Professional Documents
Culture Documents
ABSTRACT
The failure mode should be taken into account in the analytical solution for pile base resistance. In order to
determine the pile’s failure pattern in soft rock, the graphical model test method is used. That the compression
zone below the pile base in soft rock is compatible with the spherical cavity expansion is shown. An improved
calculation model for the ultimate base resistance may be developed by considering the force balance of the
rigid cone under the pile-end in the ultimate stress state and the limit pressure required to expand a spherical
cavity. To account for the influence of socket length, the limiting pressure for the spherical cavity expansion
in infinite space is adjusted to that in semi-infinite space. The ultimate resistance of a base is computed by
factoring in the length effect. The normalization of the base resistance-displacement relation is provided based
ARTICLE
on the findings of pressure chamber experiments on pile load-settlement, which are independent of socket
length and overburden pressure. This semi-empirical technique is verified via the use of pressure chamber
experiments and field testing of piles in soft rock.
KEYWORDS: Rock-Socketed Pile, Failure Mode, Soft Rock, Spherical Cavity Expansion, Base Resistance.
176 Sci. Adv. Mater. 2023, Vol. 15, No. 2 1947-2935/2023/15/176/011 doi:10.1166/sam.2023.4408
Cai et al. An Application of Spherical Cavity Expansion Theory in Soft Rock Pile-Base Resistance
rock mass. Theory models for base resistance were con- 2.2. Instrumented Model Pile Test Results
structed using the spherical cavity expansion theory in infi- The model pile tests were carried out on the seventh day
nite space, analogous to the way piles in soil are formed. after sample preparation. The parameters of model pile
The theoretical solution of spherical cavity expansion in tests in soft rock are depicted in Table II. The relation-
infinite space and in semi-infinite space should be same ships between pile base resistance and displacement in
due to the enormous length-to-diameter ratio of piles in instrumented pile tests are shown in Figure 3. The basic
soil. However, soft rock has restrictions on the length of response curves are obviously work-hardening. As the set-
the pile socket. This work provides a semi-empirical solu- tlement advances, the base resistance first rises dramati-
tion to depth effect and explains the failure mechanism of cally, but then its rate of increase begins to reduce, and it
piles in soft rock. The novel design technique of pile base eventually tends to level out.
resistance in soft rock is validated with certain examples. Figure 4 depicts the averaged connection between base
resistance and motion. It seems that normalizing base
2. SOFT ROCK SIMULATED AND PRESSURE resistance and displacement for a variety of overburden
CHAMBER TESTING FOR MODEL PILE pressures and rock socket depths is a smart idea. The anal-
2.1. Material Properties and Testing Apparatus ogous pile base displacement at 50% mobilization of base
resistance is about 0.04 d, and a pile tip displacement up
In order to get the necessary strength and modulus value
to about 0.25 d is required for full mobilization of base
for the soft rock simulation sample, a mix ratio test of
resistance.
cement, gypsum, medium sand, early strength agent, and
As a point of comparison, the comparable displacement
water is conducted. Additionally, it is more practical for
modelling testing of the rock-socketed pile in soft rock in soft rock at 50% and 100% mobilization of base resis-
with the need for consistent strength after 7 days. When tance is much greater than that in clay or sand, which is
testing piles, use a simulated soft rock with a uniaxial suggested by the American API Specification [17] to be
ARTICLE
compressive strength of 1.5 MPa. Cement-to-water ratio of 0.013 d and 0.1 d, respectively. The two ratio values in soft
0.50, early strength agent-to-cement ratio of 0.23, and den- rock obtained in pressure chamber tests are in rough agree-
sity of 1.80 g/cm3 make up the material mix ratios. Table I ment with the field experiment outcomes [3], in which
and Figure 1 show the results of triaxial tests conducted the displacement ratios are 0.035 and 0.20, respectively,
on model soft rock with a uniaxial compressive strength and quite close to displacement ratios of 0.04∼0.06 and
of 1.5 MPa. ∼0.25, respectively [18, 19]. The maximum base resis-
For instrumented pile testing, a 40 mm diameter model tance is mobilized at a settlement exceeding 25% of the
pile is used. Model pile, nylon model cylinder, pile top- diameter [20].
loading system, pile-surrounding soil/rock loading system,
response system, and measurement system are all part of 2.3. The Failure Mode of Pile in Soft Rock
the pressure chamber used for single pile testing, as illus- Ladanyi [9] pointed out that the deformation bubble
trated in Figure 2. The equivalent modulus of the solid beneath the pile base following loading in sandy soil was
model pile is 11.4 GPa, and it is based on a cylindrical comparable to the spherical cavity expansion in the infinite
aluminum tube with an outside diameter of 40 mm and a medium. The identical finding was uncovered by Shaaban
wall thickness of 2 mm. In order to measure shaft friction et al. [10] using excavation methods. The failure mech-
and base resistance, strain gauges are fastened to the inte- anism of the pile foundation in soft rock is explored by
rior of a 40 mm model pile. CT scanning was utilized to CT, and Figure 5 displays several scanning photographs of
observe the deformation around the pile and disclose the pile testing with and without markers in soft rock. Defor-
failure mode of the pile, allowing for direct acquisition of mation distribution around the pile base cannot be iden-
the bearing mechanism of the rock-socketed pile in soft tified in unmarked soft rock, as illustrated in Figure 5(a),
rock. The pile employed in CT scanning was 24 mm in although a lighter color may be observed at around 1.2–
diameter and was made of solid aluminum. The CT scan- 1.3 d beneath the pile base. A higher density, represented
ner can X-ray the whole chamber, including the nylon and by a lighter color, suggests that the soft rock in this loca-
aluminum testing pile, with no problems [14]. tion has been crushed and densified. For the identical test
settings, as shown in Figure 5(b), only the marked layers
Table I. Triaxial test results for soft rock simulation samples. are positioned in a predetermined range at the base of the
pile, with no markers present on the pile side. Deforma-
Specimen 3 : 1 − 3 f : Ei : c:
tion of the soft rock close to the pile base is seen to be
no. MPa MPa 1f : % vmax : % MPa MPa :
spherical, with an effect radius of about 1.2–1.3 d and a
1 0.01 1.51 1.13 0.33 162 0.34 43.4 little lighter color, as shown in Figure 5(a). By comparing
2 0.10 2.12 1.55 0.42 174 Figures 5(a) and (b), we can infer that the indicated layers
3 0.30 3.13 1.92 0.53 231
4 0.50 3.61 2.20 0.72 243 are arranged in a way that is conducive to studying the
deformation distribution and failure mechanism of the pile
Fig. 1. Triaxial test results for soft rock simulation samples: (a) Deviator stress-axial strain curve, (b) Volumetric strain-axial strain curve and (c)
Mohr envelope.
Table II. The parameters of model pile tests in soft rock. The photos show that soft rock under the pile base was
compressed due to the pile penetration and then crushed
Test c : c: Em : Ec : v :
no. MPa MPa :
MPa K0 GPa kPa L: mm d: mm and denser rock area resembles a hemisphere. The spher-
ical cavity expansion under the pile base reflected by the
T1 1.5 0.34 43.4 206 0.31 0.24 11.4 300 120 40 disassembly photos of the pile end and the CT scanning
T2 1.5 0.34 43.4 184 0.31 0.24 11.4 100 200 40
T3 1.5 0.34 43.4 184 0.31 0.24 11.4 100 240 40
photos of the modelling test are consistent.
T4 1.5 0.34 43.4 218 0.31 0.24 11.4 500 240 40 The CT scanning model experiments of rock-socketed
pile in soft rock in diverse load pressure, different rock-
socketed depth and different pile roughness were also
36 carried out [14]. There is no heave deformation and slip
32 surface appearing, which is different from the failure mode
28 of shear slip surface put forward by Terzaghi, Vesic [21]
24
and Meyerhof [22]. The process of failure of a model pile
socketed into simulated soft rock was studied by Johnston
qb: MPa
20
and Choi [23] using stereophotogrammetric methods. The
16
research reveals that a fan-shaped wedge eventually forms
12 beneath the pile base with a ring fracture developing from
8 Embedment=3d, overburden=300kPa the margins of the pile. The shape of the wedge would
Embedment=5d, overburden=100kPa
4 Embedment=6d, overburden=100kPa
be spheric while the pile is actually axisymmetric. Around
0
Embedment=6d, overburden=500kPa 40% settlement ratio was recorded by Radhakrishnan and
0 3 6 9 12 15 Leung and Ko [19], at which point a conspicuous failure
s : mm
bulb/stress plug was seen at the pile base. Deformation of
ARTICLE
Fig. 3. The relation between the base resistance and the displacement. soft rock beneath the pile foundation is the spherical out-
ward expansion, according to CT scanning data. To accu-
1.0 rately determine the base resistance of rock-socketed piles
in soft rock, the spherical cavity expansion theory should
0.8 be used.
qb s/d
=
qb,ult 0.81·(s / d)+0.048
0.6
3. BASE RESISTANCE DESIGN APPROACH
qb/qb,ult
Fig. 5. CT scanning photos of pile testing under the overburden pressure of 300 kPa, (a) without the marker and (b) with the markers.
ARTICLE
Fig. 6. CT scanning photos of pile at different penetration depth: (a) Pile penetration of 0.4 d and (b) pile penetration of 2 d.
Fig. 7. CT scanning photos of cross section beneath pile end: (a) The layout of cross Section 1 and (b) cross Section 1.
Upon the simplification of the formula (1), there is: The limit pressure pu can be calculated by the equations
deduced by the Mohr-Coulomb plastic yield model [25]:
qb,ult = pu 1 + tg · tg + c · tg (2) 31 + sin
pu = q + c · ctan I 4 sin /31+sin
3 − sin rr
where c and are the cohesion and friction angle of
soil/rock respectively; is the horizontal angle of rigid − c · ctan (4)
cone under pile base; pu is the limit/ultimate pressure for Ir 1 +
ARTICLE
the spherical cavity expansion. Irr = (5)
1 + Ir
Figure 5 depicts CT scanning images showing that when
3 − sin
the pile is under its ultimate weight, a sphere expands = (6)
3 cos
below the pile base, with an influence depth of 1.2∼1.3 d.
Em
The cone depth beneath the pile foundation is 1.2 d when Ir = (7)
the horizontal angle of the rigid cone is equal to 66.7. 21 + c + q · tan
This demonstrates that the following connection holds true where Em and are soil/deformation rock’s modulus and
for the rock-socketed pile in soft rock, just as it does for Poisson’s ratio, respectively; Mean initial stress in the soil
the stiff cone in soil [12]: or rock, denoted by q; The index of stiffness, or Ir ; Com-
pressibility owing to compression and shear is reflected in
the reduced rigidity index Irr , is the average volumetric
= + (3)
4 2 strain in plastic zone.
Ladanyi and Johnston [24] summed up the test data of
sand in different places and with various density, and sug-
gested a relationship between and Ir :
= 50Ir −18 (8)
1.0
0.9 PMT method (CGS)
0.8
Socket depth factor λ
0.7
0.6
0.5
Cavity expansion method
0.4 y = 0.7737ln(x) - 0.5807
0.3
0.2
0.1
0.0
0 1 2 3 4 5 6 7 8 9 10
Socket ratio L/d
Fig. 10. Relationship between the factor of socket depth and the
Fig. 9. Relation of cavity expansion limit pressure end-bearing capacity. socket ratio L/d.
The average volumetric strain of the plastic zone calcu- pressure between semi-infinite space and infinite space or
lated through Eq. (8), is 0.41%∼0.63% under the vertical equivalent to the semi-infinite space. The cavity expansion
stress of 100∼500 kPa. From the triaxial test depicted in in the half-space is only limited to the analytical solu-
Table I and Figure 1, it is known that the volumetric strain tion of the elastic problem, but the elastoplastic analytical
due to plastic yielding, i. e., maximum shearing volumetric solution in the soft rock half-space is very complicated.
change, under confining pressure of 54∼270 kPa (corre- The limit pressure for spherical cavity expansion in infinite
sponding to the vertical stress of 100∼500 kPa in chamber space may be changed using the semi-empirical technique.
test) is 0.37%∼0.53%, which is consistent with the calcu- A socket depth factor is required in front of the pu in
lated value. Thus, the empirical proposed Eq. (8) is also Eq. (2). This factor is certainly less than or equal to 1.
applicable to soft rock [26]. The equation of the pile base resistance can be obtained
by taking into account the modification factor :
3.2. Socket Depth Modification
Williams et al. [3] found that the base resistance does qb,ult = pu 1 + tg · tg + c · tg (9)
not rise indefinitely with the rock-socketed depth. There
Calculate the socket length factor by substituting the
is a depth effect and depth limit. Equation (2) of pile
pile base resistance qb,ult measured by the chamber tests
base resistance based on spherical cavity expansion the-
and the limit pressure pu for spherical cavity expansion
ory does not take into account the influence of socket
calculated through Eq. (4) into Eq. (9). As the deduced
depth, but from the test results, the influence is greater
results are seen in Figure 10, the socket depth factor can
when the socket depth is less than 6 d. This is due to
be expressed in the following equation:
the fact that the rock-socketed pile in soft rock, in actu-
ality, has a specific socket depth, i.e., the rock is not of = 07737 · lnL/d − 05807 (10)
infinite space (Eq. (4)), and therefore the limit pressure
ARTICLE
for spherical cavity expansion cannot be determined. How- Equation (10) is applicable to the socket depth of 3∼6 d
ever, the effect of the soil above should mean that it is not and the appropriate extrapolation. The limit pressure for
entirely semi-infinite. It can be regarded as the overburden spherical cavity expansion is changed from infinite space
(a) (b)
14 25
12
20
10 Measured
15 FHWA method
qb: MPa
qb: MPa
8
Williams, Johnston & Donald
Cavity expansion method
6
10
Measured
4
FHWA method
5
2 Williams, Johnston & Donald
Cavity expansion method
0 0
0 2 4 6 8 10 12 14 0 2 4 6 8 10
s: mm s: mm
(c) 35 (d)
40
30 35
25 30
Measured 25
20 Measured
qb: MPa
qb: MPa
FHWA method
20 FHWA method
15 Williams, Johnston & Donald Williams, Johnston & Donald
Cavity expansion method 15 Cavity expansion method
10
10
5
5
0 0
0 2 4 6 8 10 12 0 2 4 6 8 10 12 14
s: mm s: mm
Fig. 11. Measured and calculated base resistance versus displacement in pressure chamber tests: (a) T1; (b) T2; (c) T3; (d) T4.
Table III. The parameters of field pile tests (Williams et al., 1980; Pells & Turner, 1980).
Rock type Test no. c : MPa c: MPa : : kN/m3 Em : MPa K0 Ec : GPa L: mm d: mm Ref.
Mudstone M14 1.83 0.47 36 20 291 0.41 0.29 30 2000 100 Williams et al., 1980
Mudstone M16 2.12 0.54 36 20 313 0.41 0.29 30 3000 1000
Mudstone M18 1.53 0.39 36 20 266 0.41 0.29 30 2000 1000
Sandstone D3 6 1.40 40 21 527 0.36 0.26 30 530 290 Pells & Turner, 1980
Sandstone E1 6 1.40 40 21 527 0.36 0.26 30 500 290
Sandstone E2 6 1.40 40 21 527 0.36 0.26 30 370 710
Notes: c = c (1 − sin )/2cos, and Em = 680 pa (c /pa )05 (Rowe and Armitage, 1987) [4].
to semi-infinite space, and the value of the factor will no changes with the depth, which implies that the small
certainly decrease. Therefore, when is greater than 1.0, socket depth has little effect on the maximum pressure for
take 1.0. The Canadian Foundation Engineering Manual spherical cavity expansion. As socket depth is over 2.5 d,
(CGS-1985) [27] proposed a PMT technique to evaluate the factor increases significantly with the depth, suggesting
the base resistance of rock socketed pile, giving an equiv- that the socket depth has a great effect. However, when
alent socket depth factor of 0.15∼1.0 while socket ratio the rock socket depth is over 8 d, the factor reaches the
is 0∼7 (Fig. 10). For the rock-socketed depth less than maximum value of 1.0, reflecting the limit pressure’s depth
2.5 d, the factor is smaller than 0.15, which is the lower effect. The maximum depth is similar to that proposed by
limit referring to CGS. Therefore, a factor value of 0.15 is CGS.
taken during socket depth of 0∼2.5 d. It is worth mentioning that the empirical relation-
The relationship between the socket depth factor and ship between socket depth factor and socket ratio con-
ARTICLE
the socket ratio L/d is depicted in Figure 10. It shows that tains the influence of pile diameter, implying size-effect
when the socket depth is small, the socket depth factor has of pile diameter. Since the ultimate base resistance is
(a) (b)
35 14
30 12
25 10
20 8
qb: MPa
qb: MPa
Measured
15 FHWA method 6
Williams, Johnston & Donald
10 Cavity expansion method 4 Measured
FHWA method
5 2 Williams, Johnston & Donald
Cavity expansion method
0 0
0 5 10 15 20 0 50 100 150 200
s: mm s: mm
(c) 12
10
8
qb: MPa
4
Measured
FHWA method
2
Williams, Johnston & Donald
Cavity expansion method
0
0 50 100 150 200
s: mm
Fig. 12. Measured and calculated base resistance versus displacement in mudstone: (a) M14; (b) M16; (c) M18; open symbols in the plot from
Williams et al. (1980).
(a) (b)
16 16
14 14
12 12
10 10
qb: MPa
qb: MPa
8 8
6 6
Measured Measured
4 FHWA method 4 FHWA method
2 Williams, Johnston & Donald 2 Williams, Johnston & Donald
Cavity expansion method Cavity expansion method
0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30 35
s: mm s: mm
(c) 7
4
qb: MPa
2 Measured
FHWA method
1 Williams, Johnston & Donald
Cavity expansion method
0
0 3 6 9 12 15
s: mm
Fig. 13. Measured and calculated base resistance versus displacement in sandstone: (a) D3; (b) E1; (c) E2; open symbols in the plot from Pells &
Turner (1980).
Turner) [2]. The parameters of field pile tests are given socket length empirical factor approach. Comparing
in Table III. The test results and the comparisons between observed and computed base resistance-displacement
calculated and measured base resistance versus displace- relations reveals the cavity expansion approach is milder
ment in mudstone and sandstone field tests are displayed and more reasonable than the FHWA and Williams et al.
in Figures 12 and 13, respectively. methods.
In Figure 12, The cavity expansion and Williams et al.’s (f) Triaxial tests are straightforward to perform and pro-
methods provided a relatively close match with the base vide the ultimate base resistance model values. The
resistance-displacement relation obtained experimentally model is more sensitive to deformation modulus, socket
with the socket ratio L/d of 20, but the FHWA method length ratio, and friction angle. Base resistance design
gives much lower base resistance which ultimate base considers pile size.
resistance is less than 30% of the measured. While the
socket ratio L/d is no more than 3, all methods give
Ethical Compliance
lower base resistance than the measured, and the cav-
Research experiments conducted in this article with
ity expansion and FHWA methods’ results are more con-
animals or humans were approved by the Ethical Com-
servative. It is clear in Figure 13 that both the cavity
mittee and responsible authorities of our research organi-
expansion and FHWA methods provide curves close to
zation(s) following all guidelines, regulations, legal, and
the base resistance-displacement relation obtained in sand-
ethical standards as required for humans or animals.
stone field tests (Pells & Turner, 1980). However, Williams
et al.’s method gives a much higher base resistance which
is about 1.5 times the measured. The cavity expansion Conflicts of Interest
method appears to be relatively moderate based on obser- There are no conflicts to declare.
vations of the measured and calculated base resistance-
Acknowledgments: The first author gratefully
ARTICLE
displacement relations discussed above, while the FHWA
method is slightly underestimated when the socket ratio acknowledges the State Key Laboratory of Subtropical
L/d is greater than 5, and Williams et al. method’s is occa- Building Science (Grant No. 2018ZB08). The work is
sionally risky and uncertain. also supported by Undergraduate Online Courses Com-
mittee of University in Guangdong Province (Grant No.
2022ZXKC437).
5. CONCLUSIONS
In this study, a base resistance design approach is sug-
gested in a semi-infinite area, and the failure mechanism NOTATION
of the pile in soft rock is disclosed by CT visualization. c Cohesion
Results show: d Diameter of pile
(a) Base resistance and displacement may be normalized Ec Young’s modulus of pile
under various overburden pressures and socket lengths. Ei Initial tangent modulus
Hyperbolic functions fit normalizing curves. At 50% and Em Deformation modulus
100% of ultimate base resistance, pile base displacement Ir Rigidity index
is 0.04 d and 0.25 d. I rr Reduced rigidity index
(b) CT scanning in pressure chamber testing proves K0 Coefficient of lateral earth pressure at rest
spherical cavity expansion for piles in soft rock inde- L Socket length
pendent of overburden pressure, socket length, or shaft pu Limit pressure for the spherical cavity expan-
roughness. The compression zone below the pile base is sion
1.2∼1.3 d with no heave deformation or slip surface. q Mean initial stress
(c) The spherical cavity expansion hypothesis, which qb Base resistance of pile
matches the collapse mechanism, should calculate pile qbult Base resistance of pile in ultimate stress state
base resistance in soft rock. For socket lengths more than s Displacement of pile end
8 d, ultimate base resistance is calculated using the limit Unit weight
pressure based on spherical cavity expansion and friction Average volumetric strain of the plastic zone
that follows the Mohr-Coulomb failure criteria. v Volumetric strain
(d) Depth impact affects ultimate base resistance for vmax Maximum volumetric strain
sockets under 8 d. In semi-infinite space, the spherical 1 Axial strain
cavity expansion theory solution of ultimate base resis- 1f Axial strain at failure
tance must be adjusted. The socket length factor is semi- Socket length factor
empirical. Poisson’s ratio
(e) Pressure chamber and field experiments verify the c Uniaxial compressive strength
semi-empirical spherical cavity expansion theory and v Overburden pressure
1 Major principal stress of rock-socketed piles in soft rock based on micro X-ray CT
3 Minor principal stress analysis. Rock Mechanics and Rock Engineering, 53(8), pp.3395–
Friction of the rigid cone 3416.
16. Gharsallaoui, H., Jafari, M. and Holeyman, A., 2020. Pile end bear-
Friction angle ing capacity in rock mass using cavity expansion theory. Journal
Horizontal angle of the rigid cone of Rock Mechanics and Geotechnical Engineering, 12(5), pp.1103–
1111.
17. API, 2007. API RP 2A-WSD: Recommended Practice for Planning,
References and Notes Designing and Constructing Fixed Offshore Platform-Working Stress
1. Luo, Y.P., Yan, Z.X., Deng, Y.J. and Lu, X.C., 2017. Application Design. 21st Edition, Washington, DC, API, Errata and Supplement
of charge coupled device spectrometer in oil-bearing rock detection. 3.
Journal of Nanoelectronics and Optoelectronics, 12(12), pp.1391– 18. Radhakrishnan, R. and Leung, C.F., 1989. Load transfer behavior of
1396. rock-socketed piles. Journal of Geotechnical Engineering, 115(6),
2. Pells, P.J. and Turner, R.M., 1980. Endbearing on Rock with Partic- pp.755–768.
ular Reference to Sandstone. International Conference on Structural 19. Leung, C.F. and Ko, H.Y., 1993. Centrifuge model study of
Foundations on Rock, Sydeney, Australia, 1, pp.181–190. piles socketed in soft rock. Soils and Foundations, 33(3),
3. Williams, A.F., Johnston, I.W. and Donald, I.B., 1980. The Design pp.80–91.
of Socketed Piles in Weak Rock. International Conference on Struc- 20. Yang, Q., Cheng, S. and Zhou, B., 2020. Monitoring study on verti-
tural Foundations on Rock, Sydeney, Australia, 1, pp.327–347. cal bearing capacity of pile foundation in soft rock of Lhasa human
4. Rowe, R.K. and Armitage, H.H., 1987. A design method for drilled settlements. Journal of Intelligent & Fuzzy Systems, 38(6), pp.7639–
piers in soft rock. Canadian Geotechnical Journal, 24(1), pp.126– 7650.
142. 21. Vesic, A.S., 1967. Ultimate loads and settlements of deep foun-
5. Zhang, L. and Einstein, H.H., 1998. End bearing capacity of drilled dations in sand, edited by, A.S. Vesic, Bearing Capacity and Set-
shafts in rock. Journal of Geotechnical and Geoenvironmental Engi- tlement of Foundations. Durham, North Carolina, Duke University,
neering, 124(7), pp.574–584. pp.53–68.
6. Serrano, A. and Olalla, C., 2002. Ultimate bearing capacity at the 22. Meyerhof, G.G., 1976. Bearing capacity and settlement of pile
tip of a pile in rock—Part 1: Theory. International Journal of Rock foundation. Journal of the Geotechnical Engineering Division, 102,
ARTICLE