Surface Roughness Measurement on a Wing Aircraft by Speckle Correlation

Transcription

1 Senor 13, 13, ; do:1.339/ Artcle OPEN ACCESS enor ISSN Surface Rougne Meaurement on a Wng Arcraft by Speckle Correlaton Félx Salazar * and Alberto Barrento Departamento de Fíca Aplcada (FARN, ETSI Mna, nverdad Poltécnca de Madrd, C/Río Roa 1, 83 Madrd, Span; E-Mal: alberto_barrento@otmal.com * Autor to wom correpondence ould be addreed; E-Mal: felxjoe.alazar@upm.e; Tel.: ; Fax: Receved: 1 July 13; n reved form: 6 Augut 13 / Accepted: 3 Augut 13 / Publed: 5 September 13 Abtract: Te tudy of te damage of aeronautcal materal mportant becaue t may cange te mcrocopc urface tructure profle. Te modfcaton of geometrcal urface properte can caue mall ntablte and ten a dplacement of te boundary layer. One of te rregularte we can often fnd urface rougne. Due to an ncreae of rougne and oter effect, tere may be extra momentum loe n te boundary layer and a modfcaton n te parate drag. In t paper we preent a peckle metod for meaurng te urface rougne on an actual unmanned arcraft wng. Te reult ow an nomogeneou rougne dtrbuton on te wng, a expected accordng to te anotropc nfluence of te wnd over te entre wng geometry. A calculaton of te uncertanty of te tecnque gven. Keyword: rougne meaurement; peckle; non-contactng metod; mcro-detecton 1. Introducton Numerou example are known n wc te urface rougne play an mportant role n te operaton and performance of all type of tecnologcal ytem. In aeronautcal engneerng te ue of pecalty pant (coatng, woe am to mnmze te frcton of te object relatve to te flud n wc t move, and te ce buld-up and polluton on te wng of an arcraft are ome example n wc te control of te rougne mportant. In all tee and many oter cae, te ncreae n

2 Senor 13, rougne can lead to gnfcant negatve effect. For tee reaon, among oter, te rougne and t effect ave been nvetgated n-dept for a long tme. One pecfc cae wc we are ntereted to tudy of te eroon (damage on te wng of an actual arcraft, due to te mecancal nfluence of te wnd and oter factor tat contrbute to te ncreae of rougne. Takng nto account te geometrcal caractertc of a wng (lengt and cro-ecton, flgt condton and erove effect of te wnd, damage wll be dtnct on dfferent part of te arfol. A a conequence, a non-omogeneou rougne dtrbuton mut appear. An arfol-aped body wc move troug a flud produce an aerodynamc force [1]. Te component of t force tat perpendcular to te drecton of moton called lft, and te component parallel to te drecton of moton, wc oppoe body velocty, called drag. Tu, te latter force (drag drectly nfluenced by te body urface rougne and lft affected n turn. In general, for flyng t derable to reduce all drag (parate and nduced to a mnmum. In flud mecanc te layer of flud n te mmedate vcnty of a boundng urface known a te boundary layer. In t area te factor tat create te majorty of drag experenced by te boundary body te flud vcoty. One problem wen ncreang wng urface rougne te cange n te boundary layer flow of te arfol, leadng to a premature tranton of t layer, wc reult n early turbulent eparaton. Any cange n t contour wll caue a detacment of te boundary layer at lower angle of attack. A a reult a greater overall drag value, tat to ay, te force oppong te advance of te plane, wll be obtaned. Terefore, becaue an arcraft avng roug wng mut fly wt ger angle of attack to mantan lft (te boundary layer eparaton occur before, drag force wll be greater tan n te cae of te moot arfol. Conequently, fuel conumpton wll ncreae, a te arplane t wll ave to counter ncreaed drag. Notwttandng, t negatve nfluence not only lmted to te wng of an arcraft. Drect and ndrect effect on te dfferent mecancal element ave been extenvely tuded to date, o a mlar effect may be found wen tudyng wnd turbne and oter macne. Tu, te rougne modfcaton of te rotor blade arfol becaue of contamnaton alo lead to power lo and to an ncreae of te boundary layer tckne [ 4]. In general, frctonal drag due to urface rougne an mportant tak n flud engneerng reearc [5]. In t artcle te meaurement of te dfferent degree of urface rougne on a wng of an unmanned arcraft (MILANO unmanned arcraft-av preented. Tradtonally te mot ued metod a been te proflometer tecnque. However, t a been revealed tat t procedure not ueful to apply to te cae of an entre arfol wng conderng tat t ytem baed on te regtraton of te urface topograpy of a mall ample by mean of a type tylu. In order to avod t drawback, an optcal metod, wc baed on te peckle penomenon, a been employed. Tere are dtnct peckle procedure for meaurng rougne [6 11]. For te preent nvetgaton te angular peckle correlaton (ASC coen [7,11]. T tecnque, even toug requre an accurate algnment of te optcal lgt beam, and t very entve to envronmental vbraton, allow u ealy and accurately to meaure te non-contact rougne on all part of te wng.

3 Senor 13, Teoretcal Concept and Bac Equaton In te feld of te peckle tecnque, te two mot mportant et-up ued are te followng: (1 te angular peckle correlaton metod (ASC and ( te pectral peckle correlaton tecnque (SSC [7,11]. Te frt one obtaned by ung an optcal ytem n wc a laer beam under an angle drected onto te roug urface tat to be examned. In order to carry out te peckle correlaton, a econd expoure obtaned n te ame manner, but te ncdence angle of te lgt beam canged by amount. Te econd opton mlar but ntead to vary te ncdence angle, te beam reman contant wle only cangng te wavelengt of te laer. T artcle deal wt te ASC metod, tu we wll refer only to t tecnque. Aumng a plane urface (macrocopcally and a Gauan dtrbuton for t urface egt, and neglectng multple catterng and adowng effect, t can be demontrated tat, for a focal lengt gven te correlaton of two peckle pattern n te Fourer plane correpondng to two ncdence angle of te laer beam [7,11]: C,,,, exp n n k exp f o o f Lk f co co (1 were k te wave number of te monocromatc beam, f te focal lengt of te len, te varance of te urface, L te alf-wdt of te laer, and and o te angle of ncdence and obervaton, repectvely. and refer to te coordnate ytem on te detector (Fourer plane. By ntroducng all parameter correpondng to a pecfc et-up n Equaton (1, te meaurement of te rougne may be conducted. However, n order to carry out te meaurement a ealy a poble and, on te oter and, mnmzng te uncertanty, we can mplfy t equaton f te followng condton apply: and, co co Subttutng bot reult n Equaton (1 lead to: C f, (. (3,,, exp n tan co, o o k (4 In t paper we coe te condton, ten expreon for te correlaton become: C o k, exp( n, (5 wc eay to manpulate becaue t depend drectly on. Terefore, by meaurng te correlaton of two peckle pattern fted by an amount f, te rougne of a urface may be determned a follow: lnc,. (6 4 n

4 Senor 13, ncertanty by te ASC Metod In t ecton te ytematc uncertanty of te meaurement by te ASC tecnque computed. Wt t am te bac teory of uncertante [1] appled wt te formula, wc for t partcular cae adopt te form: N I ( I x (x j, (7 j j 1 lnc were x j are, and. Takng nto account tat, applyng Equaton (7 to t 4 n equaton, we obtan: ln C ln C co ln C I,, I( I ( (, I (8 4 n 4 ( n 4 (n and te relatve uncertanty: I ( I ( I ( I ( tg. (9 For te cae analyzed I ( = rad. On te oter and, aumng I ( =.1 nm, 45º and I ( = rad, Equaton (7 yeld: I ( %. It may be oberved tat te econd term contrbute more to te uncertanty tan te ret. It correpond to te angular dfference between ncdent beam. To reduce t uncertanty and mprove te meaurement, a ger precon wen meaurng te angle neceary. However, wt t et-up and expermental value te precon of te tecnque.1 μm. Te above calculaton only vald f te angle of ncdence and obervaton are te ame. However, f one of tem dfferent, ten te Equaton (8 mut be modfed. In t cae, followng te ame way a before we can wrte: (1 N ( (,,, x (x j, (11 j were more varable appear. By calculatng partal dervatve t lead to: j 1 ( ( ( ( n tg co co tg n ( co n tg co co. (1 T expreon may be mplfed to gve:

5 Senor 13, ( ( ( ( n tg co tg n ( tg co tan co. Let u uppoe tat te obervaton angle (45 mantaned, but te ncdence angle cange. For t pecfc cae te equalty become: ( ( ( ( n 45co tg co 45n tg co ( 45 tan tg45 (13 ( ( ( n co co n (. 1 tan By comparng t reult wt Equaton (9 we oberve tat te two frt term are te ame. Tu, n order to nvetgate weter t uncertanty ( greater tan I (, we can focu our attenton to te lat ummand of Equaton (13. In fact, a ( ( I (, for te frt uncertanty Equaton (9 to be maller tan te econd one, te followng nequalty mut apply: 1 tg n co 1 tan (14 co n However, even toug t nequalty dffcult to reolve, a te frt term of te econd member alway add, t relaton alway fulflled f: 1 1 tan 1 tan45 1 tan 1, (15 tat to ay, ( /4. T reult ow tat te bet dpoton for te layout n order to reduce te uncertanty of te metod to ue an angle of Expermental Reult In order to nvetgate te effect of te wnd on te urface of an actual arcraft wng, meaurement of te rougne on dfferent part of te wng were carred out. To perform te experment a devce a own n Fgure 1 wa employed (ee alo Fgure and 3. A He-Ne laer LB wa drected to a beam pltter (BS for eparatng te ntal beam n two part. Te frt one (lb1 pae troug a varable reflectve mrror (M and trke on te wng (ample. Te econd (lb frt reflected by

6 Senor 13, te mrror M1 and alo tranmtted troug te varable mrror M and, after tat, mpactng on te ample. Te ue of a graduated mrror allow u to adjut te ntente of bot peckle pattern on te detector avodng error wen calculatng te correlaton. Te lgt cattered by te roug urface n te two expoure eparately collected by te len and focued on te CCD array (Fgure 5. Fgure 1. Layout. (a LB, laer beam; (b BS, beam pltter; (c M1, Mrror; (d M, gradual mrror; (e W, wng (ample; (f L, len; (g CCD, detector; (, dfference angle of te two beam; ( f, focal lengt; (j and, CCD coordnate ytem; (k, lateral dplacement of te CCD camera. CCD f L He-Ne Laer LB M1 BS lb lb1 M o f W Fgure. Expermental et-up employed. Te pcture correpond to te meaurement at a pont located on te extreme of te wng flap (te lengt of te wng 1.3 m, approxmately. Owng to te dfferent angle of ncdence, te two peckle pattern are dplaced [ee Equaton (]. It mean tat for calculatng te maxmum correlaton of te two feld we mut dplace te detector by a quantty of f [7,11]. In prncple t procedure may be carred out drectly wt tee two pattern only. However, to be ure tee Equaton ( and (3 are met, t advable to regter a ere of peckle pattern for te econd expoure, eac of wc taken by movng te CCD camera n mall tep. Tu, troug t procedure we ave confdence tat te maxmum correlaton found. Fgure 4 ow te correlaton correpondng to te ere of two peckle pcture obtaned for one of te examned pont of te arfol. Oberve tat te maxmum found far away of te frt expoure.

7 Senor 13, Fgure 3. Expermental devce. In t cae te potograp ow a pont located on te mddle of te wng near te leadng edge. Fgure 4. Expermental plot of te correlaton C veru dplacement (n μm of te CCD camera. Notce tat te maxmum reaced away of te frt expoure were te camera wa not yet dplaced. C(.5 Correlaton (m 1. Fgure 5. Speckle pattern correpondng to te maxmum correlaton. (a Frt expoure for te laer beam (lb1 for an ncdence of 45. (b Speckle feld pcture of te econd beam (lb wen te CCD camera dplaced by an amount f. Te pcture correpond to a camera of pxel ( mm. (a (b All experment were made for 63.8 nm, f = 8.5 cm, o = 45, and = 1.9. Te expermental reult obtaned for te rougne are repreented n Table 1. In order to reume t table, a pcture of te wng wt all pont examned and ter rougne repreented n Fgure 6.

8 Senor 13, Table 1. Expermental rougne obtaned for te elected pont of te arcraft wng. Loc Poton Rougne R q (μm 1 Extrado 9. Intrado Intrado Extrado 8. 5 Intrado Intrado Intrado Extrado 6. 9 Extrado 6. Fgure 6. T pcture ow te rougne at eac pont meaured on te arfol. Pont 3 a pecal cae becaue t doe not ave coatng. 9 (6. m 8 (6. m Extrado 4 (8. m 1 (9. m Intrado 7 (6.7 m (1.1 m 3 (1.4 m 5 (6.9 m 6 (6.3 m Te ntal rougne of te wng wa m, and ten tee reult ow tat after many our of flgt all part deterorate, but dfferently, and on place near te leadng edge, bot above and below te arcraft wng, te wear le tan on te flap. Te nterpretaton of all tee numercal value can be undertood by mean of flud mecanc. In effect, near te leadng edge of te wng up to te oulder, a favorable preure gradent ext and te arflow lamnar [1]. T way of nteractng for te ar molecule wt t part of te arfol almot otropc, tu te mecancal acton agant te wng urface mut be alke (pont 5, 6, 7, and 8. A a reult te nduced rougne due to wear reman mlar for upper and lower part of te arfol (about 6 m. On te contrary, wen te flow dplace from te oulder to te rear urface of te wng, an unfavorable preure gradent appear, wc lead to a cange n te arflow and n te boundary layer, wc become turbulent. A t well known, wen turbulence preent te partcle of te flud decrbe a caotc moton n very complcated trajectore. Tee partcle wtn te wake fluctuate quckly, bombardng te flap urface of te arfol. T fact explan te reaon wy te rougne on t area of te wng greater tan on oter part tuded. On te oter and, wen examnng te reult along te flap we oberve tat te rougne at pont (ntrado greater tan at pont 1 and 4 (extrado. Agan, from te pont of vew of flud mecanc t may be jutfed. So, t well known tat te part of an arcraft (and not only n t geometry were urface jon, ncreae

9 Senor 13, turbulence. If we examne pont we notce tat located near te nterecton of te fuelage and wng, were bot urface jon, an angle form between tem. A a conequence a turbulent wnd regme mut appear and, by te ame reaonng aforementoned, te eroon on t regon mut be greater tan for oter flap area. In te ame way, comparng te pont 1 and 4 a grater rougne at 1 oberved, ndcatng tat te occurrence of arfol wngtp vortce a a greater nfluence n t part tan n te mddle of te flap. A pecal reult refer to pont 3 were te greatet rougne detected. In t cae, becaue of te erodng wnd acton, te area tuded dd not ave coatng, and uc te rougne mut be greater (1.4 m. T value confrm te above explanaton for emergng turbulence on regon were urface jon. To conclude t ecton t nteretng to oberve tat, due to te precon of te ASC tecnque (ee Secton and te value of te rougne obtaned, te metodology eem to be qute adequate for meaurng te problem tuded. 5. Concluon In t paper a tudy of te rougne on an actual unmanned arcraft wng preented. By ung te angular peckle correlaton tecnque (ASC, te degree of damage at dfferent mall area of te arfol meaured. Te tudy reveal a non-omogeneou reultant mecancal effect of te wnd, and oter factor, on te urface mcrotructure. A expected, becaue of te apparton of turbulence over te flap and near to te nterecton between te wng and te fuelage, te eroon greater over tee area tan on te leadng edge, were te arflow almot lamnar. Bede, a tudy of te uncertanty of te tecnque made, and a recommendaton for te bet expermental et-up gven. Acknowledgment Te autor w to expre ter grattude to INTA for t materal upport and epecally to te drector of LINES Belenguer for elpful dcuon. Conflct of Interet Te autor declare no conflct of nteret. Reference 1. Landau, L.D.; Lfctz, E.M. Flud Mecanc; Pergamon Pre: New York, NY, SA; 1966; pp Flack, K.A.; Scultz, M.P.; Roe, W.B. Te onet of rougne effect n te trantonally roug regme. Int. J. Heat Flud Flow 1, 35, van Rooj, R.P.J.O.M.; Tmmer, W.A. Rougne entvty conderaton for tck rotor blade arfol. J. Solar Energy Eng. 3, 15, Nanxn, R.; Jnpng, O. Dut Effect on te Performance of Wnd Turbne Arfol. J. Electromagn. Anal. Appl. 9, 1, Kang, S.-H.; Kang, Y.-S.; Han, K.-H. Numercal Study on Blade Rougne Effect on te Performance of Turbomacne. In Proceedng Internatonal Ga Turbne Congre, Tokyo, Japan, 7 November 3; TS-33.

10 Senor 13, Glo, M.; Muazz, S.; Pern,. Surface meaurement by mean of peckle wavelengt decorrelaton. Opt. Commun. 1979, 8, Ruffng, B. Applcaton of peckle-correlaton metod to urface-rougne meaurement: A teoretcal tudy. J. Opt. Soc. Am. A 1986, 3, Lemann, P.; Patzelt, S.; Scöne, A. Surface rougne meaurement by mean of polycromatc peckle elongaton. Appl. Opt. 1997, 36, Tay, C.J.; To, S.L.; Sang, H.M.; Zang, J. Wole-feld determnaton of urface rougne by peckle correlaton. Appl. Opt. 1995, 34, Yamaguc, I.; Kobaya, K.; Yarolavky, L. Meaurement of urface rougne by peckle correlaton. Opt. Eng. 4, 43, Ruffng, B. Non-Contactng Rougne Meaurement of Tecncal Surface by Speckle-Correlaton Metod. P.D. Dertaton, Inttut für Me- und Regelungtecnk,, nverty (TH Karlrue, Karlrue, Germany, Taylor, J.R. An Introducton to Error Analy; nverty Scence Book, Mll Valley: Saualto, CA, SA, 198; pp by te autor; lcenee MDPI, Bael, Swtzerland. T artcle an open acce artcle dtrbuted under te term and condton of te Creatve Common Attrbuton lcene (ttp://creatvecommon.org/lcene/by/3./.