Volume 4, Issue 2 June 2019
Special Contribution
Rapid Load Test of Piles
Tatsunori Matsumoto
Professor, Graduate School of Science and Technology
Kanazawa University, Kanazawa, Japan
Rapid load test of piles (RLT) is an alternative to the conventional static load test (SLT) and is widely used in Japan because
of its time and cost effectivenss. In this article, history of the standardization of RLT in Japan, interpretation methods of
RLT signals, newly dveloped RLT device, and a case of comprison of th results of RLT and SLT on an open-ended steel pipe
pile are introduced.
Keywords: load-displacement relation, falling mass rapid load test method, interpretation method, standards
1. History of Rapid Load Test in Japan
The first rapid load test (RLT) in Japan was carried out by Takenaka Corporation in 1992 on a cast-in-place concrete pile.
The Statnamic was employed in the test. The well-known method of rapid load testing, Statnamic, was developed by the
collaboration of TNO in the Netherlands and Berminghammer Co. in Canada (Middendorp et al., 1993) [1].
A private research group for RLT was launched in 1993 led by Prof. Osamu Kusakabe (Prof. of Tokyo Institute of Technology
at that time, and the president of IPA at present). The research group was composed of general contractors, piling
contractors, pile load test companies, pile manufactures and academics. The objectives of the research group were 1) to
compile the existing knowledge of RLT, 2) to examine basic characteristics and applicability of RLT to Japanese soils, and
3) to establish interpretation methods of RLT. The outcomes of the activity of the research group were presented in the
1st International Statnamic Seminar (1995) [2] held in Vancouver, Canada.
The research committee for RLT was formed in Japanese Geotechnical Society (JGS) in 1997, based on the activity of the
research group, with research targets of 1) definition of rapid load test, 2) interpretation methods, 3) preparation of
testing manual and 4) standardization of testing method. The outcomes of the research committee were presented in
the 2nd International Statnamic Seminar (1998) [3] held in Tokyo, Japan. The research committee was upgraded to the
standardization committee of JGS in 1998. JGS 1815-2002: Method for Rapid Load Test of Single Piles [4] was standardized
in 2002. JGS 1815-2002 is the first standard for RLT in the world. After the standardization of RLT, use of RLT has been
widened in Japan.
Force
2. Definition of Rapid Load Test in JGS 1815-2002 Loading duration of RLT, tL
ILT
JGS 1815-2002 clearly defines RLT. Fig. 1 shows an RLT
(DLT)
illustration of loading duration and characteristic curves
of a stress-wave propagating in a pile having a length, L.
Loading duration, tL, in impact load test (ILT) is 5 to 10 ms,
while tL in RLT ranges typically from 50 to 100 ms. Stress- Time
wave caused at the pile head propagates up and down in Time
the pile with a propagation speed, c. Relative loading
Characteristic
Length, L
duration, RT, is defined as
Pile
curves of
a stress-wave
RT = tL/(2L/c) (1)
RT is the number of return traveling of the stress-wave in
Time
the pile during the loading duration, tL. If RT is equal or
2L/c
greater 5, the influence of the wave propagation in the
pile could be negligible (Nishimura et al., 1998) [5] so that 2L/c: return travelling time of stress-wave in pile
the pile body could be assumed as a rigid, although the c: longitudinal wave propagation velocity
inertial force of the pile body needs to be considered.
Dynamic load tests with the condition of RT ≥5 is defined Fig. 1. Comparison of loading duration [4]
as RLT.
3
Volume 4, Issue 2 June 2019
3. Rapid Load Test Devices
Loading mechanism of the well-known Statnamic device is launching a
reaction mass placed on the pile head by the gas explosion pressure.
At the same time, the gas explosion force pushes the pile gently
downward. The maximum acceleration of the reaction mass caused
during loading is about 5g. Hence, weight of the reaction mass is about
20% of the planned maximum load. An advantage of the Statnamic is
high loading capacity up to 60 MN. However, repeated loading is
difficult in the Statnamic.
Another type of RLT device is a falling-mass type device. Fig. 2 is an
example of falling-mass type device, called Hybridnamic, developed by
Jibanshikenjo Co. Ltd. In the falling-mass type device, a hammer mass
is free-dropped from a certain height onto the pile head through a soft
cushion on it. Relatively wide ranges of the loading duration and the
maximum force can be realised by adjusting combination of the
stiffness of the cushion and the hammer mass. A great advantage of
the Hybridnamic is that repeated loading is conducted very easily. It is
possible to conduct 10 blows (tests) on a pile in a day. It is a common
practice in the Hybridnamic to apply several blows on a pile with
Fig. 2. Falling-mass device with soft cushion
increasing the drop height of the falling hammer.
(provided by Jibanshikenjo Co., Ltd.)
4. Interpretation method of RLT signals
Generally, force applied to the pile head, Frapid, is measured
via a load cell, and acceleration at the pile head, , is Frapid
measured via accelerometers. The velocity of the pile, v, is
obtained from time integration of the measured . An Inertia of pile ，
optical displacement meter is employed to measure the Ra = M
pile head displacement, w.
If Tr = tL/(2L/c) ≥5, the pile body could be treated as a rigid Static soil Dynamic
body having a mass of M. Fig. 3 shows a modelling of pile resistance, Rw soil resistance,
and soil during RLT. Frapid is the sum of inertia of the pile, （Non-linear Rd = C v
Ra, and the soil resistance, Rsoil (Eq. 1). It is assumed that spring） （Linear dashpot）
soil resistance, Rsoil, is the sum of the static soil resistance,
Rw, and the dynamic soil resistance, Rv (Eq. 2). Rd is
assumed to be proportional to the pile velocity, v, with a Fig. 3. Modelling of pile and soil during RLT [1, 6]
constant value of damping factor, C. Hence, Rsoil is readily
obtained from the measured Frapid and (Eq. 2).
Frapid = Ra + Rsoil = M + Rsoil (1) Force (MN)
Rw ULP Rsoil max
Rsoil = Rw + Rv = Rw + Cv = Frapid − M (2) 0 1 2 3 4 5
0
Pile head disp., w (mm)
C = ( Rsoil max − Rw ULP ) / v * (3) 10 Rsoil
Rw = Rsoil − Cv (4) 20 Rw
Fig. 4 is an example of thus obtained Rsoil vs w. The point at 30
the maximum displacement, wmax, is call Unloading Point wmax
Cv*
(ULP). The velocity of the pile, v, is zero at ULP. Hence, Rsoil 40
at ULP is regarded as the maximum of the static soil Unloading point, ULP
(v = 0)
resistance, Rw ULT. The damping factor, C, is estimated from 50
the difference of Rsoil max and Rw ULT (Eq. 3). Finally, Rw vs w
Fig. 4. Unloading Point Method (ULPM) for obtaining
is obtained using C and v (Eq. 4). Rw vs w is called "derived
static load-displacement curve". static load-displacement curve
4
Volume 4, Issue 2 June 2019
If multiple blows with increasing the drop height of hammer are carried, interpretation of RLT signals becomes easier, as
shown in the next section.
5. Comparative SLT and RLT on an Open-ended Steel Pipe Pile
Results of comparative SLT and RLT on an open-ended steel pipe pile (Hoshino et al., 2012) [7] are briefly introduced in
this section.
The pile having a length of 15.5 m, an outer diameter of 1.0 m and a wall thickness of 14 mm was installed in the sandy
ground by a water jet vibratory installation method (Fig. 5). The bottom 4.5 m section of the pile was embedded into the
sand rock layer having SPT N-values greater than 50.
The conventional SLT was carried out for the pile with a loading sequence shown in Fig. 6. It took 6 hours to complete five
loading cycles of the SLT. Each load step was maintained for one hour, and finally the maximum load of 9000 kN was
applied.
SPT N-value Test
0 10 20 30 40 50 60
pile 10000
0
Fill
Pile head load (kN)
2 8000
4
Sandy silt 6000
Depth, z (m)
6
8
Sandy silt
4000
10
Silty sand 2000
12 Sand rock
14 0
16 0 1 2 3 4 5 6 7 8
Sand rock Elapsed time (hr)
18
20 Fig. 6. Loading sequence in SLT
Fig. 5. Soil profile at the test site
RLT of the pile was carried out subsequently to the SLT. A 10000
Pile head force, Frapid (kN)
total of 8 blows were applied to the pile using a hammer
mass of 22 tonnes. Fig. 7 shows Frapid vs time in the last 8000
blow. The loading duration, tL, was 60 ms that resulted in
the relative loading duration Tr = tL/(2L/c) = 9.7 (c = 5000 6000
m/s). As mentioned above, Tr =9.7 ≥ 5 satisfies the
criterion for RLT.
4000
Fig. 8 shows Fsoil vs w in all the 8 blows. The red line is the
2000
connection of ULPs which can be regarded as the static
load-displacement curve.
0
60 80 100 120 140 160
Static load-displacement curves from the SLT and the RLTs
Time, t (ms)
are compared in Fig. 9. Note here again that the RLTs were
conducted after the 5th loading cycle in the SLT. Fig. 7. An example of RLT load
Load-displacement curve in the 5th loading cycle of the SLT and Fsoil vs w from the RLTs are shown in Fig. 10, in which the
pile head displacement at the start of loading was zeroed for comparison purpose. It is seen that the curve derived from
the RLT is almost equal to that obtained from the SLT.
5
Volume 4, Issue 2 June 2019
1200 0
Soil resistance, Fsoil (kN)
1000 0
8000
6000
4000
2000
0
0 10 20 30 40
Pile head displacement (mm)
Fig. 8. Fsoil vs w from RLTs with different falling height of hammer, and ULPs
10000
Static pile head force (kN)
from SLT
8000 from RLT
6000
4000
2000
0
0 20 40 60 80 100
Pile head displacement (mm)
Fig. 9. Load-displacement curve in SLT and Fsoil vs w from RLTs
10000
Static pile head force (kN)
8000
6000
4000
from SLT
(5th loading cycle)
2000
from RLT
0
0 10 20 30 40 50 60
Pile head displacement (mm)
Fig. 10. Comparison of static load-displacement curves from SLT and RLTs
6
Volume 4, Issue 2 June 2019
Fig. 11 shows the axial forces in the pile measured in the SLT and the RLTs at the maximum pile head load of 9000 kN, and
the distributions of the shaft resistance estimated from the axial forces in the pile. The results from the SLT and the RLT
are comparable, showing that RLT is an alternative to the conventional SLT.
2 from SLT from SLT
from RLT from RLT
4
Depth, z (m)
10
12
14
0 2000 4000 6000 8000 10000 0 200 400 600 800
Pile axial force (kN) Shaft resistance (kPa)
(a) Axial forces in pile (b) Shaft resistance
Fig. 11. Axial forces in pile and shaft resistance
The above case study was carried out for the sandy ground. Brown (2009) [8] pointed out that the static load-
displacement curve derived using UPM (Unloading Point Method) is still influenced by the strain rate effects of the
surround soils, especially in clay ground. The initial parts of the static curves (working load range) from SLT and RLT are
comparable, while the ultimate load derived from ULM is overestimated by 50% in maximum. This information is useful
when RLT is used for design purpose of piles in clayey grounds.
6. Concluding Remark
The author's belief is that the performance of a pile installed using Press-in Method is higher than that of the same pile
installed using the other installation methods. This belief, however, needs to be verified through load testing on
constructed piles. A handy load test may be possible using the Press-in machines, although the maximum load is limited
to 3 MN. The rapid pile load testing is one of useful tools to conduct load tests on many piles, because of its time and cost
effectiveness as well as a reasonable reliability.
References
[1] Middendorp, P., Bermingham, P. and Kuiper B.: Statnamic testing of foundation pile, Proc. 4th Int. Conf. on Application
of Stress-Wave Theory to Piles, The Hague, pp.585-588, 1993.
[2] Proc. 1st Int. Statnamic Seminar, Vancouver, Canada, 1995.
[3] STATNAMIC Loading Test (Kusakabe, Kuwabara & Matsumoto Ed.), Proc. 2nd Int. Statnamic Seminar, Tokyo, Japan,
1998.
[4] Standards of Japanese Geotechnical Society for Vertical Load Tests of Piles, JGS 1815-2002: Method for Rapid Load
Test of Single Piles, 2002.
[5] Nishimura, S., Shibata, A., Yamashita, K., Kita, N. Ogita, N. and Ishida, M.: One dimensional stress wave simulation
analysis of Statnamic load test, evaluation of boundary between Statnamic and dynamic loading tests, Proceedings
the 2nd Statnamic Seminar, Tokyo, pp.337-344, 1998.
7
Volume 4, Issue 2 June 2019
[6] Kusakabe, O. and Matsumoto, T.: Statnamic tests of Shonan test program with review of signal interpretation, Proc.
1st Int. Statnamic Seminar, Vancouver, Canada, pp.113-122, 1995.
[7] Hoshino, M., Sato, K., Takano, K., Kamei, S. and Nishimura, S.: A comparison between static load test and rapid load
test for a steel pipe pile installed by water jet vibratory technique, Proceedings of the 9th International Conference on
Testing and Design Methods for Deep Foundations (IS-Kanazawa 2012), Kanazawa, pp.581-585, 2012.
[8] Brown, M.: Chapter 2 Recommendations for Statnamic use and interpretation of piles installed in clay, Rapid Load
Testing on Piles, pp.26 - 36, Taylor & Francis Group, London, UK, 2009.
◆ A brief CV of Prof. Tatsunori Matsumoto
Professor Tatsunori Matsumoto obtained his Bachelor of Engineering and Master of Science
from Kanazawa University, Japan. He joined the Department of Civil Engineering of Kanazawa
University in 1981 as research associate. He became an Associate Professor in 1991 and
promoted to a Professor from August in 1999.
He retains an active involvement in research into pile dynamics and deformation of pile
foundations including piled rafts subjected to load combinations. He has published more than
200 technical papers including more than 45 Journal papers.
He was the Chairman of IS-Kanazawa 2012: The 9th International Conference on Testing and
Design Methods for Deep Foundations held in Kanazawa, Japan, from 18 to 20 September,
2012.
