https://ntrs.nasa.gov/search.jsp?R=19880017231 2018-11-19T05:11:19+00:00Z " IV .. ,I(, "'I\ !1-0/J T 11 774J{) NASA-TM-77480 19880017231 NASA TECHNICAL TM 77480 MEMO~ANDUM 'to ,.'t. EXPERIMENTAL INVESTIGATION OF THE TURBULENT FLOW IN SMOOTH ~'.'. ,I , J~',_ AND LONGITUDINAL GROOVED TUBES ,J P. Nitschke • I, • ""I "\) -~ • '4" ~ranslation of "Experimentelle Undersuchung der turbulenten Stroemung in glatten und laengsgerillten Rohren;'Max-Planck Institue fuer ... Stromungsforschung, Goettingen, West Germany, -1983 . L[./ .. _.. ... '\ r; r'"I/I',/ :;; ":":)11 t, 1/1 I ~. .... ,\.. ~I • /' . J LANGL[Y f<[S:c.'1f'C!f CEf,TER ., ,.. Lfl:f?MIY, ,',,\S,\ HAf'v1i'TUfJ, VIRG!NIA 'I. I "J1~ ,.. NATIONAL AERONAUTICS AND SPACE ADMINISTRATION WASHINGTON, D.C. 20546, May 1984 11111/1111111 1111 11111 1111111111 11111/1111111 NF00401 STANDARD TITLE PA-CE 12. J. R.porf No. Co .... nua.nf Acc .. aion No. ~ R.clpl.nl·,·Colo/og No • . ' NASA TM77480 4.. TII/ ....d ~.liC/. S. R.po;' 001. EXPERIMENTAL INVESTIGATION OF TURBULENT FLOW IN SMOOTH & LONGITUDINAL GROOVED •• p.,lonal.n g Orgcnizolion Cod. TURFC::: 7. Au~cr(" .. P • .coraJng Orgonlulion R.po .. Ho. P. Nitschke '1 JO. 1'0.1& Unit No. 9. P.rfoll,,:ng Or;enilcloon Hema end A,Ur ... 11. Contrac, or ercn' No. NASw 3541 Leo Kanner Assoc. Redwood City, CA. a. J3. T" . R.po,f ClId P .. iod Co ..... .1 Translation I 12. ~onlcrin; Ag.ncy Hem. and Add" .. ,. J(. ~onlOrln:l 1o.;lncy Cc~. I 15. 5.:;;I ..: anlcry Holu Experimentelle Untersuchung der turbulent Stroemung in glatten und laengsgertllten Rohren, Max-Planck Institut fuer Stromungsforschung, Goettingen, West Germany, 1983 !'If 16. J.~lt,a~. ... , See attached. ,r< , , 17. K.y l\·Ofd. (selected by AuthOr(S») 18. Dlalrlbulion 5101.cunl Unclassified - Unlimited 19. 51C1lnl, ClolOlf. (r:' .hl. ';$10." :20. S.CV1I1, C/clIl'. (0' Ihi. pog.) :11. Ho •• , Pag •• 22. qb Unclassified Unclassified ]J-15, bb f'J8 t.{- -/ ~: S-J 011-CD 'YC(C iA ii Ng)r-~Co{ S-1t ( , - II - ". " '1 Translation of title: Experimentat investigation of the turbutent ftow in smooth and tongitudina t grooved tubes Abstract: '. This report gives a short survey of an investigation of W. E. Reif. He discovered that shark-scates have fine grooves (characteristic breadth: ,.., o. 1 mm) making a sort of stream line pattern on the shark's body. As a result, both of this investigation, and of some other works on structures .... in a turbutent boundary tayer, in the Max-ptanck-Institut fUr Stromungs , forschung a modet has been devetoped, which teads to the expectation, that grooves on a surface reduce the momentum exchange, and thus the drag of this surface compared to that of a smooth surface. To test this thesis, drag taw, ve to city profile and the profile of the ve to city ftuctuation were determined for tubes both, with and without tongitudina t grooves. A comparison of these resu tts shows, that for moderate Reyno tds numbers (8000 ~ ReD <. 45000) the drag coefficient A measured for the grooved tubes is, about 30;0 smatter than that of the smooth tubes. At higher Reynotds 't numbers, however, the drag coefficient for grooved tubes becomes larger than that for smooth tubes. For the above range of moderate Reyno tds numbers the dimension tess breath of the g:-ooves b+ = b ~ lies between 8 and 30, thus having the same order of magnitude as that, catculated for the grooves on the shark-scates. No significant"differences in the vetocity profiles between grooved tubes and smooth tubes have been found. This, however, may be due to the comparitivety bad spatiat resotution of the '. " hot-wire used here. iii CONTENTS Page ~. List of Symbols 1. Introduction 1 " 2. Fundamental Cpnsiderations 2 2.1 General Information on the Flow Resistance of Rough Walls 2 2.2 Comments on the Grooves on the Scales of the Shark .2 2.3 Model Conceptions of the Influence of Grooves on Longitudinal Structures in Near-Wall Turbulent Flow 7 3. Test Set-up 8 , 3.1 Tube Channel 8 3.2 Tubes 11 3.3 Pressure Holes 13 3.4 Electronic Section 15 ~ 4. Flow Resistance of the Investigated Tubes 16 4.1 General Fundamentals on the Resistance Law 16 4.2 Measurement Methods and Accuracy 19 4.2.1 Measurement of the Flow-Through 19 4.2.2 Determination and Accuracy of the Average Pressure Drop 20 4.2.3 Measurement of Diameter and Its Definition for Grooved Tubes 28 4.2.4 Determination of the Density and Kinematic Viscosity 30 4.3 Measurement Accuracy for the Coefficient of Resistance A and the Reynolds Number Re 30 I <' 4.4 Results and Discussion 31 4.4.1 Smooth Tubes 31 4.4.2 Grooved Tubes 34 u 5. Profile of Average Velocity and of the Effective Value of the Velocity Fluctuations u' 40 rms 5.1 Introduction 40 5.2 Measurement Method and Accuracy 43 iv Page 5.3 Results and Discussion 47 5.3.1 Smooth Tubes 47 5.3.2 Grooved Tubes 56 6. Summary and Discussion of Future Work 66 '~ 7. References 67 , ," '~ '.' v List of Symbols '. ce Proportionality constant between mixing path and wall separation Coefficient of resistance A ',; Dynamic viscosity ;U Kinematic viscosity y Density S Shear Stress 'l: Wall shear stress ... 7: w Wall shear stress over the groove crest 'l:'B 't'T Wall shear stress over the groove valley ~ A Cross-sectional area between two grooves -of the grooved tube a Position on the micrometer screw b Spacing of groove crests Dimensionless spacing of groove crests b+=b·~\ ))' C Integration constant d Diameter of the tube f General function h Groove height, roughness height + u h = h· ~ Dimensionless groove height i Y i Continuous numbering of pressure holes .. ' 1 Various lengths n Exponent vi p Pressure (generally pressure difference compared to atmo~ph~ric pressure) PiA} I Pressure at drill hole i = 1 - 9 Pm I Side A or B, respectively I Atmospheric pressure Po '~ Over-pressure in the spin flowmeter (DDM) PDDM Q Flow-through q Stagnation pressure r Radius, r , r , r , r see figure 4/9 I H F A Re Reynolds number , u Velocity in the flow direction~ u u + u l u Average local velocity in the flow direction ," u U in the tube middle M '. u l ,Alternate component of u , _. , F.2' V I urms - u' Effective value of the alternate component in flow direction =~ U Q 2 Average velocity '1rr . ,[T;; VT u'L = I Wall velocity of shear stress + U u. = Dimensionless average local velocity urr Change in Resistance AW x Tube longitude coordinate in flow direction, x=O: Tube front y Wall separation coordinate, y = 0: Wall contact u'(' y+ = Y • ~ Dimensionless wall separation z Coordinate perpendicular to x and y, i.e. perpend~cular '.- to the flow direction and parallel to the wall + u't' z = z·- Dimensionless coordinate perpendicular to x and y 'II vii 1. Introduction . .' The reduction of flow resistance is one of the most important research tasks in the area of flow research. In addition to 'shaping objects exposed to the flow, the structure of the surface plays an important role in the occurring energy losses. For this reason, numerous investigations have been conducted on the flow resistance of smooth and rough surfaces (e.g. Nikuradse 1933, Hopf 1923, " Fromm 1923, Schlichting 1936, Wieghardt 1943). These investigations show that the smooth wall has the smallest flow resistance, compared to all investigated roughnesses. Proceeding from these results it is assumed that the smooth wall in general offers the lowest flow resistance compared to any other surface structure. Various observations do indicate however, that this assumption should be re-examined. Investigations by Reif showed for example, that the scales of sharks are covered with fine, longitudinal grooves. The scales overall produce a type of flow-line configuration on the fish (Reif 1978). According to model conceptions developed by Dinkelacker (Reif and Dinkelacker 1981), these grooves can affect the flow so much that longitudinal structures are generated which .. are smaller than those observed on smooth walls. This diminution could reduce the pulse exchange near the wall and thus reduce the friction resistance. ~ It is the goal of this paper to investigate whether the resis tance and the effective value of the velocity fluctuations change due to a longitudinal-grooved surface and whether this change exhibits any dependence on the Reynolds number and on the groove width b+ = b ~ made dimensionless by the flow quantities ~ and v. The mechanism by which the grooves exert their potential influence will be of secondary importance. As flow form, the flow in a tube was chosen, due to the simple geometry and the easily measured flow resistance. In the first part of the inves tigation, the resistance law will be compared on smooth tubes and tubes with longitudinal grooves. In the second part of the inves ~igation, velocity profiles and profiles of the effective value of the velocity fluctuations in the flow direction will be compared for smooth and grooved tubes and the results will be related to other measurements. 1 2. Fundamental, 'Considerations 2.1 General Information on the Flow Resistance on Rough Walls There are numerous investigations on the friction losses from rough walls; these cover a broad range of Reynolds numbers and were conducted on very different roughnesses, like sand, geometric elements in regular and irregular arrangement, wall holes etc. (Nikuradse '" 1933, Schlichting 1936, Wieghardt 1943). Nikuradse investigated the sand roughness in great detail;'its tube walls in this case were packed as tightly as possible with sand grains of the same, defined size. According to these measurements and according to discussions by Prandtl (Schlichting 1965, p. 573),there are three I different regions in which the friction resistance can be variously affected. The decisive parameter is the roughness height h+ = h· ~ made dimensirjifss by the characteristic length of the viscous Y I· under-layer u If h+ is less than 5, then the roughness has no measurable ii"u."I1uence on the resistanc~, since the roughness dis appears entirely in the viscous under-layer which was formerly viewed as a laminar under-layer. If h+ lies between 5 and 70, then the resistance is also dependent on the Reynolds number and on h+. In the third region, h+ > 70, the influence of h+ exceeds that of the ,- Reynolds number, so that the resistance depends only on h+ and is very much greater than that of the smooth wall. According to .. measurements by Schlichting (1936), a random roughness, e.g . less densely packed sand grains or roughnesses in the shape of spheres, hemispheres, cones etc., can be assigned an equivalent sand rough- ness and applied in this m~nner to the relationships found by Nikuradse. All these roughnesses have in common that they increase the resistance of the overflowed surface essentially by means of two mechanisms (Prandtl 1969, p. 191, Rotta 1972, p. 159). Due to the resistance of the shape, an additional pressure loss occurs in the flow. At the enlarged surface, due to additional, viscous friction, the near-Iw all velocity profile and thus the wall shear stress 't w =)l dd u ,changes. In general this thesis states that a smooth Y y=OI wall exposes the flow to the smallest resistance. Investigations in biology and new findings on coharent structures in turbulent flow provide an impetus to rethink this thesis. 2.2 Comments on the Grooves on the Scales of the Shark According to observations by Reif (Reif 1978), scales of several very-fast swimming sharks are covered with longitudinal grooves. The photos 2/1 to 2/7 show several examples which were also prepared by Reif (personal report). Figure 2/1 presents a larger section of the skin surface of a hammerhead shark. The grooves running uniformly over the skin are clearly seen on the indistinct scales. In the middle of the photo there is pit-organ, probably a sensory organ, whose function is still unexplained. 2 It is striking that the grooves run toward this pit. Figure 2/8 shows a shark with the run of the grooves represented by arrows. The dotted surfaces indicate flat scales without grooves. The overall impression resembles the flow line picture of a flow about the body of the shark. Figure 2/2 shows the scales of the velvet shark (highly enlarged). Again, in the middle of the photo we see the pit organ toward which the grooves run. In contrast to this, the grooves in fig. 2/3 run around the collateral line organ which approximately corresponds to ·" the human ear. "\ Figure 2/4 shows the skin in the tail region near the stagna tion edge. We clearly see the very minutely formed grooves here. As we see in figure 2/5, the scales of embryos are also pro vided with grooves. There seems to be a close relationship between the formation of the grooves and the flow: -shape of the flow line of the groove pattern -convergence of grooves to the pit organ, divergence of grooves before a collateral line organ -low groove depth near the stagnation edges, e.g. front edge of fins where the boundary layer is still quite thin -unchanged grooves in the course of development of the shark to . (see also Reif, Dinkelacker 1981). Figures 2/6 and 2/7 indicate the shape of the grooves: Narrow, peaked "crests" and definitely round "valleys." Measurements by Reif showed a distance b of groove crests of between 40 and 105 pm, an average value of 60 ~m and a ratio of groove height h to groove separation b of 1:2 to 1:3. In. ... order to be able to norm these quantities with the characteristic length of the viscous underlayer !~ i U-z: several assumptions must be made since the velocity of wall shear stress Uy for the shark is not known. In an initial estimation, the flow around the shark can be compared with the conqitions of a plate exposed to a flow at a speed uo:>' so that we have: Iu ~ 0,04 U 'l: 00 Wi~h a temperature of T = 20 °c and a kinematic viscosity of V = 10- m2/s the values given in table 2/1 result for the dimension less groove width b+ = b .-~ and. the dimensionless groove height v- u h + = h· t' for three streaming velocities between 5 m/ sand 20 m/ s . V 3