Comparison between various flume tests used for hydraulic-fill studies
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Comparison between various flume tests used for hydraulic-fill studies
ANGELAA. G. KUPPER,~ NORBERT R. MORGENSTERN, AND DAVIDC. SEGO
Department of Civil Engineering, University of Alberta, Edmonton, Alta., Canada T6C 2C7
Received April 30, 1991
Accepted February 22, 1992
Several flume deposition tests carried out in different parts of the world to study hydraulic fills are compared and
discussed. The results of all test programs are coherent and consistent with field observations of hydraulic fills and
natural alluvial deposits, which suggests that, at least qualitatively, flume tests are adequate t o simulate the physical
phenomena associated with hydraulic deposition.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Depository Services Program on 06/03/13
Key words: hydraulic fill, flume tests, sand, grain size, profile, slope, density.
L'on compare et discute plusieurs essais de dkposition en canal rkalises dans differentes parties du monde. Les risultats
de tous les programmes d'essai sont cohkrents et consistants avec les observations sur le terrain de remblais hydrauliques
et de dkp6ts alluvionnaires naturels, ce qui indique que les essais en canal simulent adkquatement, du moins qualitative-
ment, les phknomttnes physiques associks a la dkposition hydraulique.
Mots clPs : remblai hydraulique, essais en canal, sable, granulometrie, profil, pente, densitk.
[Traduit par la redaction]
Can. Geotech. J. 29, 418-425 (1992)
Introduction difference in testing procedure between these tests and
Flume tests are a convenient tool to study hydraulic fills. hydraulics-sedimentology aggrading-bed tests is that in the
Under laboratory conditions it is possible to control and first case water and solids are previously mixed and fed to
isolate variables in a simpler and more economical manner the flume as a slurry, whereas in the second case excess sand
For personal use only.
than would be possible in the field. Flume tests also permit is dropped onto an already established flow (that is usually
one to study the hydraulic deposition process of a certain much deeper than the flow in slurry deposition tests).
material at an early stage of the project when information In this paper, flume tests performed specifically to study
is necessary but field data are still not available. hydraulic fill are described and their results are summarized
Owing to these advantages, several flume tests programs and compared.
have been carried out to study hydraulic deposition. In fact,
there is a large number of flume tests presented in the liter- Summary of flume tests presented in the literature
ature involving the flow of water and sediment. However,
the majority of these tests study hydraulic deposition from A brief description of several flume tests to study hydraulic
a hydraulics or sedimentology point of view. In these cases, fills is presented in this section. A summary of the most
a bed of the selected sediment is usually placed on the flume important features and parameters utilized in each test pro-
bottom and water flows over it at specified flow rates and gram is presented in Table 1.
velocities. The water flow may cause the sediment to be Porto Primavera Dam, Brazil (UPP)
transported, defining particular bedforms and stratification. The flume tests reported by Ferreira et al. (1980) were con-
The sediment that is carried out at the downstream end of ducted according t o Soviet technology as part of the pre-
the flume is usually collected and replaced at the upstream liminary studies for the design and construction of Porto
end, so the bed may change shape but it is not aggrading Primavera Dam in Brazil, which was planned initially to be
or degrading. In only a few of these flume tests (Allen 1963; a hydraulic fill. The Soviets report to have used flume tests
Bhamidipathy and Shen 1971; Soni et al. 1977; Soni 1981; extensively in the past when the hydraulic fill technology
Garde et al. 1981; Torres and Jain 1984; Yen et al. 1989) was being developed in their country. Nowadays the hydro-
is there net deposition, with additional sediment being placed mechanization process is standardized in the former U.S.S.R.,
at the flume head. Typical measurements during these tests and laboratory deposition tests are no longer used routinely.
include water flow rate, velocities, depth, sediment transport For the Porto Primavera Dam studies, fine sand from one
rate, and slope. Bedform and stratigraphy are sometimes of the borrow areas was deposited in an 11-m-long flume
also recorded. using an independent sand and water feeding system
Relatively few flume tests found in the literature deal with (Table 1). Sand was added at a controlled flow rate to a
hydraulic fill from a geotechnical perspective. The objectives water flow just before being fed to the flume. Turbulence
of these tests are, in general, the determination of the beach ensured the formation of a uniform slurry. The flume was
slopes and (or) the study of physico-mechanical character- provided with piezometers installed along its bottom and
istics of the deposited material. Flume tests for hydraulic- sides. The bottom of the flume had a drainage system with
fill studies obviously have to involve an excess sediment valves that allowed drainage of the fill at a rate of 4 cm/h
being deposited and therefore an aggrading bed. The main after the deposition stopped. The slurry was discharged from
a vertical pipe with a series of outlets installed at different
'present address: HBT AGRA Ltd., 4810-93 Street, Edmonton, heights. Each outlet was a short piece of pipe pointing
Alta., Canada T6E 5M4. upwards (possibly to minimize the flow velocity) and
Printed in Canada / lrnprirnc au CanadaET AL.: I1 419
plugged. As the fill rose, the outlets were successively disturbed the flow in the flume. After the deposition was
o~ened. completed, the deposit was partially drained and Shelby tube
Three tests were carried out, with slurry concentration samples were taken at designated distances along the length
around 10% by weight and specific flow rates (flow rate of the deposited tailings. The samples were used to deter-
divided by flume width) varying from 3.3 to 13 cm3/(s ecm). mine shear strength, permeability, and grain-size distribution.
The formation of meanders is reported in some cases. Mea- The average beach slope was also measured for each test.
surements during the tests included concentration of the
slurry being discharged and concentration and composition Delft, The Netherlands (DS, DLI, and DL2)
of the outflow. After the tests, the final sand profile was Flume tests were carried out in Delft, The Netherlands,
recorded and undisturbed samples were taken for determina- as part of experimental and theoretical studies related to the
Deltaworks in the southwestern part of the country, where
tion of density and grain-size distribution of the sand along
various sea arms were closed using hydraulic-fill techniques
the flume at two depths. Horizontal and vertical undisturbed
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Depository Services Program on 06/03/13
samples were used to determine permeability, compressibility, (de Groot et a/. 1988; Winterwerp et al. 1990).
and shear strength on directions parallel and perpendicular Two flumes of different sizes were used to deposit three
to the stratification. The sand deposited in the flume had gradations of fine to medium sand (Table 1). Slurry concen-
a relative density in the range of 50-65% and some trations between 32 and 68% and flow rates varying from
anisotropy of the permeability, with Kh/Kv varying 7 to 35 L/min were utilized for the tests on the small flume
between 1 and 10. Remolded samples showed lower perme- (DS). The large-scale flume tests (DL1 and DL2) covered
ability and less compressibility than the undisturbed flume a wide range of slurry concentrations (0-64%) and used very
samples. Shear strength was similar for samples from all large flow rates (180-2700 L/min), which is out of the range
of the values used in other flume tests presented in the
three tests.
literature.
Lakefield (LK) The equilibrium slope was defined in these studies as the
Lakefield Research performed flume tests in 1983 to slope at which sedimentation and erosion are in balance.
obtain slope data for the design of East Kemptville tailings It was determined by decreasing the slope of the tilting flume
dams in Nova Scotia, Canada (P.C. Lighthall, personal in small steps of about 0.001 rad until sand bars started
communication, 1987). A series of tests was conducted with growing on the bottom of the flume. The previous slope was
varying flume slopes and slurry concentration. However, the then defined as the equilibrium slope. Therefore these tests
For personal use only.
only data available are the grain-size distribution curve of differ from the others presented in the literature by the fact
the coarse tailings used for the tests and the profiles of two that no material was deposited in the flume. It is of relevance
tests using slurry concentrations of 20 and 45%. The flow to note that the flumes had sand grains glued to the bottom;
rate is not known, and there is no information about the however, for the large-scale tests (DL1 and DL2) the sand
test procedure, except that there was no water ponding at glued on the bottom was much coarser than the sand being
the slope toe. The slopes obtained were reported to be much tested.
steeper than the field slopes of the same material. Sand concentration, flow rate, slope, flow depth, and
slurry temperature were measured in the small-scale flume
South Africa (B) tests. In the larger scale tests the following parameters were
The flume tests conducted by Blight et al. (1985) in South determined: flow rate, temperature, concentration, slope,
Africa had the objective of studying tailings beach profiles flow depth, flow velocity, and sand concentration profiles
based on the concept of master profile proposed by Melent'ev at various locations. These tests did not deal with the geo-
et al. (1973) in which a normalized profile exists independent technical characteristics of hydraulically deposited materials,
of the test scale. Three different gradations of silty tailings but they involved sophisticated hydraulic measurements.
were deposited at three different slurry concentrations each More details on these tests can also be found in Mastbergen
(50, 60, and 70% by weight) in a small flume (Table 1). The et al. (1988) and Bezuijen and Mastbergen (1988).
feeding system consisted of a 220-L drum with a bottom
discharge, but no details are provided on how the slurry was University of Queenslund, Australia (FB, FN, FFC, and
kept homogeneous in the drum during the test, nor on the
FCC)
The flume tests reported by Fourie (1988) were carried
testing procedure or flow rates. The only data reported are
the final normalized profile for each test. out using a slurry tank adapted with an electrical agitator
to feed a small flume. The discharge device consisted of a
United States Bureau of Mines (USA and USB) horizontal pipe to spread the flow across the flume and an
Flume tests were also performed at the United States energy dissipator to minimize the formation of a plunge pool
Bureau of Mines (USBM) Research Centre in Spokane, at the discharge point. A very small flow rate was adopted
Washington (Boldt 1988). Tailings obtained from two mine for all tests (Table 1). Three types of tailings were tested:
sites were deposited in a wooden sloping flume (Table 1). bauxite from North Queensland (FB), nickel ore slurry from
Tailings A consisted of a fine mill waste from a copper- New Caledonia (FN), and two gradations of coal tailings
silver mine with an average diameter of 0.0135 mm. Tail- from southeast Queensland, a fine (FFC) and a coarse
ings B was a slightly coarser tailings (& = 0.097 mm) (FCC). These tests will not be compared with the others as
from a silver-lead-zinc mine. The bulk tailings were diluted most of them deposited a nonsegregating slurry, thus having
with water in a 6400-L mixing tank to form the slurry. The a distinct rheology. In nonsegregating slurries the solids and
slurry was then pumped into the flume at controlled flow the carrier fluid do not behave independently but act as a
rates and concentrations, varying from 58 to 130 L/min and viscous fluid. Therefore the physical phenomena involved
from 20 to 57%, respectively. Boldt (1988) reports accen- are distinct from the cases of segregating slurries, which were
tuated wall effects with the formation of side eddies that used in the other experimental programs.420 CAN. GEOTECH. J. VOL 29, 1992
TABLE1. summary of flume tests for hydraulic
Test No. Flume length, Flume width, Flume depth, Feeding
Ref. code of tests l(m) w (m) I/ w d(m) system
Ferreira et al. (1980) UPP 3 11.O 0.8 14 >0.8 Independent
P.C. Lighthall, personal communication (1987) LK >2 >1.5 NA" NA >0.15 NA
Blight et al. (1985) B >9 1.8 0.3 6 0.6 220-L drum
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Depository Services Program on 06/03/13
Boldt (1988) USA 12 12.2 0.6 20 0.6 Mixing tank
USB
de Groot et a/. (1988) DS >50 1.5 0.118 13 0.045 Mixing tank
Winterwerp et al. (1990) DL1 >20 9.0 0.3 30 0.3 Recirculation
DL2
Fourie (1988) FB 2 2 0.6 3 0.6 Mixing tank
FN 2
FFC 4
FCC 1
Fan (1989) F 8 4.9 0.3 16 0.5 Independent
For personal use only.
Kiipper (199 1) SS 5 6.1 0.3 20 1.2 Independent
TS 37
KS 19
NOTE: The b in DSPbstands for bottom.
ONA, data not available.
University of Alberta, Department of Chemical Engineering weight and the flow rate from 50 to 350 cm3/s (3-20 L/min)
(F) (see Table 1). Each test was performed on a bed of sand
The objectives of the flume tests carried out by Fan (1989) to allow a certain degree of underdrainage and was carried
were to study the variation of beach profile with time and out for long enough to develop an equilibrium slope. The
distance and the effects of slurry concentration and discharge data obtained during each test consisted of the input param-
on these profiles. eters, beach profiles at various time intervals during the test,
These tests were performed using an experimental proce- and flow description. Remolded and undisturbed samples
dure common to flume tests for hydraulic or sedimentology were taken at various locations along the flume for grain-
studies, where sand is fed independently to a flume where size distribution analysis, density measurement (Kupper
a flow of water is already established. However, a very 1991; Kupper et al. 1992), fabric analysis (Kupper 1991; Law
shallow flow depth was adopted, as occurs in hydraulic fills. 1991), and triaxial tests (Law 1991).
These tests covered a relatively narrow range of low con-
centrations and low flow rates. The data produced consisted Comparison of results
of sand profiles at four different times during the tests, The various flume tests described in the previous section
which had a duration of 20 min each. The flume dimensions had the general objective of studying hydraulic fill, however,
and the sand and flow characteristics utilized are presented the specific objectives were slightly different in each case.
in Table 1. As a result, the variables that were studied and the param-
eters that were measured were not necessarily the same for
University of Alberta, Department of Civil Engineering all test programs. Moreover, there were differences in exper-
(SS, TS, and KS) imental procedure, some of which may have influenced the
Three different sands were tested by Kiipper (1991) to results obtained. In particular, the experiments carried out
study the influence of flow rate, slurry concentration, and by the Delft group (de Groot et al. 1988; Winterwerp et al.
grain-size distribution on the properties of hydraulic fills. 1990) and by Fan (1989) must be singled out for using very
These tests used a feeding system where water and sand were distinct techniques, as described here.
fed independently in a chute and formed the slurry on the (1) An equilibrium slope, defined as the slope at which
way to the flume. The discharge device consisted of a flow no deposition or erosion occurs, was adopted for the tests
spreader capable of distributing the flow uniformly across performed in Delft, since this slope is considered to be
the flume. An automatic system was designed to keep the similar to the field slopes (Winterwerp et al. 1990). Conse-
flow spreader at a constant height from the rising fill. The quently, no material is deposited in the flume and no data
slurry concentration was varied between 1.5 and 40.5% by are available on fill properties. Also, the slurry flows on theCan. Geotech. J. Downloaded from www.nrcresearchpress.com by Depository Services Program on 06/03/13
For personal use only.CAN. GEOTECH. J. VOL. 29, 1992
+ USB
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Depository Services Program on 06/03/13
0.01 0.1
Grain size (mrn)
FIG. 1. Grain-size distribution curves of materials used in various flume-test programs (see Table 1).
For personal use only.
0.01 0.1
Grain size (mm)
FIG. 2. Grain-size distribution curves of materials used in various flume tests compared with Soviet criteria. Legend as in Fig. 1.
When comparing tests carried out using different mate- divided by the distance from the point of maximum height
rials, it is important to note that grain shape, angularity, to the end of the beach and expressed as a percentage. Slurry
and surface features may affect the results and may be concentration is defined as the weight of solids in the slurry
responsible for some apparent scatter of the data. divided by the total weight.
The effect of the mean grain size and the slurry concentra-
Geometry tion on the average slope of the fill is shown in Fig. 6. This
All flume tests that utilized segregating slurries, and for graph compares results of tests described in Kiipper (1991)
which profiles were reported, formed fills with a similar con- and Kiipper et al. (1992) on TS tailings sand with the results
cave shape as shown in Figs. 3-5. The influence of the input from Fan (1989) on a commercial sand. All of these tests
parameters (such as particle size, slurry concentration, and were carried out using the same flow rate (Q = 250 cm3/s)
flow rate) on the fill geometry is evident in these figures. and the same specific flow rate ( q = 8 cm3/(s-cm)) and
In many cases only overall slopes were presented rather show that the average slope increases with the mean grain
than actual profiles. The overall slope is (or is assumed to size and with slurry concentration.
be) defined as the maximum height (at the discharge point) Figure 7 presents the variation of average slope with con-KUPPER ET AL.: I1
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Depository Services Program on 06/03/13
0.5
0 2 4 6 8 10
Distance from discharge point (m)
Fro. 3. Profiles of TS, UPP, and F tests for C , = 10% and various specific flow rates.
1.1
--
E
1 .o
........
aa.nn..a KS13 - qz4.7 cm3/(s.cm)(DSo=0.466mm)
+ 0.9
M
.-
2 - ..
For personal use only.
0.8
2% 0.7 1
......*-..-
-.-..-....... '.."
0.6 r .... ........
-'8
'l.*l,l., 0
0 5 10 15 20
Slurry Concenwation C, (%)
0.5 " " " ' " ' " " " " ' " " '
0 1 2 3 4 5
Distance from the discharge point (m)
FIG.6. Effect of grain size and concentration on the slope.
FIG.4. Profiles of TS, KS, and LK tests for C, = 20%.
"
0 20 40 60 80
Slurry concentration C, (%)
0.7 FIG.7 . Variation of slope with C , for various flume tests (see
0 1 2 3 4 5
Distance from discharge point (m) Table 1).
FIG. 5. Comparison of profiles of TS (D,, = 0.178 mm) and
F (D,, = 0.267 mm) tests.
use of materials containing a high percentage of fines (90
and 45% finer than 74 pm, respectively). Limiting this anal-
centration of the slurry being deposited in various flume ysis to sandy fills deposited at flow rates under 2500 cm3/s
tests. The relatively large scatter in this graph is partially and specific flow rates less than 40 cm3/(s.cm) to better
due to the inclusion of points corresponding to a wide range isolate the effect of the slurry concentration, a much smaller
of grain sizes and flow rates. The flat slopes of the large- scatter is observed (Fig. 8), and the increase in average slope
scale flume tests performed in Delft (DL1 and DL2) can be with concentration becomes more evident. The shaded zone
explained as being caused by the much higher flow rates in Fig. 8 includes more than 90% of the experimental points.
utilized in this case compared with the remainder of the An increase in flow rate or specific flow rate causes the
flume tests. The flat slopes obtained in the tests carried out slopes to become flatter, and this tendency is shown in
by Boldt (1988) (USA and USB) can be attributed to the Figs. 9 and 10. The small-scale tests carried out in Delft (DS)CAN. GEOTECH. J. VOL. 29, 1992
0 0
0 20 40 60 80 0 50 100 150 200
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Depository Services Program on 06/03/13
Sluny concentcation C , (%) Total flow rate Qt (Urnin)
FIG. 8. Variation of slope with C, for sandy fills with Q < FIG. 9. Variation of slope with flow rate for various flume
2500 cm3/s. tests.
For personal use only.
0
0 10 20 30 40 50
Specific flow rate q (cm3/(s. cm))
FIG. 10. Variation of slope with specific flow rate for various flume tests.
yielded more scattered results featuring relatively steep slopes adopted for UPP tests in relation to the values used for SS,
for the correspondent values of flow rate and, especially, TS, and KS tests.
specific flow rate. The results of the companion large-scale Results from SS, TS, KS, and UPP tests show the same
flume tests are not presented, as the much larger flow rates trend of variation of density with flow rate and slurry con-
utilized in these tests would have obscured the other results. centration: density decreases as slurry concentration increases,
Density and density tends to increase as specific or total flow rates
The only values of density of flume-deposited materials increase. This is consistent with the trends reported by Yufin
which were found in the literature were the results presented (1965).
by Ferreira et al. (1980) and Kiipper (1991) (see Table 1).
Figures 11 and 12 compare the results measured for UPP Summary and conclusions
(Ferreira et al. 1980) with those obtained for the SS, TS, Several flume tests carried out in different parts of the
and KS series (Kiipper 1991). The scatter of the data is par- world to study hydraulic fills were compared in this paper.
tially because the plot of density versus slurry concentration These tests were performed using slightly different testing
includes tests using a range of flow rates, and similarly the procedures and covered a wide range of values of slurry con-
points in Fig. 12 correspond to different slurry concentra- centration and flow rate. Most of the tests deposited sand,
tions. Also, other factors associated with the data scatter but some of them involved very fine tailings materials.
include difficulties in obtaining accurate density measure- The results of all test programs show a consistent trend
ment of undisturbed samples of relatively clean sands, varia- of fill slopes becoming steeper as the flow rate decreases and
tions in grain-size distribution curves, and effects of grain as the slurry concentration and the mean grain size increase.
shape, angularity, and surface features. Although more limited, the density data point to an increase
Compared with SS, TS, and KS values, the densities of in fill density as the flow rate increases and a decrease in
LTPP tests seem a little high for a material with a lower D50. density for higher values of slurry concentration.
This fact could have been caused by several factors including Generally, these conclusions are consistent with observa-
(i) a difference in mineralogy and (or) grain shape among tions of hydraulic fills and natural alluvial deposits which
the materials; (ii) the higher coefficient of uniformity (Cu) suggest that, at least qualitatively, flume tests are adequate
of UPP material compared with TS and KS sands (voids to simulate the physical phenomena associated with hydraulic
can be better filled); and (iii) the lower discharge velocities deposition in the field.KUPPER ET AL.: 11 425
1.7 Boldt, C.M.K. 1988. Beach chara'cteristics of mine waste tailings.
U.S. Bureau of Mines Report of Investigations RI9171.
W
-
de Groot, M.B., Heezen, F.T., Mastbergen, D.R., and Stefess, H.
E 1.6 1988. Slopes and density of hydraulically placed sands. Hydraulic
2.-
2.
Fill Structures, ASCE Geotechnical Special Publication 21.
Edited by D.J.A. Van Zyl and S.G. Vick. American Society of
g 1.5 Civil Engineers, New York.
E; Fan, X. 1989. Laboratory modeling of beach profiles in tailings
b disposal. M.Sc. thesis, Department of Chemical Engineering,
1.4
University of Alberta, Edmonton, Alta.
Ferreira, R.C., Peres, J.E.E., and Monteiro, L.B. 1980. Geotechni-
cal characteristics of hydraulic fill scale models. 13th Brazilian
1.3
o National Seminar on Large Dams, Rio de Janeiro, April.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Depository Services Program on 06/03/13
10 20 30 40 50
Sluny concentration (%) pp. 496-516. (In Portuguese.)
Fourie, A. 1988. Beaching and permeability properties of tailings.
FIG. 11. Variation of density of the fill with C, for various Hydraulic Fill structures, ASCE ceotechnical special publics-
flume tests. tion 21. Edited by D.J.A. Van Zyl and S.G. Vick. American
Society of Civil Engineers, New York, pp. 142-154.
1.7 Garde, R.J., Ranga Raju, K.G., and Mehta, P.J. 1981. Bed level
variation in aggrading alluvial streams. 19th Congress of the
- . International Association for Hydraulic Research, New Dehli,
rn
E 1.6
& = 0
February 2-7, vol. 2, pp. 247-253.
i "5- g o Kupper, A.A.G. 1991. Design of hydraulic fill. Ph.D. thesis,
.? 1.5
o . 8 e
Department of Civil Engineering, University of Alberta,
Edmonton, Alta.
I
E ; :
8 1.4 :
* L o K S ( D ~ ~ = O W I W-
*TS(D~sa=0.178mm)
~)
Kupper, A.A.G., Morgenstern, N.R., and Sego, D.C. 1992. Lab-
oratory tests to study hydraulic fill. Canadian Geotechnical Jour-
8 nal, 29. This issue.
+ SS(Dso=O536mm)
a UPP(D~=O.~~~) Law, D.J. 1991. The effect of fabric on the behaviour of a tail-
1.3 ' ' ' ' I . . . . ings sand. M.Sc. thesis, Department of Civil Engineering, Uni-
For personal use only.
o 5 10 15 versity of Alberta, Edmonton, Alta.
Specificflow rate q (cm3/(s .cm)) Mastbergen, D.R., Winterwerp J.C., and Bezuijen, A. 1988. On
FIG, 12. variation of density of the fill with specific flow rate the construction of sand fill dams. Part 1: Hydraulic aspects.
for various flume tests. In Modelling Soil-Water-Structures Interactions, Edited by
P.A. Kolkman, J . Lindenberg, and K.W. Pilarczyk.
A.A. Balkema, Rotterdam, pp. 353-362.
Melent'ev, V.A., Kolpashnikov, N.P., and Volnin, B.A. 1973.
Acknowledgements - Hydraulic Fill Structures. Energya, Moscow. (In Russian.)
SN~P-11-53-73 1974. Standard specifications for construction.
Financial support from Natural Sciences and Engineering
Chap. 2. Gosstroi Stroiizdat, Moscow. (In Russian.)
Research Of Canada Operating grants and from an Soni, J.P. 1981. Laboratory study of aggradation in alluvial chan-
ESSO university research grant are gratefully acknowledged. nels. Journal of Hydrology, 49: 87-106.
Soni, J.P., Garde, R.J., and Ranga Raju, K.G. 1977. Aggrada-
tion in alluvial channels due to increase in sediment load. 17th
Allen, J.R.L. 1963. Internal sedimentation structures of well- Congress of the International Association for Hydraulic
washed sands and sandstones in relation to flow conditions. Research, Baden-Baden, August 15-19, vol. 1, pp. 151-155.
Nature (London), 200: 326-327. Torres, W.F.J., and Jain, S.C. 1984. Aggradation and degrada-
Bezuijen, A., and Mastbergen, D.R. 1988. On the construction tion of alluvial-channel beds. Iowa Institute of Hydraulic
of sand fill dams. Part 2: Soil mechanical aspects. In Modelling Research Report No. 274.
Soil-Water-Structures Interactions. Edited by P.A. Kolkman, Winterwerp, J.C., de Groot, M.B., Mastbergen, D.R., and
J. Lindenberg, and K.W. Pilarczyk. A.A. Balkema, Rotterdam, Verwoert, H. 1990. Hyperconcentrated sand-water mixture flows
pp. 363-371. over flat bed. Journal of Hydraulics Division, ASCE, 116(HY1):
Bhamidipathy, S., and Shen, H.W. 1971. Laboratory study of 36-54.
degradation and aggradation. Journal of Waterways, Harbors Yen, C., Lee, H., and Chang, S. 1989. Recovery of channel bed
and Coastal Engineering, ASCE, 97(WW4): 615-630. in aggradation-degradation process. 23rd Congress of the
Blight, G.E., Thomson, R.R., and Vorster, K. 1985. Profiles of International Association for Hydraulic Research, Ottawa,
hydraulic-fill tailings beaches, and seepage through hydraulically August 21-25, vol. B, pp. B323-B329.
sorted tailings. Journal of the South African Institute of Mining Yufin, A.P. 1965. Hydromechanization. Gosstroi Stroiizdat,
and Metallurgy, 85(5): 157-161. Moscow. (In Russian.)You can also read
