DEEP FOUNDATIONS 629

IMPROVED METHODS FOR RAPID LOAD TESTS OF DEEP

FOUNDATIONS

Takaaki Miyasaka, P.E.1, Garland Likins, P.E., M.ASCE2, Fumio Kuwabara3, Frank
Rausche, P.E., M.ASCE4, Masayuki Hyodo5
1
Director, Jibanshikenjo Co., Ltd., 4-29-5 Midori, Sumida-Ku Tokyo 130-0021, Japan; ph: 81-3-5600-
2911; fax: 81-3-5600-2952; email: taka1010@jibanshikenjo.co.jp
2
President, Pile Dynamics, Inc., 4535 Renaissance Parkway, Cleveland, Ohio 44128 USA; ph: 216-
831-6131; fax: 216-831-0916; email: garland@pile.com
3
Professor, Nippon Institute of Technology, Miyashiro-machi Saitama-ken 345, Japan; email:
kuwabara@nit.ac.jp
4
President, GRL Engineers, Inc., 4535 Renaissance Parkway, Cleveland, Ohio, USA,ph: 216-831-
0916; fax: 216-831-0916; email: frausch@pile.com
5
Professor, Yamaguchi University, Tokiwadai 2-16-1, Ube 755-8611, Japan; email:
hyodo@yamaguchi-u.ac.jp

ABSTRACT

Rapid Load Tests have been promoted as an alternative for Static Pile Load Tests
since the late 1980s. Rapid Load Tests create a relatively long duration force pulse in
comparison with Dynamic Load Tests. However, a reliable prediction of pile static
capacity with this test based on the so-called Unloading Point Method (UPM) has
been subject to debate. Estimates of static capacity by UPM have overestimated static
load test results when the applied force pulse produced too little axial movement. To
resolve these issues, a drop mass Rapid Load Test called the Hybridnamic Test (HT)
was developed with a ram mass up to 700 kN to reliably estimate the pile static
ultimate bearing capacity up to a 35 MN. This paper discusses using a multi-cycle
Hybridnamic Test in soil of mostly sand. A new interpretation method called Fully
Mobilized UPM, with a required minimum net penetration per test cycle, was applied
to this case study and comparison of estimated soil resistance was made between
conventional UPM and Fully Mobilized UPM, and with signal matching of the
dynamic force pulse data using CAPWAP. Finally a practical method to prepare a
Static Load-Displacement Curve based on multi-cycle HT results is proposed.

INTRODUCTION

Rapid Load Tests (RLT) have been promoted as an alternative for Static Pile Load
Tests since the late 1980s. Compared with Dynamic Load Tests, Rapid Load Tests
create a relatively long duration force pulse by either a combustion process (i.e.

1
629
630 DEEP FOUNDATIONS

Statnamic Test, Bermingham & Janes, 1989) or an impact of a cushioned drop mass
(i.e. Pseudo Static Test. Schellingerhout & Revoort, 1996). In 2002, “Method for
Rapid Load Test of Single Piles,” standardized by the Japanese Geotechnical Society,
adopts the Unloading Point Method (UPM), which regards the test pile as a lump
mass with a spring and a dashpot as shown in Figure 1. The force at the point when
displacement is maximum (e.g. velocity is zero) is chosen to eliminate dynamic
damping effects so the force at that time is then the ultimate static resistance. The
UPM underestimates the static capacity compared to static load test results when the
force pulse test has too little energy. On the other hand, estimates of static capacity
usually significantly overestimate static load test results when the force pulse test has
produced too little permanent axial movement, particularly in cohesive soil conditions
due to the rate effect. Reliable prediction of static capacity with this test may be
questionable because of the reasons below:
1. The UPM model lacks a slider element and therefore it is unable to separate the
static resistance effect from the dynamic resistances, particularly in clay.
2. Longer piles cannot be considered a single lump mass for a short force pulse.
3. If end bearing is not fully mobilized, it is impossible to predict ultimate capacity.

FRap id
LumpMass

Spr ing Dashpo t

FIG.1. Model for UPM FIG. 2. Single Hybridnamic cushion sheet

In addition, nonlinear soil properties increase the difficulty in estimating the

ultimate soil resistance. Generally soil resistance assessable by UPM in one testing
cycle is closely associated with the maximum applied force and energy transferred to
the test pile during the force pulse and resulting net permanent movement. Therefore,
a multi-cycle testing method may be necessary to assess ultimate soil resistance.
To resolve the above issues, a variation of the drop mass Rapid Load Test called
the Hybridnamic Test (HT), developed in 2005, is equipped with a modular ram of up
to 700 kN and a hybrid cushion with low resilience. HT is able to transfer sufficient
energy to the test pile to totally activate and to estimate reliably the pile static
ultimate bearing capacity up to 35 MN. The cushion block is assembled by stacking
in parallel and/or in series single cushion sheets (primarily shaped rubber), as shown
in Figure 2. The practical maximum force pulse duration is about 0.2 second.
A case study presents a multi-cycle Hybridnamic Test in soil consisting primarily
of sand, with some thin layers of silt and stiff clay. A new interpretation method
called Fully Mobilized UPM, in which a minimum net penetration per test cycle is
required, was applied to this case study to estimate the ultimate soil resistance.
Estimated soil resistances from conventional UPM and Fully Mobilized UPM were

630 2
 DEEP FOUNDATIONS 631

compared with results of a static load test. Comparison is also made with signal
matching of the dynamic force pulse data using the CAPWAP® software program.

CASE STUDY OF HYBRIDNAMIC TEST

In this case study, the testing frame and equipment shown in Figure 3 was used.
This equipment consists of a ram with a mass of 70 kN, a stacked cushion with spring
constant of 74.0 MN/m and with dimension of 0.72 x 0.72 x 0.15 m, and a 6.5 m high
frame, which allows for a 2.0 m ram drop. Thus, this equipment can apply a force
pulse with duration greater than 31.0 ms. This duration represents the time required
for a stress-wave to travel more than six times upward and downward along the pile,
exceeding the Japanese standard requirement (five round trip passages) for RLT.

FIG. 3. Test frame and equipment (70 kN ram).

The test pile was a Type-10H steel sheet pile. This pile type has dimensions of
900 x 230 mm, a thickness of 9.2 to 10.8 mm, a perimeter of 2.3 m, pile length of
12.5 m, and has a cross-sectional area of 110 cm2. The test pile was installed using a
vibratory hammer.
Soils consist of a 2.9 m thick sand layer (N=10), followed by 2.7 m of silt (N=0),
then 5.0 m coarse to gravelly sand N = 10 to 20), followed by a 1.2 m thick layer of
stiff, gravelly clay, then dense fine sand beyond (N = 50+). Pile tip is located almost
at the bottom of the stiff, gravelly clay layer with N value greater than 50.

631 3
632 DEEP FOUNDATIONS

Testing Procedures

A Static Load Test (SLT) conducted one month after installation, shown in Figure
4, indicates a yield point at 1.65 MN and ultimate load of 1.8 MN with a permanent
pile top displacement of 99.3 mm. The summary of test results is shown in Table-1.

2000

1600

) SLT
N
k( 1200
d
a 800
o
L
400

0
0 20 40 60 80 100 120
Displacement (mm)
FIG. 4. Static load-displacement curve.

Table 1. Summary of Static Load Test Results

Yield Resistance Ultimate Resistance Pile-Soil Stiffness

Load Top Disp. Tip Disp. Load Top Disp. Tip Disp. Ko
(kN) (mm) (mm) (kN) (mm) (mm) (kN/mm)
1,650 11.7 4.8 1,800 106.5 99.3 140

Table 2. Summary of Hybridnamic Test

Drop Pile Top Displacement Soil Resistance

Height (Total) (Permanent) (Pile Top)
(m) (mm) (mm) (kN)
0.20 4.8 1.0 1,064
0.60 10.4 2.0 1,994
1.00 15.7 4.0 2,578
1.50 (1) 25.8 16.0 1,762
1.50 (2) 28.6 19.0 1,821
1.50 (3) 29.4 19.0 1,927

The Hybridnamic Test was conducted 8 weeks after the Static Load Test. The ram
was dropped from heights of 0.2 m, 0.6 m, and 1.0 m, and then was dropped three
times from a height of 1.5 m. Measuring sensors consisted of two strain gages, two
accelerometers, one optical displacement measuring device and one permanent
displacement checking device. HT results are shown in Table 2.

632 4
 DEEP FOUNDATIONS 633

The measured forces (F; also called the soil resistance Rsoil) and displacement (Δ),
obtained from HT and SLT and shown in Figure 5, are normalized by the static
ultimate soil resistance (Ru=1800 kN) and pile dimension (900 mm). Figure 5 reveals
the following:
1. The curves of the 0.2m drop, 0.6m drop, and 1.0m drop are significantly different
in general shape from the curves of the 1.5m drops.
2. When the applied force is less than the ultimate static resistance, the net permanent
displacement will be near zero and the UPM underpredicts the ultimate capacity.
3. The soil resistance of the 0.2m drop has greatly underestimated the static soil
resistance as compared with the SLT.
4. The soil resistances of the 0.6m drop and the 1.0m drop have overestimated the
static soil resistance as compared with the SLT. The result from the 1.0 m drop
(e.g. significant overprediction by factor of 1.43) is not necessarily the maximum
that would be predicted from UPM if a slightly different drop height were selected.
5. The overprediction possibility can be even larger when the soil is cohesive.
6. The soil resistance of the 1.5m drops are close to the SLT static soil resistance.
7. The normalized maximum and net residual displacements for 1.5m drops of about
3% and 2% respectively are suggested minimum requirements for improved UPM
3.0
0.2 m
Norm alized Resistance (F/Ru)

0.6 m
2.5
1.0 m
1.5 m-1
2.0
1.5 m-2
1.5 m-3
1.5 UPM
SLT
1.0

0.5

0.0
0% 1% 2% 3% 4% 5%
Normalized Displacement ( Δ /900mm)
FIG. 5. Rsoil-displacement curve

INTERPRETATIONS

Because the static soil resistance obtained from drop heights under 1.5m (with low
net permanent displacement) cannot be accurately extracted from the total response
which includes dynamic damping effects, another evaluation method is required. The
reason why the soil resistance values from the 1.5 m drop heights are close to that of
SLT can be simply explained by the added slider element model shown in Figure 6,
compared to the model shown in Figure 1. The Figure 6 model, named “Fully

633 5
634 DEEP FOUNDATIONS

Mobilized UPM,” allows the lump mass to penetrate the bearing stratum and then
retain significant permanent displacement. Subjected to significant permanent
displacement and full soil resistance mobilization, the dynamic effect can be
evaluated and the ultimate static resistance assessed more precisely. From Table 2,
the net permanent displacement difference between the drop heights less than 1.5 m
and that of 1.5 m is significant. The improved prediction reliability accompanying
significant net permanent displacement is obvious from Figure 5.

FRap id
LumpMass
S l ider

Spr ing Dashpo t

FIG. 6. Fully mobilized UPM model

To separate the dynamic component from the static component, signal-matching

analysis of the measured data from various drop heights was also investigated by
using CAPWAP (Rausche, 1972). Since full mobilization for a dynamic test requires
only 2 mm net permanent set, data from even the 0.60 m drop should be useful. The
load – displacement curves obtained from CAPWAP of the 0.60, 1.00 and 1.50 m
drops are shown in Figure 7, with an average estimated capacity of 2150 kN. This
result is 18 percent larger than the Full Mobilization UPM result. The CAPWAP
solutions for low net permanent displacements avoid the large overpredictions of
UPM (e.g. for the 1.0 m drop) and are therefore preferred when net displacements do
not achieve “full mobilization” during a RLT. Since RLTs often have relatively low
net penetration, signal matching may be useful.

2500 Axial Force (kN)

0 1000 2000 3000
2000 0
Load (kN)

1500 2
Depth (GL-m)

0.6 m 4
1000
1.0 m
6
1.50 m
500
8
0 SLT
10
0 20 40 60
Displacement (mm) 12 DLT

FIG. 7. CAPWAP Load–displacement curves FIG. 8. Axial force distributions

634 6
 DEEP FOUNDATIONS 635

The axial force distribution along the pile obtained from both SLT and signal
matching for a 1.5m drop are shown in Figure 8. The two distributions are reasonably
similar. Differences are attributed to the relatively long “rise time” in the HT force
and velocity data that makes more exact distributions more difficult to obtain.

Preparation of Static Load-Displacement Relationship

A practical method to prepare a curve for static load-displacement from an HT test

series is proposed in this study. The detailed procedures are described below:
1. Plot soil resistance versus the maximum displacement obtained from the first low
height drop where supposedly the soil resistance remains in elastic domain.
2. Plot soil resistance at the Fully Mobilized UPM versus the displacement of D/10.
3. Complete the curve based on Weibull’s Formula (NIST 2006).
The static load-displacement curve prepared using the above proposed method for
this case history (Figure 9) matches well that obtained from the SLT.

2000

1600
)
N
( 1200
k
d
a 800 Proposed
o
L
400 SLT

0
0 40 80 120
Dis placement (mm)
FIG. 9. Static load-displacement curve prepared based on proposed method.

CONCLUSIONS

An improved method for rapid load testing, called the “Hybridnamic Test” was
introduced. A new interpretation method called Fully Mobilized UPM was applied to
this multi-cycle Hybridnamic Test case study to estimate the ultimate soil resistance.
The soil resistance from the Fully Mobilized UPM, requiring sufficiently large net
permanent displacement, was compared to a conventional UPM. Comparison was

635 7
636 DEEP FOUNDATIONS

also made with signal matching of the dynamic force pulse data using CAPWAP.
Finally a practical method to prepare a static load-displacement curve based on multi-
cycle HT results was proposed. Through these studies, the following conclusions
were reached:
1. Conventional UPM underestimates the ultimate static resistance from insufficient
energy and applied maximum force less than the capacity (e.g. 0.2 m drop).
2. For applied force greater than the ultimate soil resistance, but with low net
permanent displacement, the conventional UPM may greatly overestimate the
ultimate static soil resistance due to rate effects, even in non-cohesive soils. At
low net permanent displacement, even larger overpredictions are likely in cohesive
soils.
3. Subjected to sufficient permanent displacement, the total soil resistance can be
fully mobilized, and the ultimate static soil resistance can be assessed more
accurately. This case history advocates a net permanent displacement of at least
three percent of the pile diameter as the suggested minimum limit.
4. When the net permanent displacement is less than three percent of the pile
diameter when Full Mobilization UPM does not apply, the most accurate method
to determine ultimate soil resistance is to analyze the Rapid Load Test data (for the
drop weight method of application) with signal matching such as CAPWAP.
5. The static load-displacement curve prepared based on the proposed method
matches well that obtained from the Static Load Test.
6. Base on the authors’ experience, it is recommended that the number of loading
cycles should not exceed three cycles while the soil resistance remains in the
elastic domain, and should not exceed two cycles after the soil resistance is fully
mobilized and is in the plastic domain. The optimum number of loading cycles is
suggested as three to five.

REFERENCES

Rausche, F., Moses, F., and Goble, G., (1972), “Soil resistance predictions from Pile
Dynamics,” Journal of Soil Mechanics and Foundations, ASCE, Vol. 98, No. SM9
Bermingham, P., Janes, M., (1989), “Innovative Approach to Load Testing of High
Capacity Piles”, Proc. of the Int’l Conf. on Piling and Deep Foundations, London,
Middendorp, P., Bermingham, P., Kuiper, B., (1992), “Statnamic load testing of
foundation piles”, The 4th International Conference on the Application of Stress
Wave Theory to Piles, Balkema, Sao Paulo, Brazil
，
Schellingerhout, A.J.G. Revoort, E., (1996), “Pseudo Static Pile Load Tester”,
Proc. of 5th Int’l Conf. on the Application of Stress Wave Theory to Piles, Orlando
Japanese Geotechnical Society, (2002), “JGS Standards 1815”, “Standards for
Vertical Load Test of Piles, Method for Rapid Load Test of Single Piles”
NIST/SEMATECH, (2006), e-Handbook of Statistical Methods, Engineering Statistic
Handbook, Chapter 1.3.6.6.8; Weibull Distribution

636 8