William Froude and The Froude Number
William Froude and The Froude Number
William Froude and The Froude Number
Hydraulic Engineering
EDITOR ASSOCIATE ED¡TORS Jacques W. Delleur, Ph.D., P.E. Bruce Melville, Ph.D.
Thanos Papanicolaou, Ph.D. Francesco Ballio, Ph.D.. Purdue Un¡versity Univers¡ty of Aucklond
Jnivers¡ty of Tennessee, Knowille Aff.M.ASCE Subhasish Dey, Ph.D. Gregory B. Pasternak, Ph.D.
Pol¡tecnico di Milano lndian lnstitute of Technology Un¡versity of Colifornio, Dovis
Brian D. Barkdoll, Ph.D., P.E. Cristian Escauriaza, Ph.D. Colin Rennie, eh.O.je.tng.
n Technologicol U n ¡vers¡ty
M ich i go Pontific¡o Universidad Cotólico University of Ottowa
Jean E. Berlamont, Ph.D. de Chile Wolfgang Rodi, Ph.D.
Katholieke Un¡versite¡t, Leuven Rui Miguel Lage Ferreira, Ph.D. Un¡versity of Karlsruhe
Nani Bhowmik, Ph.D., P.E. Universidade de Lisboo José F. RodrÍguez, Ph.D.
lllinois State Woter Suruey James F. Fox, Ph.D. Univers¡ty of Newcostle
Koen Blanckaert, Ph.D. Un¡versity of Kentucky Scott A. Socolofsky, Ph.D.
École Polt¡echnique Fédérate Philippe Frey, Ph.D. Texos A&M University
de Lousonne IRSTEA, University of Grenoble Alpes Luca Solari, Ph.D.
Fabian A. Bombardelli, Ph.D. Heide Friedrich, Ph.D. University of Florence
Un¡vers¡tyof Colifornio, Davis University of Auckland Thorsten Stoesser, Ph.D.
Scott F. Bradford. Ph.D. Mohamed S. Ghidaoui, Ph.D. Cardiff University
Noval Research Loboratory Hong Kong Un¡versity of Science and Kyle B. Strom, Ph.D.
Bruno Brunone, Ph.D., M.ASCE Technology Virginio Tech
University of Perugio Kevin Haas, Ph.D. Daniele Tonina, Ph.D.,
Herve Capan, Ph.D. Georgio Tech, Savonnah
P.E., M.ASCE
Notionol Toiwon University David F. Hill, Ph.D. University of ldoho
Oscar Castro-Orgaz, Ph.D. Oregon State Un¡versity
Blake P. Tullis, Ph.D.
nst ituto de Ag ri cu ltura Sosten i ble Abdul A. Khan, Ph.D.,
I
P.E. utah stote univers¡ty
M. Hanif Chaudhry, Ph.D., P.E. Clemson University
Bas van Maren, Ph.D.
University of South Carolina Martin F. Lamben, Ph.D. Deltares
Yee-Meng Chiew Ph.D. Un¡versity of Adeloide
Jose G. Vasconcelos, Ph.D.
Nonyo ng Technolog ico I U n ¡vers¡ty Eddy Langéndoen, Ph.D. Auburn University
Sung-Uk Choi. Ph.D. USDA.ARS
Daniel Griffin Wren, Ph.D.
Yonsei University Joseph Hun-Wei Lee. Ph.D., P.E. USDA-ARS, Erosion Lob
George Constantinescu, Ph.D., University of Hong Kong
Weiming Wu. Ph.D.
M.ASCE Pengzhi Lin, Ph.D. Clarkson Un¡versity
University of lowa S¡chuon University
Joanna Crowe Curran, Ph.D. Xiaofeng Liu, Ph.D., P.E.
Northwest Hydraulic Consultonts Pen n sylvo nio Stote U n i ve rcity
ASCE
ÁMEBICAN SOCIETY OF CIVIL ENGINEEHS
Steve Starrett, Ph.D., P.E., D.WRE, F.EWRI, F.ASCE, President
David D. Dee Jr., P.E., D.WRE, M.ASCE, Post President
Cristiane Q. Surbeck, Ph.D., P.E., ENV SP, M.ASCE, President-Elect
Glenn E. Moglen, Ph.D., P.E., F.EWRI, F.ASCE, Vice President
Eric Loucks, P.E., D.WRE, M.ASCE, Treqsurer
Gregory F. Scott. P.E., M.ASCE, ASCE President¡ol Appo¡ntee
Jeanne M. VanBriesen, Ph.D., P.E., F.EWRI, M.ASCE, Institute Operations Area Representqtive
Victor Cook, A.M.ASCE, Technicol Areo Representqt¡ve
JOURNAL OF
Hydraulic Engineering
Postmaster. Send address changes
lo Journal of Hydroulic Engineering, General information
ASCE, 1801 Alexander Bell Dr.,
Reston, VA 20191-43a2.
Aims and scope. The Journal of Hydraulic Engineering accepts original contributions
Copyright @ 2017 by American that describe the analysis and solutions of problems in hydraulic engineering. Technical
Society of Civil Engineers.
Notes may present a problem, without solution, of common interest. Topics range
This journol is pr¡nted on ocid-free from flows in closed conduits to free-surface flows (canals, rivers, lakes, and estuaries) to
poper, which meets the ANS/
environmental fluid dynamics. Topics include transport processes involving fluids
requ i rements for permo n ence.
(mult¡phase flows) such as sediment and contaminant transpon, and heat and gas
D¡sclaimer. Any statements transfers. Emphasis is placed on the presentation of concepts, methods. techniques, and
expressed in these materials are
those of the individual results that advance knowledge and/or are suitable for general application in the
authors and do not necessarily hydraulic engineering profession.
represent the views of ASCE, which
The Journal of Hydroulic Engineering (ISSN 0733-9429) is published monthly by the
takes no responsibility for any
statement made herein. No American Society of Civil Engineers. The publications office is at ASCE, 1801 Alexander
reference made in this publication Bell Dr., Reston, VA 20191-4382. All editorial correspondence should be directed to the
to any spec¡f¡c method, product, Journals Department. Periodicals postage is paid at Herndon, Virginia, U.S.A., and at
process or service constitutes or additional mailing offices. The online version of lhe Journdl can be found at http://
implies an endorsement,
o sce I i b ra ry. o rg /j o u r n o I s.
recommendation, or warranty
thereof by ASCE. The materials are Submission of manuscripts. Access the Journol's welcome page at http://
for general information only and do vtww.editorialmanager.com/jrnhyeng to register as a new user, view the authors'
not represent a standard of ASCE,
nor are they intended as a reference tutorial, log onto the system, and submit papers. Manuscripts under review accepted for
in purchase specifications, publication, or published elsewhere are not accepted. Go to http://oscelibrary.org/page
contracts, regulations, statutes, or /outhors for complete instructions on manuscript preparation.
any other legal document. ASCE
makes no representat¡on or ASCE Open Access. Authors wishing to make their published papers available for free
warranty of any kind, whether and under a Creative Commons Attribution 4.O License will have an opportunity to opt
expressed or implied, concern¡ng in and make payment after papers are accepted for publication. For more information, go
the accuracy, completeness, to http :/ / o sce I i b ra ry. o rg / pa g e / o pe n a cce ss o pt¡ o n sa n d r¡ g hts.
suitability, or utility of any
¡nformation, apparatus, product, or Subscription inquiries and address changes. Address all member subscription inquiries
process discussed in this to Member Records, @OO) 548-2723 or memrec@qsce,org. Nonmembers should contact
publication, and assumes no liability
the Customer Service Department, (7O3) 295-621O or [email protected]. Allow six
therefore. This information should
not be used without first securing weeks for all changes to become effective. Subscriptions are filled for the calendar year
competent advice with respect to its and must be prepaid according to rates posted at http://ascelibrory.org.
su¡tab¡lity for any general or specific
application. Anyone utilizing th¡s
information assumes all liability
arising from such use, including but
not l¡mited to infringement of any
patent or patents.
l"l"l 1,0"",*,,
This journal is a member of and subscribes to the principles of the Committee on Publication Ethics
JOURNAL OF
Hydraulic Engineering
APRIL 2017 . VOLUME 143 . ISSUE 4
Contents
r¡r]JcATtot{s FORUM
&rure Gcasen, Monoging William Froude and the Froude Number 0251 600s
Muond Publisher Willi H. Hoger ond Oscor Castro-Orgoz
ñI¡ITII¡II < DEPARTMENT TECHNICAL PAPERS
4r,nqee Cochran, Air Movement lnduced by Water Flow with a Hydraulic Jump in Changing Slope pipes 04016092
r{ssoc¡,ote Publisher Yu Q¡on, David Z. Zhu, Wenming Zhang, Nollomuthu Rajoratnom, Stephen Edwini-Bonsu,
ü.e|tu hderson, ond Peter Steffler
lMrcging Editor Time Scale of Local Scour around Pipelines in Current, Waves, and Combined Waves 0401 6093
-utltu topps¡, and Current
$Ítowging Editor Qín Zhong, Scott Draper, Liong Cheng, and Hongwei An
Elme¡ Guertin, Peak Pressure Surges and Pressure Damping Following Sudden Air o4016094
$t€,-ager, Journals Systems Pocket Compression
ú'/,ninistrotor Thomos M. Hatcher and .Jose G. Vosconcelos
rcq¡rfer Chapman, Stabil¡ty of Unstructured Block Ramps 0401609s
4ssrbfont M a nag i ng Ed ito r Volker Weitbrecht, Simono Tamogni, ond Roben M. Boes
--r.rn Dollings,
Three-Dimensional Simulation of Local Scour around a Weir-Type Structure: 0401 6096
Editorial Coordinqtor Hybrid Euler-Lagrange Model for Bed-Material Load
lrcke G. Eccles, Kozuyuki Oto, Tokoh¡ro Soto, Hojíme Nakogawa, and Kenji Kawoike
Z.ditorial Coordi notor
calculating the Exp¡icit Probability of Entrainment Based on lnert¡al Acceleration 04016097
-.aJ,e Hertneky, Measurements
z,¡blications Coordi notor
Georg¡os Maniat¡s, Trevor B. Hoey, Morwon A. Hosson, Joseph Sventek, Rebecco Hodge,
FNOOUCTION DEPARTMENT Tim Drysdole, ond Manousos Volyrokis
$Íact Boyle, Evolution of Local Pier-Scour Depth with Dune Migration in Subcritical Flow Conditions 0401 6098
Director J¡on-Hao Hong, Yee-Meng Chiew, Po-Hung Yeh, and Hsun-Chuan Chan
Misrael Gentry, Effect of sediment Transport Boundary conditions on the Numerical Modeling of Bed 0401 6099
Workflow Monoger Morphodynamics
i4¿shree Ranganathan, Alejandro Mendozo, Jorge D. Abod, Eddy J. Longendoen, Dongchen Wong, pobto Tossi,
*nior Monoger ond Komol EI Kodi Abderrezzok
-.eather DiAngelis. Detecting Local Scour Using Contact lmage Sensors 04016100
Assistont Monoger Hongwei An, We¡dong Yao, Liong Cheng, Scott Drdper, M¡ng Zhao, Guoqiong Tong, yu Zhong,
bcia Jones Kershaw, ond Philip Hortin
*nior Production Editor Formulas for the Transportation of Bed Load o401 61 01
)e"Andréa Johnson, Chong-Hung Zee and Roymond Zee
Production Editor Multiphase Mesh-Free Particle Method for Simulating Granular Flows and 04016102
-annah Patterson, Sediment Transport
Production Editor Mohommod Amin Nobion ond Leilo Farhodi
{,ffie Thacher,
TECHNICAL NOTES
?roduction Editor
Comparison of Current Methods for the Evaluation of Einstein,s lntegrals 0601 6026
r,: Uan Fleet, Senior Monoger, Koveh Zamoni, Fobión A. Bombardelli, and Bobok Komroni-Moghaddom
lnformation Services
Exact Procedure for Einstein-Johnson's Sidewall Correction in Open Channel Flow 060'16027
Junke Guo
PUBLISHING OFFICE
ASCE
-1.I.I:RICAN SOCIEry OF CIVIL ENGINEERS
Journals Department
ASCE
18O1 Alexander Bell Drive
Reston, VA20191-4382
Telephone: (703) 295-6290
E-mail: journal-services6asce.org
JOURNAL OF
Hydraulic Enginedring
APRIL 2017 . VOLUME 143 . lssuE 4
Examining the Generality of 6/TKE for Gravel and |ñIlbldB with sand Fill 0601 6028
Daniet G. Wren, Michoet E. Ursic, Roger A. Kuhnle, and üI|qfut
Rosette Diffusers for Dense Effluents ,06016029
Ozeoir Abessi, Philip J. W. Roberts, ond Varun Gondhi
Subtract¡on Approach to Experimentally Assess drcÚfri of the Complex Pier 0601 6030
Components to the Local Scour Depth
Mario Moreno, Rodrigo Moio, Lúcio Couto, ond Antón'p.trllÑ
Forum
'F7p* d.
5.a.. LD _ :.t *
tstt-;a:t
(a) (b)
Fig. 1. Wiltiam Froude: (a) around 1860; (b) in rhe late 1870s [image (a) counesl' of Antiquariat, Zunch, Switzerland]
entire resistance, and if it were strictly true, the resistance would has devoted this number to Froude, thereby relating two similar
vary as the square of the speed times the surface area on which mechanical processes governed by gravity. In contrast to the cument
it acts. Froude concludes that the model and prototype data substan- statement, Weber proposed for the Froude number the square
tially verify the law of comparison. From the present perspective, Fw.b".: V,/(sL).
his law states that the ratio V /(gL)l/z applies between a prototype
and its model if gravity effects are dominant, with Z as the ship
length.
Attribution of Froude Number
Given its current imponance in open channel hydraulics, the notion
From Law of Conservat¡on to Froude Number Froude number is considered a misnomer. Froude's merits in naval
engineering are enorrnous, but his research had nothing to do with
Weber (1919), a notable German naval engineer, presented a long the commonly used number F. Instead, the studies of both Barré De
paper on the mechanical similitude of model and prototype data. Saint-Venant (1797-1886) and Joseph V. Boussinesq (L}4Z_I929)
Various laws were attached to individuals having worked on the have definitely contributed to this subject matter. De Saint_Venant
topic. For gravity flows, Froude's name was selected, resulting (1870) deals with the elementary propagation velocity, the so_called
in the well-known expressions for the time, velocity, or acceleration celerity, of a flow in a prismatic channel. Application of the mo_
scales. Similar suggestions for flows in which viscosity and surface mentum theorem to a stagnant water column leads to the celerity c,
tension dominate a flow resulted in the Reynolds and today's as originally found by the French maüematician Joseph_Louis
Weber numbers. As to Froude's law, his observations relating to Lagrange (1736*1813)
ship resistance are mentioned. To apply his law the skin friction
force is subtracted from the total force measured, resulting in
the wave and ship form forces, to which the law only applies. c: (sh)t/z (1)
Although Weber noted that he did not find a proof of the correct-
ness of the law in Froude's papers, Froude's particular merits were In Eq. (1), h = water depth and g = Sravit\ acceleration. De Saint_
stated to be Venant refers shortly to the solitary wave first observed by John
. Subdivision of total ship resistance into skin and form effects; ScottRussell (1808-18S2) in 1837: Itresults by pushing a siagnant
. Separate determination of skin resistance using model tests; and water body in a rectangular channel by a plane immersed into the
. Attribution of form resistance to gravity effect. fluid. Its celerity is described by Eq. (1) if the wave height
Weber (1930) then proposes the Froude number instead of compared with the water depth remains small.
law of conservation as earlier retained (Weber 1919). It is men- Based on observations of Henry Bazin (1829-1917), the abso-
tioned that Froude was the first who determined ship resistance lute propagation velocity of open-channel flow is equal to the mean
using the similitude procedure and therefore is considered the velocity V plus or minus the celerity. With ¡ as streamwise coor-
founder of the science of similitude. Weber further states that he dinate and / as time, Bazin (1865) found
qrrr4a¡fcr s.
4rltqry'crs
\ L Utt brbga O O T@u3 b@,holáhaJhrc adql SFüq
t, 3, ¿ú ñü@ of *hidt ú a k ruitzá P BMs Ap, aLoú*4útvb b a ¿.rflÍhiqo
z ! tuqqnea, bluod.L análtudúV Mütu qq .B@úit;nt hal, oJf fowúq B@ ad A¿, X t"¡w
J* tub rpilry bJ.m tlaopd'C.M¿ng-Zü¿hh Cd'rüq AnúAp-
t t &r rqiaú¿,¿ú¿¿bu,bvLüv of tu.utubú ñ-@yer!:a)¿"8fon&ú R A. ?elf C,¿nk Jre eA6tg qdna 6f Twwn *eiat*s
L a bad, *tÚv st.¿ fo"óan;üü49Jv.á"t bü,'abtu át"¿aeheí " " Oymmzdu Zplzuafu tuDg otu ú o tho¿bJ" t^drg Suild
t,* d, ¿ lbnnzúbtu gf BelL 6 anl' t ttlL 8úúg
1 i b ñ,t@Jb¡ tudlvg tleatptu, b*').gd¿atu L ued'uu towng -Tfodetr¿ e fi)ght gútag inüiaL aa:búútub S""ingfomtng
5-,pry
Í-tsñ? wh;b @4o/n ¡peá. ú beag obtu"E¿. aeo of Scalb
t eals¿n oJ*ñ,úr"n;w?s ?6is'tuM S Í*r$*avq mwof Spring
A t 5¡eia @"úan:c.ahg qtuwitu gf ,|IDmf b g.t':ndz E v n*ol;üt td:t az"*t,;$¡ig*Á
T .b--bh7tubItráa,¿m" regiwel,wabru?a k,qf "h,"]"*
i !ú-'m gtln¿a¿rtu wdd,lybo bb f 'Í I't@ @.rú¿g qbh¿D &o.uaa&c At tcaul
ú- X. k Mübú;V tu:n ef5P"úg blar|eaAtu ¿r¿ 7¿Arú? cb'aúLÍü;.onfa¿ aa¡rrtt *á " "
t :r-^tu.Inlc tudtun oftnmñ;ú¿,tugrtuüb 1/ llnuluu*okeá,byüo&,*orlc
d::?j; bf"b¿Am' g g C@Jtb baqr*in! nbbtugf Cdñato b Qbrdzrr e
I \
i l
*.3.e
Fig. 2. Dynamometric apparatus used in towing ship models (reprinted from Froude 1874)
4r¿Lu",
Iag -hu
Fig. 3. Plan of Her Majesty's Ship (H.M.S .) Activetowing the Greyhound (reprinted from Froude t874)
dxldt:Vtc (2) . V l@h)r/z > 1 any disturbance is felt only downsrream of the
source.
De Saint-Venant states that an elementary disturbance of water The case VlGh)tlz < 1 was referred to as the river regime,in
ilorv is only propagated in the streamwise direction íf V > c, which the flow velocity is smaller than the wave celerity, whereas
ot Vl@h)I/2 > l, resulting in a hydraulic jump if the tailwater the case wirh V lbh)t/2 > 1 was referred to as the torrent regime.
:! characterized by VlQh)'/t <t. The limit of the approach De Saint-Venant stated that neither a river nor a toffent are defined
:lou, velocity generating a hydraulic jump corresponds to by only the absolute flow velocity or flow depth,.but by the ratio
1,' \gh)t/z > 1. De Saint-Venant differs between the following V I @h)1/2, and he noted that flows wíth V I fuh)tl2 > 1 are hardly
:¡o fundamental flow regimes: observed in nature.
. \/ljh)1/2 < I any flow disturbance is felt both upsffeam and In a second paper, De Saint-Venant (1871) derived the unsteady
downstream; and flow equations currently refeffed to as the shallow-water equations.
Fig. 4. Resistance versus speed of Greyhound model at various conditions (reprinted from Froude 1874)
Let V again be the cross-sectional average flow velocity, /¿ is the the number V l@h)1/2 is not explicitly mentioned, but rhe paper was
flow depth, and Iy' is the energy head or flow depth of stagnant revolutionary because Scott Russell's observations were theoretically
water. For horizontal rectangular channels, a frictionless fluid then described for the first time. Boussinesq (187lb) considered
satisfies the two equations gradually-varied open-channel floW thereby following De Saint-
Venant in adopting the regimes of river and tonent flows. As to
v : 2(sq)tt2 _ Z(s\t tz (3) the Froude numbe¡ De Saint-Venant thus clearly has prioriqz in
having distinguished between these two regimes.
A fuli account on free-surface flows was given by Boussinesq
, :3(gH)1/2 - 2(sh)t/z (4)
(1872). First, the fundamental flow equations including the boun-
These are culrently refered to as solutions of Ritter's dambreak dary conditions are specified. Next, 2D flows are considered result-
wave (Ritter 1892). ing for the simplest case in the linear wave equation. A higher
De Saint-Venant (1871) generalized his findings to arbitrary chan- approximation involves the solitary wave as fundamental solution.
nel geometries, resulting in the continuity and energy equations Its stability is discussed along with a description of its generation
by a piston-moving fluid at rest in one direction of the channel,
daldt -r d(av)lds :0 (s) thereby pointing at the role ot V /(gh)1/2. Neither De Saint-Venanr
nor Boussinesq defined explicitly a dimensionless group V/c
d(/ds : (tls)dv ldt + (v I g)dvlds + (yl"l)Fl@s) (6) but worked with similar dimensionless numbers. For example,
Boussinesq 877) based most of his reasoning of backwater curves
(I
In Eqs. (5) and (6), u = cross-sectional area; tl = time; s = streamwise on the group U -
V2 lGh)1. Given the value of Boussinesq's pa-
coordinate; ( = vetical elevation ofthe free surface relative to a refer- pers, the 1871 Poncelet Price waS awarded to him by a Commission
ence level; X = wetted perimeter; p = fluid density; and Flfug) = chaired by De SaintVenant ("Prix" 1872), honoring the French
boundary friction per unit channel length. Applying Eqs. (5) and hydraulician Jean-Victor Poncelet ( 1 788* I 867).
(6) to the rectangular, frictionless, and horizontal channel, Eqs. (3)
and (4) result as solutions. A comparison of the results with the data
of Bazin (1 865) for the celerity of solitary waves indicates agreement Discussion
if the wave height is not too large. Given these two impressive papers
on general 1D open-channel flows forwhich the cuüentFroude num- No attempt is made in this paper to change the current Froude num-
ber is essential, De Saint-Venant certainty is much more linked to this ber by proposing another name. However, hydraulic engineers
phenomenon than Froude. should recognize that this notion is a misnomer if applied to
Boussinesq was closely acquainted with De Saint-Venant, given free-surface flow because its main features, namely a distincúon
that the latter presided the Mechanics Section of the French Academy between the subcritical and supercritical regimes, is not given based
of Sciences, Paris, and Boussinesq had to submit to him the papers to on Froude's work. In the early 1870s, supercritical or torrential
be published inits Comptes Rendus. The first account of Boussinesq flows were hardiy observed in nature given the few manufactured
(1871a) on open channel flow involves the solitary wave, including conveyances. This is reflected by the few hydraulic jumps o
its celerity, its free surface profile, and the wave energy. The effect of bserved, except at drops or wave breaking along estuaries. It is
)nny's Tank of rectangular shape was 85 m long, 6.7 m wide, and time, he was not aware that the original Hydraulic Tank was set
-: m deep. The test techniques of Denny's Thnk were identical to up there and accessible for visitors. So, step into the building in
i¡:e of the Torquay Tank. The front wall of the building contained which the Froude Tank is on show! At the entrance hall, Sir
¡ nemorial bust of Froude. The south end of the building was de-
William Denny (1847-1887), an intimate friend of William
;¡"-r¡'ed by fire in 1924 it was lengthened by 15 m including an ob-
Froude, greets you (Fig. 6). His contact with Froude allowed
slr ation plafform. This Tank remained in use until 1983; cunently it
;: üe core of the Scottish Maritime Museum at Dumbarton. him for securing a replica of Froude's Tank, allowing for model
tests of ships manufactured by Denny & Bros.
Once a ship had been designed, its model was prepared from
lfaritime Museum plans shown in Fig. 7(a), in which the hull,s contour lines are trans-
mitted onto the model made of paraffin tFig. 7(b)1. The raw hull
)-rring his stay at Glasgow University in 2009, the first author was finally smoothed by scratching all joints and seams, resulting
:.ited the Maritime Museum at Dumbarton (Fig. 5). Until this in a smooth ship model body ready for Tank testing [Fig. 7(c)].
FlS. 8' (a) View at tank with ship model attached to platform drawing it through water (image by Wilti H. Hager,
exhibit photographed with permis-
' ::- of Scottish Maritime Museum); (b) test drive along Denny's Tank in 1950s (image by Willi H. Hager, exhibit photographeO wlth permission of
S,: ,:rish Maritime Museum)
the model data. A number of additional items are exhibited at "Memoir of the late William Froude, LL.D., F.R.S." (1819). Trans. Inst.
Nav. Archit., 20, 264-269.
the Museum including mathematical and physical instruments for
"The late Mr. W. Froude, F.R.S." (1879). Eng.,2j, 462463.
model preparation, tools to finish test hulls, photographs ofthe en-
"Mr. William Froude, LL.D., F.R.S." (1880). Proc. Inst. Civ. Eng., 60,
gineers involved in the Tank, and awards collected by the Dennys
395404.
for the numerous successful boats launched. In summary, if you "William Froude." (1933b). Eng., 156, 110-111.
spend a joumey in Scotland, a visit to the Scottish Maritime "Institution of Naval architects ar Torquay." (1954). Eng., i98,
Museum is highly recommended. 453454.
Baker, G. S. (1911). "The William Froude national tank}' Trans. Inst. Nav.
Archit., 53(1),37-52.
Conclusions Bazin, H. (1865). "Recherches hydrauliques [Hydraulic.researches]."
Mémoires présentés par divers Savants i¡ I'Académie Impériale de
This Forum article describes the origin and the current application
France, 19, 1494 (in French).
of the Froude number. Although William Froude counts among the
Boussinesq, J. (1871a). "Théorie de I'intumescence liquide appelée onde
greatest naval engineers based on his theoretical and particularly
solitaire ou de translation, se propageant dans un canal rectangulaire
experimental research, his name is not at all associated with the
[Theory of liquid intumescence referred to as solitary wave or as
number describing open-channel flows. This misnomer originates translation propagating in rectangular channel: part 11." Comptes
from Weber, a German naval enginee¡ who overlooked the main Rendus de I'Académie des Sciences,72,755-759 (in French).
fields in which the Froude number was and is applied. Boussinesq, J. (1871b). "Sur 1e mouvement permanent varié de l,eau dans
This Forum is evidence that even a misnomer of this importance les tuyaux de conduite et dans les canaux découvefts [On steady,
survives for a century. The authors do not at all suggest a name gradually-varied flow in pipes and open channels: Part 2)." Comptes
change of this number but attempt to highlight the origins of both Rendus de I'Académie des Sciences.73.34-38 (in French).
the naval engineer Froude and the two who would be destined to Boussinesq, J. (1 872). "Théorie des ondes et des remous qui se propagent le
receive the merit in name, De Saint-Venant for his shallow-water long d'un canal rectangulaire horizontal. en communiquant au liquide
contenu dans ce canal des vitesse sensiblement pareilles de 1a surface au
equations and shortly later, Boussinesq who generalized these. This
fond [Theory of waves and back*'ater propagating in a rectangular
Forum thus highlights the meandering paths of a key expression
horizontal channel with almost uniform velocity distribution from
in hydraulic engineering because of the misunderstanding of the the free surface to bottoml." J. MathémniEtes Pures et Appliqués Série
historical background, aiming to clarify the background for educa- 2, 19, 55-108 (in French).
tional and conceptual reasons. Boussinesq, J. (1877). "Essai su¡ la théorie des eaux courantes [Essay on
the theory of flowingwaterf." Métnoíres présentées par divers savants it
I'Académie des Sciences,23. 1460 (in French).
Notation Castro-Orgpz, O., and Hager. W. H. (2010). "Critical flow: A historical
J. Hydraul. Eng., 10.i06I(ASCE)HY.1943-7900
perspective."
The following symbols are used in the paper: .0000158,3-11.
c = celerity (m/s); Chow, V. T. (1959). Open charutel hydraulics, Wiley, New york.
D = length scale (-); Crichton, A. T. (1990). "William and Robert Edmund Froude and rhe
dxf dt = absolute velocity (m/s); evolution of the ship-model experimental tank." Trans. Newcomen
F = Froude number (-); Soc., 6I(l),3349.
r/bd = boundary friction per unit channel tength (N/m); Dana, R. W, ed. (1905). "William Froude, LL.D., F.R.S." Zrans. Inst. Nav.
g - gravity acceleration (m/t2); Archit., 47(l).
H = energ! head (m); De Saint-Venant, A. Barré (1870). "Démonstration élémentaire de la for-
h = flow depth (m); mule de propagation d'une onde ou d'une intumescence dans un canal
Z = length (m); prismatique; et remarques sur les propagations du son et da la lumiére,
R= resistance force (N); sur les ressauts, ainsi que sur la distinction des riviéres et des torrents
s= streamwise coordinate (m); [Elementary demonstration of theformula of propagation of a wave or
a disturbance in a prismatic channel; and remarks on the propagations
r= time (s);
of sound, light, and bores, as also on the distinction between a river and
C= vertical elevation of free surface relative to reference
a torrent: Part 1]." Comptes Rendus de I'Académie des Sciences, TI,
level (m); 186-195 (in French).
p = fluid density (kg/m3); De SainlVenant, A. Barré (1871). "Théorie du mouvement non permanent
X = wetted perimeter (m); and des eaux, avec application aux crues des riviéres et á f introduction des
c¿ = cross-sectional area (m2). marrées dans leur lit [Theory of unsteady water movement, appiied to
floods in rivers and the effect of tidal flows: Part 21." Comptes Rendus
de I'Académie des Sciences,73, 147-154 (in French).
References Duckwoth, A. D., ed. (1955). The papers of William Froude, M.A., LL.D.,
F.R.S, Institution of Naval A¡chitects, London.
Abeli, W. (1934a). "William Froude." Eng., 158,39. Froude, W. (1874). "Experiments with H.M.S Greyhound." Trans. Inst.
Abell, W. (1934b). "William Froude." Trans. Inst. Nav. Archit., 76, Nav. Archit., 15(l), 3613.
242-256. Heller, V. (201 1). "Scaie effects in physical hydrauiic engineering models."
"Prix Ponceiet, fondée par Mme V" Poncelet [Poncelet Price, founded by J. Hydraul. Res., 49(3),293-306.
Mrs. Ponceletl." (1872). Comptes Rendus de I'Académie des Scíences, Johns, A. W. (1924). "The work of W. and R.E. Froude." Eng., 138,
75, 1347-1348 (in French). 66c*661.
Abstract: This paper presents experimental and numerical results of ai¡ movement in sewer pipes. Both the airflow in a straight pipe and
pipes with changing slopes were studied. The result from the straight-pipe model suggests that the air pressure gradient is built up in the pipe
even when it is kept at atmospheric pressure at both ends. The combined effects of water drag and the pressure gradient were analyzed.
A general method for estimating the airflow rate was proposed. Further, physical and numerical experiments were performed on'air movement
induced by water flow with a hydraulic jump in pipes with a changing slope. The ratios of air and water flow rates were found to be sub-
stantially higher than published values because of different air transportation mechanisms. The rough surface and airlwater bubbly flow in the
roller of the hydraulic jump can affect the momentum flux of the air phase, and this additional momentum needs to be incorporated in airflow
modeling. DOI: 10.1061/(ASCE)HY.1943-7900.0001252. @ 2016 American Society of Civil Engineers.
Author keywords: Circular pipe; Hydraulic jumps; Numerical model; Physical model; Sewers.
rrr{Liron in sewers with headspace, it is necessary to study the air of Exp. Bl at Q.:6.5,9.88, 13.58, and 15.66Lls. The airflow
rxr:tr,ii:r€nt in pipes with hydraulic jumps with headspace. Edwini- rate, water depth upstream of the jump, and the water flow rate were
&:m:s: ,1004) suggested that the hydraulic jump would generate a measured for 0:30' (Exp. A1). In addition to túese measure-
iis:s*: ¿rnount of airflow because of the potential momentum trans- ments, air velocity and water sudace profile in the center plane
rer m üe roller by the airlwater bubbly flow, but a detailed analysis of the pipes were also measured for Exp. BI for Q,:9.88 L/s.
Js :iot presented.
¡lr
A rotating vane anemometer (VelociCaIc 5125, TSI, Shoreview,
Tiis research presents experimental and numerical studies on Minnesota) was mounted on the air inlet for measuring the airflow
tm ¡:dlow in a straight pipe and pipes with a hydraulic jump. The rate. An additional experiment was conducted for examining the
'rr-tn5}t-pipe
model was used for assessing the combined effect of measurement reliability. At the same airflow conditions, the read-
ruat:r drag and pressure gradient on airflow in the sewer. The pur- ing of the rotating vane anemometer was compared with the read-
nl*e of the hydraulic-jump model was to measure the airflow rate ing of two hotwire anemometers (HHF 42, Omega, Laval, euebec
mu¡ced by water flow with a hydraulic jump in pipes of changing and Velocicheck 8330, TSI, Shorevieq Minnesota). Over a total of
:tr:rr.e, It was also used to explore the momentum transfer effect of a 22 readings, the rotating vane anemometer gave an airflow rate
r"-,';r-¡ulic jump on airflow. This study can be used to predict the reading of 15.63 L/s with a standard deviarion of t0.60 L/s. The
oür:]o$' in a sewer pipe with known flow conditions, and thus it air velocity was measured at 18 locations across the air vent, and
* i¡-.efuI in guiding activities in sewer odor control and pipe the airflow rate through the air vent was estimated to be 14.84 Ll s
:::rr:trsion prevention. by the hotwire. With a difference of 57o, the airflow rate measure-
ment was considered reliable at the air vent.
The air velocity inside the pipe was visualized by dry ice mist
Experimental Des¡gn
created by mixing dry ice with water (Merzkirch 1987) and cap-
T:e physical experiment of the air movement in changing slope tured using a high-speed camera (Phantom v211, Vision Research,
r¡¡nes u,ith a hydraulic jump was conducted at the T. Blench Hy-
Wayne, New Jersey) for a period of 3 min at 1,000 frames/s. The
¡r'¿nlic Lab at the University of Albefta. Fig. 1 shows a schematic local air velocities were obtained from the analysis of the videos,
rr' úre experimental apparatus. Two transparent Plexiglas pipes and each local velocity was an ensemble average of 15 velocity
¡, :ü an inner diameter D :
0.28 m and a length of 2.0 m each were
calculations. For the air velocity measurement inside the pipe, the
rrritnected. The upstream pipe was set at an angle of á : nature offlow visualization limited the accuracy because the front
10 or 30'
r,:'h üe horizontal, while the downstream pipe was kept horizontal. of the mist was not as well defined sometimes and large discrep-
4. :eilgate was used to control the downstream water level and the
ancies were expected. An average deviation of f0.28 m/s was ob-
r;rüion of the hydraulic jump. Water was supplied from a pump served, with the measured ai¡ velocity ranging fiom 0. 1 1 to 2.09 m I s.
,üm': the water discharge was measured by an electromagnetic However, at this stage, there are no obviously better ways to conduct
ir,rqmeter. A hole with a diameter of l0 cm was cut above the measurement because of the restriction of pipe walls and splashes of
nuer inlet. water drops in the pipe headspace. The water flow rate from the
Table I provides a summary of the experimental conditions. magnetic flowmeter was checked with the traditional method of
]ñ-erent downstream flow conditions as well as pipe slopes were using a bucket and stopwatch, and the difference was less than 3Va.
:l-.rnined. The downstream conditions were changed so that the Two methods were used to rheasure the water flow depth in the
0
u:jlow at the pipe outlet included free floW partially restricted pipes. Method 1 was by measuring the wetted perimeter of the pipe
ü
ir:'a. or fully blocked flow. Exp. 51 was a numerical simulation using a measuring tape accurate to I mm. The effect of the pipe wall
thickness was corrected since the water level was culr/ed upstream
of the jump. Three readings were. taken at the desired locations,
d
Tóle 2 cm upstream and 2 cm downstream. This method was used to
1. Summary of Experiments
d obtain the water depth for Froude number calculation. Method 2
c irperiments Q" (L/s) Downstream condition involved a total of 7 small holes drilled into the pipe top. A wooden
d 6.50-13.58 Hydraulic jump with headspace stick was inserted into one hole where the measurement was made,
f 6.50-15.66 while the other holes were kept sealed. The distance between the
d 9.90 Hydraulic jump with no headspace water surface and the pipe invert was the measured water depth. A
It 15.84 Conduit jump comparison was made between these two methods I m downstream
n 5.01-15.00 Tailgate fully open, supercritical flow of the pipe jointfor Q,: 9.88 L/s. The direct measurement gave a
t- 6.50-15.66 Numerical simulation for Exp. Bl reading of 19.5 cm, while the indirect method yielded a water depth
¡s 5,:,te: Angle for 'A" experiments, 0 :3O'; angle for "B" experiments, of 19.7 cm. Therefore, these two methods compared well with each
re ;: l0'. other and were considered reliable.
I
Opening
Fig. 2. Geometry and boundary conditions of numerical models: (a) straight-pipe model; (b) hydraulic-jump model
The node number and element size were detemined based on a .. lro ¡u¡bulence damping reml
,: -.::ivity analysis and on mesh independence. P,!999u!4_1ry9.!j!¿
A series of meshes . -a '
' .:: Senerated by decreasing the node number by 1.3 times. There_
': Ihe node numbers tesred were 2.70,2.0g, and 1.60 million.
-'' :r-odel became mesh independent at 2.70 million nodes where
n :rsnif,cant improvement on the water velocity profile was
:'..:-.ed. Therefore, the rest of the simulations were based on é o.os
= 3.0 rd
- - nillion nodes.
i " r : ar Exp. C0 in Table 2, a straight-pipe model of pipe diam- Fig. 3, Comparison of simulated air velocity profiles at pipe center
, .- J:0.30 m, zero slope, and length of 6.0 m was first built plane with Pescod and price (1982)
Illlr
I Simulated supercritical flow
flow, the V olv, are close to previous studies. As shown in Fig. 5,
O Simulated subcritical flow
O Pescoda:rdPrice (1982) even when the upstream inlet and downstream outlet were both set
- - Edwini-Bonsu md Steffle¡ (2004 at atmospheric pressure, a positive air pressure gradient was always
generated in the pipe headspace, because the drag ofwater produ-
ces a negative pressure at the pipe inlet for air to move in. At the
outlet, the pressure would be slightly higher than atmospheric pres-
\ o
a
sure for air to move out so that the pressure increases from the inlet
to the outlet. For example, for D :0.60 m and Q* :40
Lls, the
o average pressure gradient was 0.045 Pa/m.
With respect to Exp. C2, where the pressure at the pipe exit was
above atmospheric pressure, most of the airflow rate decreased with
the increase in pipe exit pressure except for Q* :
80 L/s, with a
slight increase in airflow rate as stated in Table 2. If ¡he pressure
0.0 gradient was high enough, the airflow rate was negative. A typical
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 air velocity contour under this condition is plotted in Fig. 6. The
h/D maximum air velocity occurred at the water surface. The air veloc-
Fig. 4. Plot of the air and water velocity ratio with filling ratio for ity decreased with the vertical distance to the water surface. At the
Exp. C0 and Cl top of the cross section, the air was moving opposite to the water
flow because of the positive pressure gradient.
For Exp. C3, in which the air was purely driven by the pressure
gradient, the effect of pressure gradient on the airflow is theoreti-
The simulated airflow rate was L}.lZLls, which was 387o cally calculated using the energy equation. Considering a pipe
higher than the value of 1.33 L/s measured by Pescod and Price. D :0.60 m in diameter, 10 m in length, with a f,lling ratio
In their study, the airflow rate was obtained by integrating the h/D (where lz is the water depth) of 0.14 and a pressure gradient
point velocity measurements that inherently had large unceftainties. of 0.045 Paf m, the airflow rate can be estimated at 300 L/s from
energy equations, and the estimated airflow rate was 16.J7o smaller
Also, no detailed discussion on ambient pressure conditions was
than the simulated value of 350 L/s.
provided in their study. In the numerical model, the momentum
Associating with the control volume shown in Fig. l, assuming
transferred from water to air was amplified because of the finite
that the air and water velocities have uniform proflles, the force
amount of mesh at the air-water interface. Therefore, large discrep-
balance equation for a control volume is written as
ancies were noticed, as expected. The coefñcient E ÁEq. (4) was
determined at 150 and no more improvement was observed for l^f
a higher -E value. - V,)¿BLx - LpA, tiO"ViLxP" : 0 (5)
tCop,(V,
Further experimental conditions and simulation results are sum-
marized in Table 2. For Exp. Cl, the pipe exit pressure was fixed at where C¿ = drag coeffrcient; Ar = length of the control volume;
atmospheric pressure while the pipe diameter and water flow rate B = water surface width; and Po = perimeter of headspace for
were changed. For Exp. C2, the pipe exit pressure varied to study airflow.
the effect of the pressure gradient inside the pipe. For Exp. C3, a Looking at the balance of the lust two terms in Eq. (5), the pres-
single-phase model was conducted where airflow was only driven sure gradient can be normalized with water depth and water veloc-
by the pressure gradient. It was found that for the same pipe diam- ity as (Ap/A*)(D-n)lQ",V'*). The term C, is obtained from
eteq with the increase in water discharge, the Froude number the normalized pressure gradient and the filling ratio h I D . 'fhe re-
[defined as F: V*/(gh")0's, where h" = hydraulic depth and lation of drag coefficient C¿ with h/D atd the normalized pressure
Vw = dyetai9a water velocity] decreases as a result of the increase gradient is fitted from software Oügín
in water depth, in Table 2.
as shown
For Exp. the ratio of air to water flow rates, p, decreases
Cl,
with the increase in water flow rate. The ratio of the air to water
C¿ -ter x ro5
" (X'#)0" " (*)'"+oo43 (6)
velocity values, V"/V., are plotted in Fig. 4 with the study of
Edwini-Bonsu and Steffler (2004) and Pescod and Price (1982). with R2 :0.98. The value of C¿ estimared from Eq. (6) is
From the figure it can be seen that the V o/ V * values for supercriti- of the order of 0.1. The drág coefficient between the air and
cal flow are substantially higher than published values. While for water surfaces is commonly given as 0.13 for wind force over a
the supercritical flow with lower Froude number and subcritical lake (Fischer et al. 1979). In a confined space such as in a pipe,
Pressure [Paj
-.r
o9s odb o|6 6rt o*of pf s$ !9u of $S
Fig. 5. Typical pressure distribution in straight-pipe model for D:60 cm, Q,:40L/s, pexit - 0 Pa, I : 10 m
li
Similarly, the water surface change at the hydraulic jump was also
a" (rn/s)
predicted accurately by the numerical model.
i 1.50
Preliminary simulation suggests that the numerical model
,::1
126 was not able to accurately model the extra momentum transfer
:,;.i 1.02 between the air and water because of the rough surface and
oze the bubbly flow at the roller of the hydraulic jump. While the
ffi momentum transfer from the hydraulic jump to the air phase
#$
ffi
0.54
did not affect much on the water flow, it induces a pressure change
ffi u.JU
that significantly affects the airflow given the low density of the
K
€ 0.06 air. Therefore, a momentum source term was incorporated on air-
flow in our numerical model. The momentum source term (S.)
specified in the CFX code has a dimension of ML-2T 2, which
meant momentum flux per unit volume (ANSIS). In the hydraulic-
I."' jump model, the source term was added to the pipe joint with a
volume of 0.0015 m3. Therefore, the actual momentum flux
added M, is the value of momentum source ,S" added in the
numerical model multiplied by 0.0015 m3. The measured airflow
rate was used as a calibration quantity and the strength of the
momentum source was adjusted accordingly. The air velocity
profile and theoretically calculated pressure were used as valida-
Fig. 6, Typical velocity contour at pipe location of x
- 5 m showing tion criteria.
reversed airflow at top; D:0.60 m, Q.:20L/s, p."it:5 Pa The change in airflow rate with added M" for Q, :13.58 L/s
is plotted in Fig. 8(a). A clea¡ growing trend of Qo is observed,
although the growth was small for a wide range of M".The M,
the drag coefficient is expected to change given the unique air value was determined when the simulated airflow rate was within
lelocity profiles. l}Vo of that measured in this case. As shown in Fig. 8(b), it was
also found that for a higher water flow rate, a higher M" value was
needed to match the simulated airflow with experiment results.
Hydraulic-Jump Model: Water Surface Profile
The M, value increases rapidly if the water flow rate exceeds a
andAir Velocity Profile certain value. It can be explained by the fact that higher water flow
The measured and simulated water surface profiles are plotted rate would generally produce a hydraulic jump with a higher
in Fig. 7. The model was able to simulate the change in the water Froude number that had a rougher surface and more vigorous sur-
iurface profile. The bump in the sloping pipe at x ¡y
-1.0 m was a face movement. Therefore, more momentum would be transferred
:esult of the sloshing of water from the circular water inlet. to the air phase.
i o.r
-¿-ltll2
x {¡n}
(b)
frg. 7, Comparison between measured and simulated water surfaces for Q.:9.88 L/s, 0 - l0': (a) water surface in sloping pipe;
: neasured and simulated water surface
l5
l4 É
.9
6
J
oi' 13
Water surface
12
Pipe invert
0.0
-1.0 0.5 L0 1.5 2_0
tt zr, (m/s)
(a)
t\
l\
t\
?
-a
I
l#
k o.z
ú
(b) 0_ (L/s)
the hydraulic
Fig. 8. Plot of addiúonal momentum source induced by
juáp: (a) change of airflow rate different M" values for
'e*: L/s; (b) different Ml
-with
.ú9:gttl0,
13.58
Air velocity profiles at sections 1'0 and 0'5 m upstream of the >'
m downstream of the pipejoint were measured in
pip"¡oint *¿i.'O
;h; pb. center plane for Exp. Bl at Q.:.9'99 L/s' and these re-
;;id;*" with those simulated in Fig' 9' The error bars
"ompared measure-
in tt ngu." highlight the uncertainties of air velocity
"
ment. Tñe simulateá air velocity profiles are overall in agreement
J. Hydraul. Eng.
0401 6092-7
O ASCE
lPal
Ws;fifiii::r'r l
Fig. 10. Simulated air pressure distribution for B1; 0. :9.88 L/s,0:10", M,:0.0015 kg.m/s2
u,(mls)
uo(mlss) ,o?
- 2.84 z3a
.):.,,¡ 2.46
it¡i 2.17
#,¡ 2.08 ";& 1.79
ffi r.zo i#ffi t.+t
ffi r.az ¡ffirn"
r.ve
@
ffi o.s+
0.66
m^--
/
Éffi u-5
ffi
m o.2B
g u.lv
H-.- H -o.lo
8.U,19
II
r 13t
IiÍl
(a) (b)
u, (mls)
=,8. 11 . Simulated air velocity contour at different locations, looking from downstream with Q. : 9.88 L/t, 0 : l0o, M,: 0.0015 kg . m/s2:
, . :n upsfream of the pipe joint; (b) 0.5 m upstream of the pipe joint; (c) I m downstream of the pipe joint
Itesimulated cross-sectional air velocity contours are shown in pipe and at the top of the pipe headspace. At 1 m downstream of
'' , , l. At 1 m upstream of the pipe joint [Fig. 11(a)], the air veloc- the pipe joint, positive air velocity was found in the entire pipe
: rltour was somehow symmetric to the center plane. In the pipe headspace, and the highest air velocity was at rhe top right of the
" :. ' rDace near the water suface, a maximum air velocity was com- pipe headspace. The reason for the asymmetry distribution of air
: -::i. In the upper region of the headspace, negative airflow was velocity was that the pressure upstream of the jump was not evenly
' ::i. At 0.5 m upsffeam of the pipe joint, a similar flow pattern distributed.
' :: !-omputed [Fig. 11(b)], but the simulated air velocity distribu-
I '¡ as asymmetric because of the uneven distribution of the water
Hydraulic-Jump Model: Bulk Airflow Rate
--,..-e obtained in the numerical model. The maximum simulated
"l: '' llocity at this location was computed at the left bottom of the The measured and simulated airJlow rates and B values are pre-
: :t headspace where water flowed downstream at a relatively high sented in Fig. 12 with water discharge and the Froude number
: ^-'itv. Reversed airflow was observed at the right bottom of the in the upstream pipe. The airflow rate was used as a calibration
0.5 Conclusions
Three-dimensional, multiphase numerical models were built to
0.0 study the effect of water drag and pressure gradient on airflow in
partially filled sewer pipes. The straight-pipe model suggested that
(b) for a circular pipe with the inlet and outlet open to the atmosphere,
an air pressure gradient was automatically generated in the pipe. A
Fig. 12. Plot of airflow rate in hydraulic-jump model: (a) airflow rate positive pressure gradient reduces the airflow rate. If the pressure
versus water flow rate; (b) ratio of airlwater flow rate B versus Froude gradient is high enough, the air can move in the opposite direction
number at the toe of jump or in supercritical flow (for Case 84) of the water flow. For the airflow affected by both water drag and
pressure'gradient, the drag coefFrcient was found to be a function
of the fiiling ratio and normalized pressure gradient. A general
criterion in the hydraulic-jump numerical model as mentioned method was proposed to predict the airflow rate in a straight sewer
previously. The airflow rate increases with the water flow rate. The pipe with known flow conditions. This work provides a way to pre-
trend of air and water flow rate in Fig. 12(a) is in good agreement dict the airflow in a certain sewer network given the upstream and
for Exp. A1 and B1, suggesting that the effect of d on B is relatively downstream pressure and wastewater flow conditions. For sewer
small. The airflow rate in Exp. 84 shows a slightly faster growth networks, the pressures in pipes are not necessarily atmospheric.
rate with the water flow rate because of the fact that there was no With the proposed method, one can estimate the airflow condition
sudden reduction of headspace. The airflow rate for the conduit (e.g., airflow rate and flow direction) in sewer systems and there-
jump (Exp. 83) was lower than other cases because the air was fore have a general understanding ofthe odor transportation in such
mainly transported in the forms of bubbles and air pockets, while a system.
headspace existed for air movement in other cases (except for Exp. The hydraulic-jump model was focused on a more complex flow
B2). Exp. 83 was similar to most of the previous studies (Kalinske regime. The momentum transferbf a hydraulic jump on airflow was
and Robertson 1943; Mortensen et aL 20ll; Escarameia 2007). proposed from the hydraulic-jump model. For the hydraulic-jump
Finally, no air was transported in Exp. 82 where the tailgate was experiment and numerical model, when a hydraulic jump existed
closed from the top and with no headspace. the airflow rate was larger than published values and the ratio of air
In Fig. 12(b), most of the present p values were found to be to water flow rates p was of the order of 1.0 when the downsffeam
substantially higher than the published values in which the roller pipe was not full. By forming a hydraulic jump with a Froude num-
of the hydraulic jump filled the pipe and the flow became a conduit ber large enough to partially seal the headspace, the airflow rate
jump. Because ofthe blockage ofheadspace, the transportation of was found to decrease and approach published values. No air
bubbles induced by the hydraulic jump was the dominant mecha- was transported if the tailgate was closed from the top, leaving
nism of air transportation. Therefore their p values were lower than no headspace downstream of the jump. The proposed method from
those in the present study in which headspace existed and air the straight-pipe model also works well with the hydraulic-jump
moved freely from upstream to downstream. tn Exp. 83, a conduit model. The additional momentum flux caused by the roller of a
jump was created and the B value was found to be fairly close to the hydraulic jump on airJlow was discussed. It was found that the
published values [Fig. 12(b)]. numerical model was not able to accurately simulate the airflow
The method proposed previously from the straight-pipe model induced by a hydraulic jump without adding additional momentum
can be used to estimate the airflow in the sloping pipe. Considering sources in the model. The strength of this momentum flux source
ir = rneoll velocity components; Mofensen, J. D., Barfuss, S. L., and Jhonson, M. C. (2011). "Scale effects
¡, = fluctuating velocity components; of air entrainment by hydraulic jumps within closed conduits."
tr-, = averaged air velocity; J. Hydraul. Res., 49(7), 90-95.
1-," = averaged water velocity; Origin version b9.3.226 [Computer software]. Originlab, NoÍhampton,
: (i : l, 2, 3) = Cartesian coordinates MA.
-r = length of conffol volume; Pescod, M. B., and Price, A. C. (1982). "Major factors in sewer
:
= volume fraction of fluid; ventilation." J. Water Pollut. Coerol Fed., 54(4),385-397.
,i = ratio of airflow rate and water flow rate; Rajaratnam, N. (1965). "Hydraulic jump in horizontal cond:uit." Water
Power Dam Constr.,17(2), 80-83.
: = model constant 0.075;
Sharma, H. R. (1976). 'Air-entrainment in high head gated conduits."
: = angle between pipesi J. Hydr. Div., 102(ll), 1629-1646.
- - viscosity of fluid; Speerli, J., and Hager, W. H. (2000). "Air-water flow in bottom outlets."
-r = density of fluid; Can. J. Civ. Eng.,27(3),454462.
;o = density of air; Tullis, B., and Larchar, J. (2011). "Determining air demand for small- to
;., = density of water; and medium-sized embankment dam lowlevel outlet works." J. Itig.
:... mean viscous stress tensor components.
= D rain. En s., 1 0. 1 06 I (ASCE)IR.l9 43 -47 7 4.0000345, 793-800.
USACE (U.S. Army Corps of Engineers). (1980). "Engineering and
design: Hydraulic design of reservoir outlet works." Rep. EMlI10-2-
1ó02. Washingron. DC.
ffibrences U.S. EPA. (1985). "Design manual: Odor and conosion control in sanitary
sewerage systems and treatment planfs." Rep. epa/625/1 -85/01,
rrut .c"r*triare Engineering). (20081. "Odor control program report." City Washington, DC.
m Enlmonton, AL, Canada. Volkart, P. U., and Speerli, J. (1994). "Prototype investigation of the high
ilüü-ñr"15 .rJ.0 lcomputer software]. ANSYS, Canonsburg, PA. velocity flow in the high head tunnel outlet of the Panix Dam." proc.,
J. Hydraul. Eng.
0401 6092-1 1
@ ASCE
Time Scale of Local Scour around Pipelines in Current,
Waves, and Combined Waves and Current
Qin Zhangl; Scott Draper2; Liang Chengs; and Hongwei Ana
Abstract: This paper presents results of an experimental investigation into the time scale of local scour around subsea pipelines in current-
only, wave-only, and combined wave and current flow conditions. To cover a range of geometric scales, the experiments were conducted in
a small and a large recirculating (Otube) flume with model pipelines ranging from 50 to 196 mm in diameter. Based on the experimental
results and existing data available in the literature, it is shown that the time scale of the scour process is mainly dependent on the maximum
instantaneous Shields parameter in all three flow conditions. Using this result, a new predictive formula for the time scale is presented, which
is applicable universally to current-only, wave-only, and combined wave and cufient conditions. In addition to the time scale analysis,
equilibrium scour depth measurements are presented for all three flow conditions. A new method of predicting the equilibrium scour depth
in combined wave and cuffent conditions is proposed based on the current component of equilibrium scour depth and the wave component of
equilibrium scour depth. Collectively, the empirical formulas developed for time scale and equilibrium scour depth may be used to improve
predictive engineering models of scour beneath subsea pipelines. DOI: 10.1061(ASCE)HY.1943-7900.0001247. @ 2016 American Society
of Civil Engineers.
Author keywords: Subsea pipeline; Time scale; Equilibrium scour depth; Combined waves and cunenr.
Ph.D. Candidate, Schooi of Engineering, Ocean Univ. of China, where D = diameter of the pipeline; s = relative density of the sedi-
i38 Songling Rd., Qingdao, Shandong 266100, China (corresponding
ment; d56 = mean sediment grain size; and g - acceleration due to
..",- . :r. E-mail: [email protected]
- gravity. Based on an empirical fit to experimental data available in
S:nior Lecture¡, School of Civil, Environmental and Mining Engineer-
-: - niv. of Western Australia, Crawley, WA 6009, Australia. the literature for current-only and wave-only scour beneath pipe-
'Pofessor, School of Civil, Environmental and Mining Engineering, lines, Fredspe et al. (1992) found that Eq. (1) could explain the
..r. of Western Australia, Crawley, WA 6009, Australia. scour process reasonably well when the nondimensional time scale
-Lecturer, School of Civil, Environmental and Mining Engineering, was calculated according to
. -:.. of Western Australia, Crawley, WA 6009, Australia.
li:te. This manuscript was submitted on April 25,2015; approved on
, .9. 2016; published online on October 26,2016. Discussion period rrr !u-' t
-50
(3)
::- -nti1 March 26,2017; separate discussions must be submitted for
i - .lual papers. This paperis pafioffheJournalofHydraulic Engineer- where d = normalized free field shea¡ stress, known as the Shields
.ri ASCE. ISSN 07.1.1-9429. parameter, and is given by