall
alignment, the model is deemed suitable
acceptable for the purposes of this study. for subsequent numerical analyses.

Distance from boundary of the ring foundation [m] Horizontal displacement [mm]
2.5 7.5 12.5 22.5 32.5 42.5 -4 0 4 8 12 16 20
0 0
Numerical value_L1
Monitored soil
-1 Monitored value_L1 S1
Numerical value_L2 -10
-2
Settlement of ground surface [mm]

Monitored value_L2
Numerical value_L3 -20
-3
Monitored value_L3
Monitored pile
-4 P1 Boundary of ring
-30
Depth [m]

foundation
-5
-40
-6

-7 -50
L2
L3 Pile_numerical value
-8
Boundary of ring L1 -60 Pile_monitored value
-9 foundation Soil_numerical value
-70 Soil_monitored value
-10

(a) (b)
Figure5.5.Comparison
Figure Comparisonofofnumerical
numericalvalues
valueswith
withmonitored
monitoredvalues:
values:(a)(a) surface
surface settlement;
settlement; (b)(b) hori-
horizon-
zontal displacements of piles and soils along the depth.
tal displacements of piles and soils along the depth.

4. Results
The soil displacement’s influence on the piles results in a similar trend in the horizontal
displacement
4.1. Mechanicalvariations
Response inofthe
both piles of
Vicinity andthesoil along the depth. Notably, the maximum
Shaft
horizontal displacement
The vertical positionofof
Pile
the1 bottom
(P1) andofthe
themonitored
reinforcedsoil layer
zone (S1) along
is denoted by the vertical
H1 (−24 m),
axis occurs at a depth of approximately 40 m, reaching
while the final excavation depth is denoted by H2 (−51 m). a magnitude of 15 mm (Figure 5b).
Conversely,
Figure the numerical
6 shows model predicts
the deformations a maximum
of the displacement
surrounding towards
soils through half ofthe
theshaft of
model
13 mm at a depth of roughly 45 m.
in various directions at the final depth (Step 24). This visualisation highlights the locations
Below this
of maximum depth, a noticeable
displacements. gap emerges
Computational between
analysis the monitored
has determined thatresults and the
the horizontal
numerical results, widening further as depth increases. Interestingly, the
displacement of the soil beneath the reinforced zone is considerably greater than monitored values
that
for the piles and soil indicate a displacement away from the shaft below the final excavation
above it. Specifically, the maximum horizontal displacement of the surrounding soil to-
wards the shaft is approximately 24 mm. Concurrently, the maximum upward displacement
occurs on the excavation surface, reaching a significant value of 50 mm.
However, it is noteworthy that the largest settlement does not occur at the ground
surface but rather at the H1 level. This phenomenon is attributed to the softer soil condi-
tions beneath the reinforced zone and the relatively weaker supporting structure, which
lead to increased horizontal deformation that subsequently drives vertical displacement
Buildings 2024, 14, 2402 13 of 25

depth (−51 m), while the numerical simulations indicate the opposite. However, this does
not entirely exclude the possibility of minor errors in the monitored data.
Despite the observed discrepancies between the monitoring and numerical results, the
overall similarity in trends suggests that the difference is acceptable. Given this overall
alignment, the model is deemed suitable for subsequent numerical analyses.

4. Results
4.1. Mechanical Response in the Vicinity of the Shaft
The vertical position of the bottom of the reinforced zone is denoted by H1 (−24 m),
while the final excavation depth is denoted by H2 (−51 m).
Figure 6 shows the deformations of the surrounding soils through half of the model in
various directions at the final depth (Step 24). This visualisation highlights the locations
of maximum displacements. Computational analysis has determined that the horizontal
displacement of the soil beneath the reinforced zone is considerably greater than that above
it. Specifically, the maximum horizontal displacement of the surrounding soil towards the
Buildings 2024, 14, x FOR PEER REVIEW
shaft is approximately 24 mm. Concurrently, the maximum upward displacement 15 of 27
occurs
on the excavation surface, reaching a significant value of 50 mm.

(a) (b) (c)

Figure 6.
Figure 6. Displacements
Displacements(m) (m)in
inthe
thevicinity
vicinityofofthe
theshaft
shaftinin(a)
(a)horizontal
horizontal direction
direction (x-axis);
(x-axis); (b)(b) hori-
horizon-
zontal direction (y-axis); (c) vertical direction (z-axis).
tal direction (y-axis); (c) vertical direction (z-axis).

Figure 7 shows
However, the peak vertical
it is noteworthy that thedeformation at various
largest settlement doeslocations
not occurthroughout
at the groundthe
sinking process. As excavation depth increases, vertical displacements at
surface but rather at the H1 level. This phenomenon is attributed to the softer soil conditions these locations
demonstrate
beneath a linear trend.
the reinforced zoneWhen
and thethe relatively
excavationweakerface progresses
supportingpast the reinforced
structure, whichzone,
lead
theincreased
to surrounding soft soil
horizontal experiencesthat
deformation greater deformation,
subsequently resulting
drives verticalin accelerated settle-
displacement of the
mentConversely,
soil. on the surface
dueoftothethetop ring foundation.
stabilising Nonetheless,
effect of the until ansettlement
piles, the overall excavation depththe
within of
42 m is reached, the rate of uplift
reinforced zone remains relatively minimal.at the excavation face remains relatively constant, influ-
encedFigure
primarily
7 showsby thethedigging process.deformation
peak vertical At this depth, at the influence
various rangethroughout
locations of vertical the
de-
formation encounters the firmer sand layer beneath, causing a decrease
sinking process. As excavation depth increases, vertical displacements at these locations in the uplift rate
of the excavation
demonstrate face.trend.
a linear Before When
this point, due to the soil’s
the excavation uniformity, past
face progresses the uplift increases
the reinforced
linearly
zone, thewith depth, with
surrounding each
soft soil2experiences
m excavation step leading
greater to a 1.5resulting
deformation, mm rise.in accelerated
settlement on the surface of the top ring foundation. Nonetheless, until an excavation
depth 60of 42 m is reached, the rate of uplift at the excavation face remains 5000 relatively constant,
Ring foundation surface
influenced primarily by the digging process. At this depth, the influence 4500 range of vertical
50 Excavation face
vertical displacement [mm]

deformation encounters the firmer sand layer beneath, causing a decrease in the uplift rate
Volume
of the excavation of plastc
face. zonethis point, due to the soil’s uniformity,
Before 4000 the uplift increases
40 Depth of 42m
linearly with depth, with each 2 m excavation step leading
(excavation face) to a 1.53500
mm rise.
30
3000
Volume [m3]

20 2500

The excavation face reaches 2000

10
the bottom of reinforced zone
 ment on the surface of the top ring foundation. Nonetheless, until an excavation depth of
42 m is reached, the rate of uplift at the excavation face remains relatively constant, influ-
enced primarily by the digging process. At this depth, the influence range of vertical de-
formation encounters the firmer sand layer beneath, causing a decrease in the uplift rate
Buildings 2024, 14, 2402 of the excavation face. Before this point, due to the soil’s uniformity, the uplift increases
14 of 25
linearly with depth, with each 2 m excavation step leading to a 1.5 mm rise.

60 5000
Ring foundation surface
4500
50 Excavation face

Maximum vertical displacement [mm]

Volume of plastc zone 4000
40 Depth of 42m
(excavation face) 3500
30
3000

Volume [m3]
20 2500

The excavation face reaches 2000

10
the bottom of reinforced zone
1500
0
1000
-10
500

-20 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Step
Figure
Figure 7.
7. Vertical
Vertical displacements
displacements (m)
(m) and
and plastic
plastic zone
zone volume (m33))in
volume (m inthe
thevicinity
vicinityof
ofthe
theshaft.
shaft.

Examiningthe
Examining thegrowth
growthrates
ratesofofdisplacement
displacement and
and plastic
plastic zone
zone volume
volume reveals
reveals the the
re-
reinforcement zone’s significance in maintaining structural stability. Only a
inforcement zone’s significance in maintaining structural stability. Only a few plasticfew plastic
zonesare
zones arepresent
present on
onthe
theexcavation
excavationsurface
surfacebefore
beforecrossing
crossingthe
thereinforced
reinforcedzone.
zone.However,
However,
once the excavation depth surpasses this zone, the focus shifts primarily to the lateral soil
beneath it. Consequently, there is a marked increase in top deformation and plastic zone
volume, significantly compromising stability.
Figure 8 shows the horizontal and vertical stresses along the depth at S1 (shown in
Figure 5). Due to the larger self-weight of concrete ring foundation from 0 to −3 m, the rate
of increase in both vertical and horizontal stresses is higher. Before step 6, the excavation
depth does not exceed the reinforced zone (0 m to −21 m), causing minimal disturbance
to the soil, and the distribution of vertical and horizontal stresses still follows a roughly
linear increase. After exceeding the reinforced zone (step 9), irregular horizontal stress
distributions begin to appear in the soil near the excavation face due to the formation of a
plastic zone. Below the reinforced zone, the reduction in support strength leads to greater
stress release in the surrounding soil, resulting in significantly reduced horizontal and
vertical stresses compared to before. More stress release indicates increased deformation
and decreased safety.
Figure 9a illustrates the vertical displacements of the piles. Despite variations in
the horizontal movements of each pile due to uneven load distribution on the top ring
foundation, they all exhibit a similar trend. Notably, the significant pressure on the top
load area results in the greatest settlement in Piles 1 and 2, while Piles 8 and 9 experience
the least. Conversely, in terms of horizontal displacement, Piles 8 and 9 exhibit the largest
movements, while Piles 1 and 2 display relatively minor shifts (Figure 9b). As depth
increases, the vertical displacements of all piles steadily diminish, peaking at 13 mm
at the pile tops. Furthermore, a reverse bending zone appears at H1 for the horizontal
displacements, where the movement sharply rises with depth and then sharply drops
below H2 , ultimately resulting in minimal variation in horizontal displacement among
the piles.
The variation in axial force shown in Figure 9c exhibits a similarity to the horizontal
displacement patterns displayed in Figure 9b, particularly within the range of H1 to H2 ,
where values are comparatively high. Among the piles, Pile 3 experiences the highest axial
force, reaching 7520 kN, closely followed by Pile 4. This underscores the complexity of pile
interactions, revealing that the greatest deformation does not always correspond to the
highest axial force. Turning to Figure 9d, the differences in bending moments (Mv ) among
 once the excavation depth surpasses this zone, the focus shifts primarily to the lateral soil
beneath it. Consequently, there is a marked increase in top deformation and plastic zone
volume, significantly compromising stability.
Figure 8 shows the horizontal and vertical stresses along the depth at S1 (shown in
Buildings 2024, 14, 2402 Figure 5). Due to the larger self-weight of concrete ring foundation from 0 to −3 m, the 15 rate
of 25
of increase in both vertical and horizontal stresses is higher. Before step 6, the excavation
depth does not exceed the reinforced zone (0 m to −21 m), causing minimal disturbance to
the piles
soil, and the distribution
are relatively of vertical
small. There and distinct
are three horizontal stresses
peaks in thestill follows
bending a roughly
moment lin-
profiles:
ear increase.
at the bottomAfterof theexceeding
reinforced the reinforced
zone, zonedepth,
at the final (step 9),
and irregular
notablyhorizontal
at a depth stress
of −58dis- m,
tributions
termed H3begin, which tomarks
appeartheininterface
the soil near the excavation
between the hard and face due
soft soiltolayers.
the formation
This peak ofata
plastic zone.
H3 arises from Below the reinforced
the abrupt change in zone, the reduction
horizontal in support
displacement as thestrength leads tofrom
piles traverse greater
the
stress
clayey release
silt layerintothe
thesurrounding soil,Consistent
fine sand layer. resulting with
in significantly reduced horizontal
their large horizontal and
displacements,
vertical stresses
Piles 8 and 9 havecompared
relativelyto before.
high totalMore stress
bending release indicates
moments. The maximum increased deformation
recorded bending
moment
and recorded
decreased is approximately −660 kN·m.
safety.

Horizontal stress [MPa] Vertical stress [MPa]

-0.2 0 0.2 0.4 0.6 0.8 1 1.2 -0.2 0 0.2 0.4 0.6 0.8 1 1.2

-5 Step1 -5 Step1
Step6 Step6
Step12 Step12
-15 -15
Step18 Step18
Step24 Step24
Depth [m]

Depth [m]
-25 -25

-35 -35
Buildings 2024, 14, x FOR PEER REVIEW 17 of 27

-45 -45

peak at H3 arises from the abrupt change in horizontal displacement as the piles traverse
-55 -55
from the clayey silt layer to the fine sand layer. Consistent with their large horizontal dis-
(a)
placements, (b)
Piles 8 and 9 have relatively high total bending moments. The maximum rec-
orded bending
Figure moment
8. Stress recorded
distribution is approximately
with depth −660 kN·m.
at S1: (a) horizontal stress (b) vertical stress.
Figure 8. Stress distribution with depth at S1: (a) horizontal stress (b) vertical stress.

Figure [mm]
Vertical displacement Horizontal
9a illustrates the vertical displacements of displacement [mm] variations in the
the piles. Despite
-15 -12 horizontal
-9 -6 movements
-3 0
0 of each pile due to uneven3 load6 distribution
9 12 15 the18top ring foun-
on
0 0
dation, they all exhibit a similar trend. Notably, the significant pressure on the top load
Pile 1 Pile 2 Pile 1 Pile 2
area results in the greatest settlement in Piles 1 and 2, while Piles 8 and 9 experience the
Pile 3 Pile 4 Pile 3 Pile 4
-10
least. Pile
Conversely,
5
in terms of horizontal
Pile 6 -10
displacement, Piles 8 and Pile
Pile 5
9 exhibit
6
the largest
movements,
Pile 7 while
Pile 8Piles 1 and 2 display relatively minor shifts
Pile 7 (Figure 9b).
Pile 8 As depth in-
creases, the
Pile 9 vertical
Pile 10displacements of all piles steadily diminish,
Pile 9 peaking
Pile 10at 13 mm at the
-20 Pile 11Furthermore, a reverse bending
pile tops. -20 zone appears at H Pile 11 the horizontal displace-
1 for
H1 H1 depth and then sharply drops below H2,
ments, where the movement sharply rises with
ultimately resulting in minimal variation in horizontal displacement among the piles.
-30 The variation in axial force shown -30 in Figure 9c exhibits a similarity to the horizontal
Depth [m]
Depth [m]

displacement patterns displayed in Figure 9b, particularly within the range of H1 to H2,
where values are comparatively high. Among the piles, Pile 3 experiences the highest axial
-40 -40
force, reaching 7520 kN, closely followed by Pile 4. This underscores the complexity of
pile interactions, revealing that the greatest deformation does not always correspond to
-50 H2 the highest axial force. Turning to Figure 9d, the differences in bending moments (Mv)
-50 H2
among the piles are relatively small. There are three distinct peaks in the bending moment
H3 profiles: at the bottom of the reinforced zone,H3at the final depth, and notably at a depth of
-60 −58 m, termed H3, which marks the interface -60 between the hard and soft soil layers. This

-70 -70
(a) (b)
Axial force [kN] Bending moment [kN·m]
0 40009. Cont.
2000 Figure 6000 8000 -1000 -500 0 500 1000
0 0

-10 -10 Pile 1

Pile 2
Pile 3
-20 -20 Pile 4
 H3 H3

-60 -60

Buildings 2024, 14, 2402 16 of 25

-70 -70
(a) (b)
Axial force [kN] Bending moment [kN·m]
0 2000 4000 6000 8000 -1000 -500 0 500 1000
0 0

-10 -10 Pile 1

Pile 2
Pile 3
-20 -20 Pile 4
H1 H1 Pile 5
Pile 6
Pile 7
-30 Pile 1 Pile 2 -30
Pile 8
Depth [m]

Depth [m]
Pile 3 Pile 4
Pile 9
Pile 5 Pile 6
Pile 10
-40 Pile 7 Pile 8 -40 Pile 11
Pile 9 Pile 10
Pile 11
-50 H2 -50 H2

H3 H3
-60 -60

-70 -70
(c) (d)
Figure 9. Mechanical
Figure 9. Mechanicalresponse of theofpiles:
response (a) vertical
the piles: displacement;
(a) vertical (b) horizontal
displacement; displacement;
(b) horizontal displacement;
(c) axial force; (d) bending moment.
(c) axial force; (d) bending moment.

4.2. The Influence of the Reinforced Zone Strength

The effect of reinforced zone strength on the stability of the shaft will be investigated
in this section. The reinforced zone is formed by incorporating cement into natural soils
through grouting or mixing methods, resulting in a hard, semi-rigid, and durable material
formed via cement particle hydration. This material possesses excellent compressive and
shear strengths but exhibits brittleness and limited tensile strength, making it susceptible
to cracking.
The unconfined compressive strength test is a crucial assessment tool for evaluating
the effectiveness of in situ soil-stabilizer mixtures in practical applications. Although
laboratory-measured compressive strength moduli may differ from those validated in field
tests, they remain an essential component in the design of soil stabilization mixtures. In
our project, the unconfined compressive strength is typically maintained within a range of
0.6 MPa to 1.4 MPa. Falling below this lower limit poses safety concerns, while exceeding
the upper limit becomes uneconomical.
However, when it comes to numerical modelling, there is no direct option to set
the unconfined compressive modulus. To address this, we use equivalent conventions for
numerical analysis. Based on experimental data from previous studies, the unconfined com-
pressive modulus, qu , can be correlated with elastic parameters, enabling us to incorporate
this crucial parameter into our numerical simulations [39]:

E50 = 142qu (17)

where E50 is the secant Young’s modulus. Based on soil properties and engineering expe-
rience in Shanghai, it is suggested that the Young’s modulus of soil-cement, E, could be
4–5 times the secant Young’s modulus [40,41]. However, it should be noted that in practical
engineering, there is a considerable variability in the relationship between the compressive
strength and the soil modulus. Therefore, the equivalent conversion in this paper can only
be used as a reference for research into the relationship between the soil-cement strength
Buildings 2024, 14, 2402 17 of 25

and mechanical response, while the shear and tensile strengths given in Table 5 are derived
from some experimental studies [39,42,43].

Table 5. Equivalent parameters for unconfined compressive strength.

Qu [Mpa] E [Mpa] c [kPa] φ [◦ ] σ t [kPa]

0.6 426 131 27 110
0.8 565 177 28 137
1.0 710 204 27 162
1.2 852 285 28 185
1.4 994 290 34 208

Monitoring points have been strategically positioned along the L1 direction (Figure 5a)
to track the ground surface subsidence variations after sinking, particularly those influ-
enced by varying reinforced zone strengths. The concrete ring foundation, with a width
matching
Buildings 2024, 14, x FOR PEERthe thickness of the reinforced zone, is primarily the area most affected by this19 of 27
REVIEW
variable, while the influence on other regions is negligible (Figure 10a). Notably, the vertical
displacement at the most affected location shows a mere decrease of 0.9 mm. Furthermore,
it becomes evidenton thethat the impact
deformation of reinforced
results zone
for the piles, strengtheven
as expected, on vertical displacement
with varying the soil-cement
diminishes as the qu increases.
strength, the deformation and internal forces within the piles undergo minimal changes.

Distance to the shaft wall [m] Horizontal displacement [mm]

0 5 10 15 20 25 30 35 40 45 50 0 3 6 9 12 15
0 0
Boundary of the
-4 qu=0.6
-2 concrete ring foundation
-8 qu=0.8
qu=1.0
Vertical displacement [mm]

-4 -12
qu=1.2
-16
-6 qu=1.4
-20
qu=0.6 H1
Depth [m]

-24
-8 qu=0.8
qu=1.0 -28
-10 qu=1.2 -32
qu=1.4 -36
-12 -40
-44
-14
-48
H2
-16 -52
(a) (b)

Figure 10. Displacement Displacement

Figure 10.change change atpoints
at monitoring monitoring points by
by varying qu :varying qu: (a)displacement
(a) vertical vertical displacement
along
along the L1 direction; (b) horizontal displacement with depth at S1
the L1 direction; (b) horizontal displacement with depth at S1.
Despite the limited influence of the reinforced zone’s strength on soil deformation,
Figure 10b illustrates the horizontal displacement of the soil at S1 (Figure 5b) across
as shown in Figure 11, the total volume of the plastic zone surrounding the shaft exhibits
various depths. In contrast to vertical deformation, the horizontal displacement change
a reduction after the completion of the sinking process. While the plastic zone’s volume is
is relatively minor, yet its influence is confined to the depth range of the reinforced zone,
not a direct safety indicator, it does provide insight into the degree of soil disturbance
with a maximum caused byofexcavation.
shift 1.15 mm. Its reduction, therefore, contributes to improving the structural sta-
Based on bility.
the computational
Notably, the soilanalysis, it is apparent
in the vicinity that enhancing
of the reinforced thefree
zone remains soil-cement
of plastic zones,
strength is notindicating
an effective means of controlling
its reinforcing effect. the deformation around the shaft. Based
on the deformation results for the piles, as expected, even with varying the soil-cement
strength, the deformation and internal forces within the piles undergo minimal changes.
Despite the limited influence of the reinforced zone’s strength on soil deformation,
as shown in Figure 11, the total volume of the plastic zone surrounding the shaft exhibits
a reduction after the completion of the sinking process. While the plastic zone’s volume
is not a direct safety indicator, it does provide insight into the degree of soil disturbance
caused by excavation. Its reduction, therefore, contributes to improving the structural
Buildings 2024, 14, 2402 18 of 25

Buildings 2024, 14, x FOR PEER REVIEW Notably, the soil in the vicinity of the reinforced zone remains free of plastic20zones,
stability. of 27
indicating its reinforcing effect.

Changes in
Figure 11. Changes
Figure in volume of plastic zone by varying quu..
volume of

As qquu increases
As increases from
from 0.6
0.6 MPa
MPa toto 11 MPa,
MPa, the
the plastic
plastic zone’s
zone’s volume
volume decreases
decreases by
by 7%.
7%.
When qu rises further to 1.4 MPa, the plastic zone diminishes by an even greater margin of
When qu rises further to 1.4 MPa, the plastic zone diminishes by an even greater margin of
8.6%. However, interestingly, at q = 1.2 MPa, the plastic zone’s volume slightly increases
8.6%. However, interestingly, at quu = 1.2 MPa, the plastic zone’s volume slightly increases
compared to qu = 1 MPa. Considering both economic factors and the reinforcement effect,
compared to qu = 1 MPa. Considering both economic factors and the reinforcement effect,
it is advisable for this project to maintain the unconfined compressive modulus of the
it is advisable
reinforced soilfor this project to maintain
at approximately 1 MPa. the unconfined compressive modulus of the re-
inforced soil at approximately 1 MPa.
4.3. The Influence of the Ring Bottom Beam
4.3. The
To Influence
effectively of distribute
the Ring Bottom Beam soil pressure around the cutting edge, mitigate
the excessive
To effectively
the convergent distributeof
deformation thetheexcessive soilresist
shaft wall, pressure around
bending the cutting
moments, and edge, mitigate
strengthen the
the convergent
interface between deformation
the bottomofconcrete
the shaft wall,
and resist
plain bending
concrete for moments, and strengthen
bottom sealing, the
a ring bottom
interface
beam design between the bottom
has been concreteHowever,
implemented. and plaintheconcrete
actual for bottom
efficacy sealing,
of this a ring
design bottom
remains to
beam designthus
be verified, hasnecessitating
been implemented. However,
the application the actualsimulation
of numerical efficacy of methods
this design remains
to assess its
to be verified,
supportive rolethus
andnecessitating
scrutinise thethe application of numerical simulation methods to as-
results.
It is
sess its worth noting
supportive that,scrutinise
role and althoughthe supplementary
results. sections are added to the shaft as
excavation
It is worthprogresses
noting and
that,the shaft settles,
although the numerical
supplementary model
sections areinadded
this study
to theassumes
shaft asa
continuousprogresses
excavation shaft wall,and overlooking potential
the shaft settles, thejoints between
numerical prefabricated
model in this studysegments.
assumesIn a
contrast, real-world projects often employ a prefabricated shaft wall,
continuous shaft wall, overlooking potential joints between prefabricated segments. In introducing potential
disparities
contrast, in the internal
real-world force
projects distribution
often employ aofprefabricated
the concrete lining, particularly
shaft wall, at thepoten-
introducing joints.
Nevertheless,
tial disparities the mechanical
in the response
internal force results from
distribution this
of the model provide
concrete valuable insights
lining, particularly at the
into the
joints. mechanical behaviour
Nevertheless, the mechanical of cast-in-place shafts and
response results fromserve
this as a basis
model for comparing
provide valuable
trends with
insights intoprefabricated
the mechanical shaft walls. of cast-in-place shafts and serve as a basis for com-
behaviour
paring trends with prefabricated shaft walls.with and without the ring bottom beam, as
The comparison of shaft deformation
depicted in Figure 12a,b,
The comparison clearly
of shaft highlights
deformation withtheand
effects of this
without thedesign choice.
ring bottom When
beam, the
as de-
shaft reaches its final depth, the top of the shaft exhibits an offset of
picted in Figure 12a,b, clearly highlights the effects of this design choice. When the shaftapproximately 6 mm,
increasing
reaches gradually
its final depth,towards
the top of thethe
cutting edge where
shaft exhibits the maximum
an offset offset occurs.
of approximately 6 mm, in-By
analysing the data from monitoring lines ML1 and ML2 set at
creasing gradually towards the cutting edge where the maximum offset occurs. By ana- the shaft wall’s centre on
lysing the data from monitoring lines ML1 and ML2 set at the shaft wall’s centre on the
symmetric plane, with the ring bottom beam, the centre axis of the shaft offsets 8.5 mm
 towards Pile 1, while without it, the offset increases to 9.6 mm. Additionally, a slight ab-
rupt change can be seen in deformation near the top of the initial support structure, as
shown in Figure 12c. This observation suggests that the initial support structure, including
Buildings 2024, 14, 2402 the ring beam does not significantly mitigate the overall offset but may have a localised 19 of 25
impact on shaft deformation patterns.
The most significant impact of the ring bottom beam, however, lies in reducing con-
vergent deformation
the symmetric within
plane, the ring
with the depth rangebeam,
bottom of thethe
initial support
centre axis ofstructure.
the shaft As shown
offsets in
8.5 mm
Figure
towards the1,presence
12d,Pile of the it,
while without ring
thebottom beam leads
offset increases to mm.
to 9.6 a notable reduction
Additionally, in conver-
a slight abrupt
change
gent can be seen
deformation, in deformation
achieving near the top
a 50% reduction in of
thethe initial
case support
study. This structure,
reductionas is shown
crucialin
forFigure
shafts12c. This observation
excavated suggests
using vertical shaft that the initial
sinking machinessupport
(VSM) structure, including
as excessive the ring
convergent
beam does can
deformation not significantly
significantly mitigate
hinder thethelifting
overalland
offset but may
lowering of have a localised
the machine. impact
It is worthon
shaft that
noting deformation
the extentpatterns.
of this reduction may vary depending on the specific geological co

M1
ML1
M2
ML2
M1
ML1 ML2
M2

Z
Z
X
X
Y
Y
Zone Displacement Magnitude
Deformed Factor: 200 Zone Displacement Magnitude
2.0169E-02 Deformed Factor: 200
2.0000E-02 2.0110E-02
1.9000E-02 2.0000E-02
1.8000E-02 1.9000E-02
1.7000E-02 1.8000E-02
1.6000E-02 1.7000E-02
1.5000E-02 1.6000E-02
1.4000E-02 1.5000E-02
1.3000E-02 1.4000E-02
1.2000E-02 1.3000E-02
1.1000E-02 1.2000E-02
1.0000E-02 1.1000E-02
9.0000E-03 1.0000E-02
8.0000E-03 9.0000E-03
7.7100E-03 8.0000E-03
Zone Group 7.6623E-03
Bottom Beam (No disp) Zone Group
Shaft wall (No disp) Lining (No disp)

(a) (b)
Horizontal displacement [mm] Convergent displacement [mm]
-12 -9 -6 -3 0 3 6 9 12 -1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5 4
-1 -1
Convergent displacement
-6 -6 (with ring bottom beam)
Convergent displacement
-11 -11 (without ring bottom beam)

-16 ML1(with ring bottom beam) -16

ML2(with ring bottom beam)
-21 ML1(without ring bottom beam) -21
Depth [m]
Depth [m]

-26 ML2(without ring bottom beam) -26

-31 -31
Top of the initial
Top of the initial
-36 support structure (-41m)
-36 support structure (-41m)

-41 -41

-46 Depth range of the sidewall -46 Depth range of the sidewall
(-50m~-44m) (-50m~-44m)
-51 -51
(c) (d)
Figure 12. Comparison of shaft wall displacement with and without the ring bottom beam. (a) de-
Figure 12. Comparison of shaft wall displacement with and without the ring bottom beam. (a) defor-
formation of shaft wall with ring bottom beam (deformed factor: 200); (b) deformation of shaft wall
mation of shaft wall with ring bottom beam (deformed factor: 200); (b) deformation of shaft wall
without ring bottom beam (deformed factor: 200); (c) horizontal displacement of monitoring lines;
(d) convergent displacement with depth.

The most significant impact of the ring bottom beam, however, lies in reducing
convergent deformation within the depth range of the initial support structure. As shown in
Figure 12d, the presence of the ring bottom beam leads to a notable reduction in convergent
 The analysis of the bending moment (Mv) at different depths of the cross-sections
where M1 and M2 are located provides further insights into the structural behaviour of
the shaft. As seen in Figure 13b, the difference in Mv between M1 and M2 begins to emerge
below the reinforced zone, but it remains relatively small. However, below the top of the
Buildings 2024, 14, 2402 initial support structure, the ring bottom beam exerts a significant influence on the distri- 20 of 25
bution of Mv. The substantial variation in Mv near the top of the initial support structure
can be attributed to the larger bending deflection of the connection due to the gravity load
of the ring bottom
deformation, beam and
achieving a 50% thereduction
sidewall. in the case study. This reduction is crucial for
Within the support range
shafts excavated using vertical shaft of the sidewall
sinkingand the cutting
machines (VSM)edge,as the direction
excessive of Mv on
convergent
the
deformation can significantly hinder the lifting and lowering of the machine. It is beam.
shaft wall with the ring bottom beam is opposite to that without the ring bottom worth
This reversal
noting in direction
that the extent of indicates a change
this reduction mayinvary
the curvature
depending ofonthethe
shaft wall, likely
specific due
geological
to the presence of the ring bottom
conditions and structural design of the project.beam. In terms of the magnitude of values, the ring
bottomThe beam
axialdoes
forces notofappear
the shaft to wall
improve
alongthe ML1vertical distribution
and ML2 of Mv in the
(Fv ) are presented in lower
Figurepart
13a.
of the shaft wall. Instead, it increases the local bending moment
Due to the deviation of the shaft wall, the Fv acting on M1 and M2 is not symmetrical. in some regions due to
changes in the thickness of the concrete structure. This increase
Below a depth of −11 m, the difference in Fv between ML1 and ML2 begins to increase in local bending moment
may be a concern,
significantly. as it can
The largest lead to occurs
disparity higher at stresses and potential
the cutting edge, where for cracking
ML2 exceeds or failure
ML1 by in
those regions. It is important to monitor and analyse these changes
approximately 1370 kN in both cases with and without the ring bottom beam. However, in bending moment
distribution
within the depthcarefully
rangetoofensure
sidewall the(−structural
50 m to −integrity
44 m), Fv and stability
is notably of the in
reduced shaft. Further
the presence
design optimisation
of the ring bottom beam, or reinforcement
decreasing by measures may be35needed
approximately kN~654 tokN.
mitigate the negative
Nonetheless, Fv at
impact of the
the cutting ringreturns
edge bottomtobeama levelonsimilar
the bending
to thatmoment
withoutdistribution.
the ring bottom beam.

Axial force [kN] Bending moment [kN·m]

-500 0 500 1000 1500 2000 2500 -80 -60 -40 -20 0 20 40 60 80
-1 -1

-6 -6
ML1(with ring bottom beam)
-11 ML2(with ring bottom beam) -11
ML1(with ring
ML1(without ring bottom beam) bottom beam)
-16 ML2(without ring bottom beam) -16
ML2(with ring
bottom beam)
-21 -21
Depth [m]
Depth [m]

ML1(without ring
bottom beam)
-26 -26
ML2(without ring
bottom beam)
-31 -31

-36 -36

-41 -41

-46 -46

-51 -51

(a) (b)

Figure
Figure 13.
13. Interactions
Interactions in
in the
the shaft
shaft wall
wall elements along ML1
elements along ML1 and
and ML2:
ML2: (a)
(a) vertical
vertical axial
axial force
forceFFvv,, and
and (b) vertical
(b) vertical bending
bending moment
moment Mv . Mv.

The analysis of the bending moment (Mv ) at different depths of the cross-sections
where M1 and M2 are located provides further insights into the structural behaviour of the
shaft. As seen in Figure 13b, the difference in Mv between M1 and M2 begins to emerge
below the reinforced zone, but it remains relatively small. However, below the top of
the initial support structure, the ring bottom beam exerts a significant influence on the
distribution of Mv . The substantial variation in Mv near the top of the initial support
structure can be attributed to the larger bending deflection of the connection due to the
gravity load of the ring bottom beam and the sidewall.
Within the support range of the sidewall and the cutting edge, the direction of Mv
on the shaft wall with the ring bottom beam is opposite to that without the ring bottom
beam. This reversal in direction indicates a change in the curvature of the shaft wall, likely
due to the presence of the ring bottom beam. In terms of the magnitude of values, the ring
bottom beam does not appear to improve the vertical distribution of Mv in the lower part
of the shaft wall. Instead, it increases the local bending moment in some regions due to
changes in the thickness of the concrete structure. This increase in local bending moment
may be a concern, as it can lead to higher stresses and potential for cracking or failure in
those regions. It is important to monitor and analyse these changes in bending moment
 versely, without the ring bottom beam, these values drop to 1.5 MPa and 1.8 MPa. Notably,
at the same section position, the hoop stress without the ring bottom beam is 1.2 MPa to
2.3 MPa higher than that with the beam.
The radial stress, oriented perpendicularly to the hoop stress (Figure 14a), reveals
Buildings 2024, 14, 2402 21 of 25
that the ring bottom beam does not mitigate but potentially augments this stress, resulting
in an uneven distribution (Figure 14c). This unevenness stems from pressure variations
caused by the pillars connected to the sidewalls, though the differences are relatively mi-
distribution
nor, rangingcarefully
from 0.04toMPa
ensure the MPa.
to 0.18 structural integrity and stability of the shaft. Further
design optimisation or reinforcement measures may be needed to mitigate the negative
When considering the bending moments along the studied section, the majority of
impact of the ring bottom beam on the bending moment distribution.
values with a ring bottom beam are negative, indicating an opposite orientation compared
After comparing the interactions between M1 and M2, we discovered a significant
to that shown in Figure 14a. In the absence of the ring bottom beam, the bending moments
discrepancy near the top of the sidewall between the conditions with and without the ring
reach negative peaks at 45° and 135°, and positive peaks at M1, 90°, and M2, spanning a
bottom beam. Consequently, we chose the cross-sectional elements of the shaft wall as
range of −40.8 kN·m to 30 kN·m (Figure 14d). In contrast, with the ring bottom beam, the
the focus of our study to analyse the mechanical alterations induced by the ring bottom
bending moments range from −14.8 kN·m to 8.4 kN·m. Clearly, the inclusion of the ring
beam, as depicted in Figure 14. This analysis was conducted by converting rectangular
bottom beam significantly improves the distribution of bending moments in the horizontal
coordinates to polar coordinates, utilising the average stress across the section to capture
direction of the shaft wall, effectively reducing the extreme values by approximately 65%.
the interactions.

100° 80°
4.0
120° 60°

140° 3.0 40°

σθ [MPa]
Mh
σρ
Z σρ Mh 2.0
160° 20°
X 1.0
σθ σθ
Y
180° 0.0 0°
(M1) With ring bottom beam X (M2)
Without ring bottom beam Y

(a) (b)
100° 80° 100° 80°
0.8 60
120° 60° 120° 60°

140° 0.6 40° 140° 30 40°

Mc [kN·m]
σρ [MPa]

0.4 0
160° 20° 160° 20°
0.2 -30

180° 0.0 0° 180° -60 0°

(M1) With ring bottom beam X (M2) (M1) With ring bottom beam X (M2)
Without ring bottom beam Y Without ring bottom beam Y

(c) (d)

Figure 14. Interactions of the studied section of the shaft wall with and without the ring bottom
beam: (a) hoop stress σθ , radial stress σρ and bending moment Mh and their positive directions; (b) σθ
distribution in the studied section; (c) σρ distribution in the studied section; (d) Mh distribution in the
studied section.

The hoop stress is oriented tangentially along the shaft wall (Figure 14a), revealing
that with the presence of the ring bottom beam, the internal forces within the shaft wall
exhibit a predominantly symmetrical pattern, albeit with minor variations in intensity along
the polar axis (Figure 14b). However, in the absence of the ring bottom beam, a notable
discrepancy emerges; the hoop stress is considerably higher on the M1 side (180◦ ) compared
to the M2 side (0◦ ), reaching values of 3.9 MPa and 2.7 MPa, respectively. Conversely,
without the ring bottom beam, these values drop to 1.5 MPa and 1.8 MPa. Notably, at the
same section position, the hoop stress without the ring bottom beam is 1.2 MPa to 2.3 MPa
higher than that with the beam.
The radial stress, oriented perpendicularly to the hoop stress (Figure 14a), reveals
that the ring bottom beam does not mitigate but potentially augments this stress, resulting
in an uneven distribution (Figure 14c). This unevenness stems from pressure variations
caused by the pillars connected to the sidewalls, though the differences are relatively minor,
ranging from 0.04 MPa to 0.18 MPa.
Buildings 2024, 14, 2402 22 of 25

When considering the bending moments along the studied section, the majority of
values with a ring bottom beam are negative, indicating an opposite orientation compared
to that shown in Figure 14a. In the absence of the ring bottom beam, the bending moments
reach negative peaks at 45◦ and 135◦ , and positive peaks at M1, 90◦ , and M2, spanning a
range of −40.8 kN·m to 30 kN·m (Figure 14d). In contrast, with the ring bottom beam, the
bending moments range from −14.8 kN·m to 8.4 kN·m. Clearly, the inclusion of the ring
bottom beam significantly improves the distribution of bending moments in the horizontal
direction of the shaft wall, effectively reducing the extreme values by approximately 65%.

5. Discussion
Soft clay is extensively distributed in Shanghai. Due to its high viscoplasticity and low
permeability, the construction speed of projects for underground space structures signifi-
cantly impacts soil deformation. The longer the construction period, the greater the soil
deformation. Compared to other methods, the on-site assembly technology and excavation
speed of the VSM method in soft clay are significantly faster. Furthermore, the VSM method
is an undrained operation; it minimizes disturbance to the surrounding environment, as
demonstrated by our numerical analysis and monitoring data. Consequently, the safety
and stability of the shaft during construction are higher, and surrounding soil collapse or
excessive shaft tilting was observed in this project after completion. It is advisable to raise
the water table in the shaft or control the density of the internal slurry to reduce settlement,
and to adjust the parameters of the bentonite grout promptly to mitigate lateral deformation
of the soil when excavating to the stratum interface, thereby ensuring construction safety.
However, since underwater conditions are not visible, encountering boulders or hard rock
layers may cause damage to the cutting drum or the suction pump or lead to blockages,
resulting in extended construction duration. Such issues are a technical limitation of the
VSM method.
Numerical results indicate that while the reinforced zone effectively controls defor-
mation, significant soil deformation below the reinforced zone during subsidence makes
deformation control difficult, and conventional reinforcement methods are not always ef-
fective. Thus, the thickness and depth of the reinforced zone should be carefully calculated
prior to construction. Although the ring bottom beam improves the stress and deformation
near the cutting edge and reduces the amount of plain concrete for bottom sealing, its
economic performance in this project is not favorable. The deformation results indicate that
the improvement is relatively minor. Therefore, the application of this structure should be
considered with caution for small or medium-sized caissons.

6. Conclusions
To gain a better understanding of the mechanical responses of the adjacent soil and
shaft wall during the sinking process and to analyse the practical implications of rein-
forcement measures, this study conducted a comprehensive numerical simulation. The
simulation outcomes highlight the following key findings:
The reinforced zone demonstrates excellent control over soil deformation and stability.
Once the shaft passes through this zone, however, the rate of soil deformation and expan-
sion of the plastic zone accelerate significantly, which results in the soil strength below
the reinforced zone decreasing due to greater stress release, causing reduced soil stability.
Therefore, it is advisable to reduce the excavation speed and intensify monitoring efforts
subsequently.
While enhancing the soil strength within the reinforced zone (ranging from 0.6 MPa to
1.4 MPa in unconfined compressive strength) does not significantly impact soil and pile
deformation, it can effectively minimize the plastic zone volume surrounding the shaft. For
cost efficiency, the numerical analysis results recommend maintaining the soil-cement’s
unconfined compressive strength at approximately 1 MPa for this project.
The ring bottom beam has no impact on the upper shaft wall’s deformation or the lower
section’s overall offset. However, it significantly reduces the convergent deformation of
Buildings 2024, 14, 2402 23 of 25

the lower shaft wall (up to 50% in our case study) and vertical axial force (with reductions
ranging from 100 kN to 654 kN). Nevertheless, it does not optimise vertical bending
moment or radial stress on the shaft wall during sinking, and may even intensify local
vertical bending moment distribution. This aspect should be carefully considered during
design. On the positive side, the ring bottom beam notably reduces hoop stress and
horizontal bending moment in the adjacent shaft wall, and the volume of plain concrete
used is also reduced. However, achieving a balance between mechanical optimisation and
economic feasibility remains a design challenge.
The success of this project demonstrates that the VSM method is highly suitable for
large and deep shaft sinking in soft soils. Additionally, it has the advantages of requiring
less space and generating less noise, making it particularly advantageous and promising
in congested urban areas. The VSM method stands apart from conventional methods
in several key aspects, and these features yield a unique stress mechanism, but current
research in this domain is still limited. Future research could involve detailed studies on
optimal pile structure, the thickness of plain concrete, and the stress effect on segment joints.
Based on the numerical results, this project strictly controlled the sinking rate during
construction, increased the frequency of monitoring and measurements, and managed the
tilting of the shaft. The timely replenishment of slurry inside the shaft effectively controlled
the deformation of both the soil and the piles. It is recommended that for similar projects,
a detailed design of the thickness and depth of the reinforced zone and the structural
parameters of the ring bottom beam can be completed before construction, and finite
element analysis should be performed if necessary.

Author Contributions: C.M.: Conceptualization, Methodology, Writing—original draft; H.H.: Inves-

tigation, Writing—review and editing; L.Y.: Validation, Supervision, Writing—review and editing;
K.L. and J.H.: Conceptualization, Methodology. All authors have read and agreed to the published
version of the manuscript.
Funding: This research was funded by Science and Technology Project of Guangzhou Municipal
Construction Group Co., Ltd. (BH20230317604, BH20230315563); Research Project of Department of
Housing and Urban Rural Development of Guangdong Province (2023-K9-240562).
Data Availability Statement: Some or all data, models, or code that support the findings of this study
are available from the corresponding author upon reasonable request.
Conflicts of Interest: Author Chunjing Ma was employed by the company Guangzhou Municipal
Engineering Group Co., Ltd. Authors Haijian Hong and Jiadong Huang were employed by the
company Guangzhou Municipal Construction Group Co., Ltd. The remaining authors declare that
the research was conducted in the absence of any commercial or financial relationships that could be
construed as a potential conflict of interest.

References
1. Luo, W.; Lin, H.; Huang, W.; Zhou, X.; Zhang, Z. Undrained Sinking Caisson Construction Technology of a 28.4 m Deep Well in
Soft Soils. Spec. Struct. 2022, 39, 113–119.
2. Krauze, K.; Boloz, L.; Wydro, T. Mechanized Shaft Sinking System. Arch. Min. Sci. 2018, 63, 891–902. [CrossRef]
3. Neye, E.; Burger, W.; Rennkamp, P. Rapid Shaft Sinking. Min. Rep. 2013, 149, 250–255. [CrossRef]
4. Abdrabbo, F.; Gaaver, K. Challenges and Uncertainties Relating to Open Caissons. J. Deep. Found. Inst. 2012, 6, 21–32. [CrossRef]
5. Peng, F.; Dong, Y.; Wang, H.; Jia, J.; Li, Y. Remote-Control Technology Performance for Excavation with Pneumatic Caisson in Soft
Ground. Autom. Constr. 2019, 105, 102834. [CrossRef]
6. Matsumoto, T.; Gomi, N.; Ikemi, Y. Process for Improving Open Caisson, SS Caisson Process; Open Keson No Kairyo kohoSS
Keson Koho. Denryoku Doboku Electr. Power Civ. Eng. 1997, 5, 100–102.
7. Tanimura, D.; Ueda, J.; Tani, Y. Large-Scale Vertical Shaft Excavation Using Super-Open Caisson System. Construction of Tamasato
Vertical Shaft (Ishioka Vertical Shaft No. 5); Jidoka Open Caisson Koho Ni Yoru Daikibo Tatekui No Kussaku. Tamasato Tateko
(Ishioka Dai5 Tateko) Shinsetsu Koji. Kensetsu No Kikaika 1998, 9–14.
8. Wang, Y.; Liu, M.; Liao, S.; Yi, Q.; He, J.; Liu, L.; Gong, Z.; Li, K. Investigation on the Stratigraphic Response and Plugging Effect
Induced by Press-in Open Caisson in Mucky Soil. KSCE J. Civ. Eng. 2023, 27, 1928–1941. [CrossRef]
Buildings 2024, 14, 2402 24 of 25

9. Tani, Y.; Nakano, M.; Okoshi, M.; Maeda, S.; Isa, H. Research & Development of Automatic System for Open Caisson Method. In
Proceedings of the 13th International Symposium on Automation and Robotics in Construction, Tokyo, Japan, 11–13 June 1996;
pp. 323–332.
10. Peila, D.; Viggiani, G.; Celestino, T. Tunnels and Underground Cities. Engineering and Innovation Meet Archaeology, Architecture
and Art: Proceedings of the WTC 2019 ITA-AITES World Tunnel Congress (WTC 2019), 3–9 May 2019, Naples, FL, USA; CRC Press:
Boca Raton, FL, USA, 2019.
11. Kwong, J.; Sandefur, K.; Nishimura, J. Geotechnical Engineering Aspects of Deep Microtunneled Pipeline Crossings of Honolulu
Harbor in Very Soft Sediments. In Proceedings of the Pipelines 2017: Condition Assessment, Surveying, and Geomatics, Phoenix,
AZ, USA, 6–9 August 2017; ASCE Press: Reston, VA, USA, 2017; pp. 482–493.
12. Frenzel, C.; Delabbio, F.; Burger, W. Shaft Boring Systems for Mechanical Excavation of Deep Shafts. In Proceedings of the Second
International Symposium on Block and Sublevel Caving, Australian Centre for Geomechanics, Perth, Australia, 20–22 April 2010;
pp. 289–295.
13. DeMerchant, M.R.; Kierstead, R.D. Evolution of the ARC-100 Reactor Building Design; ARC Clean Technology Canada Inc.: Saint
John, NB, Canada, 2023.
14. Jiang, H. A Summary of Precast Technology Application in Tunneling. Proceedings of Geo Shanghai 2018 International Conference:
Tunnelling and Underground Construction, Shanghai, China, 27–30 May 2018; Springer: Berlin/Heidelberg, Germany, 2018;
pp. 749–756.
15. Huang, M.; Zhang, Z.; Xu, J.; Jiang, H.; Bao, H.; Liu, X. Measuring of Shaft Wall Stress in Caisson Construction Process Based on
Vertical Shaft Sinking Machine: A Case Study of Caisson-Type Underground Intelligent Parking Garage Project in Nanjing, China.
Tunn. Constr. 2022, 42, 1033–1043.
16. Jiang, H.; Bao, H.; Lin, Y. Research and Application of Design and Construction Technology of Assembled Shaft: A Case Study of
a Sinking Shaft Underground Parking Garage in Nanjing, China. Tunn. Constr. 2022, 42, 463–470.
17. Zhou, J.; Liu, C.; Xu, J.; Zhang, Z.; Li, Z. Deformation Mechanism and Control of In-Situ Assembling Caisson Technology in Soft
Soil Area under Field Measurement and Numerical Simulation. Materials 2023, 16, 1125. [CrossRef]
18. Stevens, F.; Weiner, S. Deep Tunnel Sewerage Systems Singapore’s Success Story. Water Sanit. Afr. 2022, 17, 19–23.
19. Fu, H.; Xu, Y.; Cheng, L. Overview of the Research Hotspots of China’s Shaft Boring Machine. Mine Constr. Technol. 2023, 44,
83–88.
20. Ma, C.; Hong, H.; Wen, W.; Yu, W.; Pei, H. VSM Technology and Equipment for Shaft Sinking and Its Engineering Applications.
Mine Constr. Technol. 2024, 45, 10–15.
21. Xu, G.; Xiao, W.; Zhao, F.; Qi, Z.; Lyu, D.; Zhao, Z. Structural Design of a Kind of Caisson Submerged Shaft Boring Machine.
Tunn. Constr. 2022, 42, 145–153.
22. Schmäh, P. Vertical Shaft Machines. State of the Art and Vision. Acta Montan. Slovaca 2007, 12, 208–216.
23. Schmäh, P.; Berblinger, S. VSM Shaft Sinking Technology—Mechanized Shaft Sinking with the VSM in Different Projects for
Subway Ventilation Shafts. In Proceedings of the World Tunnelling Congress, Foz do Iguaçu, Brazil, 9–15 May 2014.
24. Rahimi, M.; Rajmohan, A.; Spasojevic, A.; Scheele, J.; Cabot, E.; Ho, K.-S.; Ball, S. Geotechnical Considerations of Deep Mine
Shaft, Sinking by Combined VSM and Drill & Blast Construction Methodology. In Proceedings of the GeoCalgary 2022: The 75th
Annual Canadian Geotechnical Society Conference, Calgary, AB, Canada, 2–5 October 2022.
25. Royston, R.; Sheil, B.B.; Byrne, B.W. Monitoring the Construction of a Large-Diameter Caisson in Sand. Proc. Inst. Civ. Eng.
Geotech. Eng. 2022, 175, 323–339. [CrossRef]
26. Templeman, J.O.; Phillips, B.M.; Sheil, B.B. Cutting Shoe Design for Open Caissons in Sand: Influence on Vertical Bearing Capacity.
Proc. Inst. Civ. Eng. Geotech. Eng. 2023, 176, 58–73. [CrossRef]
27. Guo, M.; Dong, X.; Li, J. Study on the Earth Pressure during Sinking Stage of Super Large Caisson Foundation. Appl. Sci. 2021, 11,
10488. [CrossRef]
28. Chu, J.; Ma, J.; Jiang, B.; Li, M.; Zhang, K. Experimental Study on Side Friction Distribution of Caissons during Sinking in Water.
Chin. J. Geotech. Eng. 2019, 41, 707–716.
29. Yan, F.; Shi, G. Analysis of Limiting Soil Resistance beneath Cuttingcurb during Sinking Ofopen Caisson. Rock. Soil. Mech. 2013,
34 (add 1), 80–87.
30. Royston, R.; Sheil, B.B.; Byrne, B.W. Undrained Bearing Capacity of the Cutting Face for an Open Caisson. Géotechnique 2022, 72,
632–641. [CrossRef]
31. Chavda, J.T.; Mishra, S.; Dodagoudar, G.R. Experimental Evaluation of Ultimate Bearing Capacity of the Cutting Edge of an Open
Caisson. Int. J. Phys. Model. Geotech. 2020, 20, 281–294. [CrossRef]
32. Mu, B.; Bie, Q.; Zhao, X.; Gong, W. Meso-Experiment on Caisson Load Distribution Characteristics During Sinking. China J.
Highw. Transp. 2014, 27, 49–56.
33. Gaudio, D.; Madabhushi, S.G.; Rampello, S.; Viggiani, G.M. Experimental Investigation of the Seismic Performance of Caisson
Foundations Supporting Bridge Piers. Géotechnique 2022, 74, 892–906. [CrossRef]
34. Wang, W.; Wang, H.; Xu, Z. Experimental Study of Parameters of Hardening Soil Model for Numerical Analysis of Excavations of
Foundation Pits. Rock. Soil. Mech. 2012, 33, 2283–2290.
35. Yang, L.; Zhou, L. Experimental Study of Parameters of HS Model for Numerical Analysis of Foundation Pits in Ningbo Soft Soil
Area. Tunn. Constr. 2018, 38, 954–962.
Buildings 2024, 14, 2402 25 of 25

36. Itasca Consulting Group, Inc. Fast Language Analysis of Continua in 3 Dimensions, Version 6.0, User’s Mannual; Itasca Consulting
Group, Inc.: Minneapolis, MN, USA, 2017.
37. Brinkgreve, R.B. Selection of Soil Models and Parameters for Geotechnical Engineering Application. In Soil Constitutive Models:
Evaluation, Selection, and Calibration, Proceedings of the GeoFrontiers 2005, Austin, Texas, 24–26 January 2005; American Society of
Civil Engineers: Reston, VA, USA, 2005; pp. 69–98.
38. Schanz, T.; Vermeer, P.; Bonnier, P.G. The Hardening Soil Model: Formulation and Verification. In Beyond 2000 in Computational
Geotechnics, 1st ed.; Routledge: London, UK, 1999; pp. 281–296.
39. Ma, J.; Wang, Y.; Li, H.; Wang, G. Study on the relations of cement-Soil parameters with compressive strength. Build. Sci. 2009, 25,
65–67.
40. Li, J.; Liang, R. Research on compression strength and modulus ofdeformation of cemented soil. Rock. Soil. Mech. 2009, 30,
473–477.
41. Liang, Z.; Li, Z.; Liu, J.; Weng, X.; Li, W. Strength analysis and experimental research on soil-cement mixed pile made by triaxial
stirring machine. Chin. Joumal Undergr. Space Eng. 2009, 5, 1562–1566.
42. Li, J.; Wang, S. Indoor experimentation and analysis of soil-and-cement tensile strength. J. Hebei Inst. Archit. Sci. Technol. 1998, 15,
35–38.
43. Sun, Y.; Shi, D. Test study to determine the tensile strength of cement-Soil. Struct. Eng. 2003, 6, 70–73. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.