Ultrasonics 93 (2019) 107–111

Contents lists available at ScienceDirect

Ultrasonics
journal homepage: www.elsevier.com/locate/ultras

Short communication

An acoustic emission characterization of the failure process of shallow T

foundation resting on sandy soils
⁎
Wuwei Maoa,b, , Yang Yangb, Wenli Linc
a
Department of Geotechnical Engineering, College of Civil Engineering, Tongji University, Shanghai 200092, China
b
Key Laboratory of Geotechnical and Underground Engineering, Ministry of Education, Tongji University, Shanghai 200092, China
c
Department of Civil Engineering, The University of Tokyo, Tokyo 113-8656, Japan

A R T I C LE I N FO A B S T R A C T

Keywords: Shallow foundation is a common foundation type that is usually used for small to medium size structures. The
Acoustic emission bearing ability and the failure mechanism of shallow foundation are the fundamental concerns for geotechnical
Shallow foundation engineers, and the demand for new insights into the relevant issue is still increasing. This paper presents an
Failure acoustic emission (AE) characterization of the failure process of shallow foundation, with the aim of revealing
Sand
the fundamental information on AE signals associated with shallow foundation loading as well as its connection
Kaiser Effect
with the ground bearing behavior. Experiments were carried out to model the failure process of shallow foun-
dation resting on sandy ground with different densities (i.e. loose and dense) and subjected to different loading
conditions (i.e. monotonic and cyclic loading). Comparisons between AE activities and ground bearing behavior
are presented. The feasibility of using AE for stability monitoring of shallow foundation is revealed and dis-
cussed.

1. Introduction
Shallow foundation is widely used for small to medium size struc-
tures. In order to ensure the safety of the structure, its foundation is
expected to meet the strength conditions. That is, the required foun-
dation should has sufficient ability to bear the load to resist sliding
damage due to insufficient shear strength. Accordingly, determination
of the bearing capacity of the foundation is of fundamental importance
for foundation design, and the clarification of the ground failure me-
chanism as well as the evaluation of shallow foundation bearing ca-
pacity have been of great interest among researchers [1–7]. Compared
with the damage or failure assessment of the upper structures, the
foundation failure is often difficult to be detected since it is hidden
under the ground. Up to now, accurate determination of bearing ca-
pacity of foundations remains an intractable problem in the field of
geotechnical engineering [8], and there are ongoing demands to have
more insights into the ground bearing behavior and its failure me-
chanism [9,10].
Usually, the ultimate bearing capacity of a foundation is the major
concern for the engineering designers. However, apart from the foun-
dation’s bearing behavior during the service stage, the stability of a
foundation during the construction operations also deserves close at-
tention. In a sense, the foundation during the construction period may
experience the most critical condition in its life cycle. One example is
that a 13-storey apartment building toppled over almost intact due to
the foundation failure in Shanghai, China, which is believed to be
caused by a combination of the temporary dumping of the dug-out soil
against one side and the excavation of an underground garage on the
other [11]. Therefore, the status of foundation during construction
process should also be properly monitored.
Load and displacement are routinely used parameters for the se-
verability monitoring of upper structures. Regarding shallow founda-
tions, it should be noted that the bearing load monitoring on site is
relatively difficult, while the displacement monitoring of ground sur-
face or foundation itself may also not reveal the ground failure me-
chanism. Accordingly, this study introduces the acoustic emission
method to monitoring the failure process of shallow foundations. As a
non-intrusive method, the AE method has been demonstrated to have
the ability to detect the impending failure of rock, concrete and com-
posite materials [12–14]. However, it has seldom been used for mon-
itoring of granular soils due to the complex nature of the soil materials.
The AE signals could experience severe attenuation in a granular
system, which brings measurement and interpretation difficulties.
Nevertheless, the feasibility of the AE method applied to soil materials
and its benefits have long been evidenced [15–20]. The earlier works
mainly focused on revealing the qualitative relations between the AE

⁎
Corresponding author at: Department of Geotechnical Engineering, College of Civil Engineering, Tongji University, Shanghai 200092, China.
E-mail address: maowuwei@tongji.edu.cn (W. Mao).

https://doi.org/10.1016/j.ultras.2018.11.007
Received 9 June 2018; Received in revised form 22 October 2018; Accepted 20 November 2018
Available online 22 November 2018
0041-624X/ © 2018 Elsevier B.V. All rights reserved.
W. Mao et al. Ultrasonics 93 (2019) 107–111

Fig. 1. Typical failure pattern and the relevant load-displacement curves for shallow foundations resting on soil.

activity (mostly represented by the ring-down count) and different

types/conditions of stressed soils; while more recent advances have
been made on establishing a potential approach toward solving a spe-
cific geotechnical problem, and the interpretation of AE have been
extended to a variety of parameters including spectrum analysis. Ty-
pical examples of such applications include the AE monitoring system
for early warning of soil slope failure [21,22], breakage behavior of soil
particle [23], seepage of soils [24], erosion of soils [25], pile bearing
behavior in soils [26,27], and etc. This study presents a follow-up effort
of such applications regarding AE generation in a shallow foundation
situation. When a shallow foundation is subjected to load, the potential
internal damage in the ground body would generate AE signals, pro-
pagate to the ground surface and be picked up by the AE sensor, as
schematically illustrated in Fig. 1.
According to the previous studies [1,28], the failure mode of
shallow foundations can be generally divided into three categories, i.e.
general shear failure, local shear failure and punching failure, as shown
in Fig. 1. The general shear failure is usually seen in stiff or dense soil,
characterized by a well-defined failure surface extended to the ground
surface, and a pronounced peak in the load displacement curve. The
local shear failure is seen in relatively loose soils and the failure surface
is not well developed and does not expose to the ground surface. Slight
bulging of ground is expected and there is no well-defined peak in the
load displacement curve. The punching shear failure is usually seen in
very soft ground and is characterized by significant large settlement.
Ideally, no increase in load is expected when the displacement is con- Fig. 2. Layout of shallow foundation loading system equipped with AE sensor.
tinuously increasing for punching shear failure. In this study, loading
tests were examined on shallow foundations resting on two typical
types of the above sand grounds, i.e. a dense ground and a relatively
loose ground. The AE signals were detected in the respective loading
tests, and the data of AE activity and the ground bearing load were
synchronized to verify the feasibility of the AE method applied to
shallow ground monitoring.

2. Experimental arrangements

A schematic illustration of the test setup is shown in Fig. 2. The

loading tests were conducted in a conventional motor loading frame
assembly. The sands were prepared in a steel soil tank with internal
dimensions 600 mm × 600 mm × 500 mm. The model footing used for
the tests was made of rigid steel, square in shape, and measured
40 × 62 mm in the bottom surface. The tests were performed by a
displacement controlled manner with a constant loading rate of 1 mm/
min. Load cell and displacement meter were used to capture the load-
settlement behavior of the footing.
The AE signals were detected using AE sensor attached to the side of
the footing. The piezo-ceramic type AE sensor, manufactured by Fuji
Ceramics Corporation, has an optimum operating frequency range of
10 kHz–2 MHz. A main amplifier was used to enhance the AE signals
with a gain of 115 dB when cooperated with the head amplifier em-
bedded within the sensor. The sampling rate of AE data recording
system was set at 2 Mps.
Silica sand No. 5 was used as the model ground material. This sand
is classified as the poorly-graded sand according to the Unified Soils
Fig. 3. Grain size distribution of the tested sand.
Classification System [29] and has specific gravity (Gs) of 2.65, max-
imum void ratio (emax) of 1.09, minimum void ratio (emin) of 0.66,
uniformity coefficient (Cu) of 1.692, and mean grain size (D50) of
0.557 mm. The grain size distribution of the sand is shown in Fig. 3.
The air pluviation method, i.e. sprinkling the sand from the air at a
constant falling height, was used for preparing the model ground. In the
current study, two types of ground conditions were made, which were
the dense case and relatively loose case. For dense condition, the model
ground was further compacted after every 30 mm deposition by drop-
ping a 10 kg-weight at a height of about 10 cm 20 times all over the

108
W. Mao et al.
Fig. 4. AE energy and ground resistance as a function of footing displacement:
(a) loose ground; (b) dense ground.
surface. This procedure was accurate enough as the repeatability of
results can be easily realized. While for loose condition, no further
compaction was performed. The final relative density of the ground was
97% and 35% for dense and loose cases respectively.
3. Results and discussions
3.1. Ground resistance and AE activity
Fig. 4 shows the trends of ground resistance and AE energy devel-
opment with respect to the footing displacement. The AE energy here is
calculated by the integration of squared AE signal voltage within the
signal duration [30]. Distinct differences between the dense and loose
ground conditions can be noticed. For loose case, the ground resistance
increased gradually with the depth of footing settlement throughout the
whole loading process. In view of AE, it was detected throughout the
above process as well, and consequently the accumulative AE energy
kept rising as the footing displacement increased. In general, the AE
energy rate, represented by the summation of AE energy of each 10 s
interval, increased with the footing displacement. For dense case, the
ground resistance was firstly increased rapidly to a peak value, followed
by a substantial reduction and then increased again, but with a lower
rate. The AE activity was also detected throughout the whole loading
process, except that at the very beginning of loading, the AE was much
less significant. The trend of AE rate shared similar tendency with the
development of ground resistance. For example, the parabolic feature of
ground resistance in the first 15 mm displacement can be well captured.
A rising or decreasing in the ground resistance corresponded to the
rising or decreasing of AE energy rate. In particular, notable rising of AE
is detected when the ground resistance is approaching its peak at
around 5 mm displacement. This suggests that the AE monitoring could
109
Ultrasonics 93 (2019) 107–111
Fig. 5. AE count, AE amplitude and ground resistance as a function of footing
displacement: (a) loose ground; (b) dense ground.
be potentially used for investigation of the ground bearing behavior.
Apart from the AE energy, the AE count and amplitude are also
regarded as important parameters for analysis of AE signals. The mean
AE amplitude, calculated by averaging the AE event amplitude of each
10 s interval, is used to represent the AE amplitude behavior in the
current study. It can be seen from Fig. 5 that for loose ground, the AE
amplitude is rather fluctuating, while for dense ground, it is more sta-
bilized. Overall, the values of the mean AE amplitude do not show
much difference between the loose and dense conditons. However, the
AE count in case of loose ground is almost two orders of magnitude
smaller than that of dense ground. In particular, it is notable that the AE
count was not significant until the footing displacement was larger than
∼25 mm for loose ground. Meanwhile, the corresponding mean AE
amplitude also appeared to be less fluctuated after similar depth of
footing settlement.
On the other hand, the AE amplitude and event count in case of
dense ground demonstrated more obvious relevance with the ground
resistance, especially during the first 15 mm footing settlement where a
rising or decreasing in the ground resistance corresponded to the rising
or decreasing of AE count and mean amplitude. However, when the
ground displacement was larger than 15 mm, the mean amplitude did
not show any significant change, while the AE count kept increasing
continuously, similar with the increasing trend of the ground resistance.
Compared with AE energy, it seems that the AE count and amplitude
have less correspondences with the ground resistance, especially when
the ground is relatively loose. Therefore, it is recommended to use AE
energy as the identification parameter of AE analyses.
The two loading tests described above demonstrated distinct AE
characteristics for different ground conditions. For dense ground ex-
pected to experience general failure mode, a pronounced peak in the
load and AE curve was observed. While for relatively loose ground
expected to experience local failure mode, the load and AE kept
W. Mao et al. Ultrasonics 93 (2019) 107–111

Fig. 6. Ground resistance and AE activity in case of cyclic loading condition.

increasing throughout the loading process. It is therefore suggested that

the AE could capture the failure mode of the shallow foundation as
traditionally defined by the load-displacement relations.
Fig. 7. Accumulative AE history plot demonstrating Kaiser Effect.

3.2. AE activity during cyclic loading

which consequently generates AE signals. Such process is different from
the mechanism of AE generation in metal or rock materials where the
Regarding the AE activity of stressed materials, it is generally re-
crack formation or growth is the main AE sources. If the applied load is
cognized that when the material is loaded and then unloaded, the
smaller than the previous maximum load, the crack will not suffer
corresponding AE response is different upon reloading compared to the
further growth and the AE is expected to be dim.
original loading. This is due to the well-known Kaiser Effect.
It is worth noting that after global failure of the ground, the ground
Consequently, shallow foundation subjected to historical load may re-
resistance experienced a gradual reduction period. However, the AE
sult in different AE activity. Therefore, an additional test was performed
during this period remained quite active. This is possibly due to the fact
to investigate the AE behavior of shallow foundation under pre-loaded
that after the global failure, the shear band is well formed below the
condition. 19 cycles of loading with maximum ground resistance of
foundation, and the particles within the shear band are continually
about 40 kPa (Loading Stage 1) and subsequent 7 cycles of loading with
subjected to relative displacement during loading process.
maximum ground resistance of about 60 kPa (Loading Stage 2) were
Consequently, despite that the ground resistance is reducing, the AE
applied, followed by a monotonic loading until ground failure (Loading
signals are constantly generated. Therefore, the Kaiser Effect is not
Stage 3). The ground was prepared under the same condition of the
applicable for characterization of ground situation after global failure.
above dense case, and accordingly, the applied cyclic loading stresses
Considering that the ground is usually functioning before the general
were believed to be smaller than the peak failure stress of the ground.
ground failure, it is suggested that AE could still be used as a useful tool
Fig. 6 shows the time history of the ground resistance and AE activity.
for evaluation of ground status.
One notable issue is that there existed stress plateau zones when the
It is worth noting that the current study presents the results of AE
ground resistance approached around 20–25 kPa. This might be at-
monitoring in a laboratory modelling scale. For potential applications
tributed to the local failure of the ground. Another important issue is
of the AE instrumentation in the field, the foundation could have a
that the Kaiser Effect did demonstrate its influence on AE activity. Al-
more significant length and may have other interactions with the soil in
though the AE activity did not turn to be zero during reloading steps,
place, including the existence of water, ground anisotropy, considerable
the amount of the detected AE during the reloading steps reduced
attenuation of AE signals and etc. These involve the problems of many
dramatically. When the stress exceeded the previous maximum value,
new aspects and further investigations deserve to be proceeded.
e.g. during the 20th loading cycle or after 26th loading cycle, the AE
According to the existing field applications, the above issues may not
activity became significant again.
change the validity of AE monitoring [21,31]. Although the intensity or
Fig. 7 shows the accumulative AE and the ground resistance re-
property of the AE signals might be easily affected by different field
lationship. It is clearly demonstrated that during unloading and re-
conditions, the evolution tendency could remain similar, which is more
loading process, the accumulative AE did not increase notably if the
important for field engineers to assess the initiation of damage or the
ground resistance did not exceed the previous maximum value (i.e.
impending failure of the ground.
from 1st cycle to 19th cycle during Loading Stage 1 and 20th cycle to
26th cycle during Stage 2). On the other hand, notable increment of AE
4. Conclusion
can be observed once the applied stress was larger than the previous
maximum value (i.e. from Loading Stage 1 to Stage 2 and from Stage 2
This study presents an experimental investigation of acoustic
to Stage 3).
emission during the failure process of shallow foundation. The major
For materials such as metal, rock, and concrete, the AE activity is
findings of this study can be summarized as following:
generally considered to be faint if the current stress is smaller than the
historical maximum stress due to the Kaiser Effect (or similarly the
(1) The AE is well observed throughout the failure process of the
Felicity Effect). Nevertheless, for granular soils in the current study,
shallow foundation. The rate of AE activity is related with the
although the AE become less significant during reloading process, it did
ground density and the loading procedure. In general, the evolving
occur throughout the whole process. This is because the soils consist of
trend of AE rate shared similar tendency with the development of
numerous granular particles, and any disturbance due to foundation
ground resistance.
settlement could result in the rearrangement of the granular assembles

110
W. Mao et al.
(2) The energy of AE signals can be used as a better identification
parameter for characterizing the behavior of the shallow foundation
compared with AE count or amplitude.
(3) The influence of Kaiser Effect is observed during cyclic loading of
shallow foundation. AE becomes less significant during reloading
process when the applied stress does not exceed the previous
maximum stress. However, it did occurred throughout the whole
loading process, demonstrating certain difference compare with
those observed from metal or rock materials.
(4) After the global failure of the ground, the Kaiser Effect is not ap-
plicable since the AE may continuously increase upon decreasing
ground resistance.
Acknowledgment
This work is supported by the Shanghai Sailing Program (Grant
number 18YF1424000), the Fundamental Research Funds for the
Central Universities of China (Grant number 22120170118), and
Shanghai Education Commission (Peak Discipline Construction, Grant
number 0200121005/052).
References
[1] K. Terzaghi, Theoretical Soil Mechanics, John Wiley and Sons, New York, 1943.
[2] G.G. Meyerhof, The ultimate bearing capacity of foundations on slopes, Proc. 4th
Int. Conf. on Soil Mechanics and Foundation Engineering, vol. 1, 1957, pp.
384–386.
[3] L.Y. Wu, Y.F. Tsai, Analysis of earth pressure for retaining wall and ultimate bearing
capacity for shallow foundation by variational method, J. Mech. 20 (1) (2004)
43–56.
[4] M.T. Omar, B.M. Das, S.C. Yen, V.K. Puri, E.E. Cook, Ultimate bearing capacity of
rectangular foundations on geogrid-reinforced sand, Geotech. Test. J. 16 (2) (1993)
246–252.
[5] M. Abu-Farsakh, Q. Chen, R. Sharma, An experimental evaluation of the behavior of
footings on geosynthetic-reinforced sand, Soils Found. 53 (2) (2013) 335–348.
[6] M. Chakraborty, J. Kumar, Bearing capacity of circular foundations reinforced with
geogrid sheets, Soils Found. 54 (4) (2014) 820–832.
[7] S. Saha Roy, K. Deb, Effects of aspect ratio of footings on bearing capacity for
geogrid-reinforced sand over soft soil, Geosynth. Int. 24 (4) (2017) 362–382.
[8] C. Vulpe, S. Gourvenec, M. Power, A generalised failure envelope for undrained
capacity of circular shallow foundations under general loading, Geotech. Lett. 4 (3)
(2014) 187–196.
[9] E. Conte, A. Donato, A. Troncone, Progressive failure analysis of shallow founda-
tions on soils with strain-softening behavior, Comput. Geotech. 54 (2013) 117–124.
[10] Q. Chen, M. Abu-Farsakh, Ultimate bearing capacity analysis of strip footings on
reinforced soil foundation, Soils Found. 55 (1) (2015) 74–85.
111
Ultrasonics 93 (2019) 107–111
[11] J. Chai, S. Shen, W. Ding, H. Zhu, J. Carter, Numerical investigation of the failure of
a building in Shanghai, China, Comput. Geotech. 55 (2014) 482–493.
[12] Z. Moradian, H.H. Einstein, G. Ballivy, Detection of cracking levels in brittle rocks
by parametric analysis of the acoustic emission signals, Rock Mech. Rock Eng. 49
(3) (2016) 785–800.
[13] A. Behnia, H.K. Chai, T. Shiotani, Advanced structural health monitoring of con-
crete structures with the aid of acoustic emission, Constr. Build. Mater. 65 (2014)
282–302.
[14] V. Malolan, G. Wuriti, A.S. Gopal, T. Thomas, Comparison of acoustic emission
parameters for fiber breakage and de-lamination failure mechanisms in carbon
epoxy composites, J. Eng. Technol. Res. 8 (3) (2016) 21–30.
[15] R.M. Koerner, A.E. Lord Jr, Acoustic emissions in stressed soil samples, J. Acoust.
Soc. Am. 56 (6) (1974) 1924–1927.
[16] A.E. Lord Jr, R.M. Koerner, Acoustic emissions in soils and their use in assessing
earth dam stability, J. Acoust. Soc. Am. 57 (2) (1975) 516–519.
[17] R.M. Koerner, W.M. McCabe, A.E. Lord, Acoustic emission behavior and monitoring
of soils, Acoustic Emissions in Geotechnical Engineering Practice, a symposium, 750
ASTM STP, 1981, pp. 93–141.
[18] R.M. Koerner, A.E. Lord Jr, W.L. Deutsch, Determination of prestress in granular
soils using AE, J. Geotech. Eng.-ASCE. 110 (3) (1984) 346–358.
[19] K. Tanimoto, Y. Tanaka, Yielding of soil as determined by acoustic emission, Soils
Found. 26 (3) (1986) 69–80.
[20] J. Chen, H. Wang, Y. Yao, Experimental study of nonlinear ultrasonic behavior of
soil materials during the compaction, Ultrasonics 69 (2016) 19–24.
[21] A. Smith, N. Dixon, P. Meldrum, E. Haslam, J. Chambers, Acoustic emission mon-
itoring of a soil slope: comparisons with continuous deformation measurements,
Geotech. Lett. 4 (4) (2014) 255–261.
[22] A. Zaki, H.K. Chai, H.A. Razak, T. Shiotani, Monitoring and evaluating the stability
of soil slopes: a review on various available methods and feasibility of acoustic
emission technique, C. R. Geosci. 346 (9–10) (2014) 223–232.
[23] W. Mao, I. Towhata, Monitoring of single-particle fragmentation process under
static loading using acoustic emission, Appl. Acoust. 94 (2015) 39–45.
[24] M.H. Hung, G.C. Lauchle, M.C. Wang, Seepage-induced acoustic emission in gran-
ular soils, J. Geotech. Geoenviron. 135 (4) (2009) 566–572.
[25] Z. Lu, G.V. Wilson, Acoustic measurements of soil pipe flow and internal erosion,
Soil Sci. Soc. Am. J. 76 (3) (2012) 853–866.
[26] W. Mao, S. Aoyama, I. Towhata, Feasibility study of using acoustic emission signals
for investigation of pile spacing effect on group pile behavior, Appl. Acoust. 139
(2018) 189–202.
[27] W. Mao, Y. Yang, W. Lin, S. Aoyama, I. Towhata, High frequency acoustic emissions
observed during model pile penetration in sand and implications for particle
breakage behavior, Int. J. Geomech. 18 (11) (2018) 04018143.
[28] A.S. Vesic, Analysis of ultimate loads of shallow foundations, J. Soil Mech. Found.
Div. 99 (sm1) (1973) 45–73.
[29] ASTM, Standard Practice for Classification of Soils for Engineering Purposes
(Unified Soil Classification System), D2487-11, West Conshohocken, PA, 2011.
[30] R.K. Miller, E.V.K. Hill, P.O. Moore (Eds.), Nondestructive Testing Handbook, third
ed., vol. 6, Acoustic Emission Testing, 2005.
[31] N.W. Xu, C.A. Tang, L.C. Li, Z. Zhou, C. Sha, Z.Z. Liang, J.Y. Yang, Microseismic
monitoring and stability analysis of the left bank slope in Jinping first stage hy-
dropower station in southwestern, China, Int. J. Rock Mech. Min. 48 (6) (2011)
950–963.