• Aucun résultat trouvé

Intermittency as a transition to turbulence in pipes: A long tradition from Reynolds to the 21st century

N/A
N/A
Protected

Academic year: 2021

Partager "Intermittency as a transition to turbulence in pipes: A long tradition from Reynolds to the 21st century"

Copied!
19
0
0

Texte intégral

(1)

HAL Id: hal-02130924

https://hal.archives-ouvertes.fr/hal-02130924

Submitted on 16 May 2019

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Distributed under a Creative Commons Attribution - NonCommercial - NoDerivatives| 4.0 International License

�10.1016/j.crme.2017.06.004�. �hal-02130924�

(2)

Contents lists available atScienceDirect

Comptes Rendus Mecanique

www.sciencedirect.com

A century of fluid mechanics: 1870–1970 / Un siècle de mécanique des fluides : 1870–1970

Intermittency as a transition to turbulence in pipes: A long tradition from Reynolds to the 21st century

Les intermittencies comme transition vers la turbulence dans des tuyaux : Une longue tradition, de Reynolds au XXI

e

siècle

Christophe Letellier

NormandieUniversité,CORIA,avenuedel’Université,76800Saint-Étienne-du-Rouvray,France

a rt i c l e i n f o a b s t ra c t

Articlehistory:

Received21October2016 Accepted3April2017 Availableonline3July2017 PresentedbyFrançoisCharru Keywords:

Pipeflows Laminarregime Turbulence Intermittency Frictioncoefficient Mots-clés :

Écoulementdanslesconduitescylindriques Régimelaminaire

Turbulence Intermittences Coefficientdefrottement

Intermittencies are commonly observed in fluid mechanics, and particularly, in pipe flows. Initially observed by Reynolds (1883), it took one century for reaching a rather fullunderstandingofthisphenomenonwhoseirregulardynamics(apparentlystochastic) puzzledhydrodynamicistsfordecades.Inthisbrief(non-exhaustive)review,mostlyfocused on the experimental characterization of this transition between laminar and turbulent regimes, we present some key contributions for evidencing the two concomittant and antagonistprocessesthatare involvedinthiscomplextransitionand weresuggestedby Reynolds. It is also shown that a clear explicative model was provided, based on the nonlineardynamicalsystemstheory,theexperimentalobservationsinfluidmechanicsonly providinganappliedexample,duetoitsobviousgenericnature.

©2017Académiedessciences.PublishedbyElsevierMassonSAS.Thisisanopenaccess articleundertheCCBY-NC-NDlicense (http://creativecommons.org/licenses/by-nc-nd/4.0/).

ré s u m é

Les intermittencies sont communément observées en mécanique des fluides et, plus particulièrement, dans les écoulements dans des conduites cylindriques. Initialement obervéespar Reynoldsen 1883, il afallu unsièclepour parveniràunecompréhension plutôtcomplètedecephénomènedontladynamiqueirrégulière(apparemmentstochas- tique) déconcertaleshydrodynamiciensdurant plusieurs décades. Parcette brèverevue (nonexhaustive), essentiellementfocalisée surlacaractérisation expérimentale de cette transition entrerégimes laminaire etturbulent,nous présentonsquelques contributions clésayantconduitàmettreenévidencelesdeuxprocessusconcomittantsetantagonistes impliquésetquiavaientdéjà étésuggérés par Reynolds. Ilestégalement montréqu’un modèleexplicatifclairfutproposé,surlabasedelathéoriedessystèmesdynamiquesnon linéaires,lesobservationsexpérimentalesenmécaniquedesfluidesayantserviuniquement d’exemple,etceenraisondesoncaractèregénériqueévident.

©2017Académiedessciences.PublishedbyElsevierMassonSAS.Thisisanopenaccess articleundertheCCBY-NC-NDlicense (http://creativecommons.org/licenses/by-nc-nd/4.0/).

E-mailaddress:Christophe.Letellier@coria.fr.

http://dx.doi.org/10.1016/j.crme.2017.06.004

1631-0721/©2017Académiedessciences.PublishedbyElsevierMassonSAS.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense (http://creativecommons.org/licenses/by-nc-nd/4.0/).

(3)

In thispaper,I willshow that the transitionto turbulencevia intermittencyfollows moreor lessthisexample.First, thereisalongtraditionininvestigatingtheroutetoturbulence,andintermittencieswerealreadyobservedintheearlyex- perimentsmadebyReynolds.Roughlyonemaysaythatacenturywasneededforclarifyingthistransition(see,forinstance, Barkley’sreview[16]).Second,inthe1980s,anotherapproachtothisproblemcamefromnumericalexperimentsandpro- videdsimpleandgeneric modelstodescribe such transition.Ifit contributedto popularizedtheidea ofintermittencyas a route toturbulencein fluid mechanics,thistheoretical explanationhad amuch moresignificant impact inthenonlin- ear dynamical systemstheory andopened ourminds to thepossibility tohave intermittentbehaviorsinwell-conducted experiments,thatis,withoutanyexternalperturbations.

2. Thehistoricalapproach 2.1. Earlyexperimentalevidences

HenryDarcy(1803–1858),preoccupiedbytheengineeringofwatersupply,investigatedindetailsthelawprovidingthe headloss hper unit oflength ina tubeof diameterd.Darcystartedfroma generallawforthehead lossh per unit of lengthmadeofthefirsttwopowerofthemeanvelocity V,thatis,from

dh

4 =A V +B V2 (1)

Heclearlyunderstoodthatthesecondterm,B V2,wassufficienttoexplaintheflowwithlargevelocities(thuscorresponding to the turbulentregime inour presentterminology). This“turbulent” component was forhim themost importantsince directlyrelatedtotheroughnessofpipes.DarcyunderconsideredthecaseswherethetermA V wasnegligible,thatis,when thevelocity was small(lessthan 0.10 m s1) orwhenthe roughnessofpipeswasimportantasmostly encountered ina workingsystemforwatersupply, pipesare “very quicklycoveredby slits, tubersor limescale”[17,p. 91].Consequently, Darcy mostly focused his attention on the turbulent regime, arguing that “the experiments for which the slope or the velocitywas smallare ingeneralthelessaccurate onesand, contrarytothis, thoseforwhichslopeandvelocityarelarge providemorereliableresults.”Nevertheless,Darcyremarkedthatthereisathresholdvelocitybeyondwhich“removingthe firstterm(A V)doesnotaffectthevalueoftheflow”[17,p. 120].Heevenaddedthat“thefirsttermwasthereforerelevant onlyforvelocities(V<0.10 m s1)andforpipeswithsufficientlysmoothwalls”[17,p. 120]:inthiscase,thesecondterm canberemoved.Darcymentionedthatforintermediatevelocities,thefrictionofwateragainstwallsbecomesproportional to a binomial made ofthe first and the second power of the velocity” [17, p. 121], thus suggesting that there are two concomittantunderlying phenomena. He was convinced that beyond a velocity equal to 0.10 m s1,“a new phenomena occurs”[17,p. 214].Whenhekeptthetwoterms,Darcyrecommendedtouse

⎧⎪

⎪⎩

A=0.000 031 655+0.000 000 007 5112 D

B=0.000 442 939+0.000 012 402 D

(2)

In1854,GotthilfHagen(1797–1884)investigatedtheinfluenceoftemperatureonwaterflowinpipes[18].Inorderto dothis,Hagenuseddifferentpipesandvariedthetemperatureofwater:indoingso,hewasinfactvaryingveryslowlythe propertiesofthewaterflow.Hethusremarkedthat,forcertainconditions(pressureanddiameterofthepipe,forinstance), thewaterflow, whenplottedversus thetemperature,was presentingamaximumfollowedbya minimum(Fig. 1).Hagen observed the evolution ofthe waterflow by looking atthe jet at the end of the tube. He remarked that betweenthe maximumandtheminimumvelocity(17<T<30C)inthe caseofFig. 1,thejet wasjerking. Initially, hebelievedthat sucha featurewasduetoabadlyconductedexperiment, butfinally convincedhimselfthatthisphenomenonwasinfact

“normal”. He also understood that this was a transition from one regime to another one. Since he developed a theory

(4)

Fig. 1.Watervelocityinapipe(d=0.28 cm,l=47.2 cm)withaloadof11.08waterinch(2934.7 PawhentheoldFrenchinchistakenequalto27 mm) whenthetemperatureisvaried.ThevaluesforthevelocitywhereassessedwiththehelpofPoiseuille’slaw.RedrawnfromHagen,1854.

Fig. 2.“Theflasheswouldoftencommencesuccessivelyatonepointinthepipe.Theappearancewhentheflashessucceededeachotherrapidly[ishere]

shown.”FromReynolds,Plate72,Fig. 16,1883.

basedonthehypothesisthatthevelocitywasproportionaltothedistancefromthewall,hewas notabletoprovideafull agreementbetweenhistheoreticalrulesandobservationaldata.

Osborne Reynolds(1842–1912) ismostly knownforhaving evidencedthat there aretwo very differenttypesofflow, namelythelaminarandturbulentflows[19,p. 935].Heunderstoodthat

thegeneralcharacterofthemotionoffluidsincontactwithsolidsurfacesdependsontherelationbetweenaphysicalconstantof thefluidandtheproductofthelineardimensionsofthespaceoccupiedbythefluidandthevelocity.

In ordertodistinguishlaminarfromturbulent flows,hethusderived a dimensionlessnumber,the“Reynoldsnumber”as namedbyArnoldSommerfeld[20].ItismorerarelyknownthatReynoldsalreadydescribed“theintermittentcharacterofthe disturbance[...]givingtheappearanceofflashes”(Fig. 2).ForReynolds,theconceptofthecriticalReynoldsnumberwasclear, inspiteoftheexperimentaldifficultiesencounteredformanagingtheinitialconditionsoftheflow[19,p. 955]:

TheonlyideathatIhadformedbeforecommencingtheexperimentswasthatatsomecriticalvelocitythemotionmustbecome unstable,sothatanydisturbancefromperfectlysteadymotionwouldresultineddies.

[...]Ihadnotbeenabletoformanyideaastoanyparticularformofdisturbancebeingnecessary.Butexperiencehavingshown theimpossibilityofobtainingabsolutelysteadymotion,Ihadnotdoubtedbutthatappearanceofeddieswouldbealmost simultaneouswiththeconditionofinstability. Ihadnot,therefore,consideredthedisturbancesexcepttotryanddiminish themasmuchaspossible.Ihadexpectedtoseetheeddiesmaketheirappearanceasthevelocityincreased,atfirstinaslowor feeblemanner,indicatingthatthewaterwasbutslightlyunstable.Anditwasamatterofsurprisetometoseethesuddenforce withwhichtheeddiessprangintoexistence,showingahighlyunstableconditiontohaveexistedatthetimethesteadymotion brokedown.

Thus, in1895,heproposed thattheflow inround tubeswas stablefor ρV d

μ <1900 and unstablewhenitisgreater than 2000 [21].But theresearch fora “criticalReynoldsnumber” characterizingthe transitionfromlaminarto turbulentflow wasmadedifficultbythedependencesofthetransitiononinitialdisturbances.

Maurice Couette(1858–1943) investigatedwhetherthe twotermsinDarcy’s lawwere duetodifferentphenomena or not.IfhediscoveredReynolds’resultsjustbeforethepublicationofhispaper(hethankBoussinesqforhavingshowedhim Reynolds’results),thereisnomentionofHagen’sworks.Couetteusedthreedifferentexperimentstocheckthatthereisa discontinuitybetweenthetwo(laminarandturbulent)regimes[22]:i)hemeasured thefrictionappliedtothewallsofa cylinder bya fluidentrained by therotationofa secondcoaxialcylinder, ii)hemeasured theheadloss producedby the flow ofa liquidthrough a pipeandiii) heobserved thejet comingout froma long tube.He showedthat foraflow Q lessthanathresholdflow Q0,thejet wasalwayssmooth;contrarytothis,whentheflowwasgreater than Q1>Q0,the jet was alwaysrough. Whenthe flowwas such Q0<Q <Q1,heobservedajet with“suddenvariations,withoutapparent regularity,initsshapeaswellasinitsamplitude,thelatterbeingalwayslongerforthesmoothjetthanfortheroughjet.”

Inhisbookpublishedin1907,MarcelBrillouin(1854–1948)madeaquitecomprehensivedescriptionofHagen’sobser- vations [23]: whenthe temperatureisincreased, theviscosity decreases,andirregularmotions areno longer“sufficiently damped”forallowingwatermotionstoremainalmostrectilinear.Brillouinused“Poiseuilleregime”fordesignatingthelam- inar regime and“hydraulicregime”fortheturbulent regime.We systematicallyreplaced histerminologyby the“modern”

one betweenbrackets.Brillouinalsodistinguished “ondulatorymotion”fromturbulentmotion,adistinctionwhichwasnot later followed: wewill thereforeuseturbulent motion.So, forBrillouin,turbulent motion“appearintheliquid,amplifythe

(5)

Fig. 3.AcapillarytubeBCisadaptedhorizontallytoameasuringtubeAfilledwithmercury.TheendBofthecapillarytubehasafunnelshapetoobtain moreeasilyalaminarregime.ThejetflowingoutinCisobservedwhiletheheightHofmercuryisdecreasing.

Table 1

ObservationsmadebyBrillouinwithmercuryflowinginacapillarytube(l=18.9 cmandd=1.0 mm).TheReynoldsnumbers(notprovidedbyBrillouin) wereherecomputedaccordingtoPoiseuille’slaw.

Mercury height Regime Reynolds number

H>160 mm [Turbulent]regime Re>11131

H=136 mm Smalloscillations Re=9461

H=126 mm Largeoscillations Re=8766

73.9<H<122 mm Hugeoscillations 5141<Re<8487

H<73.9 mm [Laminar]regimedominateswithfewsuddenlossinamplitude(sudden returnstoturbulentregime)

Re=8766 H=68.1 mm Endoftheblurredregime.Beginningofthestationarylaminarregime. Re=4737 inequalitiesinvelocityanddampproportionallymoreenergythan[laminar]regime,foragivenflow”.Brillouinaddedthat“when thetemperatureisstillincreased, [turbulence]isfirst,moreandmoreimportant,andthenbecomesstationary;theregimeismore regularandtheflowincreasesagain.”Brillouinunderstoodthatwhatisobservedatagivenpressurewhenthetemperatureis variedistheimageofwhatoccurswhenthetemperatureiskeptconstantandthepressureisprogressivelyincreased.

Inordertoevidencethisphasewherethelaminarregimebecomes turbulent,Brillouinuseda verysimpleexperiment (Fig. 3)hepresentedinhislectureon4February1899.Brillouindescribeshisobservationsasfollows[23]:

Atfirst,theflowproducedunderahighpressure,andthe[turbulent] regimeisobserved:thejetflowingoutisregularbutnot smooth.Sincemercuryisflowing,thepressuredecreases;atagiventime,the“blurredphase”starts;the[laminar]regime(smooth jet)startstooccurbyintermittencies;eachtimeitappears,thejetislonger,showingthattheflowincreasesandthattheresistance decreases;butthe[turbulent]regimerestartsquickly,thejetisoncemoretimesuddenlyblurredand,simultaneously,thejetis morecurvedanditsamplitudedecreases.

Thejetthusjerks,fromoneamplitudetotheother.First,theseoscillationsarerare,thesmallestamplitudebeingthemost observed;the[turbulent]regimedominates.Then,theheightyetdecreasing,theoscillationsaremorefrequent,beforebeingrare again,butthelargeamplitudebeingnowthemostoftenobserved;the[laminar]regimedominates.Finally,belowacertainvalue ofthedrivingpressure,thesole[laminar]regimeremains,thejetissmoothandtheoscillationsdisappeared.

Withagiventube(l=18.9 cmandd=1.0 mm,Brillouinprovidedadetaileddescriptionofhisobservations(Table 1).

Herepeatedthisexperimentwithothertubesandwithwater,andgotsimilarobservations.Hewasthus abletoconclude that “foragivenglasstube,thetransitionfromoneregimetotheotherdoesnotoccursuddenly,atathresholdvelocity,butthere isa blurred phaseduringwhichthetworegimesarepossibleandalternatewithamoreorlesslargefrequency”.Intermittencies were thus alreadyclearly identified in1907! It seems that theseworks by Hagen,Couette andBrillouin remainedquite unknown...

2.2. Thefrictioncoefficients

In his book, Julius Weisbach (1806–1871) proposed a lawfor the head loss h per unit of length as [24] (quoted by Flamant[25,p. 143])

dh 4 =

a+√b

V

V2 (3)

(6)

where d is the pipe diameterand V the mean velocity of thewater. The parameter values are a=0.0007336 and b= 0.0004828.Fewyearslater,inhisexperimentalresearchesonwaterflowinpipes,HenriDarcy(1803–1858)expressedthe headlosshperunitoflengthas[17,p. 14](aswrittenbyFlamant[25,p. 142])

dh 4 =

α+β

d

V2 (4)

wheretheparametervaluesare α=0.000507 andβ=0.00001294[17].Tothisbackground,Jean–Louis–MariePoiseuille’s (1797–1869)contribution[26]shouldbeadded.Heobtainedaresistanceproportionaltothevelocity[26]undertheform

Q=kP d4

l (5)

wheredisthepipe’sdiameter,litslengthandk aconstantcoefficientforthesametemperatureandthesamegravitational acceleration.Poiseuilledesignatedby Q whathenamedthe“productcorrespondingtopressures”andwhatisnamedtoday theflow.Reynoldsrewroteitas[19,p. 972]

α d

3

μ2h=β

d

μV (6)

whereweexpressedthepressurehintermsofheightofwaterperunitofpipelength.Introducingexplicitlythepressure lossP andthelengthl ofthepipe,Reynolds’equation(6)canberewrittenas

P=γμl

d2 V (7)

that is,undertheformproposedby EduardHagenbach-Bischoff(1833–1910)who usedthemeanvelocity V andnot the flow Q [27]: the resistance to the flow is thus proportional to the velocity. A law similar to Poiseuille’s one was also obtained by Hagen [28]. Hagenbach-Bischoff, who was aware of these two different contributions, choose to designate equation(7)asPoiseuille’slaw.

Reynoldswasthusawarethat[19,p. 995]

therelationsbetweentheresistanceencounteredby,andthevelocityof,asolidbodymovingsteadilythroughafluidinwhich itiscompletelyimmersed,orofwatermovingthroughatube,presentthemselvesmostlyinoneorotheroftwosimpleforms.

Theresistanceisgenerallyproportionaltothesquareofthevelocity,andwhenthisisnotthecaseittakesasimplerformandis proportionaltothevelocity.

Nevertheless,hisexperimentsconvincedhimthatratherthanusingthelawhV2 orhaV +bV2as“propoundedbyany ofthepreviousexperimenters”,nootherlawthanhV1.723shouldbeused[19,p. 975],alawthathedesignatedasthe“law ofpressures”.Reynoldsthusproposedalawforthepressurelossas

αd

3

μ2h= βd V

μ 1.723

(8) which ishererewrittenusingour notations,h beingthepressure loss(expressed asaheight ofwater)betweenthetwo endsofoneunitoflengthoftheconsideredpipe, μbeingthedynamicviscosity.Formakingsimplerthecomparisonswith laterresultsthatwewilldiscussbelow,letusrewritethisexpressionas

h=0.00078V1.723

d1.277 (9)

wherethenumericalcoefficientisobtainedfromthenumericalvaluesprovidedbyReynolds.

Slightlylater,AlfredFlamant(1839–1915)madeadetailedreviewofthedifferentlawsproposedfortheresistancehead inpipesandproposedtouse[25,p. 149]

dh=0.00092 4

V7

dh=0.00092V1.75

d1.25 (10)

whichisthereforenottoodifferentfromequation(9)proposedbyReynolds(whenrewritteninasimilarform).

In1901,AugustusV.SaphandErnestW.SchoderstartedajointdoctoralthesistoperformbetweenFebruary24,1902, andFebruary25,1903approximately800experiments,testing23kindsofpipeandhoses,coveringa200-foldrangeof velocities[29].Theyobtainedfortheheadloss[30]

h=0.00054V1.75

d1.25 (11)

Références

Documents relatifs

At small Reynolds numbers, before the flow transition from convective to absolute instability, fluctuations start rising at the edge, within a region where the flow is no longer

A spectral analysis shows that this transition is dominated by two kinds of transfer: first, the shear instability induces a direct non-local transfer toward horizontal wavelengths

At small Reynolds numbers, before the flow transition from convective to absolute instability, fluctuations start rising at the edge, within a region where the flow is no longer

the stator boundary layer (lower part) is already turbulent while the rotor boundary layer (upper part) remains stable until very high Reynolds number Re = U/νR then shows an

paradigm, wherein an essential factor in the functioning of individuals will transition from value to values: from profit as a decision-making and motivation vector, achieving a

To create signs, ceramists used technologies of working with clay and stone, found in a number of artifacts of Neolithic cultures, the highest achievements of sculptors are

Le terme de bilinguisme inclut celui de bilingualité qui réfère à l’état de l’individu mais s’applique également à un état d’une communauté dans laquelle

It was firmly established that the important criterion was the length of the flow field which was perturbed and this enabled the uncovering of a ½ scaling law for the amplitude