Constriction-Based Retention Criterion

for Granular Filter Design

Buddhima Indraratna, F.ASCE1; Ashok K. Raut2; and Hadi Khabbaz3

Abstract: The filter design criteria in practice are currently based on laboratory tests that were carried out on uniform base soil and filter
materials. These criteria mostly involve specific particle size ratios, where the system of base soil and filter is represented by some
Downloaded from ascelibrary.org by University of Tokyo on 12/25/20. Copyright ASCE. For personal use only; all rights reserved.

characteristic particle sizes. Consequently, these criteria have limitations when applied to nonuniform materials. In filters, it is the
constriction size rather than the particle size that affects filtration. In this paper, a mathematical procedure to determine the controlling
constriction size is introduced, and subsequently, a constriction-based retention criterion for granular filters is presented. The model also
incorporates the effect of nonuniformity of base soil in terms of its particle size distribution, considering the surface area of the particles.
The proposed retention criterion is verified based on experimental data taken from past studies plus large-scale filtration tests carried out
by the authors. The model successfully and distinctly demarcates the boundary between effective and ineffective filters in the case of
cohensionless base soils.
DOI: 10.1061/共ASCE兲1090-0241共2007兲133:3共266兲
CE Database subject headings: Filters; Design; Construction; Material properties.

Introduction
Although known to be conservative and originally developed for
cohensionless uniform base soil and filter materials, the well-
known Terzaghi retention criterion 共USACE 1953兲 is given by the
relationship, D15 / d85 ⬍ 5. Here, for the filter material, 15% by
mass of particles are finer than the size denoted by D15, and for
the base soil, 85% by mass of particles are finer than the size
denoted by d85. Several past studies 共e.g., Sherard et al. 1984a;
Bertram 1940兲 revealed that filters even with higher values of
retention ratio, D15 / d85, can be effective, especially in the case of
uniform base soils. In contrast, studies conducted by Lafleur
共1984兲 demonstrated that some filters with retention ratios smaller
than four involving nonuniform or well-graded base soils were
ineffective. To address this effect of diminishing filter effective-
ness in the case of well-graded base soils, the current design
practice 共NRCS 1994兲 recommends the use of d85 after regrading
the base soil PSD for particles larger than #4 sieve size
共4.75 mm兲, i.e., d85R rather than the conventional d85 without re-
grading. ICOLD 共1994兲 also suggests the use of a smaller base
soil representative size.
Honjo and Veneziano 共1985兲 carried out a statistical analysis
1
Professor of Civil Engineering, Univ. of Wollongong, Wollongong
City, NSW 2522, Australia 共corresponding author兲. E-mail: indra@
uow.edu.au
2
Senior Water Engineer, Snowy Mountains Engineering Corporation,
Level 6, 76 Berry St. North Sydney, NSW 2060, Australia.
3
Lecturer, School of Civil Engineering, Univ. of Wollongong,
Wollongong City, NSW 2522, Australia.
Note. Discussion open until August 1, 2007. Separate discussions
must be submitted for individual papers. To extend the closing date by
one month, a written request must be filed with the ASCE Managing
Editor. The manuscript for this paper was submitted for review and pos-
sible publication on June 14, 2005; approved on September 26, 2006.
This paper is part of the Journal of Geotechnical and Geoenvironmental
Engineering, Vol. 133, No. 3, March 1, 2007. ©ASCE, ISSN 1090-0241/
2007/3-266–276/$25.00.
on various test data and found that the reliability of filters dimin-
ished for nonuniform base soils. However, such statistical analy-
ses do not explain the fundamental physics of filtration, and are
not always free from bias inherent in experimental procedures.
For example, consider three different base soils 共B1, B2, and B3兲
having the same d85 tested against three different filters 共F1, F2,
and F3兲 having the same D15 共Fig. 1兲. These base soil and filter
arrangements have identical D15 / d85 ratios. The question is
whether these base soil-filter systems have similar filtration char-
acteristics in terms of mass retention and flow rates. In other
words, is the D15 / d85 ratio on its own adequate to describe the
filter effectiveness? Locke et al. 共2001兲 highlighted that the evalu-
ation of filter effectiveness based on the constriction size distri-
bution is more appropriate than the sole use of particle sizes. This
study is an attempt to present a constriction-based retention cri-
terion valid for both uniform and well-graded materials supported
by experimental evidence.
Theoretical Concepts
The main theoretical concepts describing the nature of constric-
tion size distribution 共CSD兲 of a filter have been addressed by
Indraratna and Locke 共2000兲, Raut and Indraratna 共2004兲, and
Locke et al. 共2001兲. In this study, the authors have extended the
above principles and have developed a computational procedure
to determine the filter constriction size distribution 共CSD兲 for a
given particle size distribution 共PSD兲 and relative density 共Rd兲.
Some salient features of the constriction model are elucidated
below.
Constriction Sizes in the Densest and Loosest Particle
Arrangements
In a real granular filter, particles exist in a group of three or four,
representing the densest and loosest arrangements, respectively.
Humes 共1996兲 assumed that in a filter at maximum density, only

266 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / MARCH 2007

J. Geotech. Geoenviron. Eng., 2007, 133(3): 266-276

Downloaded from ascelibrary.org by University of Tokyo on 12/25/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 3. A typical filter PSD and CSD showing passing probability

p共=1 − Pc兲

Fig. 1. Base soils and filters with various uniformity coefficients 共Cu兲
but having the same retention ratio 共D15 / d85兲 striction size in the loosest arrangement based on equivalent di-
ameter DcL is given by

the densest arrangements exist, and defined the constriction size

DcD as the diameter of the largest circle that can fit within three DcL = 冑 4Sc,max
␲
共3兲
tangent filter particles 关Fig. 2共a兲兴, which can be given by

冉 冊 冉 冊 冉 冊 冉 冊
2 2 2 2
The probability of occurrence of DcD and DcL depends upon the
2 2 2 2 probabilities of individual particles constituting the arrangements,
+ + +
D1 D2 D3 DcD and can be calculated statistically 共Silveira 1975兲. If a filter PSD

冋冉 冊 冉 冊 冉 冊 冉 冊册
= 0.5
2
D1
+
2
D2
+
2
D3
+
2
DcD
2
共1兲
is divided into a number of particle sizes 共Fig. 3兲, in the manner
explained above, DcD and DcL, and their corresponding probabili-
ties can be determined for all possible unique combinations of
However, a real filter is not always compacted to its maximum particles in the densest and loosest states, resulting in the densest
density, which implies that the densest constriction model is con- and loosest CSD models, i.e., CSD 共D兲 and CSD 共L兲.
servative. For any general particle arrangement, the constriction
space Sc between four particles 关Fig. 2共b兲兴 is given by 共Silveira Particle Surface Area and Filter Compaction
1975兲
Most researchers have used the densest CSD, i.e., CSD 共D兲 for
1
Sc = 8 共共D1 + D2兲共D1 + D4兲sin ␣ + 共D2 + D3兲共D2 + D4兲sin ␥ simplicity, where the filter PSDs either by mass or by number of
particles have been used. However, as explained by Locke et al.
− 共␣D21 + ␤D22 + ␥D23 + ␦D24兲兲 共2兲
共2001兲, although PSD by mass obtained through sieve analysis is
where the angles ␤, ␥, and ␦ can be related to ␣ by plane geom- accepted as a good representation of constriction size distribution
etry. The angle ␣ varies between ␣min and ␣max 关Figs. 2共c and d兲兴. 共CSD兲 for uniform filters, the use of PSD by mass introduces
For a particular value of ␣ between these two extreme values, errors in well-graded filters. This is because large particles with a
when the value of Sc is maximum, then the corresponding con- high individual mass but low in number will be over-represented,
as it is unlikely that these few large particles will meet together to
form a large constriction. In a similar manner, the PSD by number
over-represents the finer constrictions. Humes 共1996兲 suggested
that although there are only a small number of large particles,
they impose significant contact with other particles due to their
larger surface area, and showed that the filter PSD based on sur-
face area is the best option for filtration analysis.
If a filter material is composed of n diameters,
D1 , D2 , D3 , . . . , Dn and their mass probabilities of occurrence are
pm1 , pm2 , pm3 , . . . , pmn, respectively 共Fig. 3兲, then their respective
probabilities of occurrence by surface area 共pSAi兲 can be obtained
by 共Humes 1996兲

Fig. 2. Filter particle arrangements in 共a兲 densest state; 共b兲 loosest

pSAi = 共pmi/Di兲 冒 冉兺 冊 n

i=1
pmi
D
共4兲

state; 共c兲 minimum ␣; and 共d兲 maximum ␣ 共adapted from Silveira et Similarly, their probabilities of occurrence by number 共pNi兲 can be
al. 1975兲 obtained by 共Raut and Indraratna 2004兲

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / MARCH 2007 / 267

J. Geotech. Geoenviron. Eng., 2007, 133(3): 266-276

 and the loosest CSDs are plotted as boundaries, and two specific
CSD curves corresponding to relative densities 共Rd兲 of 50% and
70% are determined and plotted.
For a relative density of 70%, Fig. 5 illustrates the calculated
PSD and CSD curves for uniform and well-graded filters based on
mass, number of particles, and surface area. For the uniform filter
关Fig. 5共a兲兴, the difference between the above three methods is
insignificant, whereas the difference in the PSD and CSD curves
for the well-graded filter is considerable depending on the method
used. For the reasons explained earlier, the authors have used the
surface area method for subsequent constriction analyses.

Constriction Model
Downloaded from ascelibrary.org by University of Tokyo on 12/25/20. Copyright ASCE. For personal use only; all rights reserved.

Experimental Investigations

Fig. 4. A typical program output and effect of filter density on CSD Based on the probabilistic filtration theory proposed by Silveira
共1993兲, Soria et al. 共1993兲 carried out several experiments to

冒 冉兺 冊
back-analyze the filter CSD from a given PSD by mass. Relative
n density 共Rd兲 of all filters was in the order of 90%. Humes 共1996兲
pmi calculated the CSD by surface area 共SA兲 for the filters in the
pNi = 共pmi/D3i 兲 共5兲
i=1 D3i densest state 共D兲 for the same data 共i.e., denoted as CSD共D兲SA in
Fig. 6兲. The CSDs computed by the authors 关Eq. 共6兲兴 are also
Real filters are likely to exist in between the two extreme shown for comparison. Fig. 6 illustrates that these model predic-
states, densest and loosest. Irrespective of whether the CSD is tions are in good agreement with the densest model of Humes
determined by mass, number or surface area, the actual constric- 共1996兲 and the experimental findings of Soria et al. 共1993兲 at
tion size Dc for any given relative density Rd is given by Locke et Rd = 90%.
al. 共2001兲

Dc = DcD + Pc共1 − Rd兲共DcL − DcD兲 共6兲 Controlling Constriction

where Dc⫽actual constriction size for a given value of the percent
finer Pc; DcD and DcL⫽constriction sizes in the densest and loos-
est models, respectively, for the same Pc. The authors have incor-
porated these theoretical concepts in a comprehensive computer
program to compute the filter CSD, and as a worked-out example,
a typical output is shown in Fig. 4 for a uniform filter. The densest
Kenney et al. 共1985兲 used a multilayered one-dimensional con-
striction model to analytically investigate the size of controlling
constriction in a filter, defined as the size of the largest base soil
particle that can potentially infiltrate through the filter. Although
this model is a good approximation for uniform filters and pro-
vides a sound understanding of the fundamental filtration mecha-

Fig. 5. Filter PSDs and CSDs by mass 共M兲, by number 共N兲 and by surface area 共SA兲: 共a兲 uniform filter, F1 共Cu = 1.2, Rd = 70%兲; 共b兲 nonuniform
filter, F2 共Cu = 3.8, Rd = 70%兲

268 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / MARCH 2007

J. Geotech. Geoenviron. Eng., 2007, 133(3): 266-276

Downloaded from ascelibrary.org by University of Tokyo on 12/25/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 7. Probability of forward movement and predicted depth of

infiltration
Fig. 6. Model prediction of filter constrictions versus experimental
and analytical results
1
Dm = n 共12兲
pSA,i
nisms, it considers the flow channels to be independent. Consid-
ering the three-dimensional 共3D兲 pore network model and the
兺
i=1 Di
possible sideways exits available for the base soil particles, Locke
Fig. 7 represents the relationships given by Eqs. 共9兲 and 共10兲 for
et al. 共2001兲 found an increased value of probability of forward
movement P共F兲 corresponding to the value of passing probability a confidence level 共P̄兲 of 95%. Indraratna and Locke 共2000兲 and
p 共percentage of larger constrictions= 1 − Pc兲, as shown earlier in Locke et al. 共2001兲 found a constriction model less sensitive to
Fig. 3. The relevant algebraic relationship between P共F兲 and p is the choice of P̄, and adopted P̄ = 95%. It is noted that the value of
given by n becomes exceedingly high when P共F兲 in Eq. 共10兲 approaches
⬁
unity as Pc ⬍ 35%. In other words, the rapidly increasing nature
of the n-curve for Pc ⬍ 35% clearly indicates that any further
P共F兲 = p + 兺
i=0
关1 − 共1 − p兲4兴共1 − p兲p兵关1 − 共1 − p兲3兴共1 − p兲其i
increase in filter thickness beyond the value of 225Dm does not
contribute to base soil retention significantly. For all practical
共7兲 purposes, a base soil particle smaller than Dc35 may not be re-
By simplifying the above equation based on the limiting sum of tained by a granular filter, unless the constrictions become pro-
the above geometric series, one may obtain gressively finer by self-filtration. Therefore, the authors propose
that the controlling constriction in a granular filter can be given
兵1 − 共1 − p兲4其共1 − p兲p by the specific constriction size 共Dc35兲.
P共F兲 = p + 共8兲 The particle size distributions for five filters 共F1–F5兲 with
1 − 兵1 − 共1 − p兲3其共1 − p兲
varying Cu and D15 values are shown in Fig. 8共a兲. The corre-
By replacing p by 共1 − Pc兲 in Eq. 共8兲 and further simplification sponding CSDs are computed and presented in Fig. 8共b兲. In the
gives controlling constriction analysis of Kenney et al. 共1985兲, the uni-
form Filters F1, F4, and F5 were compacted to a relative density
共1 − Pc兲共1 + P4c − P5c 兲
P共F兲 = 共9兲 of about 70%, whereas the nonuniform Filters F2 and F3 were
共1 − Pc + P4c 兲 compacted to Rd = 90%. Witt 共1993兲 experimentally measured the
constriction sizes from imprints of filter particles on silicon rub-
For a confidence level P̄, the number of layers 共forward exits兲 n
bers. The controlling constrictions Dc35 of these filters have been
that a base particle with a passing probability p can infiltrate in
calculated by two different procedures and compared with the
the filter before it is captured by a smaller constriction is given by
current model in Table 1. In general, the authors’ Dc35 values
except for F3 are similar to the Kenney et al. 共1985兲 recommen-
ln共1 − P̄兲
n= 共10兲 dation of Dc* 艋 0.20 D15, but they are consistently smaller than
ln P共F兲 the value of dp* 共effective opening size兲 given by Witt 共1993兲.
The corresponding distance S that the same base particle travels One possible reason for this deviation is that Witt’s approach for
in the filter having a mean particle size Dm can now be deter- calculating dp* does not include the role of Rd. The controlling
mined as constriction size in well-graded Filter F3 共Cu = 7兲 is significantly
smaller than that in F2 共Cu = 3.8兲, and this explains the discrep-
ln共1 − P̄兲 ancy between Kenney’s Dc* = 0.210 mm and the authors’
S= · Dm = n · Dm 共11兲 Dc35 = 0.178 mm for the Filter F3 共Table 1兲. It is of interest to note
ln P共F兲
that for less uniform or well-graded filters, as in the case of F2
where 共Cu = 3.8兲 and F3 共Cu = 7.0兲, Kenney’s second recommendation

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / MARCH 2007 / 269

J. Geotech. Geoenviron. Eng., 2007, 133(3): 266-276

Downloaded from ascelibrary.org by University of Tokyo on 12/25/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 8. Controlling constrictions in filters with varying uniformity coefficients 共Cu兲 and D15 sizes

based on D5 共i.e., Dc* 艋 0.25 D5兲 is in perfect agreement with the
authors’ Dc35 as indicated in Table 1. This also verifies that as
expected, the controlling constriction sizes in nonuniform filters
are smaller than those in uniform filters for the same D15 and for
a given level of compaction 共Sherard et al. 1984a兲.
Foster and Fell 共2001兲 measured the size of base particles
washed through the filters, and found an upper bound value of
0.20D15 with a median size of 0.16D15. Using the molten wax
technique, Sherard et al. 共1984a兲 measured the dimension of mini-
mum flow channel and found 0.18D15 as an upper bound for the
effective opening size. Both these studies used highly compacted
filters with a relative density 共Rd兲 approaching 100%. The Dc35
data presented in this study by authors 共see Table 1兲 include Rd of
about 70% for uniform filters and 90% for well-graded filters,
similar to those used by Kenney et al. 共1985兲. However, the au-
thors’ filter data based on Rd = 100% can be directly compared
with those of Foster and Fell 共2001兲 and Sherard et al. 共1984a兲. In
Table 1, the values of Dc35 at Rd of 100% are also tabulated and
compared with the median size 共0.16D15兲 of eroded base particles
proposed by Foster and Fell 共2001兲, and the upper bound
共0.18D15兲 of Sherard et al. 共1984a兲. It is seen that the values of
Dc35 共at Rd = 100%兲 are only slightly larger than the median size
of eroding base particles 共0.16D15兲, and very close to the above
stated upper bounds.
Modeling of Base Soils
Eroded base soil particles are transported to the filter, making the
filter constrictions smaller upon retention, and thereby initiating
self-filtration. Like filters, the base soil can also be modeled by
the PSD based on surface area. Hereafter, all filter parameters
D15, d85, and PSD, based on surface area, will be denoted by
D15SA, d85SA, and PSDSA, respectively. Consider three base soils
having the same d85 by mass of 0.80 mm and different Cu values
共see the base soil PSDs in Fig. 1兲 filtered through a mechanical
sieve of aperture equal to 0.8 mm 共Fig. 9兲. Except for the very
uniform base Soil B1 where d85SA = d85, for less uniform Soils B2
and B3, the values of d85SA are less than that of d85 or the sieve
aperture. In other words, although the three base soils have the
same d85 共Fig. 1兲, only the base Soil B1 has effectively 15% base
soil particles larger than the sieve opening 共Fig. 9兲. As the base
soil becomes less uniform, increasingly smaller amounts of base
particles remain larger than the sieve aperture of 0.80 mm 共i.e.,
d85兲. In general, it may be concluded that d85SA should become
considerably smaller as the uniformity coefficient 共Cu兲 of the base
soil increases from 1.4 共d85SA = 0.80兲 to 4.0 共d85SA = 0.37兲. This
explains why the filter effectiveness tends to decrease as the base
soil becomes increasingly nonuniform as observed in various past

Table 1. Controlling Constrictions by past Procedures and the Current Model

Kenney et al. 共1985兲 Authors
Witt 共1993兲 Foster and Fell 共2001兲 Sherard et al. 共1984a兲
Filters 0.20D15 0.25D5 dp* Median 共0.16D15兲 0.18D15 Dc35 Dc35 共100%兲
F1 0.213 0.255 0.240 0.167 0.191 0.215 0.181
F2 0.209 0.200 0.210 0.167 0.188 0.200 0.188
F3 0.210 0.177 0.198 0.168 0.189 0.178 0.167
F4 0.095 0.115 0.108 0.076 0.086 0.096 0.081
F5 0.023 0.028 0.026 0.019 0.021 0.024 0.020
Note: dp* = effective opening size, ⫹ all dimensions in mm.

270 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / MARCH 2007

J. Geotech. Geoenviron. Eng., 2007, 133(3): 266-276


Downloaded from ascelibrary.org by University of Tokyo on 12/25/20. Copyright ASCE. For personal use only; all rights reserved.
Fig. 9. PSDSA of base soils of different uniformity coefficients
studies 共Honjo and Veneziano 1989; Lafleur 1984; Foster and Fell
2001兲.
Take another example of well-graded base soil as studied by
Lafleur 共1984兲. The study found that the base soil particles larger
than the sieve no. 4 size 共4.75 mm兲 do not influence filtration.
The PSDSA of the base Soil B-3 is computed and plotted in Fig.
10. It is noted, based on the PSDSA curve, the larger base soil
particles have insignificant representation. In effect, as these
larger particles “float” in the matrix of finer grains, the influence
of these large base particles on self-filtration is insignificant.
Development of Constriction-Based Retention
Criterion
Several past studies including Honjo and Veneziano 共1989兲 inves-
tigated the filtration process using mechanical sieves as filters,
Fig. 10. PSDSA of well-graded base Soil B-3
JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / MARCH 2007 / 271
J. Geotech. Geoenviron. Eng., 2007, 133(3): 266-276
and they revealed that the sieve can be effective in retaining the
base soils only if there are at least 15% of base soil particles
larger than the sieve aperture. These investigations were mainly
carried out on uniform base soils and filters. Lafleur 共1984兲
showed that self-filtration takes a longer time in the case of non-
uniform base soils and that the soil loss is excessive if the filter is
designed based on the Terzaghi criterion. As shown in Fig. 9, the
use of conventional d85 for nonuniform soils does not ensure that
at least 15% of the base soil particles are retained, whereas the
use of d85SA 共smaller than d85兲 invariably satisfies this condition.
Although a granular filter of randomly compacted particles is
more complex than a regular mechanical sieve, it can still be
considered as equivalent to a sieve with apertures equal to the
controlling constriction size 共Dc35兲. In this point of view, for an
effective base soil-filter combination, Dc35 must be smaller than
d85SA to ensure that at least 15% base particles are available to
initiate and sustain self-filtration, hence
Dc35
⬍1 共13兲
d85SA
The above constriction-based criterion for base soil retention is
comprehensive as it takes into consideration an array of funda-
mental filter parameters including PSD, CSD, Cu, and Rd, in com-
parison with the single filter grain size of D15 in the Terzaghi
criterion.
Verification of the Model Based on Experimental
Data
Series A: Very Uniform Base Soils and Filters
As a preliminary example, Fig. 11 presents three very uniform
sand filters 共F4, F5, and F6兲 and a base soil 共fine sand兲 all of
parallel gradation 共Cu = 1.2兲. The relevant filter and base soil pa-
rameters are also given in Table 2. It is noted that with such a
uniform base soil, the filter with a retention ratio of 5 or less is
more effective at higher relative densities. Because Dc35共F4兲
⬍ d85SA, F4 is considered effective, whereas F6 is deemed to be
ineffective as Dc35共F6兲 ⬎ d85SA. It is shown that the magnitude of
Dc35共F5兲 is almost the same as d85SA, hence, the filter F5 will
probably be effective only at higher relative densities. The above
constriction-based analysis agrees with the applicability of
USACE 共1953兲 and Sherard et al. 共1984a兲 criteria for uniform
base and filter materials, where effective filters are safely charac-
terized by D15 / d85 ⬍ 5.
Indraratna et al. 共1996兲 tested a uniform lateritic residual base
soil against three uniform sand filters 共fine, medium, and coarse兲
关Fig. 12共a兲 and Table 2兴. The filter CSDs and the PSDSA of the
base soil are computed and presented in Fig. 12共b兲. For the fine
sand filter, Dc35 共fine兲 ⬍ d85SA; hence, it is expected to be effec-
tive. The coarse sand filter is ineffective as Dc35 共coarse兲 ⬎ d85SA.
These predictions are in total agreement with the laboratory ob-
servations made by Indraratna et al. 共1996兲. However, the current
model classifies the medium sand filter also as ineffective in the
retention of lateritic base soil. Laboratory observations indicated
that a very uniform 共Cu ⬍ 1.5兲 medium sand filter often took a
much longer time to establish self-filtration compared to a uni-
form fine sand filter, while uniform coarse sand filters could not
establish a self-filtering interface at all 共Indraratna et al. 1996兲.
However, uniform medium sand filters could be made more ef-
fective when compacted to a much higher relative density exceed-
ing 70% 共Dilema 1990兲.
Downloaded from ascelibrary.org by University of Tokyo on 12/25/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 11. Series A: Analysis of very uniform filters and base soil of parallel gradations

Series B: Moderately Graded Base Soil and Filters
Data is taken from Vafai 共1997兲, where the base soil consisted of
fine beach sand from Wollongong. Two parallel-graded filters 共F1
and F2兲 consisted of a medium-grained sand and a river gravel,
respectively. The relevant filter and base soil parameters, and
laboratory results are shown in Figs. 13共a and b兲, and Table 3. In
relation to the current model, the filter CSDs and the PSDSA of the
base soil are computed and plotted in Fig. 13共b兲. It is shown that
Dc35共F1兲 ⬍ d85SA, and Dc35共F2兲 ⬎ d85SA, classifying F1 as effective
and F2 as ineffective. These predictions are in accordance with
the experimental observations reported by Indraratna and Vafai
共1997兲.

Table 2. Filter and Base Soil Parameters

Base/filter Rd D15 d85 Dc35 d85SA Laboratory
materials 共%兲 Cu 共mm兲 共mm兲 D15 / d85 共mm兲 共mm兲 Dc35 / d85SA D15 / d85R assessment
Series A: 共a兲 Very uniform base soils 共fine sand兲 and filters 共current study兲
Base soil — 1.2 — 0.116 — — 0.116 — — —
Filter F4 70 1.2 0.465 — 4.0 0.096 — 0.83 4.0 Effective
Filter F5 70 1.2 0.580 — 5.0 0.120 — 1.03 5.0 Effective
Filter F6 70 1.2 0.698 — 6.0 0.144 — 1.24 6.0 Effective
共b兲 Uniform base soils 共lateritic兲 and filters 共Indraratna et al. 1996兲
Base soil — 1.29 — 0.044 — — 0.041 — —
Coarse Filter 50 1.47 0.680 — 15.45 0.161 — 3.95 15.45 Ineffective
Medium Filter 50 1.45 0.228 — 5.18 0.054 — 1.33 5.18 Effective
Fine Filter 50 1.28 0.120 — 2.72 0.027 — 0.67 2.72 Effective
Series B: Moderately graded base soil and filters 共Indraratna and Vafai 1997兲
Base Soil — 2.86 — 0.42 — — 0.301 —
F1 90 2.87 0.75 — 1.78 0.147 — 0.49 1.78 Effective
F2 90 2.87 4.21 — 9.94 0.849 — 2.82 9.94 Ineffective
Series C: Well-graded base soil and uniform filters 共current study兲
Base soil — 8.75 — 1.8 — — 0.39 — — —
Filter F1 70 2.0 2.8 — 1.56 0.30 — 0.77 1.86 Effective
Filter F2 70 1.2 7.3 — 4.06 1.47 — 3.76 4.87 Ineffective
Series D: Well-graded base soils and filters 共Lafleur 1984兲
Base soil B-3 — 9.0 — 7.00 — — 0.159 — — —
Filter F-1 70 25.0 0.26 — 0.04 0.031 — 0.20 0.37 Effective
Filter F-5 70 1.9 15.0 — 2.14 3.236 — 20.34 21.43 Ineffective

272 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / MARCH 2007

J. Geotech. Geoenviron. Eng., 2007, 133(3): 266-276

Downloaded from ascelibrary.org by University of Tokyo on 12/25/20. Copyright ASCE. For personal use only; all rights reserved.

Fig. 12. Series A: Analysis of very uniform base soil and filters 共Rd = 50% 兲

Series C: Well-Graded Base Soil and Uniform Filters
In this series of base soil and filters, data is taken from the au-
thors’ laboratory tests. The base soil mainly consisted of well-
graded silty sand, and two selected Filters F1 and F2 consisted of
uniform sub-rounded gravel. The relevant filter and base soil pa-
rameters, and the laboratory evaluations are summarized in Table
2. As described earlier for Series A and B, by comparing Dc35
with d85SA it can be predicted that F1 is effective and F2 is inef-
fective confirming the laboratory observations. The corresponding
graphical plots are omitted for the brevity of presentation.
Series D: Well-Graded Base Soils and Filters
Finally, another example is examined based on Lafleur 共1984兲
filtration tests. They carried out several tests involving broadly
graded cohensionless tills as base soils. The filters consisted of
sand and gravel sizes 共see Table 2兲. By comparing Dc35 with
d85SA, it is clear that the Filter F-1 is effective and F-5 is ineffec-
tive. The Terzaghi criterion will classify F-5 as effective as its
retention ratio 共D15 / d85 = 2.2兲 is less than five. This again demon-
strates that for highly well-graded base soils, the Terzaghi crite-

Fig. 13. Series B: Analysis of moderately graded base soil and filters

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / MARCH 2007 / 273

J. Geotech. Geoenviron. Eng., 2007, 133(3): 266-276

 Table 3. Description of Base and Filter Materials for Various Laboratory
Base soil
Filter Laboratory
Test # Cu Type Cu D15 / d85 observation Reference
1 25.0 Broadly graded cohensionless tills 8.9 0.04 Effective Lafleur 共1984兲
2 10.0 8.9 0.14 Effective
3 1.9 8.9 2.14 Ineffective
4 25.0 Broadly graded cohensionless tills 100.0 0.02 Effective Lafleur et al. 共1989兲
5 1.9 100.0 1.36 Ineffective
6 25.0 15.4 0.09 Effective
7 2.4 15.4 2.43 Ineffective
8 1.9 15.4 5.00 Ineffective
9 25.0 100.0 0.01 Effective
Downloaded from ascelibrary.org by University of Tokyo on 12/25/20. Copyright ASCE. For personal use only; all rights reserved.

10 2.4 100.0 0.37 Ineffective

11 1.9 20.0 2.14 Ineffective
12 1.5 Lateritic residual soil 1.3 15.45 Ineffective Indraratna et al. 共1996兲
13 1.5 1.3 5.18 Effective
14 1.3 1.3 2.72 Effective
15 2.9 Moderately uniform sand 2.9 1.78 Effective Indraratna and Vafai 共1997兲
16 2.9 2.9 9.94 Ineffective
17 1.2 Uniform sand 1.2 12.86 Ineffective Current Study
18 1.4 1.2 3.25 Effective
19 1.2 1.2 6.25 Effective
20 1.3 1.3 4.01 Effective
21 1.3 1.3 5.00 Effective
22 1.3 1.3 6.02 Effective
23 1.2 Moderately uniform sand 2.6 4.71 Ineffective
24 3.1 2.6 0.90 Effective
25 1.2 Well-graded sand 5.2 5.70 Ineffective
26 4.5 5.2 0.61 Effective
27 1.2 5.8 9.73 Ineffective
28 3.1 Well-graded sand with gravel and silt 5.8 1.87 Effective
29 1.2 8.8 4.06 Ineffective
30 3.1 8.8 0.78 Effective
31 1.2 Well-graded silty sand 9.3 3.57 Ineffective
32 5.2 9.3 0.18 Effective

rion may be unreliable as this retention criterion 共D15 / d85 ⬍ 5兲
was developed on the basis of testing uniform materials 共USACE
1953兲.
The above experimental results and CSD-based analyses dem-
onstrate that the validity of the proposed constriction-based model
共Dc35 / d85SA ⬍ 1兲 is consistent with most laboratory observations.
Given the original PSDs of a filter and a base soil, and the ex-
pected level of filter compaction in the field, the current model
can reliably predict whether the filter is effective or not, for a
range of both uniform and well-graded materials.
Comparison with Existing Retention Criteria
The use of CSD and the PSD by surface area instead of PSD by
mass is a novel feature of the current model. With the increase in
Cu of the base soil, the reliability of D15 / d85 ⬍ 5 共USACE 1953兲
criterion for effective filters becomes questionable. For instance,
uniform filters can often become ineffective for highly well-
graded base soils. The proposed filter criterion eliminates this
limitation by employing d85SA instead of d85, and comparing it
with the specific constriction size Dc35. While ICOLD 共1994兲 and
NRCS 共1994兲 suggest the possible benefits of using a finer repre-
sentative particle size for well-graded base soils or regrading, the
current model incorporates the nonuniformity of a given base soil
in a more comprehensive and quantifiable manner. The existing
design practices 共NRCS 1994兲 propose considerable improve-
ments over the original Terzaghi criterion 共USACE 1953兲. It is
demonstrated in the following that the current constriction size
approach is an equally acceptable method for distinguishing be-
tween effective and ineffective filters.
Terzaghi Criterion
Thirty-two sets of experimental data from past studies for both
effective and ineffective filters including those discussed earlier in
the paper are plotted together in Fig. 14. Well-known particle-size
based retention criteria 共USACE 1953兲 are also shown in Fig.
14共a兲 to examine the demarcation between the effective and inef-
fective filters. In contrast, the authors’ constriction size based
model is shown in Fig. 14共b兲 for the same data. Fig. 14共a兲 dem-
onstrates that for highly well-graded cohesionless tills in which
the conventional 共by mass兲 d85 size is usually much larger than
d85SA, a cluster of coarse and uniform ineffective filters fall in the
predicted effective zone. This perhaps justifies the introduction of
an additional USACE 共1953兲 constraint, D15 ⬍ 0.40 mm, to ensure

274 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / MARCH 2007

J. Geotech. Geoenviron. Eng., 2007, 133(3): 266-276

Downloaded from ascelibrary.org by University of Tokyo on 12/25/20. Copyright ASCE. For personal use only; all rights reserved.
Fig. 14. Application of retention criteria to distinguish between
effective and ineffective filters 共a兲 Terzaghi criterion 共USACE 1953兲
and 共b兲 current constriction model
effective filtration. These contradictory tests are invariably related
to well-graded base materials, verifying the limitations of the
Terzaghi retention criteria. In comparison, Fig. 14共b兲 verifies that
the current model can more successfully separate the effective
from the ineffective filters for a wide range of base and filter
materials, and without being unduly conservative. It is noted that
only a few filters 共very close to the boundary兲 experimentally
evaluated to be effective fall on the predicted ineffective zone. It
is understood that for such borderline filters, where Dc35 is only
slightly larger than d85SA, the actual retention efficiency cannot be
assured with absolute confidence. For instance, some uniform fil-
ters may take a long time to establish a self-filtering layer thereby
losing a significant amount of base soil initially, but stabilizing
after a much longer period of time. It is encouraging to note that
in Fig. 14共b兲, none of the experimentally proven ineffective filters
fall in the predicted effective zone. In this respect, the proposed
criterion of Dc35 / d85SA ⬍ 1 seems to benefit from an inherent
safety factor.
Current Design Practice
The use of d85SA in the current model and d85R in the NRCS
共1994兲 design guidelines provides two alternative solutions to the
same filtration problem. The current model cannot be directly
compared with the NRCS 共1994兲 guidelines where the filter
boundaries vary depending on the percentage of fines in the base
soils. However, as the current model is developed for cohesion-
less soils 关Fig. 14共b兲兴, it can be compared to the available crite-
rion 共D15 / d85R 艋 4兲 for regraded base soils 共Fig. 15兲. Fig. 15
represents d85R on the horizontal axis, and the boundary
D15 / d85R = 4 demarcates the effective filters from the ineffective
ones. It can be seen that while the regraded boundary applies well
JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / MARCH 2007 / 275
J. Geotech. Geoenviron. Eng., 2007, 133(3): 266-276
Fig. 15. Application of retention criteria to distinguish between
effective and ineffective filters using the current design practice with
regraded base soil PSDs 共NRCS 1994兲
for cohensionless soils, the authors’ criterion based on constric-
tion size Dc35 is equally acceptable 关Fig. 14共b兲兴. However, a key
advantage of the proposed Dc35 approach is that regrading of base
soil is not needed. Moreover, this method is based on analytical
principles capturing the surface area and constriction size con-
cepts rather than a purely empirical technique.
Filter Thickness
As explained earlier 共see Fig. 7兲, the current model suggests a
minimum filter thickness of about 225Dm for a 95% confidence
level. A smaller thickness can be obtained for a reduced confi-
dence level or if a coarser base soil is considered. Given that the
computation of Dm is based on surface area principle, it varies
from D5 to D15 in most practical dam filters. In this respect, a
filter thickness of 225Dm is in agreement with the laboratory ob-
servations of 300共D5 to D10兲 as suggested by Witt 共1993兲 and
200D5 共Kenney et al. 1985兲. For typical filter gradations 共e.g.
ICOLD, 1994兲, all these values vary in the range of 40– 60 mm
and may be used as preliminary guidance in the design of filters.
In practice, the thickness of dam filters is usually much greater
than the above mentioned values. For both construction feasibility
and structural stability, the actual thickness of dam filters often
exceeds 500 mm 共ICOLD, 1994兲.
Conclusions
Filter criteria employed in practice are often based on laboratory
tests that were carried out on uniform filter and base materials.
Most of these empirical criteria invariably involve some charac-
teristic particle sizes by mass, typically the Terzaghi retention
ratio, D15 / d85. They have obvious limitations especially when
well-graded base soils are tested with coarse uniform filters. Un-
less regrading is carried out 共NRCS 1994兲, the original D15 / d85
criterion does not hold for many well-graded base soils. In this
study, in lieu of regrading, the authors have proposed an alterna-
tive filter criterion based on the controlling constriction and sur-
face area concepts applied to base soils, especially those that are
well-graded. Similar to mechanical sieves, it is the constriction
共opening兲 sizes rather than the particle sizes that influence filtra-
tion in granular filters 共Kenney et al. 1985; Locke et al. 2001兲. In
this study, the representative filter constriction size is proposed to
be Dc35. Also, by surface area consideration of the base soil, d85SA
is found to be a more appropriate representative parameter com-
 pared to the conventional d85 by mass. It is demonstrated that the
proposed criterion, Dc35 / d35SA ⬍ 1, for identifying effective filters
is more realistic, whereby the size of prescribed base particle size
共d85SA兲 is directly compared with the size of controlling filter
constriction 共Dc35兲.
For a wide range of base and filter materials, the current model
is shown to successfully separate the effective from the ineffec-
tive filters without being unduly conservative. Unlike D15 the
main advantage of Dc35 is that it is sensitive to the uniformity
coefficient 共Cu兲 of the filter particle size distribution 共PSD兲 as
well as to its relative density 共Rd兲. It can also be used to estimate
the minimum desired filter thickness as a function of the mean
filter particle size 共Dm兲, where a value of 225Dm can be regarded
as a reasonable approximation. In this respect, the design of dam
Downloaded from ascelibrary.org by University of Tokyo on 12/25/20. Copyright ASCE. For personal use only; all rights reserved.
filters in the order of 400– 800 mm thickness would usually be
adequate in practice to retain most base soils.
It is important to note that the proposed constriction-based
criterion for effective filters 共Dc35 / d85SA ⬍ 1兲 is developed for co-
hensionless base soils. For cohesive soils, granular filters with
conventional retention ratio, D15 / d85, even exceeding nine have
been found to be effective 共Sherard et al. 1984b兲. In contrast,
Foster and Fell 共2001兲 found that some filters with cohesive soils
having a retention ratio less than nine also failed. As the authors
are still in the process of conducting further tests on a variety of
cohesive base soils, the application of the proposed criterion for
cohesive base soils is cautioned. In this study, the authors have
considered both uniform and well-graded base and filter materials
in the development of the model. However, the clogging potential
of some well-graded filters 共Indraratna et al. 1996兲 has not been
evaluated here. Undoubtedly, further investigations of well-
graded filter materials will be required to extend the current
model to describe clogging.
Acknowledgments
The writers express their special thanks to Dr. Mark Locke
共SMEC-International兲, former Ph.D. candidate at the University
of Wollongong for his contribution to the development of initial
framework of the CSD model. The support provided by Snowy
Mountain Engineering Corporation 共SMEC, Melbourne兲 and Sin-
clair Knight Mertz Ltd. 共SKM, Sydney兲 in this area of research is
appreciated.
References
Bertram, G. E. 共1940兲. “An experimental investigation of protective fil-
ters.” Rep. No. 267, Harvard Graduate School of Engineering, Vol. 6.
Dilema, E. L. G. 共1990兲. “Development of permeability-floc size criteria
for granular filter design.” MSc thesis, No. GT-89-4, Asian Institute of
Technology, Bangkok, Thailand.
Foster, M., and Fell, R. 共2001兲. “Assessing embankment dam filters that
do not satisfy design criteria.” J. Geotech. Geoenviron. Eng., 127共5兲,
276 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / MARCH 2007
J. Geotech. Geoenviron. Eng., 2007, 133(3): 266-276
398–407.
Honjo, Y., and Veneziano, D. 共1989兲. “Improved filter criterion for cohe-
sionless soils.” J. Geotech. Engrg., 115共1兲, 75–94.
Humes, C. 共1996兲. “A new approach to compute the void-size distribution
curves of protective filters.” Proc., GeoFilters ’96, J. Lafleur and A.
L. Rollin, eds., Bitech Pub., Montreal, Canada, 57–66.
ICOLD. 共1994兲. “Embankment dams—Filters and drains.” Bulletin, No.
95, ICOLD Nice, France.
Indraratna, B., and Locke, M. 共2000兲. “Analytical modeling and experi-
mental verification of granular filter behavior.” Keynote Paper, Filters
and Drainage in Geotech. and GeoEnv. Eng., W. Wolski and J. Mly-
narek, eds., Balkema, Rotterdam, The Netherlands, 3–26.
Indraratna, B., and Vafai, F. 共1997兲. “Analytical model for particle migra-
tion within base soil-filter system.” J. Geotech. Geoenviron. Eng.,
123共2兲, 100–109.
Indraratna, B., Vafai, F., and Dilema, E. 共1996兲. “An experimental study
of the filtration of a lateritic clay slurry by sand filters.” Proc. Inst.
Civ. Eng., Struct. Build., 119共2兲, 75–83.
Kenney, T., Chahal, R., Chiu, E., Ofoegbu, G., Omange, G., and Ume, C.
共1985兲. “Controlling constriction sizes of granular filters.” Can. Geo-
tech. J., 22共1兲, 32–43.
Lafleur, J. 共1984兲. “Filter testing of broadly graded cohesionless tills.”
Can. Geotech. J., 21, 634–643.
Locke, M., Indraratna, B., and Adikari, G. 共2001兲. “Time-dependent par-
ticle transport through granular filters.” J. Geotech. Geoenviron. Eng.,
127共6兲, 521–529.
Natural Resources Conservation Services. 共1994兲. “Gradation design of
sand and gravel filters.” National engineering handbook, Chap. 26,
Part 633, USDA.
Raut, A. K., and Indraratna, B. 共2004兲. “Constriction size distribution of
a non-uniform granular filter.” Proc. 15th, South East Asian Geotech-
nical Conf., Bangkok, Thailand, 409–414.
Sherard, J., Dunnigan, L., and Talbot, J. 共1984a兲. “Basic properties of
sand and gravel filters.” J. Geotech. Engrg., 110共6兲, 684–700.
Sherard, J., Dunnigan, L., and Talbot, J. 共1984b兲. “Filters for silts and
clays.” J. Geotech. Engrg., 110共6兲, 701–718.
Silveira, A. 共1993兲. “A method for determining the void size distribution
curve for filter materials.” Proc., Filters in Geotechnical and Hydrau-
lic Engineering, J. Brauns, M. Heibaum, and U. Schuler, eds.,
Balkema, Rotterdam, The Netherlands, 71–74.
Silveira, A., de Lorena Peixoto T., and Nogueira, J. 共1975兲. “On void size
distribution of granular materials.” Proc., 5th Pan-Amer. Conf. Soil
Mech. and Found. Eng. 161–176.
Soria, M., Aramaki, R., and Viviani, E. 共1993兲. “Experimental determi-
nation of void size curves.” Proc., Filters in Geotechnical and Hy-
draulic Engineering, J. Brauns, M. Heibaum, and U. Schuler, eds.,
Balkema, Rotterdam, The Netherlands, 43–48.
USACE. 共1953兲. “Investigation of filter requirements for underdrains.”
Tech. Memo., No. 3–360, U.S. Waterways Experiment Station, Vicks-
burg, Miss.
Vafai, F. 共1996兲. “Analytical modelling and laboratory studies of particle
transport in filter media.” Ph.D. thesis, Univ. of Wollongong, Wollon-
gong City, Australia.
Witt, K. 共1993兲. “Reliability study of granular filters.” Proc., Filters in
Geotech., and Hydraul. Eng., J. Brauns, M. Heibaum, and U. Schuler,
eds., Balkema, Rotterdam, The Netherlands, 35–41.