DISCUSSIONS AND CLOSURES
discussers consider that vertical confining stresses equivalent to 5
Discussion of “Interactions between Three
Tropical Soils and Municipal Solid Waste
Landfill Leachate” by E. M. Frempong
and E. K. Yanful
March 2008, Vol. 134, No. 3, pp. 379–396.
DOI: 10.1061/共ASCE兲1090-0241共2008兲134:3共379兲
1
Kolawole J. Osinubi, M.ASCE ; and
Charles M. O. Nwaiwu, A.M.ASCE2
1
Prof., Dept. of Civil Engrg., Ahmadu Bello Univ., Zaria 8100001, Nige-
ria. E-mail: kosinubi@yahoo.com
2
Senior Lecturer, Dept. of Civil and Water Resources Engrg., Univ. of
Maiduguri, 600004, Nigeria. E-mail: nknwaiwu@yahoo.co.uk
The authors have presented an interesting paper that examined the
impact of municipal solid waste 共MSW兲 landfill leachate on hy-
draulic conductivity and the chemical and mineralogical compo-
sitions of three tropical soils from Ghana, West Africa.
Laboratory tests to determine the index and geotechnical proper-
ties and the physico-chemical and mineralogical compositions of
the soils, before and after leachate permeation, were conducted.
Batch sorption and diffusion tests were also conducted on the
soils.
The discussers are interested in the results of tests on soil H, a
reddish-brown, sandy clayey silt which was identified as a later-
itic soil based on a silica: sesquioxide ratio 关SiO2 / 共Al2O3
+ Fe2O3兲兴 of 1.86. The discussion will concentrate on the follow-
ing areas: index and geotechnical properties, effect of MSW
leachate permeation, as well as batch sorption and diffusion tests
with respect to the lateritic soil.
Index and Geotechnical Properties
Soil H 共lateritic soil兲, which classifies as “clay loam” based on
USDA soil classification triangle, was reported as having a clay
content of 39% and a plasticity index of 30. Hence, the activity
共plasticity index/percentage of clay-size particles兲 is 0.77 and not
0.55 as reported by the authors. Otherwise, using an activity value
of 0.55, the plasticity index is approximately 21. It is pertinent to
note that soils with plasticity indices of about 30 and above can
give field construction problems 共EPA 1989; Das 1998兲 and this
should be taken into consideration when selecting lateritic soils
for the construction of engineered landfills. The discussers have
already shown that lateritic soils with lower plasticity indices can
be employed in the construction of barrier layers for MSW con-
tainment 共Osinubi and Nwaiwu, 2002, 2005, 2006a兲.
The hydraulic conductivity testing involved the application of
a vertical confining stress of 160 kPa on 2-cm-thick specimens.
The specimen size is quite small and the application of a stress of
160 kPa seems to simulate a short-term loading that might not be
appropriate because for a constructed liner, it takes some time
for the entire landfill space to be completely filled with MSW.
Thus, the total stress from the MSW is not borne by the liner
within a short period but it is a relatively long-term process. The
JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010 / 891
to 10 m of MSW will be more appropriate for a testing program
of this nature. It is still possible to obtain hydraulic conductivity
less than 10−9 m/s at +2% water content relative to optimum,
particularly for a lateritic soil having up to 39% clay-size fraction.
Effect of MSW Leachate Permeation on Soil H
Although there were changes in the chemical, mineralogical,
index, and geotechnical properties of soil H, none of the changes
was adverse or detrimental to the integrity of the soil as a liner
material. The hydraulic conductivity decreased upon permeation
with MSW leachate. Similar trends were obtained by the discuss-
ers 共Osinubi and Nwaiwu 2006b兲 for lateritic clays from Zaria
in Northern Nigeria. According to Rowe et al. 共1995兲, MSW
leachate will normally cause a slight decrease in the hydraulic
conductivity of inactive clays with divalent cations on the ex-
change site.
The high confining stress of 160 kPa used by the authors to
simulate field stress is enough to induce low hydraulic conductiv-
ity 共 ⬍ 10−10 m/s兲 in soil H. However, the factors that might have
probably contributed to the decrease in hydraulic conductivity
upon MSW leachate permeation are diffuse double layer expan-
sion, sodium adsorption, and bacterial or microbial clogging
共Rowe et al. 1995兲 as well as formation of chemical precipitates.
The predominant factors might have been sodium adsorption 共so-
dium was present at high concentration in the MSW leachate as
reported by Frempong and Yanful 2008兲 and microbial clogging
共microbial growth is enhanced by higher ambient temperature兲.
Batch Sorption and Diffusion Tests
The transport parameters—distribution coefficient, Kd 共partition-
ing coefficient, K p兲; retardation factor, R, and effective diffusion
coefficient, De—obtained by the authors for soil H with respect to
Na+ compare well with those reported by Osinubi and Nwaiwu
共2006c兲. The discussers obtained Kd values in the range 0.26–0.3
for Na+ and R values that ranged from 2.44 to 2.54, which lie
between the authors’ experimental value of 2.2 and modeled value
of 2.6 for Na+ diffusion in compacted lateritic clay soils.
The observation by the authors that the values of effective
diffusion, De, for the various chemical species are likely influ-
enced by the particle size distribution and the amount and type of
clay minerals and oxides present in the soil is in close agreement
with results obtained by Osinubi and Nwaiwu 共2006c兲. The dis-
cussers observed that:
1. Soil composition 共in terms of particle size distribution and
plasticity兲 can be characterized by the “plasticity product,”
which is a product of percentage of soil fraction passing
75 ␮m aperture sieve and plasticity index. The plasticity
product represents the effective contribution of the plasticity
of the fines to the performance of the whole material, which
depends on the proportion of fines 共CIRIA 1988兲.
2. The Kd values as well as the soil profile De values decrease
with higher values of plasticity product, while R values in-
 crease with higher plasticity product values 共Osinubi and
Nwaiwu 2006c兲.

References

Construction Industry Research and Information 共CIRIA兲. 共1988兲. Later-

ite in road pavements, Special Publication No. 47, CIRI, London.
Das, B. M. 共1998兲. Principles of geotechnical engineering, 4th Ed., PWS
Publishing Company, Boston.
Environmental Protection Agency 共EPA兲. 共1989兲 “Requirements for haz-
ardous waste landfill: Design, construction, and closure.” EPA Publi-
cation No. EPA-625/4–89–022, Cincinnati.
Frempong, E. M., and Yanful, E. K. 共2008兲. “Interactions between three
tropical soils and municipal solid waste landfill leachate.” J. Geotech.
Geoenviron. Eng., 134共3兲, 379–396.
Osinubi, K. J., and Nwaiwu, C. M. O. 共2002兲. “Compacted lateritic soils
Fig. 1. Simulated unit-cell surface settlement with time for Case 2
as hydraulic barriers in waste containment systems.” Proc., 4th Int.
Congress on Environmental Geotechnics, Ashgate, San Diego, 225–
230. Basic Ideas
Osinubi, K. J., and Nwaiwu, C. M. O. 共2005兲. “Hydraulic conductivity of
compacted lateritic soil.” J. Geotech. Geoenviron. Eng., 131共8兲,
The plane-strain material stiffness is based on the matching of the
1034–1041.
Osinubi, K. J., and Nwaiwu, C. M. O. 共2006a兲. “Design of compacted
column-soil composite stiffness. In the paper, this average stiff-
lateritic soil liners and covers.” J. Geotech. Geoenviron. Eng., 132共2兲, ness is analyzed for the final drained situation, assuming uniaxial
203–213. 共confined兲 deformation. Imposing the equilibrium of vertical
Osinubi, K. J., and Nwaiwu, C. M. O. 共2006b兲. “Compatibility of com- stresses and the compatibility equation 共vertical stresses are di-
pacted lateritic soils with municipal solid waste leachate.” Proc., 5th rectly proportional to the correspondent confined elastic moduli兲,
Int. Congress on Environmental Geotechnics, Thomas Telford, Lon- Eq. 共3兲 of the paper is obtained. The following expressions are
don, 608–615. also of interest
Osinubi, K. J., and Nwaiwu, C. M. O. 共2006c兲. “Sodium diffusion in
pEc
compacted lateritic soil.” Proc., 5th Int. Congress on Environmental ␴zc =
Geotechnics, Thomas Telford, London, 1224–1231. Emcas + Ems共1 − as兲
Rowe, R. K., Quigley, R. M., and Booker, J. R. 共1995兲. “Clay-leachate
compatibility by measurement of hydraulic conductivity.” Chapter 4, pEs
Clayey barrier systems for waste disposal facilities, Chap. 4, Chap- ␴zs = 共1兲
Emcas + Ems共1 − as兲
man and Hall, London.

Discussion of “Simplified Plane-Strain
Modeling of Stone-Column Reinforced
Ground” by S. A. Tan, S. Tjahyono, and K.
K. Oo
February 2008, Vol. 134, No. 2, pp. 185–194.
DOI: 10.1061/共ASCE兲1090-0241共2008兲134:2共185兲
Jorge Castro1 and César Sagaseta, M.ASCE2
1 column.
Group of Geotechnical Engineering, Dept. of Ground Engineering
Materials Science, Univ. of Cantabria, Avda. de Los Castros,
39005 Santander, Spain. E-mail: castrogj@unican.es
2
Group of Geotechnical Engineering, Dept. of Ground Engineering
Materials Science, Univ. of Cantabria, Avda. de Los Castros,
39005 Santander, Spain. E-mail: sagasetac@unican.es
The authors have developed an interesting finite-element analysis
of stone-column reinforced ground, and compared it with field
data. Two different methods are considered to convert the actual
3D geometry to a plane-strain model. The discussers, who have
also developed 3D and plane-strain finite-element models of
stone-column improved field sites, would like to comment on
both methods, and propose some possible modifications.
sz p
= 共2兲
L Emcas + Ems共1 − as兲
where sz = final surface settlement; L = height of the unit cell; p
= applied load; and Em = oedometric 共confined兲 modulus.
There are infinite possible combinations of the plane-strain
parameters 共Emc , Ems , as兲pl that fulfill Eq. 共3兲 of the paper.
In Plane-Strain 1 model, the elastic modulus of the soil and the
trench width, bc, are kept unchanged, but the area replacement
ratio and the column stiffness are modified. Therefore, according
to Eq. 共1兲 of the discussion, the vertical stress on the soil is the
same in the plane-strain model but not the vertical stress on the
and
s/n, In Plane-Strain 2 model, only the trench width is changed.
Hence, the stresses in the soil and in the column are not altered.
and
s/n, Case 1
The results of Case 1 are shown in Fig. 4 共of the original paper兲.
The final settlements show a reasonably good agreement for the
different models. Although the differences are small and not im-
portant for practical purposes, they may be attributed to the con-
fined analysis assumed to obtain the column-soil composite
stiffness.
If the radial deformation of the column is included, the com-
posite stiffness for axial symmetry can be derived from the
drained elastic solution by Balaam and Booker 共1981兲. Following

892 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010

Table 1. Parameters for Unit-Cell Models
rc − bc R−B Es Ec khs ␸s⬘ ␸c⬘
Model 共m兲 共m兲 共kPa兲 共kPa兲 共m/s兲 共°兲 共°兲
Axisymmetric 0.425 1.275 3,000 30,000 3.47⫻ 10−9 22 40
Plane-Strain 1 共␸c⬘ = 40° 兲 0.425 1.275 3,000 12,000 3.38⫻ 10−9 22 40
Plane-Strain 1 共␸c⬘ = 22.5° 兲 0.425 1.275 3,000 12,000 3.38⫻ 10−9 22 22.5
Plane-Strain 2 共bc = 0.126 m兲 0.126 1.13 3,000 30,000 4.51⫻ 10−9 22 40
Plane-Strain 2 共bc = 0.425 m兲 0.425 3.825 3,000 30,000 3.50⫻ 10−8 22 40

their same calculation steps, a similar solution for the plane-strain Plane-Strain 2 model is basically the same as Van Impe and De
case has been developed by the discussers. This leads to the fol- Beer’s 共1983兲. As it is known, in the plastic range this method
lowing matching of the column-soil composite stiffness predicts final settlements on the safe side. This is well shown in

冋 册
the paper, where the final settlement is 8% higher.
as共1 − as兲共␭c − ␭s兲2 When the main feature of the improvement is to increase the
Emcas + Ems共1 − as兲 −
as共␭s + 2Gs兲 + 共1 − as兲共␭c + 2Gc兲 slope stability instead of reducing the final settlement, it is very

冋
pl
important to match the average shear strength throughout the col-
= Emcas + Ems共1 − as兲 umn and the surrounding soil 共Barksdale and Bachus 1983兲. This
needs an estimation of the stress concentration ratio and the fail-

−
as共1 − as兲共␭c − ␭s兲2
as共␭s + 2Gs兲 + 共1 − as兲共␭c + Gc + Gs兲
册 ax
共3兲
ure surface.

where G and ␭ are the Lamé’s constants. In this case, the expres-
sions for plane-strain and axial symmetry are not the same be-
cause the last factor of the denominator is different.
Using Eq. 共3兲 of the discussion and keeping the same soil
stiffness for the plane-strain models, the equivalent column stiff-
nesses for Plane-Strains 1 and 2 are 10,965 kPa and 28,167 kPa,
respectively. Finite-element analyses of the two models using
these elastic moduli give a value of the final settlement of
139 mm, exactly the same as that obtained with the axisymmetric
model. This proves that the slight differences between the final
settlements of the plane-strain models in Fig. 4 共of the original
paper兲 are due to neglecting the radial deformation of the column
in the evaluation of the composite stiffness 关Eq. 共3兲 of the original
paper兴.
Case 2
In the plastic range, the final settlement is not well-reproduced
with Plane-Strain 1 method. As the authors clearly pointed out,
this is attributed to the different yielding pattern. Plane-Strain 1
model has a larger area replacement ratio and the equivalent stiff-
ness of the column is decreased from 30 MPa to 12 MPa. Hence,
its vertical stress is reduced roughly by the same ratio 关Eq. 共1兲 of
the discussion兴, and then the column does not reach the plastic
Consolidation Rate
In the Plane-Strain 1 method, the width B of the plane-strain
model is chosen equal to the radius of the drainage zone R. There
is no need to do this because the consolidation rate will be
matched afterward, changing the soil permeability in the plane
strain model, khs. In the discussers’ opinion it is better to choose a
width B that keeps the same area replacement ratio, B = R2 / rc
= 3.825 m. This gives a wider model, similar to Plane-Strain 2,
with less number of trenches in the whole section and with the
actual soil and column parameters but the soil permeability, khs
共Table 1 and Fig. 1兲. In a complex geometry, reducing the number
of trenches and their narrowness is very convenient to avoid nu-
merical problems.
The consolidation rate is matched using any of the proposed
equations 共Tan and Oo 2005; Indraratna and Redana 1997; Hird
et al. 1992兲 for the soil permeability, khs. In the discussers’ opin-
ion, Eq. 共4兲 in the paper is not correctly defined. Neglecting the
smear zone and the finite permeability of the column, the factor
for the plane-strain case, F共N兲pl, is equal to 2 / 3 according to the
solution developed by Hird et al. 共1992兲. If the stiffness and the
area replacement ratio are the same, Eq. 共4兲 reduces to
k pl 2B2

冋 册
yielding. A possible solution of this discrepancy implies a match- = 共4兲
ing of the elastic-plastic final settlement, altering the friction
kax 2 N2 3N2 − 1
3R ln N −
angles of the soil and column. N2 − 1 4N2
The discussers have developed a solution for the elastic-plastic
settlement of a unit cell in axial symmetry 共Castro and Sagaseta where N = rc / R = 1 / 冑共as兲ax.
2009兲. Further investigation of the elastic-plastic settlement in a Using this equation, the soil permeability in Plane-Strain 2
plane-strain model would clarify the matching expression. model should be 1.3 times the axisymmetric one. There is little
For the case presented, if the friction angle of the soil is kept difference, but using that permeability does not imply further
constant, an equivalent friction angle of the column of 22.5° gives complications.
the same final settlement as in the axisymmetric case 共Fig. 1兲.
The numerical simulations shown in Fig. 1 of this discussion
were performed under the same conditions described by the au- References
thors. The parameters that were changed are detailed in Table 1.
In the paper there is no reference to the value adopted for the Balaam, N. P., and Booker, J. R. 共1981兲. “Analysis of rigid rafts sup-
gravel dilatancy. The results shown in Fig. 1 have been obtained ported by granular piles.” Int. J. Numer. Analyt. Meth. Geomech.,
for zero dilatancy. It would be useful to clarify if this has been 5共4兲, 379–403.
also assumed in the authors’ calculations. Barksdale, R. D., and Bachus, R. C. 共1983兲. “Design and construction of

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010 / 893

 stone columns.” FHWA/RD–83/026, Federal Highway Administration, LOAD (KN)
Washington, D.C. -400 -200 0 200 400 600 800 1,000

Castro, J., and Sagaseta, C. 共2009兲. “Consolidation around stone columns. 0

Influence of column deformation.” Int. J. Numer. Analyt. Meth. Geo- 2

#5
#1 #2 #3 #4
mech. 33共7兲, 851–877. 4
#6
Hird, C. C., Pyrah, I. C., and Russell, D. 共1992兲. “Finite element model- 6
#7
ling of vertical drains beneath embankments on soft ground.” Geo-

technique, 42共3兲, 499–511.
Indraratna, B., and Redana, I. W. 共1997兲. “Plane-strain modeling of smear
effects associated with vertical drains.” J. Geotech. Geoenviron. Eng.,
123共5兲, 474–478.
Tan, S. A., and Oo, K. K. 共2005兲. “Stone column FEM modeling—2D
and 3D considerations illustrated by case history.” Proc. Int. Symp. on
Tsunami Reconstruction with Geosynthetics, ACSIG, Bangkok, Thai-
land, 157–169.
Van Impe, W., and De Beer, E. 共1983兲. “Improvement of settlement be-
haviour of softy layers by means of stone columns.” Proc., 8th Euro-
pean Conference on Soil Mechanics and Foundation Engineering,
Balkema, Rotterdam, The Netherlands, 309–312.
Discussion of “Load Testing of a
Closed-Ended Pipe Pile Driven in
Multilayered Soil” by D. Kim, A. V. D. Bica,
R. Salgado, M. Prezzi, and W. Lee
April 2009, Vol. 135, No. 4, pp. 463–473.
DOI: 10.1061/共ASCE兲1090-0241共2009兲135:4共463兲
Bengt H. Fellenius, Ph.D., P.E., M.ASCE1
1
Bengt Fellenius Consultants Inc., 2475 Rothesay Ave., Sidney, British
Columbia, Canada V8L 2B9. E-mail: Bengt@Fellenius.net
The authors have presented a case history on the results of a static
loading test on a strain-gauge instrumented, 356-mm diameter,
closed-toe pipe pile, driven 17.4 m into a clayey silty soil with
layers of sand. It is appreciated that the paper includes compre-
hensive background information on the soil profile and installa-
tion data, as well as the results of the static loading test. The paper
presents the dynamic 共driving兲 data, pile-head load movement,
residual load, shaft and toe resistances, and total capacity. This
information is useful for piled foundation design in the general
area and geology of the test site. However, in my opinion, the test
results would be useful also for a wider application, if the authors
could provide more of the basic results data, and if the paper had
included a more detailed analysis. Specifically, a table showing
DPTH (m)
8
10
12
14
16
18
First Loading Test
20
Fig. 1. Distribution of shaft resistance, set to zero at 16 m depth
The authors indicate that the pile capacity in the first static
loading test is 1,500 KN, which is close to the 1,512 KN load
applied to the pile head in increment 5 of the seven increments of
load reported in the paper. The shaft resistance, determined as the
difference between the applied load and the toe resistance shown
in the load distribution curves 关Fig. 8共a兲 of the original paper兴, is
very similar for increments 4 through 6. Moreover, the shaft re-
sistance mobilized for increment 7 共i.e., when the pile experi-
enced plunging failure兲, is also quite similar to that of the
preceding three increments. As shown in Fig. 1 of this discussion,
the difference is mostly in the lowest 1.5 m length of the pile.
Indeed that similarity, the comparable magnitude of shaft resis-
tance determined in the second loading test, and the fact that the
second loading test indicates the same ultimate shaft resistance
共albeit a stiffer response兲 as the first test indicates that the soil is
neither strain hardening nor strain softening.
By fitting the shaft resistance to an effective stress distribution,
I have estimated a distribution of the ␤-coefficients that produces
a shaft resistance distribution that is equal to the measured distri-
bution. I have assumed that the pore pressures are hydrostatically
distributed below the groundwater table 共located 1.0 m below the
ground surface兲. The results are presented in Fig. 2. Beta-
coefficients of 0.6 and 1.4 are high for the silty sandy soil layers
at the site, having compressibility values 共Table 1 of the original
paper兲 that correlate to virgin Janbu modulus numbers ranging
from 2 to 4 and recompression modulus numbers ranging from 11
LOAD (KN) Cone Stress, qc (MPa)
0 500 1,000 1,500 2,000 0 10 20 30 40 50 60
0 0

the applied loads and the end-of-increment duration strain-gauge #2 #3 #4 #5 #7

2
determined loads. It would also be of value to show the full dia- ß = 0.1
4
gram from the cone sounding at the test pile location 共cone stress, ß = 0.3
5

sleeve friction, pore pressure, and friction ratio兲. With regard to 6

#6
(m)

DEPTH (m)
DEPTH(m)

the analysis, I believe users of the paper would appreciate a dia- 8

ß = 0.7
gram showing the results of analysis of the cone penetrometer 10 10
DPTH

sounding at the test pile location in terms of distribution of shaft ß = 0.4

12
resistance for the different methods used, which would be more
14
informative than the brief listing in Table 5 共of the original paper兲 ß = 0.6 15
16
of the total capacities obtained from such calculations. Moreover, Effective stress fit
I find it disappointing that the authors have not provided a refer- 18
ß = 1.4
ence to an effective stress calculation with the main reference 20 20

value being the effective stress ratio, the so-called ␤-coefficient,

which, in my opinion, is the most basic correlation of the results Fig. 2. Distribution of measured resistance, resistance fitted to that of
of a static loading test and most suitable also as correlation to increment 5, and the ␤-coefficients used to achieve the fit 共cone stress
shaft resistance in clay. and a delineation eleven soil layers added for reference兲

894 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010

 LOAD (KN) resistance is about twice that measured, and the corrected
0 500 1,000 1,500 2,000 ␤-coefficients are half those indicated in Fig. 2. Were the residual
0 shaft shear after all not fully mobilized near the pile toe, then the
residual load at the pile toe would be smaller and the residual-
2
load force equilibrium lie correspondingly higher up in the pile,
4
as indicated by the dashed curves in Fig. 3. If so, the correspond-
Assumed residual load ing ␤-coefficient immediately above the pile toe would be about
6 unity, which is representative for a very dense silt, but hardly so
for a compressible clayey silt. A more exact determination of the
(m)

8
residual load would require a direct measurement by, for example,
DEPTH(m)

From strain-gages
10
performing an O-cell test instead of the head-down test. Though
DPTH

the curves shown in Fig. 3 are only approximately close to actual

12
"True"
resistance
14
16
18
When residual shaft shear is not fully
mobilized all the way to the pile toe
20
Fig. 3. Distribution of measured resistance from strain gauges di-
rectly, residual load assumed fully mobilized, and “true” resistance
to 37. These values are representative for a highly compressible
soil. It is my experience that, when driving displacement piles
into such soils, the subsequent reconsolidation and set-up process
will give rise to significant residual loads in the pile as well as a
residual-load force equilibrium located close to the pile toe. The
authors mention analysis of residual load, but state it to be mini-
mal when compared to the “measured capacity.” They, therefore,
neglect its influence on the results of the test.
The authors compare the residual load to the pile capacity and
indicate that the ratio between the maximum residual load—
reasonably estimated—and the pile capacity is about 1 over 6,
which is not a large number. Correlating residual load to the pile
capacity is not meaningful, but including it is essential when de-
termining pile shaft and toe resistances is 共Fellenius 2000,
2002a,b, 2009兲. In a compressible soil such as at this site, it is
most likely that the residual load is built up by fully mobilized
shaft shear. For this condition, the residual shaft shear along the
pile must be in the negative direction more or less all the way
to the pile toe. Fig. 3 shows the distribution of residual load and
the corrected load distribution, the “true” resistance distribution.
The shaft resistance of the latter is half that measured, the toe
“true” distribution curves, they demonstrate that residual load is
not negligible in the analysis of the test results.
The authors report that the strain gauge readings before the
static loading test are questionable due to “drift” occurring during
the driving. Change of gauge readings due to construction is a
recurring problem with any instrumented pile and is due to many
effects, zero drift being one. Instrumentation of steel pile is usu-
ally completed above ground at temperature that is different from
the one down in the soil, and strain gauges, even if by themselves
unaffected by temperature changes, will dutifully register the
strain changes due to the change of temperature of the pile mate-
rial. For a steel pile, strain is built-in during the manufacturing
process, and can be partially released when the pile is driven. The
pile–soil interaction during the driving and the difference between
the shaft resistance 共requiring only small movements to change
values兲 and the toe resistance 共requiring larger movements兲 mean
that strain is built-in due to the driving. Once driven, the recovery
of the soil from the driving disturbance will add 共or reduce兲 the
strain already in the pile. It is not that the gauges themselves
drift—these days, the gauges are quite reliable. Rather, the pile
construction process imposes strain, and significantly so. How-
ever, the problem is not the change per se; it is that the engineer
performing the test needs to include this change in the analysis
and to separate the change that is not caused by the soil or affect-
ing the soil from those which have, and that is very difficult as
long as the measuring system uses the pile as a part of the gauge.
Normally, we have to work from the strain changes imposed by
the test, as in the authors’ case. However, this does not mean that
one can assume zero strain and zero load in the pile at the start of
the test, and conclusions neglecting the effect of residual load
may result in unconvincing conclusions.
The authors do not report the pile toe movement. However, the
pile shortening is small in relation to the pile head movement, and

600
Reloading
portion
500
Test 2
Test 1
400
LOAD (KN)

Presumed virgin curve as if

no residual load is present
300

200

100

0
-300 -200 -100 0 100 200 300 400 500
PILE TOE MOVEMENT (mm)

Fig. 4. Calculated pile toe load movements: Test 1 and Test 2

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010 / 895

 2,500
TOTAL CAPACITIES
2,000
CAPACITY (KN)
SHAFT RESISTANCES
1,500
1,000
500
0
0 30 60 90 120 150 180
DAYS AFTER INITIAL DRIVING
Fig. 5. Values of shaft resistance and total capacity from static load-
ing tests and from CAPWAP analyses versus days after initial driving
plotted from Table 3 in the original paper
the pile toe movement can be determined approximately from
subtracting calculated pile shortening from the information on
measured pile head movement for the applied load increments.
Fig. 4 shows the load movement for the pile toe in the two tests
calculated accordingly. The values are not corrected for residual
load. At about 55-mm movement, the Test 1 curve implies a
change of response that could be interpreted as a definition of an
ultimate toe resistance. However, up to that point, the response is
a reloading of the pile toe. A potential virgin pile-toe response is
indicated in the figure. In fact, of the three components of pile
head movement measured in a static loading test, only the shaft
resistance exhibits something that can be termed ultimate resis-
tance, or “capacity.” The second component is the “elastic” short-
ening of the pile, which is a smaller or larger portion of the total
movement depending on the length of the pile and on whether or
not the load generating the shortening has overcome the shaft
resistance along the full length of the pile. The third is the pile-toe
load-movement response, which in most cases is a curved line
showing no ultimate resistance or failure 共Fellenius 1999兲. There-
fore, definitions of total pile capacity, as well as toe capacity, if
meant to be ultimate values, are extraneous. Of course, the con-
cept of total pile capacity has been in general use for almost 100
years and is not going to go away. However, for the individual
components, the term is only truly applicable to shaft resistance,
when there is no definite strain-hardening or strain-softening re-
sponse, which is the case for the subject test. 共If the soil exhibits
strain hardening, the term “ultimate resistance” does not apply. If
strain softening, the term “ultimate resistance” applies only to the
shaft resistance at a certain pile element, not to a single total shaft
resistance value unless it is calculated from the individual ele-
ments兲. For the other two components 共pile shortening and pile
toe response兲, the concept of capacity has little relevance. As to
the authors’ reference to the capacity being the load that caused a
movement equal to 10% of the pile diameter, it a useless defini-
tion. This definition was first proposed by Terzaghi 60 years ago
for the analysis of small footings and laboratory tests for lack of
any more suitable definition. It is of absolutely no relevance to the
analysis of a loading test on a pile, or for that matter, on a full-
size footing.
896 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JUNE 2010
The authors comment on the fact that the two static loading
tests performed at 50 days and 90 days after the driving show
only a small increase of total capacity and no increase of shaft
resistance, in contrast to the increase shown by the CAPWAP
analyses of the restrike blows for the two other pipe piles driven
at the site. I have plotted the values in Fig. 5. However, the data
Test total
do not necessarily conflict. First, for each restrike event, not only
Test shaft
ECEP total
is the pile response stiffer for the next event without necessarily
ECEP shaft including an increase in shaft resistance 共as was the case for the
RECEP total repeated static test兲, the pile is also driven a bit deeper into the
RECEP shaft dense material at the pile toe, generating a larger toe resistance,
210 240 and therefore a larger pile capacity 共as also is implied in Fig. 4 of
this discussion兲. The authors do not report the number of restrike
blows given and the total penetration for each event. As indicated
by the pile-driving diagram 关Fig. 4共a兲 of the original paper兴, it
does not take much additional penetration for the pile toe resis-
tance to increase considerably. Moreover, as for the static test,
presence of residual load will cause a dynamic analysis 共CAP-
WAP兲 to indicate a larger than true shaft resistance and a corre-
spondingly smaller toe resistance 共Fellenius 2004兲. It may well be
that the shaft resistance values for the restrikes are overestimated
and the toe resistance values underestimated. Hypothetically, with
regard to the test piles, after about 30 days the shaft resistance
was approximately unchanged, but the toe resistance increased
due to the repeated driving.
I disagree with the authors’ conclusion that axial response of a
single pile is not well understood. It is in fact rather well under-
stood, and if the pile is driven or otherwise constructed or the soil
is uniform or multilayered makes little difference. That is, the
principles and the qualitative response are understood. Obtaining
quantitative values is a different matter, however, and this is what
the authors might have meant. To enable the profession to calcu-
late the actual values—that is, to design piled foundations—
requires reference to results of well-instrumented tests, similar to
the test and paper presented by the authors; their effort to write up
a project-related study into a full journal paper is appreciated.
References
Fellenius, B. H. 共1999兲. “Bearing capacity—A delusion?” Proc., Annual
Meeting of the Deep Foundation Institute, Hawthorne, N.J.
Fellenius, B. H. 共2002a兲. “Determining the resistance distribution in piles.
Part 1: Notes on shift of no-load reading and residual load.” Geotech-
nical News Magazine, 20共2兲, 35–38.
Fellenius, B. H. 共2002b兲. “Determining the resistance distribution in
piles. Part 2: Method for determining the residual load.” Geotechnical
News Magazine, 20共3兲, 25–29.
Fellenius, B. H. 共2004兲. “Unified design of piled foundations with em-
phasis on settlement analysis.” Proc., Geo-Institute Geo-TRANS Con-
ference, J. A. DiMaggio and M. H. Hussein, eds., ASCE Geotechnical
Special Publication, GSP 125, ASCE, Reston, Va., 253–275.
Fellenius, B. H. 共2009兲. Basics of foundation design, electronic Ed.,
具http://www.Fellenius.net典.
Fellenius, B. H., Brusey, W. G., and Pepe, F. 共2000兲. “Soil set-up, vari-
able concrete modulus, and residual load for tapered instrumented
piles in sand.” Proc., Conf. on Performance Confirmation of Con-
structed Geotechnical Facilities, GSP94, A. J. Lutenegger, ed.,
ASCE, Reston, Va., 98–114.
Copyright of Journal of Geotechnical & Geoenvironmental Engineering is the property of American Society of
Civil Engineers and its content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email articles for
individual use.