Geotechnical Testing Journal

doi:10.1520/GTJ20160069 / Vol. 40 / No. 3 / May 2017 / available online at www.astm.org

TECHNICAL NOTE

S. M. Fattahi,1 A. Soroush,2 and P. T. Shourijeh3

The Hole Erosion Test: A Comparison

of Interpretation Methods

Reference
Fattahi, S. M., Soroush, A., and Shourijeh, P. T., “The Hole Erosion Test: A Comparison of Interpretation
Methods,” Geotechnical Testing Journal, Vol. 40, No. 3, 2017, pp. 494–505, http://dx.doi.org/10.1520/
GTJ20160069. ISSN 0149-6115

ABSTRACT
Manuscript received April 6, 2016; The hole erosion test (HET) is widely performed for determining soil erosion characteristics;
accepted for publication November 30,
viz. critical shear stress, erosion rate coefficient, and erosion rate index. Refinements to
2016; published online February 16,
2017. measurement/estimation of water head drop in the hole through the specimen, which is
1
essential in accurate interpretation of HET, were proposed. Particular attention was devoted
Dept. of Civil & Environmental
Engineering, Amirkabir Univ. of to head losses in the entrance/exit of flow to/from the hole in the specimen, and any
Technology, Tehran, Iran changes in the hole’s entrance geometry. Soil erodibility parameters of 13 core soils obtained
2
Dept. of Civil & Environmental in HET through previously available methods, and the new proposed technique were
Engineering, Amirkabir Univ. of described and compared. Results suggested that, although differences are observed
Technology, Tehran, Iran
(Corresponding author),
between erosion rate coefficient and critical shear stress from different methods, the erosion
e-mail: [email protected] rate indices were close, in so far as the interpretation methods manifest an identical
3
Dept. of Civil & Environmental
classification for soil erosion propensity.
Engineering, Shiraz Univ. of
Technology, Shiraz, Iran
Keywords
internal erosion, concentrated leak, Hole Erosion Test (HET), erosion rate, critical shear stress

Nomenclature
Ce ¼ coefficient of soil erosion (s/m)
fL ¼ friction factor in laminar flow conditions (Ns/m3)
fT ¼ friction factor in turbulent flow conditions (Ns2/m4)
g ¼ acceleration of gravity (m/s2)
hd ¼ downstream hydraulic head (m)
hent ¼ head loss in hole entrance (m)
hext ¼ head loss in hole exit (m)

Copyright V C 2017 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. 494
&RS\ULJKWE\$670,QW O DOOULJKWVUHVHUYHG )UL-XQ*07
'RZQORDGHGSULQWHGE\
9LUJLQLD3RO\WHFK 6WDWH8QLY 9LUJLQLD3RO\WHFK 6WDWH8QLY SXUVXDQWWR/LFHQVH$JUHHPHQW1RIXUWKHUUHSURGXFWLRQVDXWKRUL]HG
 FATTAHI ET AL. ON HOLE EROSION TEST METHODS 495

hf ¼ friction head loss along the axial hole (m) TABLE 1 Qualitative description of erosion rate based on erosion
rate index- Wan and Fell (2002, 2004a).
hu ¼ upstream hydraulic head (m)
I ¼ erosion rate index Group No. Erosion Rate Index (I) Description
i ¼ flow gradient along the hole 1 <2 Extremely rapid
Kent ¼ entrance head loss coefficient 2 2–3 Very rapid
Kext ¼ exit head loss coefficient 3 3–4 Moderately rapid
L ¼ length of axial hole (m) 4 4–5 Moderately slow
LL ¼ liquid limit (%) 5 5–6 Very slow
 ¼ kinematic viscosity (m2/s) 6 >6 Extremely slow

p ¼ pressure (kN/m2)
PI ¼ plasticity index (%)
Q ¼ flow rate (m3/s) hole is performed prior to testing. The prepared specimen is
Re ¼ Reynolds number then subjected to head or flux controlled flow, while hydraulic
t ¼ elapsed time (s) conditions (flow rate, water head, etc.) in addition to any hole
tf ¼ test termination (end) time (s) enlargement, are monitored in order to determine erodibility
Vd ¼ mean flow velocity in downstream chamber (m/s) characteristics.
Vh ¼ mean flow velocity in axial hole (m/s) In the HET soil erosion is defined as e_t ¼ Ce(st  sc)
Vu ¼ mean flow velocity in upstream chamber (m/s) where, e_t ¼ rate of erosion per unit surface area of hole at time t
wopt ¼ optimum water content (%) (kg/s/m2), Ce ¼ coefficient of soil erosion (s/m), st ¼ hydraulic
z ¼ elevation head (m) shear stress exerted by eroding fluid to hole surface area at time
DH ¼ total energy head loss (m) t (N/m2), and sc ¼ critical shear stress, that is minimum hydrau-
Dh ¼ hydraulic head difference across the soil specimen (m) lic shear stress for erosion initiation (N/m2). As seen in Table 1,
e_t ¼ rate of erosion per unit surface area of hole at time soil erosion propensity is designated by the erosion rate index,
t (kg/s/m2) I, defined as I ¼ logCe (Wan and Fell 2002).
/d ¼ downstream flow chamber diameter (m) The erosion rate- from the hole- is estimated by (Wan and
/u ¼ upstream flow chamber diameter (m) Fell 2002):
/t ¼ estimated mean diameter of axial hole at time t (m)
h ¼ entrance funnel shape angle qd d/t
e_t ¼ (1)
qd ¼ dry density of soil (kg/m3) 2 dt
qw ¼ density of water (kg/m3) where:
sc ¼ critical shear stress for soil erosion (N/m2) qd ¼ soil dry density (kg/m3), and
st ¼ shear stress exerted by eroding fluid at time t (N/m2) d/t/dt ¼ change in hole diameter with time (m/s).
The hole diameter is calculated from pipe hydraulics as the
following equations (Wan and Fell 2002):
8 1
Introduction >
> 16QLfL 2
>
< Laminar Flow
pq gh
Internal erosion from concentrated leaks occurs when water /t ¼  w f 1 (2)
>
> 64Q2 LfT 5
>
: 2 Turbulent Flow
flows in an opening, crack, or deficiency in earth structures p qw ghf
(ICOLD 2016). Consequently, the crack transmits water akin to
a pipe, and soil particles are eroded due to hydraulic shear where:
stresses exerted (ICOLD 2013). A well-known and very danger- /t ¼ estimated mean diameter of hole at time t (m),
ous type of concentrated leak is erosion in cracks through Q ¼ flow rate (m3/s),
impervious cores of embankment dams (Bonelli 2013). In such L ¼ length of axial hole (m),
cases, the erodibility of core materials has significant influence qw ¼ water density (kg/m3),
on initiation and progression of internal erosion (Foster et al. g ¼ acceleration of gravity (m/s2),
2000). hf ¼ friction head loss along axial hole (m), and
The hole erosion test (HET), developed by Wan and Fell fL and fT ¼ friction factors of hole periphery in laminar
(2002, 2004a, 2004b), is a comparatively simple, fast, and (Ns/m3) and turbulent (Ns2/m4) flow conditions, respectively.
repeatable index test to study soil erosion from concentrated The flow is turbulent for Reynolds number, Re, higher
leaks (ICOLD 2016). This test includes a reconstituted or undis- than 2000 (Wahl et al. 2008a). Note that Re ¼ Vh/t/, where
turbed soil specimen in which a 6 mm-diameter longitudinal Vh ¼ flow velocity in hole (m/s), /t ¼ hole diameter (m), and

&RS\ULJKWE\$670,QW O DOOULJKWVUHVHUYHG )UL-XQ*07

'RZQORDGHGSULQWHGE\
9LUJLQLD3RO\WHFK 6WDWH8QLY 9LUJLQLD3RO\WHFK 6WDWH8QLY SXUVXDQWWR/LFHQVH$JUHHPHQW1RIXUWKHUUHSURGXFWLRQVDXWKRUL]HG
496 Geotechnical Testing Journal

 ¼ water’s kinematic viscosity (m2/s). Exact measurements of flow, friction factor variation, and determination of final hole
/t, are possible only at the beginning and end of tests, diameter.
from which the initial and final friction factors may be Luthi (2011) and Luthi et al. (2012) presented a new HET
back-calculated (cf. Eq 2). Throughout the test duration, logical configuration, termed HET-P, wherein incorporating a Pitot
estimations of fL and fT are prerequisite for mathematical tube provides accurate measurement of total energy head and
calculation of /t. It is assumed that fL and fT vary in linear pro- relevant losses. Fig. 1 schematically shows the Pitot configura-
portion to (Q/hf)1/3 and (Q2/hf)1/5 for laminar and turbulent tion in HET-P apparatus.
flow, respectively, in-between their initial (t ¼ 0), and final Reiffesteck et al. (2006) proposed effluent turbidity assess-
(t ¼ tf) values (Farrar et al. 2007; Wahl et al. 2008a, 2008b). ment for interpreting HET results. This approach is versatile,
The boundary shear stress, st, exerted to the hole periphery, yet requires sophisticated/expensive equipment. Researchers
is directly related to hydraulic gradient, i ¼ hf/L, along the hole; such as Chevalier (2011), Chevalier et al. (2012), Benahmed and
that is (Wan and Fell 2002): Bonelli (2012), and Haghighi et al. (2012) opted to use turbidity
assessments in studying soil erosion.
/t
st ¼ qw gi (3) In addition to laboratory proceedings, a number of
4
researchers studied HET analytically and numerically. Bonelli
The various simplifying assumptions and limitations of the et al. (2006) and Bonelli and Brivois (2008) proposed a numeri-
HET procedure and interpretation method, have led to attempts cal erosion model that correlates a dimensionless hole radius
by researchers to modify/refine the HET test. with critical shear stress, hydraulic gradient, and a dimension-
Lim (2006) pointed to some problems of the HET and pro- less test time, enabling estimation of erosion characteristics
posed modifications to specimen preparation, friction factors without the need for interpolating hole diameter. Boukhemacha
estimations, and test interpretation in case of slaking. Farrar (2009) and Boukhemacha et al. (2013) presented a mathemati-
et al. (2007) and Wahl et al. (2008a, 2008b) modified the HET cal model for piping erosion; this model is derived from pipe
apparatus and test setup, and also discussed factors effective in flow equations, soil detachment and mass conservation laws,
interpreting HET results, viz. curve-fitting procedures, erosion and may be written into an analytical radius-time equation in
regimes (delayed or non-delayed), laminar versus turbulent case of a constant pressure drop. Marot et al. (2011) and

FIG. 1
Schematic illustration of flow conditions in HET apparatus
with Pitot-tube (Not to scale); redrawn/reworked after Luthi
(2011) and Riha and Jandora (2014).

&RS\ULJKWE\$670,QW O DOOULJKWVUHVHUYHG )UL-XQ*07

'RZQORDGHGSULQWHGE\
9LUJLQLD3RO\WHFK 6WDWH8QLY 9LUJLQLD3RO\WHFK 6WDWH8QLY SXUVXDQWWR/LFHQVH$JUHHPHQW1RIXUWKHUUHSURGXFWLRQVDXWKRUL]HG
 FATTAHI ET AL. ON HOLE EROSION TEST METHODS 497

Regazzoni and Marot (2013) proposed a delicately balanced where hent, hext, and hf are head losses in the entrance, exit and
equation for fluid energy dissipation (considering identical along the hole in specimen, respectively (cf. Fig. 1). The follow-
upstream and downstream velocities) and eroded mass meas- ings provide detailed description of the head losses:
urements in HET. Kissi et al. (2012), Riha and Jondora (2014),
and Mercier et al. (2015) incorporated Computational Fluid
hent or Head Loss Upon Flow Entrance to Hole
Dynamics (CFD) software to analyze the HET.
At the hole’s entrance the cross-sectional area of flow reduces to
The techniques employed for analyzing HET data, which a minimum, that is smaller than the hole. This phenomenon is
depend on available equipment and measurements, stem from
known as “vena contrata,” after which the stream widens to fill
various fundamental theoretical assumptions that deserve
the hole. A fraction of the velocity head that is lost as a result of
precise pondering prior to application. Different methods for
stream turbulence, is defined by (Riha and Jondora 2014; Potter
interpreting HET data may have merits and limitations,
et al. 2011):
depending on soil type and erosion severity.  2
V
This treatment aims to refine the method from which hent ¼ Kent h (6)
2g
water head drop (i.e., hf) in the hole through the specimen is
determined. This head drop is essential in accurate estimation/ where:
calculation of shear stresses exerted by eroding water flow on the Vh ¼ flow velocity in hole, and
walls of the hole, and in turn determination of principal soil ero- Kent ¼ entrance head loss coefficient.
sion parameters, that are critical shear stress, sc, and erosion rate The graph in Fig. 2, presented by Okiishi et al. (2006), may
coefficient, Ce. To this end, soil erodibility parameters of 13 core be used to determine Kent. Accordingly, hent  0.5Vh2/2g and this
soils obtained in HET through previously available methods, dissipation occurs between the “vena contrata” up-until the
and the new proposed technique are described and compared. point when a uniform stream flow fills the hole (cf. Fig. 1). It is
worthwhile noting that from continuity Vh is much higher than
Flow Hydraulics in HET Vu, hence; hent > Vu2/2g (cf. Fig. 1). The upstream hole end in
erodible soil specimens usually deforms into a bevel/funnel-
shaped entry during HET. Nonetheless, hent would still be signif-
Considering the center flow line through the HET apparatus in icant since the bevel-shaped hole is small in dimension com-
Fig. 1, the conservation of energy, configured by Bernoulli’s law, pared to the upstream flow chamber diameter, hence the stream
requires that (Luthi 2011; Bonelli 2013): line would experience sudden contraction (Riha and Jondora
0 h 1 0 h 1 2014).
u d
zfflfflfflfflffl}|fflfflfflfflffl{ zfflfflfflfflffl}|fflfflfflfflffl{
B pu Vu C
2 B Vd C
2
B C B pd C
@q g þ zu þ 2g A ¼ @q g þ zd þ 2g A þDH (4)
hext or Head Loss Upon Flow Exit From Hole
w w
As illustrated in Fig. 1, at the hole’s exit, stream flow from a
|fflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl} |fflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl}
Hu Hd smaller pipe (hole) into a larger pipe (downstream chamber)
experiences sudden enlargement, accompanied by abrupt
where:
p ¼ pressure,
qw ¼ water density,
FIG. 2 Coefficient of entrance head loss, Kent; redrawn after Okiishi et al.
g ¼ acceleration of gravity,
(2006).
V ¼ mean flow velocity,
z ¼ elevation head,
h ¼ hydraulic head,
H ¼ total energy head, and
DH ¼ energy head loss from specimen’s upstream to down-
stream, and connotations of u and d represent, respectively, the
specimen’s upstream and downstream. Apparently, p/qwg and
V2/2g define the pressure and velocity heads, respectively.
The energy head loss, i.e., DH ¼ Hu  Hd, of flow passing
through the hole from the specimen’s upstream to downstream,
consists of three parts (Riha and Jondora 2014; Bonelli 2013),
hence:

DH ¼ hent þ hext þ hf (5)

&RS\ULJKWE\$670,QW O DOOULJKWVUHVHUYHG )UL-XQ*07

'RZQORDGHGSULQWHGE\
9LUJLQLD3RO\WHFK 6WDWH8QLY 9LUJLQLD3RO\WHFK 6WDWH8QLY SXUVXDQWWR/LFHQVH$JUHHPHQW1RIXUWKHUUHSURGXFWLRQVDXWKRUL]HG
498 Geotechnical Testing Journal

velocity decrease, which results in turbulence and velocity head and negligible. Hence, Vu2/2g, Vd2/2g and associated losses, i.e.,
loss, calculated by (Okiishi et al. 2006; Potter et al. 2011): hent and hex, are excluded from hydraulic calculations, resulting
   2 in DH ¼ Dh ¼ hf. In this framework, termed the Hydraulic
Vh2 /h 2 Grade Line (HGL), DH is measured from differential hydraulic
hext ¼ Kext Kext ¼ 1ð Þ (7)
2g /d heads measured by manometers at sidewalls of flow chambers,
where: just upstream and downstream of test specimen (cf. Fig. 1).
/h ¼ hole diameter,
/d ¼ downstream flow chamber diameter, and Luthi (2011)
Kext ¼ exit head loss coefficient. According to Luthi (2011) and Luthi et al. (2012), Vu2/2g and
As illustrated in Fig. 1, hext is exhausted through a horizon- Vh2/2g are both important in hydraulic considerations.
tal distance, downstream the hole exit, beyond which the turbu- The upstream velocity, Vu, is determined by continuity, whilst
lence secedes, and stream velocity becomes Vd, that from flow Vu2/2g appears insignificantly small compared to the resolution
rate continuity, is equal to Vu. Hence, Vh2/2g ¼ hext þ Vd2/2g. In of hu (cf. Fig. 1); however, neglecting Vu2/2g introduces signifi-
HET, /h is significantly smaller than /d, thus Kext  1. On the cant errors at low hu accompanied with high flow rates. Con-
other hand, Vd is insignificant considering the large /d. There- trary to this, Luthi (2011) perceived that hent is insignificant and
fore for HET interpretations that require measuring down- negligible. With the application of a Pitot-static tube, in close
stream hydraulic conditions immediately at the hole’s end, it proximity to the point where the flow stream exits the hole, see
2
Fig. 1, Hd ¼ hd þ Vh /2g is directly measured, circumventing the
would be convenient to say, hext  Vh2/2g and all the velocity
head of exiting stream jet is lost (cf. Fig. 1). need to calculate hext. This method is termed the energy grade
line (EGL), cf. Fig. 1, since DH ¼ hf is measured considering
complete conservation of hydraulic energy, i.e., accounting for
hf or Friction Head Loss From Stream Flow in Hole
Vu, and Vh immediately downstream.
The friction loss in the hole, hf, is caused by counteracting shear
stresses, i.e., st, that are exerted by the stream flow on the hole’s
Riha and Jondora (2014)
peripheral walls (note: refer to the well-known Darcy-Weisbach
Likewise, Wan and Fell (2002, 2004), Riha and Jondora (2014)
equation). As previously stated (Eq 2), hf depends on the fric-
considered Vu2/2g and Vd2/2g to be insignificant and negligible,
tion factor of hole (fT for turbulent and fL for laminar flow),
that is H ¼ h. However, CFD numerical simulations by Riha
that is governed primarily by Re and relative roughness of the
and Jondora (2014) revealed that for the turbulent flow entering
hole’s peripheral walls (Streeter et al. 1998), all which are con-
the hole in HET, hent is important and requires attention in
tinuously changing in an HET test.
hydraulic considerations. It was also observed by those
researchers that Kent, and in turn hent, depends on h that is
HET Interpretation Methods hole’s entrance bevel-shaped angle (note that, h is measured
from the line perpendicular to hole orientation, as illustrated in
Hitherto, different aspects of flow hydraulics in HET, and sev-
Fig. 6). The CFD modelling results for /u ¼ 100 mm and
eral assumptions regarding the relevancy and significance of
/h ¼ 6 mm (see Fig. 1), suggested Kent ¼ 0.33 for h ¼ 80 and
head losses have been considered by researchers. Three main
Kent ¼ 0.27 for h ¼ 60 . For the stream flow exiting the hole, it
perspectives, summarized in Table 2, are as follows:
was assumed that Kext  1. As seen in Fig. 1, in the context of
this paper, the flow-hydraulics approach of Riha and Jondora
Wan and Fell (2002)
(2014) is denoted as the modified hydraulic grade line (MHGL),
It was assumed by Wan and Fell (2002) that flow velocity, i.e.,
since DH ¼ Dh ¼ hf þ hent þ hext.
V, in the upstream/downstream flow chambers is insignificant

HET Testing Program

TABLE 2 Hydraulic considerations of HET.
APPARATUS CONFIGURATION
Reference H DH ¼ Hu  Hd Connotation
The HET apparatus designed and commissioned in this study is
Wan and Fell (2002) p hf HGL
þz schematically illustrated in Fig. 3. The apparatus consists of a
qw g
Luthi (2011) p V2 hf EGL 100 mm internal-diameter test cell (length to diameter ratio
þzþ of 1) and upstream/downstream flow chambers (100 mm inside
qw g 2g
Riha and Jandora (2014) p hf þ hent þ hext MHGL diameter and 500 mm length). Contrary to many precursor
þz
qw g studies, herein the HET setup is vertically aligned. This is
This study p V2 hf þ hent MEGL
þzþ beneficiary from the prospects of producing a symmetrical and
qw g 2g
regularly shaped post-test eroded hole, and precluding

&RS\ULJKWE\$670,QW O DOOULJKWVUHVHUYHG )UL-XQ*07

'RZQORDGHGSULQWHGE\
9LUJLQLD3RO\WHFK 6WDWH8QLY 9LUJLQLD3RO\WHFK 6WDWH8QLY SXUVXDQWWR/LFHQVH$JUHHPHQW1RIXUWKHUUHSURGXFWLRQVDXWKRUL]HG
 FATTAHI ET AL. ON HOLE EROSION TEST METHODS 499

FIG. 3
(a) Photograph, and (b) schematic illustration of
HET apparatus; note: 1-upstream conical
connector, 2-hole in specimen (6 mm), 3-
upstream flange/baseplate, 4-soil specimen
(height ¼ 10 cm, diameter ¼ 10 cm), 5-hd
measurement (from Pitot tube), 6-Hd
measurement, 7-downstream flow chamber
(length ¼ 50 cm, internal diameter ¼ 10 cm), 8-
downstream conical connector, 9-outlet pipe,
10-Pitot tube, 1-hd measurement (from
manometer), 12-downstream flange/baseplate,
13-wire grid (6.7 mm aperture, 30 mm circular
central opening), 14-poly(methyl methacrylate)
(PMMA) cylinder (2 cm wall thickness), 15-hu
measurement (from manometer), 16-
downstream flow chamber (length ¼ 50 cm,
internal diameter ¼ 10 cm), 17-inlet pipe.

accumulation of eroded mass and air entrapment in the hole. solids (TDS) of 300–400 mg/l, dissolved oxygen (DO) around
Concerns about the mechanical stability of the vertical HET 6 mg/l, electric conductivity (EC) of about 0.1 ls/cm, and pH of
specimen are mitigated by utilizing a bottom wire grid (cf. Fig. 6.5–8.5. The water temperature in all tests was from 18 to 20 C.
3).
A Pitot-static tube is installed at the downstream hole end
(cf. Fig. 3). The experimental setup uses manual flow rate (by TESTING PROCEDURE AND INTERPRETATION METHOD

stopwatch and graduated cylinder) and head (using standpipes) Soil specimens were moisturized to target water content and
measurements. The water flow is provided by a variable- stored overnight in plastic bags for curing. Test specimens were
elevation constant-head tank, with maximum applicable test reconstituted in four layers inside the cell using a standard com-
head of 3500 mm (note that; test head is neither Hu nor hu, but paction hammer, while exercising an under-compaction scheme
rather the constant-head of water provided by the variable- (Ladd 1978) to insure even compaction throughout the speci-
elevation tank during testing stages). The upstream/downstream men’s length, in order to prevent possible cavities (i.e., necking)
flow chambers are fabricated from transparent poly(methyl in the hole upon erosion (Lim 2006). A 6 mm-diameter hole
methacrylate) (PMMA), thus enabling qualitative assessment of was carefully drilled along the specimen’s longitudinal axis,
emerging effluent turbidity with guides similar to pinhole dis- using a sharp auger.
persion test (ASTM D4647/D4647M-13). The test cell is affixed in the apparatus, between the
upstream and downstream flow chambers, and manometer
pipes/connections are installed. The system is then de-aerated
MATERIALS SPECIFICATIONS very slowly by a minute upward flow of water. The test is started
The fine grained soils selected for testing were acquired from after the upstream constant head tank is adjusted to provide the
core material borrows of several embankment dams currently initial test head of 200 mm, and manometer valves are opened.
under construction in Iran. The gradation curves and properties Elapsed time, manometers readings, flow rate, and water tem-
of tested soils are presented in Fig. 4 and Table 3, respectively. perature are recorded every 1–2 minutes.
All soil samples had a maximum size of 4.75 mm (cf. Fig. 4); i.e. Whenever the hole experiences no erosion (noticed from
about 1/21 cell diameter, which fits well in limits suggested for effluent turbidity and constant flow rate), the test head is
the permeameter test (ASTM D2434-68(2006)). Erosion tests increased (by raising the upstream constant head tank)—usually
were carried out using tap water, with average total dissolved doubled unless there are reasons to believe that critical

&RS\ULJKWE\$670,QW O DOOULJKWVUHVHUYHG )UL-XQ*07

'RZQORDGHGSULQWHGE\
9LUJLQLD3RO\WHFK 6WDWH8QLY 9LUJLQLD3RO\WHFK 6WDWH8QLY SXUVXDQWWR/LFHQVH$JUHHPHQW1RIXUWKHUUHSURGXFWLRQVDXWKRUL]HG
500 Geotechnical Testing Journal

FIG. 4
Gradations of core soils samples.

conditions may occur at an intermediate level. If the flow rate Possible shortening, in length of eroded hole during testing,
increases, the test head is maintained until the test completion is assumed to be linearly related with elapsed time (Farrar et al.
(Luthi 2011). 2007; Wahl et al. 2008a).
The test is terminated (by closing the downstream valve) Test interpretations were predicated on a new method devised
upon observing one of the following conditions: several minutes by the authors, termed the modified energy grade line (MEGL),
of accelerating flow, no noticeable erosion in one hour at maxi- which in essence combines aspects of EGL and MHGL methods
mum test head, and severe erosion with hole enlargement, (see HET Interpretation Methods). As illustrated in Fig. 1 and stated
reaching walls of test cell (Luthi 2011). Finally, the apparatus is in Table 2, in this methodology: (i) hd þ Vh2/2g is directly measured
drained from the bottom and dismantled. (by Pitot-tube) at the hole’s exit, alleviating the need to consider
hext; (ii) Hu ¼ hu þ Vu2/2g where Vu is defined from flow-rate conti-
nuity; and (iii) hent is considered in the overall head losses.
TABLE 3 Properties of tested core soil samples.
Recalling the computational analysis by Riha and Jandora
Atterberg Compaction (2014), a relatively accurate estimate of hent requires considering
Limitsb Specificationsc Dispersivity
the bevel-shaped entrance (to the hole). For this reason, the
Soil USCSa LL (%) PI (%) qdmax (kg/m3) wopt (%) Dispersion %d bevel-shaped entrance angle, h, is carefully measured after each
S1 CL 24.4 8.4 1870 14.3 23.1 test; that is, after disassembling the apparatus, the hole is
S2 CL-ML 22.4 6.4 1840 13.6 23.8 carefully photographed, then filled with molten wax, which after
S3 CL 38.7 19.3 1690 18.3 18.4 setting/solidification assists in diameter/shape measurement.
S4 CL 41.8 21.2 1650 19.0 26.1 With the aide of h, a rational estimation of Kent is made, which
S5 CL 42.0 19.4 1650 20.3 25.4 defines hent and subsequent calculations.
S6 CL 35.0 15.5 1700 19.0 18.9 With reference to Fig. 5, calculation of erodibility character-
S7 CL 26.6 10.7 1820 14.8 23.3
istics for soils, i.e., sc and Ce, requires the following steps:
S8 CL 25.5 9.3 1860 13.2 26.4
S9 CL-ML 20.8 4.0 1960 12.0 21.1 • In the plot of test head and Q, versus time, initiation of
S10 CL-ML 27.0 7.1 1870 13.4 57.8 accelerating flow is determined as where Q starts to
S11 CL-ML 26.5 5.7 1780 16.7 61.2 increase in a constant test head (cf. Fig. 5a).
S12 CL-ML 22.7 5.3 1890 13.2 15.9 • A polynomial curve is fitted to portion of /t  t variation
S13 CH 73.0 39.7 1430 26.1 70.0 with accelerating flow (see Fig. 5b). Whenever the initial
a test head is insufficient to cause immediate erosion, yet
Unified Soil Classification System.
b
ASTM D4318-10e1.
progressive erosion begins at further time without any
c
ASTM D698-12e2; head increase, it is most effective to fit a third order poly-
d
ASTM D4221-11. nomial function to /t  t. This leads to a parabolic e_t  st

&RS\ULJKWE\$670,QW O DOOULJKWVUHVHUYHG )UL-XQ*07

'RZQORDGHGSULQWHGE\
9LUJLQLD3RO\WHFK 6WDWH8QLY 9LUJLQLD3RO\WHFK 6WDWH8QLY SXUVXDQWWR/LFHQVH$JUHHPHQW1RIXUWKHUUHSURGXFWLRQVDXWKRUL]HG
 FATTAHI ET AL. ON HOLE EROSION TEST METHODS 501

FIG. 5
Calculation of erosion characteristics, i.e., sc and Ce, from
HET; (a) test head and flow rate, Q, versus elapsed time, (b)
variation of hole diameter, /t, with elapsed time, and (c)
relation between erosion rate, e_t , and exerted boundary
shear stress, s.

variation. When the initial test head causes immediate ratio, i.e., qd/qdmax, from 93 to 98 %, at water content of
progressive erosion, a second order polynomial is more wopt 6 2 %.
realistic to model /t  t, resulting in linear variation for During the test, it was observed that in specimens with low
e_t  st (Wahl et al. 2008a, 2008b). erosion resistance, the effluent is highly turbid and eroded par-
• As seen in Fig. 5c, a straight line approximates the raising ticles are small flocculates, whereas for high erosion resistance
portion of e_t  st variation, i.e. progressive erosion. From specimens, the effluent is comparatively clearer and eroded soil
this, Ce is the slope and sc is the horizontal intercept. is in form of large flakes/chunks.
The holes’ bevel-shaped entrance angles, h, reported in
Table 4, are measured from the erosion pattern profiles drawn
Results and Observations
in Fig. 6. Specimens S12 and S13 completely collapsed, leaving
The HET test specifications for 13 core soils are presented in no erosion pattern, almost immediately upon flow establish-
Table 4. Test specimens were reconstituted with compaction ment. Moreover, erosion characteristics were not quantifiable

&RS\ULJKWE\$670,QW O DOOULJKWVUHVHUYHG )UL-XQ*07

'RZQORDGHGSULQWHGE\
9LUJLQLD3RO\WHFK 6WDWH8QLY 9LUJLQLD3RO\WHFK 6WDWH8QLY SXUVXDQWWR/LFHQVH$JUHHPHQW1RIXUWKHUUHSURGXFWLRQVDXWKRUL]HG
502 Geotechnical Testing Journal

TABLE 4 Summary of HET tests conditions. for specimens S9 to S13. Such an outcome was caused by one of
Specimen w (%) qd/qdmax (%) Time (min) /h @ tf (mm) h ( ) a  b
Kent c the following conditions: (i) the hole collapsed almost immedi-
ately upon flow establishment. Apparently, the initial low flow
S1 14.5 95.7 11 17 65 0.3
rate was unable to remove collapsed soil debris; hence, the hole
S2 14.1 96.1 33 13 78 0.3
was blocked and test continuation was impossible. (ii) The
S3 18.5 97.2 120 10 69 0.3
S4 19.0 95.7 105 9.7 90 0.49 hole was not blocked, yet very rapid specimen erosion occurred,
S5 19.6 96.4 131 10 90 0.49 insomuch as upstream and downstream manometer levels
S6 19.6 95.3 120 10 90 0.49 quickly converged (i.e., equalized) in short time (about
S7 15.3 97.4 90 11 90 0.49 4–5 min); hence, head difference was undetectable, rendering
S8 13.3 95.1 20 12.7 90 0.49 calculation schemes inapplicable.
S9 12.7 95.3 9 19.6 53 — Soils S10, S11, and S13 are dispersive (cf. Table 3) and this
S10 15.2 95.1 23 24 67 — is the reason for their quick disintegration. Soils S9 and S12 are
S11 18.5 94.2 10 29 63 — low plasticity silts (CL-ML), which easily soften upon contact
S12 13.3 95.8 3 17 — —
with water.
S13 26.5 93.9 3 13 — —
For specimens S1, S2, and S3, h is between 60 and 80
a
Hole diameter at test termination. (cf. Table 4), moreover, according to Fig. 2 and the final hole
b
Defined in Fig. 6. diameters, /h, reported in Table 4, little variation in Kent is
c
Entrance head loss coefficient.
expected. Therefore, from the numerical results of Riha and
Jandora (2014), Kent ¼ 0.3 would be a fair estimate (see HET
Interpretation Methods). In case of soils S4 to S8, the hole
entrance in eroded specimens was not bevel-shaped implying

FIG. 6
Erosion patterns of specimens after HET tests.

&RS\ULJKWE\$670,QW O DOOULJKWVUHVHUYHG )UL-XQ*07

'RZQORDGHGSULQWHGE\
9LUJLQLD3RO\WHFK 6WDWH8QLY 9LUJLQLD3RO\WHFK 6WDWH8QLY SXUVXDQWWR/LFHQVH$JUHHPHQW1RIXUWKHUUHSURGXFWLRQVDXWKRUL]HG
 FATTAHI ET AL. ON HOLE EROSION TEST METHODS 503

TABLE 5 Summary of HET results from different interpretation approaches.

HGL EGL MEGL

5 5
Soil sc (N/m ) 2
Ce (10 ) (s/m) I Group sc (N/m ) 2
Ce (10 ) (s/m) I Group sc (N/m ) 2
Ce (105) (s/m) I Group
S1 10.33 501.19 2.30 2 9.11 446.68 2.35 2 9.02 125.89 2.90 2
S2 20.40 19.95 3.70 3 15.93 95.45 3.02 3 13.00 70.00 3.15 3
S3 110.00 1.58 4.80 4 83.30 1.86 4.73 4 70.00 2.00 4.70 4
S4 115.00 2.24 4.65 4 68.11 2.82 4.55 4 50.00 3.00 4.52 4
S5 125.00 1.26 4.90 4 69.94 2.00 4.70 4 49.00 2.34 4.60 4
S6 160.00 1.00 5.00 5 98.52 0.30 5.52 5 83.33 0.93 5.03 5
S7 34.00 19.95 3.70 3 24.25 17.78 3.75 3 21.00 30.00 3.52 3
S8 18.21 79.43 3.10 3 12.75 44.67 3.35 3 10.95 70.79 3.15 3

Note: Group number based on Table 1.

that h  90 . Considering /h//u for such specimens, Kent ¼ 0.49 As illustrated in Fig. 7b, sc,ME/sc,E ranges from 0.70 to 0.99
is discernible from Fig. 2. with the best trend of sc,ME/sc,E ¼ 0.8, and the divergence from
The HET results from different interpretation approaches sc,ME ¼ sc,E is more pronounced at higher sc. This observation is
are presented in Table 5. Accordingly, although differences are foreseeable, since EGL and MEGL methods are identical, except
observed between Ce and sc from different methods, the erosion for the latter considering hent.
rate indices, I, are close, in so far as all interpretation methods Overall, for soils with low erosion resistance I < 3, sc,ME,
classify the soil in an identical group of Table 1. sc,E, and sc,H are very close. With the increase in specimen
The variations of critical shear stresses, sc, obtained from erosion resistance, i.e., soils with I > 3 according to Table 1, dif-
MEGL (i.e., sc,ME), EGL (i.e., sc,E), and HGL (i.e. sc,H) methods ferences of sc,ME, sc,E, and sc,H increase.
are shown in Fig. 7. Generally, speaking differences between Considering soils for which HET is unfeasible (e.g., S9 to
sc,ME and sc,H increases with increase of sc. Moreover, sc,ME/sc,H S13), Wahl et al. (2008a) state that the HET “in its current con-
ranges from 0.39 to 0.87 with the best trend of sc,ME/sc,H ¼ 0.5 figuration cannot provide a quantitative measure of the erod-
(see Fig. 7a). Interestingly, numerical computations by Riha and ibility of many materials in groups 1–2 and 5–6,” and the lowest
Jandora (2014) for critical shear stresses in the MHGL method, I reported by those authors was 2.45. In Table 5, the lowest I,
i.e., sc,MH, had suggested that sc,MH/sc,H ¼ 0.52–0.6, where calculated from the HGL method, is 2.3 observed for soil S1.
“This holds for relatively small discharges and velocities in the A guide to estimating erosion properties of soils for which
hole, and for higher discharges the difference can be even HET is unfeasible, is using precursor experiences, similar to
greater” (Riha and Jandora 2014). The results of this study those provided by Wan and Fell (2002). Accordingly, for non-
comply well with the results of Riha and Jandora (2014), since dispersive SM, SC, ML, and CL-ML soils the erosion rate index,
inherently the only difference between the MHGL and MEGL I, is about 2 to 3. It may, therefore, be concluded that when
methods, is consideration of the small Vu2/2g in the latter HET is unfeasible for highly erodible soils, the erosion rate
approach. index, I, would most probably be less than 2.

FIG. 7
Graph of critical shear stress variations; (a) MEGL
(sc,ME) versus HGL (sc,H), and (b) MEGL (sc,ME) with
respect to EGL (sc,E).

&RS\ULJKWE\$670,QW O DOOULJKWVUHVHUYHG )UL-XQ*07

'RZQORDGHGSULQWHGE\
9LUJLQLD3RO\WHFK 6WDWH8QLY 9LUJLQLD3RO\WHFK 6WDWH8QLY SXUVXDQWWR/LFHQVH$JUHHPHQW1RIXUWKHUUHSURGXFWLRQVDXWKRUL]HG
504 Geotechnical Testing Journal

Summary Remarks and Conclusions Benahmed, N. and Bonelli, S., 2012, “Investigating Concen-
trated Leak Erosion Behaviour of Cohesive Soils by Perform-
The HET test has been introduced a decade ago; nevertheless, a ing Hole Erosion Tests,” Eur. J. Environ. Civ. Eng., Vol. 16,
unified and standardized method for conducting and interpret- No. 1, pp. 43–58, http://dx.doi.org/10.1080/19648189.2012.
667667
ing HET is not yet available. This is disquieting noticing that
Benahmed, N. and Bonelli, S., 2012, “Internal Erosion of
the HET is now widely accepted as a routine test for measuring Cohesive Soils: Laboratory Parametric Study,” presented
soil erodibility characteristics, both for research and real engi- at the 6th International Conference on Scour and Erosion,
neering projects/services. Paris, France, August 27–31, ICSE, Paris, France, pp.
This paper presented refinements to calculating/measuring 1041–1047.
the water head drop, hf, in the HET. A new approach was pro- Bonelli, S., Ed., 2013, Erosion in Geomechanics Applied to Dams
and Levees, John Wiley & Sons, New York.
posed and compared with three previous HET analysis meth-
Bonelli, S., Brivois, O., Borghi, R., and Benahmed, N., 2006, “On
ods, through analyzing erosion characteristics of 13 fine-grained the Modelling of Piping Erosion,” Comptes Rendus Mécani-
soils. que, Vol. 334, Nos. 8–9, pp. 555–559, http://dx.doi.org/
Test results and observations suggest that soils with high 10.1016/j.crme.2006.07.003
erosion resistance reflect higher difference in critical shear Bonelli, S. and Brivois, O., 2008, “The Scaling Law in the Hole
Erosion Test With a Constant Pressure Drop,” Int. J. Numer.
stresses, sc, obtained by hydraulic grade line method (HGL),
Anal. Methods Geomech., Vol. 32, No. 13, pp. 1573–1595,
energy grade line method (EGL), and modified energy gradient http://dx.doi.org/10.1002/nag.683
line method (MEGL). Moreover, sc,ME/sc,H ranges from 0.39 Boukhemacha, M., 2009, “A Hole Erosion Test Model A Step
to 0.87 with the best trend of sc,ME/sc,H ¼ 0.5 and sc,ME/sc,E on Internal Erosion Modeling,” Sci. Bull. Ser.: Math. Modell.
ranges from 0.70 to 0.99 with the best trend of sc,ME/sc,E ¼ 0.8, Civ. Eng., Vol. 5, No. 3, pp. 17–24, http://dx.doi.org/10.3311/
and the divergence from sc,ME ¼ sc,E is more pronounced at PPee.2145
Boukhemacha, M. A., Bica, I., and Mezouar, K., 2013, “New
higher sc.
Procedures to Estimate Soil Erodibility Properties From a
Values of the erosion rate indices, I, obtained from HGL, Hole Erosion Test Record,” Period. Polytech. Civ. Eng.,
EGL, and MEGL are close, and the associated group number is Vol. 57, No. 1, pp. 77–82.
identical. Therefore, the erosion rate index reliably manifests Chevalier, C., Haghighi, I., Pham, T. L., Reiffsteck, P., Burns, S.
soil erosion propensity. E., Bhatia, S. K., Avila, C. M. C., and Hunt, B. E., 2011, “Two
Complementary Tests for Characterizing the Soil Erosion,”
presented at Scour and Erosion. Proceedings of the Fifth
ACKNOWLEDGMENTS International Conference on Scour and Erosion (ICSE-5),
San Francisco, CA, November 7–10, ASCE, Reston, VA,
The constructive comments, both technical and editorial, by the
pp. 152–161.
anonymous reviewers are gratefully acknowledged. The writers Chevalier, C., Haghighi, I., and Herrier, G., 2012, “Resistance to
would like to express their gratitude to Professor A. R. Zarrati Erosion of Lime Treated Soils: A Complete Parametric Study
for his comments on pipe-flow hydraulics. Also, the kind efforts in Laboratory,” presented at the 6th International Conference
of Mr. Reza Javadi for providing help and assistance during test- on Scour and Erosion, Paris, France, August 27–31, ICSE,
Paris, France, pp. 161–168.
ing is appreciated.
Farrar, J. A., Torres, R. L., and Erdogan, Z., 2007, “Bureau of
Reclamation Erosion Testing for Evaluation of Piping
References and Internal Erosion of Dams,” presented at the
Geo-Denver2007: New Peaks in Geotechnics, Denver, CO,
ASTM D698-12e2, 2012, Standard Test Method for Laboratory February 18–21, ASCE, Reston, VA, pp. 22–31.
Compaction Characteristics of Soil Using Standard Effort, Foster, M., Fell, R., and Spannagle, M., 2000, “A Method for
ASTM International, West Conshohocken, PA, Assessing the Relative Likelihood of Failure of Embankment
www.astm.org Dams by Piping,” Can. Geotech. J., Vol. 37, No. 5,
ASTM D2434-68(2006), 2006, Standard Test Method for Perme- pp. 1025–1061, http://dx.doi.org/10.1139/t00-029
ability of Granular Soils (Constant Head), ASTM Interna- Haghighi, I., Chevalier, C., Duc, M., Guédon, S., and Reiffsteck,
tional, West Conshohocken, PA, www.astm.org P., 2012, “Improvement of Hole Erosion Test and Results on
ASTM D4221-11, 2011, Standard Test Method for Dispersive Reference Soils,” J. Geotech. Geoenviron. Eng., Vol. 139,
Characteristics of Clay Soil by Double Hydrometer, ASTM No. 2, pp. 330–339, http://dx.doi.org/10.1061/(ASCE)GT.
International, West Conshohocken, PA, www.astm.org 1943-5606.0000747
ASTM D4318-10e1, 2010, Standard Test Methods for Liquid ICOLD, 2013, “Internal Erosion of Existing Dams, Levees and
Limit, Plastic Limit, and Plasticity Index of Soils, ASTM Dykes, and Their Foundations,” Bulletin 164, Volume 1:
International, West Conshohocken, PA, www.astm.org Internal Erosion Processes and Engineering Assessment, Inter-
ASTM D4647/D4647M-13, 2013, Standard Test Method for national Commission on Large Dams, Paris, France.
Identification and Classification of Dispersive Clay Soils by ICOLD, 2016, “Internal Erosion in Existing Dams, Dikes and
the Pinhole Test, ASTM International, West Conshohocken, Levees and Their Foundations,” Bulletin 164 Volumes 2:
PA, www.astm.org Case Histories, Investigations, Testing, Remediation and Sur-

&RS\ULJKWE\$670,QW O DOOULJKWVUHVHUYHG )UL-XQ*07

'RZQORDGHGSULQWHGE\
9LUJLQLD3RO\WHFK 6WDWH8QLY 9LUJLQLD3RO\WHFK 6WDWH8QLY SXUVXDQWWR/LFHQVH$JUHHPHQW1RIXUWKHUUHSURGXFWLRQVDXWKRUL]HG
 FATTAHI ET AL. ON HOLE EROSION TEST METHODS 505

veillance, International Commission on Large Dams, Paris, Reiffsteck, P., Pham, T. L., Vargas, R., and Paihua, S., 2006,
France. “Comparative Study of Superficial and Internal Erosion
Kissi, B., Vera, M. A. P., Cintas, M. R., and Khamlichi, A., 2012, Test,” Proceedings 3rd International Conference on Scour
“Modeling the Fluid/Soil Interface Erosion in the Hole Ero- and Erosion (ICSE-3), Verheij, H.J. and Hoffmans G.J.C.M.
sion Test,” presented at the MATEC Web of Conferences, (Eds.), November 1–3, 2006, CURNET, Gouda, Amsterdam,
Volume 1, CSNDD 2012 International Conference on Struc- The Netherlands, pp. 571–575.
tural Nonlinear Dynamics and Diagnosis, M. Belhaq and R. Regazzoni, P. L. and Marot, D., 2013, “A Comparative Analysis
Ibrahim (Eds.), Marrakech, Morocco, April 30–May 2, 2012, of Interface Erosion Tests,” Nat. Hazards, Vol. 67, No. 2,
Article No.: 00003, Published by EDP Sciences- France, 17 pp. 937–950, http://dx.doi.org/10.1007/s11069-013-0620-3
Avenue du Hoggar, Parc d’Activité de Courtabuf, BP 112, Rı́ha, J. and Jandora, J., 2014, “Pressure Conditions in the Hole
91944 Les Ulis Cedex A, France, http://dx.doi.org/10.1051/ Erosion Test,” Can. Geotech. J., Vol. 52, No. 1, pp. 114–119,
matecconf/20120100003 http://dx.doi.org/10.1139/cgj-2013-0474
Ladd, R. S., 1978, “Preparing Test Specimens Using Under- Streeter, V. L., Wylie, E. B., and Bedford, K. W., 1998, Fluid
compaction,” Geotech. Test. J., Vol. 1, No. 1, pp. 16–28. Mechanics, McGraw-Hill, New York.
Lim, S. S., 2006, “Experimental Investigation of Erosion in Vari- Wahl, T. L., Regazzoni, P. L., and Erdogan, Z., 2008a, Determin-
ably Saturated Clay Soils,” Ph.D. thesis, The University of ing Erosion Indices of Cohesive Soils With the Hole Erosion
New South Wales, NSW, Australia. Test and Jet Erosion Test, US Department of Interior,
Luthi, M., 2011, “A Modified Hole Erosion Test (HET-P) to Bureau of Reclamation, Technical Service Center, Washing-
Study Erosion Characteristics of Soil,” Ph.D. thesis, The Uni- ton, D.C.
versity of British Columbia, Vancouver, BC. Wahl, T. L., Erdogan, Z., and Kepler, W. F., 2008b, Results of
Luthi, M., Fannin, R. J., and Millar, R. G., 2012, “A Modified Laboratory Physical Properties and Hole Erosion Tests,
Hole Erosion Test (HET-P) Device,” Geotech. Test. J., Truckee Canal Embankment Breach, Newlands Project,
Vol. 35, No. 4, pp. 660–664, https://doi.org/10.1520/ Nevada, US Department of Interior, Bureau of Reclamation,
GTJ104336 Technical Service Center, Washington, D.C.
Marot, D., Regazzoni, P. L., and Wahl, T., 2011, “Energy-Based Wan, C. F. and Fell, R., 2002, “Investigation of Internal Erosion
Method for Providing Soil Surface Erodibility Rankings,” and Piping of Soils in Embankment Dams by the Soil Slot
J. Geotech. Geoenviron. Eng., Vol. 137, No. 12, pp. 1290–1293, Erosion Test and the Hole Erosion Test,” Report No. R-412,
http://dx.doi.org/10.1061/(ASCE)GT.1943-5606.0000538 University of New South Wales, School of Civil and Envi-
Mercier, F., Bonelli, S., Golay, F., Anselmet, F., Philippe, P., and ronmental Engineering, NSW, Australia.
Borghi, R., 2015, “Numerical Modelling of Concentrated Wan, C. F. and Fell, R., 2004a, “Investigation of Rate of Erosion
Leak Erosion During Hole Erosion Tests,” Acta Geotech., of Soils in Embankment Dams,” J. Geotech. Geoenviron.
Vol. 10, No. 3, pp. 319–332, http://dx.doi.org/10.1007/ Eng., Vol. 130, No. 4, pp. 373–380, http://dx.doi.org/
s11440-014-0349-5 10.1061/(ASCE)1090-0241(2004)130:4(373)
Okiishi, M. Y., Munson, B., and Young, D., 2006, Fundamentals Wan, C. F. and Fell, R., 2004b, “Laboratory Tests on the Rate of
of Fluid Mechanics, John Wiley & Sons, Inc., New York. Piping Erosion of Soils in Embankment Dams,” Geotech.
Potter, M., Wiggert, D., and Ramadan, B., 2011, Mechanics of Test. J., Vol. 27, No. 3, pp. 295–303, https://doi.org/10.1520/
Fluids SI Version, Cengage Learning, Boston, MA. GTJ11903

&RS\ULJKWE\$670,QW O DOOULJKWVUHVHUYHG )UL-XQ*07

'RZQORDGHGSULQWHGE\
9LUJLQLD3RO\WHFK 6WDWH8QLY 9LUJLQLD3RO\WHFK 6WDWH8QLY SXUVXDQWWR/LFHQVH$JUHHPHQW1RIXUWKHUUHSURGXFWLRQVDXWKRUL]HG