Prof. Dr.-Ing. Joseph H. Spurk TH Darmstadt Institut fiir Technische Strémungslehre Petersenstrafe 30 D - 64287 Darmstadt / Germany Translation: Professor Taher Schobeiri Texas A & M University, Dept. of Mechanical Engineering College Station, 77843 31 2 Texas/USA ISBN 3-540-61652-7 Springer-Verlag Berlin Heidelberg New York Die Deutsche Bibliothek - CIP-Einheitsaufnahme Spurk, Joseph H.: Fluid mechanics : problems and solutions / Joseph H. Spurk. With the assistance of H. Marschall. (Trans!.: Taher Schobeiri). -Berlin; Heidelberg; New York; Barcelona; Budapest; Hong Kong; London; Milan; Paris; Santa Clara; Singapur, Tokyo: Springer, 1997 Dt. Ausg. uw. d. T.: Spark, Joseph H.: Stromungslehre ISBN 3-540-61652-7 CIP data applied for This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in other ways, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer-Verlag. Violations are liable for prosecution act under German Copyright Law. © Springer-Verlag Berlin Heidelberg 1997 Printed in Germany ‘The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Product liability: The publisher cannot guarantee the accuracy of any information about dosage and application contained in this book. In every individual case the user must check such information by consulting the relevant literature. ‘Typesetting: Camera-ready by author SPIN; 10749892 60/3012 -5 43 2 1- Printed on acid-free paperContents 1_The Concept of Continuum and Kinematics 1 1L.2_ Kinematics ‘ oss 1 Problem 1.2-1 Calculation of material coordinates for giv- en pathlines ....... 1.1. 1 Problem 1,2-2 Velocity and acceleration in material and spatial coordinates with given pathlines . 2 Problem 1.2-3 | Material description of a potential vortex Problem 1.2-4 flow... Material description of an axisymmetric stag- nation point flow. ............- 7 Problem 1.2-5 _Pathlines, streamlines, and streaklines of an unsteady flow field ........... 9 =| i i i i ive! = gence free flow field ............ 14 Problem 1.2-7__ Kinematics of an unsteady, plane stagna- tion point flow ...........-.-.- 19 Problem 1.2-8 Streakline of a water jet... - 23 Problem 1.2-9 Problem 1.2-10 Streamlines and Streaklines in cylindrical coordinates ee 2 se 2 2 Streamlines and pathlines of standing grav- Problem 1.2-11 ity waves 2. ee ee ee 29 Change of material line elements in a Couette- Problem 1.2-12 flow Change of material line elements in a three- Problem 1.2-13 dimensional flow... ....... 2... 34 Angular velocity vector and the change of material li ~ flow field .................. 38x Contens Problem 1.2-14 Rate of deformation and spin tensors of an unsteady two-dimensional flow... ... 43 Problem 1.2-15 Time change of the kinetic energy of a fluid body ....-.........,...... 45 2 Fundamental Laws of Continuum Mechanics 50 2.1 Conservation of Mass, Equation of Continuity ....... 50 Problem 2.1-1 One-dimensional unsteady flow with given densityfield ................ 50 Problem 2.1-2 Plane, steady flow with a given density field 52 Problem 2.1-3 Velocity at the exit of a container... .. 54 Problem 2.1-4 Steady flow through a circular channel... 56 Problem 2.1-5 Squeeze film flow. ............. 58 Problem 2.1-6 Moving Piston ............... 60 Problem 2.1-7 Flow between two inclined flat plates... 63 Problem 2.1-8 Oscillating journal bearing ........ 65 Problem 2.1-9 Effect of boundary layer displacement thick- “MeS$ Problem 2.1-10 Flow through a diffuser with a linear ve- locity change in flow direction... 0... 7 Problem 2.1-11 Termperature boundary aver along a cold wall. sa -....... 73 Problem 21-12 Flow in a : lubrication gap. eee eee 74 2.2 Balance of Momentum ............... i. 78 Problem 2.2-1__ Principal axes of a stress tensor... .. . 78 Problem 2.2-2 Fluid forces on a manifold... ....... 80 Problem 2.23 Calculation of af rag. fo force re 82 Problem 2.3-1 Torque on pipe with slot. ........ 87 Problem 2.3-2__| Moment exerted on the inlet_guide vanes of a water turbine |... 0.0.0... 90 Problem 2.3-3_ Curvature radius of circular arc profiles of acircular cascade . 4... ...... 93 2.4 Momentum and Angular Momentum in an Accelerating Frame_96 Problem 2.4-1 Fluid sprayed on a rotating disk... .. 96 Problem 2.4-2 Velocity of a moving container with a nozzle100 Problem 24-3 Acceleration and velocity of a rocket... 107 Thrust reversal... 9 Problem 2.4-4 Problem 2.4-5 Torque on a rotating bent pipe ..... . lu Problem 24-6 Thrust of ajet engine ........... 114 Urheberrechilich geschiitztes MaterialContens XI 2.5 Applications to Turbomachines ............... 117 Problem 2.5-1__ Circulation around a blade profile in a cir- Problem 2.5-2 Axial turbine stage... 1.2... 0... 119 Problem 2.5-3 Kay turbine... we ee 121 Problem 2.5-4 Torque converter... ........... 125 Problem 2.5-5 Balancing of axial thrust... 1... 129 2.6 Conservation of Energy. ........-+-.++..-0- 131 Problem 2.6-1__ Cylinder with heat flux ........ 4. 131 Problem 2.6-2__ Energy balance in an axial turbine stage . 134 3 Constitutive equations 138 Problem 3-1 Velocity of araft.............. 138 Problem 3-2___ Energy balance in a journal bearing ... 140 Problem 3-3 Pressure driven flow of paper pulp .... 144 Problem 3-4 Flaw of WN. ian fluid 146 Problem 3-5 __Extensional flow... 2» 148 4 Equation of Motion for Particular Fluids 152 4.1 Newtonian Fluids. ............... wo... 152 Problem 4.1-2__ Temperature distribution in a Poiseuille flow156 Problem 4.1-3_ Pressure driven flow in a channel with porous walls 000 Problem 4.1-5 Mixing of streams of fluids ........ 165 Problem 4.1-6 Drag on a flat plate. 168 Problem 4.1-7 Tue. -Simensional water jet jeupining ¢ ona ee ee 173 Problem 4.1-8 ia body rotation and potential vortex 175 Problem 4.1-9 Energy balance in a potential vortex flow 180 4.2 Inviscid flow. .................. 184 Problem 4.2-1 " Pressure and energy increase of fluid in a centrifugal pump... 1... 1... 184 Problem 4.2-2 Pressure distribution within a spiral casing 189 Problem 4.2-6 ‘Two infinitely long vortex filaments near a wall... 1... ............. 194 Urheberrechilich geschiitztes MaterialXII Contens Problem 4.2-7 | Wing with an elliptic spanwise distribution of circulation... ........4 .. 198 Problem 4.2-8 Airfoil in parallel flow... ......... 201 Problem 4 42-9 Jet anele ina Bete diffuser wee eee 204 jsymmetric flow... ...+...0 5 208 Problem 4.2-12 Increase of static pressure in a Betz diffuser 210 Problem 4.2-13 Fluid flowing out ofatank ........ 212 Problem 4.2-14_ Air bubble moving in a channel... . . . 214 Problem 4.2-15 Aircraft above the ground... ..... 217 Problem 4.2-16 Flow between two rotating cylinders, cir- culation and vorticity ........... 223 Problem 4.2-17 Power of a Pelton turbine... .... .. 225 4.3 Initial and Boundary Conditions... .. 1... 2-1... 230 Problem 4.3-1 Oscillation of an elliptic cylinder in fluid . 230 Problem 4.3-2 Flat plate with a pitching and oscillating Problem 4.3-3 Rotating cylinder moving through fluid _. 232 Problem 4.3-4_ Vortical flow inside an elliptic cylinder. . 234 5 Hydrostatics 236 5.1 Hydrostatic Pressure Distribution .............- 236 Problem §.1-] U-tube manometer. 28 Problem 5.1-2 | Hydraulic safety clutch .....-...- 237 Problem 5.1-3 Rotating container filled with fluid... . 239 Problem 5.1-4 Centrifugal casting process ........ 241 Problem §.1-5 Depth gauge ........+.....0- 242 5.2 Hydrostatic Lift, Force on Walls........... oe 244 Problem 5.2-1 Force and moment on a throttle valve 244, Problem 5.2-2 Half sphere closing an orifice... ... - 246 Problem §.2-3 _Foree on adam... 1... 248 Problem 5.2-4 Half sphere cup sealing by its own weight 250 Problem 5.2-5 Cylindrical submarine... .. 1... . 252 Problem 5.2-6 Car under water... 4... ....... 254 6 Laminar Unidirectional Fl O57 Problem 6-1 Flow in an annular gap .........- 257 Problem 6-2 Crude oil transport through pipeline... 261 Problem 6-3 Oscillating pipe flow...........- 264 Urheberrechilich geschiitztes MaterialContens XI Problem 6-4 Comparison of a Couette-Poiseuille flow of a Newtonian fluid, a Stokes fluid, and a Bingham material ............- 267 Problem 7-2 Velocity distribution in turbulent Couette flow with given Reynolds number... . . 277 Problem 7-3 Turbulent pipe flow ............ 278 Problem 7-4 Crystal growth on pipe walls... .. . 280 Problem 7-5 Comparison of momentum and energy flux in laminar and turbulent flowinapipe . 282 Problem 7-6 Velocity distribution in a turbulent pipe flow resulting from the Blasius friction law 285 Problem 7-7 Location of a pipe leakage... . . . . . - 287 Problem 7-8 Cooling of superheated steam by water in- jection... eee ee 289 8 Hydrodynamic Lubrication 293 Problem 8-1 Bearing with step slider... 1... 293 Problem 8-2 Friction torque transmitted by the shaft to the journal Problem 8-3 Slider load in squeeze flow: Comparison between different slider geometries .... 299 9 Stream filament theory 302 9.1 Incompressible Flow... ees 302 Problem 9.1-1__ Rotating tube acting as pump ...... 302 Problem 9.1-2 Volume flux through an orifice . 2... . . 305 Problem 9.1-3 Injectorpump ............... 306 Problem 9.1-4 Radialpump..............-. 308 Problem 9.1-5 _Bulb turbine ................ 312 Problem 9.1-6 Coanda effect. ............... 315 Problem 9.1-7 Principle of a shaped charge ....... 316 Problem 9.1-8 __Penstock and nozzle of a Pelton turbine _ 319 Problem 9.1-9 Operating characteristic of afan .... . 321 Problem 9.1-10 Water power plant ............. 325 Problem 9.1-11 Flow through an exhaust gas analyser .. 328 Problem 9.1-12 Flow deflection through a screen... . . 329 Problem 9.1-13 Hovercraft 2.0.0... ce eee 331 Problem 9.1-14 Wind turbine... 1... 00.0.0 004 333XIV Contens Problem 9.1-15 Discharge pipe of a reservoir: Comparison between different pipe geometries .... 337 Problem 9.1-16 Vibrating system consisting of a fluid col- umn and a spring suspended piston... 339 Problem 9.1-17 Unsteady flow in a tube with flexible walls 343 Problem 9.1-18 Plungerpump ............... 346 Problem 9.1-19 Flow within an urethra prothesis .... . 350 9.2 Steady Compressible Flow ..................- 352 Problem 9.2-1 Force on a plate in subsonic flow... .. 352 Problem 9.2-3 Normal shocks in an inlet guide vane... 358 Problem 9.2-4 Blunt body in supersonic flow... ... . 363 Problem 9.2-5 Shock waves in the divergent part of a Laval noaile.................... Hf8 Problem 9.2-6 Supersonic nozzle in a spinneret ..... 367 Problem 9.2-7 Ram jet in subsonic flow ....-.... 370 Problem 9.2-8 High speed train inatunnel ....... 373 Problem 9.2-9 Labyrinth seal of aturbomachine.... . 376 Problem 9,2-10 Gas flow through an orifice ........ 379 9.3 Unsteady Compressible Flow... .. 1... 381 Problem 9.3-1 Traveling normal shock in a pipe... . . 381 Problem 9.3-2 Shock tube 383 Problem 9.3-3 Motion of a pistoninatube ....... 386 Problem 9.3-4 Reflection of a normal shock wave at the openendofatube............. 389 Problem 9.3-5 Principle of an expansion tube... . . . 392 Problem 9.3-6 Propagation of acoustic waves in a closed 10 Potential Flow 399 10,3 Incompressible Potential Flow... 1. ee ee ee 399 Problem 10.3-1_ Expanding sphere... . 1... 1... 399 Problem 10.3-2_ Sphere in a translational flow ....... 402 Problem 10.3-3 Flow near the stagnation point of a body in parallel flow .............-. 406 Problem 10,3-4 Point source in a rotationally symmetric stagnation point flow .... ee 409 Problem 10.3-§ Point source above an impermeable wall . 412 Problem 10.3-6 Source distribution in parallel flow .... 414 Problem 10.3-7 Expanding sphere in an inviscid and in a viscous flow... 0. ee ee ee 416Contens xV Problem 10.3-8 Growth of a vapor filled cavity ... .. . 420 Probl 03.9 0. : flicient f iru ifice 493 Problem 10,3-10 Sphere rising in water... 1... 427 Problem 10.3-11 Unsteady motion of a cylinder perpendic- ular to its axis 6... ee ee ee 430 Problem 10,3-12 Rotor oscillating in an inviscid fluid... 432 10,4 Plane Potential Flow ........... se . 436 Problem 10,4-1 Flow in the squeeze gap between a moving piston anda wall... . 2.2... 2.0040, 436 Problem 10.4-2 Sink distribution in a stagnation point flow 439 Problem 10.4-3 Circle theorem... ............ 442 Problem 10.4-4 Half cylinder in stagnation point flow... 447 Problem 10.4-5 Dipol flow around a circular cylinder... 451 Problem 10.4-6 Flow around a thin plate ........- 454 Problem 104-7__Airfoil fixed wall BT Problem 10.4-8 Semi infinite body inachannel ..... - 461 Prot 10.4-9 Karman’s vortex street 464 Problem 10.4-10 Joukowski mapping of a circular cylinder in auniform flow... ........000- 467 Problem 10.4-13 Schwarz-Christoffel transformation of a con- vergent channel... ....-.......% 476 Problem 10.4-14 Cavitation ina channel .......... 480 Problem 10.4-15 Representation of a slender body by a source fistributi 83 Problem 10.4-16 Distribution of vortex intensity and mean camber line of a slender airfoil . 488 Problem 10,4-17 Straight cascade... ee 492 Problem 10.4-18 Vortex distribution of a flat-plate cascade 497 Problem 10.4-19 Compressible flow over a wavy wall... . 503 11 Supersonic Flow 509 11.1 Oblique Shock Waves... ee 509 Problem 11.1-1 Wedge with a thin plate in front of it. . 509 Problem 11.1-2 Inlet of a plane channel .......... 511 11.3 Reflection of Oblique Shock Waves ............- 514 Problem 11.3-1_ Flow over a wedge in a supersonic wind funnel. ...........,....... S14 Problem 11.3-2_ Supersonic flow in a convergent channel . 516XVI Contens 11.5 Prandtl-Meyer Flow . 2.1... 0. cee eee 518 Problem 11.5-1 Centered expansion wave in a divergent channel»... 2... ee ee 518 11.6 Shock Expansion Theory. .................. 522 Problem 11.6-1 Airfoil in supersonic flow ......... 522 Problem 11.6-2 Inlet of a supersonic jet engine ..... . 526 12 Boundary Layer Theory 530 Problem 12-1 Boundary layer momentum equation... 530 Problem 12-2 Flow over a wedge... 2... 0.000. 533 Problem 12-3 Diffuser with discontinuous change of the Problem 12-4 Drag coefficient of a diamond airfoil... 543 A Tensor calculus 551 Problem A-Lo0 ce eee eee 551 Problem A-2 6. ee 551 Problem A-3 pe Problem A- wee, . pee .. 553 Problem A-6 555 Problem A-B we ee 556 Problem A-9 000 BT Problem A-10 558 Problem A-lL we ee 560 B Examination problems 562 Problem B-1 Streamlines and pathlines......... 562 Problem B-2_ Drag of a half cylinder shell... ... . 563 Problem B-3 Awninginastorm............. 564 Problem B-4 Stretching of a foil. ............ 565 Problem B-5 Single stage, axial blower... 1... 566 Problem B-6 Blade profile for given pressure distribution 567 Problem B-7 Combustion chamber of a piston engine . 568 Problem B-8 Two-dimensional oblique stagnation point flow... 5d Problem B-9 Generalized Hagen-Poiseuille flow... . 570 Problem B-10 Induced velocity of a horse-shoe vortex. 571 Problem B-11 Open channel flow through a weir... . 572 Problem B-12 Safety valve... ee. 573Contens XVII Problem B-13_ Liquid in container... 2... 0.0.00, S74 Problem B-14 Sluice gate ©. 2... ee eee 575 Problem B-15 Pressure driven flow in the radial gap be- tween two concentric ring plates .... . 576 Problem B-16 — Pressure driven channel flow with variable viscosity . 2... eee ee ee SIT Problem B-17__ Temperature induced flow. ........ 578 Problem B-18 Shock absorber... ............. 579 Problem B-19 Frequency of a Helmholtz resonator ... 580 Problem B-20 | Chamber and exhaust pipe of an internal combustion engine... wee ee ees 582 Problem B-21 | Pump-turbine storage plant... .. 1... 583 Problem B-22 Overexpanded Laval nozzle ........ 584 Problem B-23 Nozzle inlet. ................ 585 Problem B-24 Solid propellant rocket engine... ... . 586 Problem B-25 | Ramjet. ...........-...... 587 Problem B-26 Ludwieg-tube. . oo. ee ee 588 Problem B-27 Dipol above an impermeable wall... . . 589 Problem B-28 Virtual mass of a thin plate. ...... . 591 Problem B-29 Removal of liquid through a plane channel 592 Problem B-30 Unsteady flow over a wavy wall... ... 593 Problem B-31 Wing section for given source and vortex distribution. 60... ee 594 Problem B-32__ Infinitely thin plate with aileron ..... 595 Problem B-33 Supersonic inlet... ........... 596 Problem B-34__ Infinitely thin, flat plate in two-dimensional supersonic ow ..............-. 598 Problem B-35 Guide vane cascade of a supersonic com- pressor... ee ee ee ee ee 599 Problem B-36__ Boundary layer ona foil. ......... 6001 The Concept of Continuum and Kinematics 1.2 Kinematics Problem 1.2-1 Calculation of material coordinates for given pathlines The material description of a flow is given by the motion ™ = &, m = k&?+&, t = & with & as a constant having a dimension, such that the dimensional in- tegrity of both sides of the above system of equations is preserved. Show that the Jacobian determinant J = det(x;/@€;) does not vanish and obtain the inverse f= &(z,t). Solution We obtain the necessary derivatives and insert them into the Jacobian determinant: a, OG B& 1 2ké2? 0 Ox, Or, Oxy J=det| —- = = | =det] 0 1 O]=1. | G& O& a | dx, Ar, Oxy 0 0 1 OE; Bes BE2 1 The Concept of Continuum and Kinematics Since the Jacobian determinant does not vanish, the mappings == 7(€,t) and € = €(Z,t) are unique inverses of each other. We obtain: & = %, fg = t9- kx? ? : fs = a3. At the time t = 0, & = 2;. Problem 1.2-2 Velocity and acceleration in material and spatial coordinates with given pathlines The fluid motion is described by: my = &, (2) m= 5 (ts t+&)e" + ; (&-&)e™, (2) ty = 5(G +e — ; (@-&)e™. (3) a) Show that the Jacobian determinant does not vanish, b) Determine the velocity and acceleration components 1) in material coordinates u;(€;,¢), bi(€;, 4), 2) in spatial coordinates u,(z;,¢), b;(z;,t). Solution a) The Jacobian determinant is: Bay Oe, Guy 06 08 Ob 1 0 0 det re z ze =det} 0 coshat sinhat | =1, Oxy Oxy Org 0 sinhaé coshat Obs Oks AEs thus different from zero.1.2 Kinematics 3 b) Velocity and acceleration components: 1) The velocity components in material coordinates are calculated from: On; u(G 4) = (3), ’ and thus u = 0, (4) um = F(a+b)e"-F(@-&)e“, (5) us = Fet&)e+ 5 (G-b)e™. (6) Correspondingly, the acceleration components are - 2 on. written out bh = 0, (7) a at a —at b = Pathe" + 7 (a—Se™, (8) a? a? b = Slathe"-S(@-Ge™. (9) 2) We obtain the spatial description by extracting the material coor- dinates €; = €;(z;,¢) from equations (1) to (3) and insert them into uz = uz(é; , t): uy = ui(Ej(te, 0), t) = wi(ze, t) . from (1) > &=a1, (10) from (2)+ (3) => (62 + £3) e™! = +23, (11) from (2)-(3) + (@—-G)et=x-25. (12) It is not necessary to solve for £2 and & because u,(€;, 2) in equations (4), (5), and (6) contain & and & only in combined form as in (11)1 The Concept of Continuum and Kinematics and (12). As a result, the velocity field is described by: wm = 0, (13) u = $ (22+ 23) ~ 5 (a2~ 2s) = ars, (14) a a ug = g (ta + 23) + 5 (22 — 23) =azr. (15) Similarly, the acceleration field b;(z, ,¢) can be calculated from (7) — (9) and (10) - (12) and is: & = 0, a a? bh = y (t2+29) + F(e2— 2s) =a? a2, a’ a by = y (#2 +23) — 7 (e243) = aay. As an alternative, the substantial derivatives b,(z;,,t) = Du;/Dt can be applied, where the acceleration components are calculated from Du; du; ou; b= ate ous ‘= De Ge +" Se, as follows: bh o= Fe a St uy St wy SH =, bb = = set see Set salsa, bn Get ae te tae1.2 Kinematics 5 Problem 1.2-3 Material description of a potential vortex flow The motion of a fluid is given by the material description )¥? cos arclan & (+7 cog torn (2), = a = (¢ + ey? sin [ata + arctan (f )] wm = &3. a) Find the equation of the pathline in an implicit form and show that for £ at time t = 0 xz) = +6, and zz = +& holds. b) Calculate the components of the velocity u;(£;,¢) and the acceleration 5;(€;,t). c) Determine the velocity field uj(z,,#) and the acceleration field bi(xp, t). d) Explain the equation of the streamline through the point (x19, 20). Solution a) The pathlines are in the plane z3 = £3, We obtain their implicit form by squaring and adding the equations for 2, and 2 +a =G+G. (1) The fluid particles £ = const. describe circles around the z3-axis in 2, 2- plane. Dividing the equations for z2 and x, at time t = 0 yields 28 (2 1 ~&0 We present (1) in the form 2 2 #142) -e (9) 2 ei _ gt & aleed) alg). and find, using (2), #1) = +é, and ry = +&. and6 1 The Concept of Continuum and Kinematics b) The material description of the velocity and the acceleration: The desired partial derivatives are w= (28) 2 8 gf mt ran (& \a), ~ Grae" ares w= (Gr), = lprateoe (2) at & (+r aR G+ & v=), = ! o b = (=) = ee cos ts arctan (@)] , ty}, (G+ Gy? LG +G &/}' n= (F = © i t 2) ~ at 6 ~ “(eye arg tein (2), c) Velocity and acceleration in spatial coordinates: To obtain the velocity components in spatial form w;(2,,1), we replace the material coordinates in u,(€;,t) by &; = €)(2«,t). For the sake of simplicity, we use (1) and the relations following from the material description tT _ fae _ “n larg tent (®)| = Gray cos | a + arctan (2)] = a G+ & (@ + G)? The insertion leads to Nr Qe, =-— = =0. m1 zi +23’ zi +23" ws Similar procedure is applied for calculating the acceleration components y h=- b= - by =0. (ai + 23)?" (ai +29)?’1.2 Kinematics 7 Using the substantial derivative b; = Du;/Dt, the same acceleration components can be obtained. d) Streamline equation: The velocity field is steady, that means the streamline and pathline fall on the same curves. The streamline through the point x10, £29 is 2 2_ 2 2 Ty + Ty = Tig + Fp - Problem 1.2-4 Material description of an axisymmetric stagnation point flow The motion of a fluid is described in material coordinates by: q ™% = fe", st m = fem, *s wy = &o% with given a = const and € = z(t = 0). a) Calculate the velocity and acceleration components u;(£;,£) and 6;(£;,t) in material coordinates. b) Determine the spatial description of the velocity and acceleration com- ponents u,(z,,t) and b;(a,g,¢t) by eliminating the material coordinates €; = &(zx,t) in the results obtained in a). c) Find the acceleration components using the substantial derivatives of u;(zx,t). d) Is this a potential flow? If yes, find the potential function. Solution a) The material description of velocity and acceleration is determined us- ing: da; Pa; u(t) = (#) » b(t) = (=) . & & As a result, the velocity components are: w=ahke", uw=abe’, us=—2abse™,b} 1 The Concept of Continuum and Kinematics while the acceleration components take the form: bb =a? fe”, by =a? fze™, by = 4a? E307. The spatial description of velocity and acceleration: From fame", &=me™, fy = rge7 , we find from a) W=ary, Ug = azz, ug = —2ar3, and bh=a'a, b=a2, by =4a' ay. The acceleration components are the substantial derivatives of the ve- locity components u,(z;,t): Using _ Du; _ Ou; ; Ou; “De Ot Mu; Oz; and uj =a2,, uz = a2, ug = —2az3 we obtain the acceleration components as bh =a’ x, by =a? 22, by = 4a x3. Potential flow, Potential function: The necessary and sufficient condition for the existence of a potential flow is the vanishing of the vorticity vector curli in the entire flow field: Ou; curl? = V xi=0 ° Gjk po =O. Oz; The three components resulting from this condition are: dug ug _ Ou, uy _ Oug Ou Or. Oxy’ Oc, Oa, ° Or, Oe, 0. The above results show that the flow under investigation is a potential flow. In fact, all six terms of the above equations here identically vanish. From an u=— . Ox; we find for the potential function ® the partial differential equations: ae an ob ~— =u, =a, =Ug=a22, = ug = —2az3. ae, a, zs1.2 Kinematics 9 Integrating the first differential equation yields $= zy + h(z2, ra), ble the second equation then gives ab Oh * a a nan, a Ot, «Ory 5 23 +g(zs). = — Mtaas) = The arbitrary function g(<3) is determined by the last differential equa- tion: — =~ =-2az3 => g{a3) = -a zi + const . The potential function can be expressed as b= 5 (Ft + 73 — 223) + const where the constant may be omitted. Problem 1.2-5 Pathlines, streamlines, and streaklines of an unsteady flow field Given is the following unsteady velocity field: ‘1 um = zr 1 to +t Ly Uz = Vo, uz = 0 (to = const, vo = const). a) Find the equation of the streamline through the point (19, 20, t20) at time t. b) Find the pathline equation of a fluid particle with the material coordi- nate Z(t =0) = €. c) Determine the particle velocity along its pathline. d) What happens to the fluid particles with the material coordinates &; = 0,&=07 e) Find the equation for the streaklines.10 1 The Concept of Continuum and Kinematics Solution a) The streamline through the point (19, t29, £39) at an arbitrarily fixed time t: The streamlines are the solutions of the differential equations di ui ds fugu, * Instead of using the arc length s, it is appropriate to introduce the parameter 7 ds = ,fupu, dy , n(s =0)=0, and to rearrange the differential equation to dz; —— =u;(z;,t) , t=const, dy The components are: da, ry dz dry dy ott ay Integrating the above differential equations results in: Int, = rag tines > a =Cyeor | Zz = uy+ Co, zz = C3. The streamline is in the plane r3 = Cs, thus the field is plane. The above three integration constants are determined using the condition that the streamline goes through the point (19, t20, Tao). Measuring 7 from the point (219, T29, Z39) we have n=0: a =2, C2=%99, Te= Fa. Thus, the constants read: Ch=2, Cr=29, Cs3=220. The parametric solution is obtained as Tp = zype/lor4 , (1) Ze von + 220, (2) ty = 29. (3)1.2 Kinematics 11 In (1) to (3), 7 is the curve parameter and 219, 229, and Zao are the family parameters. The explicit representation of the streamline in the plane z3 = 239 can be found by eliminating the curve parameter. From (2), we obtain _ t2— £20 Uo and then a aapexp( 2222™) gp Zh oexp {—ta/tmn! ' 10 xP *P vo to(1 + t/to)/r20 , vo (to + t) For the value vo to/ #20 = 2, the fig- ure on the right shows the stream- line through the point (x19, 740, £30) at time ¢ = 0 and tf = 0.5%. Thus, we are dealing with an unsteady flow field with time dependent streamlines. b) The pathline of a fluid particle: The pathline differential equations are dz; ye 7 (et) in an extended form, they read dz, _ a drz _ dr3 _ 0. ditt’ dt ae The integration furnishes H=Cy(lott), rg=ut+C,, r3=C. The above integration constants are determined from the initial condi- tions. At time ¢ = 0, the fluid particle has the materia] coordinates t=0: m=, m=&, m=h&, resulting in Q=&/t, Gi=h, Gr=&.12 = The Concept of Continuum and Kinematics The parametric solution is a = (i+), (a) m = wtth, (5) tz = &. (6) In the above equations ¢ is the curve parameter of the pathline; &, &, and are the family parameters. As shown previously, the explicit form of the pathline equation is obtained by eliminating the curve parameter i in the plane z3 = &. We obtain from (5): in BZ & Ug and with (4) a6 (5) The above explicit form of the pathlines is a straight line for each fluid particle. An arbitrary fluid particle with the material coordinates £1 = 210, 2 = 220, £3 = £30 has the pathline Ty 7/29 —1 ; (7) Zo v9 to/ £20 This pathline has to be tangent to the streamline at time t = 0 and the space point (x19, t29, T30). The pathline of the particle is plotted in the figure as a solid line. The streamline at time ¢ = 0 is shown as a dashed line. -1 ° 1 2 3 4 6 ¥2/Xeq c) The velocity of a fluid particle is defined as the temporal change of the path coordinates at fixed € Ox; (Est) = (=) &1.2 Kinematics 13 and from the parametric representation of the pathline follow the ve- locity components ui (Ej, t) = 2 » Ua(ezt) =e, ual€;,t) = Since the velocity components are constant, the pathline is again seen to be a straight line. d) For a fluid particle with the material coordinates £, = 0 and £5 = 0 we obtain the velocity components wi(é,t)=0, urléj,t)=v0, us(&,t)=90, and the pathline t nah (+5) =0, 2=0. The above equations indicate that the fluid particle moves along the @2-axis at constant speed. e) Streaklines: A streakline at a fixed time ¢ is the connecting line or the locus of dif- ferent fluid particles, which have or will pass through a fixed location # at time ¢’. The pathlines of the particles are given by # = 2(£,t). Solv- ing these equations for = &(Z,t) and replacing @ by the coordinates of the fixed location y, and setting t = 2’, we locate the fluid particles é that were passing through the fixed location 7 at time t’. The equa- tion assumes the form € = €(7,t'). The pathlines of these particles are # = 2(E(G,t'),t). Thus, at the fixed time ¢ und variable ¢’ we obtain the streakline as a curve, which connects the fluid particles having passed through the fixed spatial location # at time t'. For this problem, all fluid particles remain in the plane z3 = €. Inserting the coordinates 2, = 4 and 2 = y in the pathline equations (4), (5) expressed in parametric form, we obtain the material coordinates of the fluid particles which passed through the above location at time t = t’: ¥ a - & = yo — vol’. Now, we insert these material coordinates into the pathline equations and obtain thus the parametric representation of the streaklines: 1L+i/to = vw , 14t/to (8)14 1 The Concept of Continuum and Kinematics t v m= n+ wto(Z-Z). (9) to to In the above equations ¢ represents the fixed actual time and ?¢’ the curve parameter. Eliminating the curve parameter t’ from (9) %2—Y2 vo Ust-—- and inserting the above equation into (8), yields the explicit form of the streakline zy = ——_4___, (10) _ T2— Y2 Up to (1 + t/to) The streakline through the point y; = r10, yz = 229 is given by ay 1 The figure shows the streakline for vo to/z20 = 2 at the fixed time t/tp = 2. The dashed line represents the pathline of the particle, which, at time t'/tp = 0 was located at point T1 = Tyo, T2 = To. At time t/ty = 2 the fluid particle is at the point £1 = 3219, 2 = 5x29. Using this, the streakline in parametric form (8), (9), gives for the parameter val- ue t/t = 0 the point z, = 3210, . tq = 5220. o 14 2 sos 4 X,/ X49 8 x,/'r, Problem 1.2-6 Kinematics of an irrotational and divergence free flow field The velocity field u,;(€;) is given by a(x, +22) ’ UL u; = a(z;—722), wy = W ————— _ : ll Toy £2/Z2—1 (11) Up tof £a0(1 + tf to1.2 Kinematics 15 with the constants a and W. Determine a) the divergence V + @ of the flow field, b) the vorticity V x @, c) the parametric representation of the pathlines 2; = x;(é;,t) with & =2;(t=0), d) nonparametric representation of the projection of the pathlines in 21, r2- plane by eliminating the curve parameter ¢, e) the projection of the streamlines in 21, 2-plane by integrating the dif- ferential equations for the streamlines. Solution a) The divergence of a vector field is defined as . 4 ~ Ou; Ou, . dug dug div@=V-d= SW OH, Oe ts we “Oa; Ox," Oa2 | Ors For the present case, it follows a an =a-a+0=0, i thus the velocity field is divergence free. b) Similarly, the three components of the vorticity vector disappear: 1 fa a wy = 3 (oen+ en) =0, 1 fdu Ou, 2 = 5 (Fens Fete) =O, 1 fa a w= 3 (2 Ga3 + Fen] =0. As a result, the velocity field can be expressed in terms of a potential, which is obtained as (see also Problem 1.2.4): @ = 5 (2}4+ 2x12, — 23) + Way. c) Parametric representation of pathlines: The pathline differential equations are dai _ ae16 1 The Concept of Continuum and Kinematics or o = t=a(x, +23), (1) ou = t2=a(s,—2), (2) tes = w=W. (3) Integrating the last differential equation (z3(t = 0) = £3) leads us to ag(t)= Witt és. (4) The coupled equations (1) and (2) are first order differential equations. They may be reduced to one equation by a further differentiation. For this purpose, we differentiate (2) Wry _ dai dee di" dt dt and replace the derivatives on the right hand side by (1) and (2). Asa result, we obtain d?zy “di? This is a second order linear differential equation with constant coeffi- cients, It is solved by =2a’x,. r2=Ce'.~ The eigenvalues are found as A=ivla and the complete solution is: z(t) = Cy e¥™! + Cpe" ¥™* (5) From equation (2) we extract a(t) = - (v2aC, - V2aC2 eV") + +C; eviat + Cae ¥2t + = mt) = (2+ 1)Cye¥**—-(V2-1)C,e-V*". (6)1.2 Kinematics 17 The two integration constants C; and C2 are obtained from the initial conditions m(t=0)=& = (v241)C,-(v2-1)C2, a(t=0=& = C+h, this yields a = 1e-vaar a, o G = 1a4vae-e. (8) Equations (4), (5), (6), together with (7) and (8) describe the pathlines. d) Nonparametric representation of the pathlines in 11, r2-plane: To arrive at the nonparametric representation of the plane curve, we eliminate from equations (5) and (6) the pathline parameter ¢. For this purpose, we multiply (5) with (72 — 1) and add to (6): (V2-l)e tay, = ((V24+1) + (V2-1)) Cre! = 220, eo" Viat _ 1 _ => eViet = WiC, ((V2 = 1)22 +21) > evt x 1 iG, ( - v2) t+ v2.21) , and insert the final result into (5) z= jl — VB) 0 + Vin) + 40,02 ((2 - VB) a + Vn) . Thus, we obtain 1 1 gtataite— scp =40iC2. (9) The above equation together with (7) and (8) is the implicit equation of pathlines. The explicit equation is tgs —2, $20? +8CO,C, = 2 (-12 y? + soit ' 1 Setting C,C, = 0 reduces the pathline equation to a linear one. tg = 2 (-14+ V2), with tana = 2/2, = —1 + V2 and therefore18 ¢) 1 The Concept of Continuum and Kinematics a = —67.5° and a = 225°, The figure shows the pathlines of a stagnation point flow in 1, 22plane tilted at an angle of ag = 22.5°. Streamlines in 21, 22-plane: The differential equations of streamlines are dei Mi ds ~ |i’ with the components in 2, and x2 direction dz, wy dzq ug ‘ds i’ ds fay Dividing the above equations by each other eliminates the streamline curve parameter 5: dr uz dz, uy Inserting the previous values for u, and uz gives us dr, _ a(x, — 22) or dz; ~ a(x) + 22) (zg — 21) dz, + (a, + 22)dz.=0. The last equation is an exact differential equation of the form av ov dv = —d. dr = Oxy w+ Or, ™2=0 with the solution Y = const, To prove this statement we form the derivatives Oo (8) 8 ay a1 dz, \8r,} O22"? ue and a (au a on (=) = On, +22) =11,2 Kinematics 19 As seen, the mixed derivatives €°W/02,8r, and &W/xr20z; are equal. This is the necessary and sufficient condition for the differential equa- tion to be exact. As a result, W/A2, and 0W/@z_ are known. For the calculation of W we first integrate av Oz, 72 71 and obtain 1 U = a2, - 5 ai + h(a) and the derivative oe a +2, = 2, +h'(22). Consequently h'(a2) = ey and thus 2 = h(a) = 3 +0, and we arrive at the final solution 1 1 v= gttrite—52i+C. The lines Y = const are the projection of the streamlines in the plane 23 = const . Compared with (9) from Problem d) we recognize that for the present steady flow case, pathlines and streamlines coincide. Problem 1.2-7 Kinematics of an unsteady, plane stagnation point flow The velocity components of an unsteady, plane flow field are given by uy (a+ bsinwt) 2, ug —(a+bsinwt) 2, with the constants a >b>0. a) Find the equation of the streamline through the point (10, 220). b) Find the equation of the pathline for a fluid particle which at the time t = 0 was at the place #(t = 0) = &. c) Find the equation of the streaklines through the origin (jf = 0). d) What is the velocity change that a probe would measure if it moved along tip = t2p = col ?20 1 The Concept of Continuum and Kinematics Solution a) Streamlines (plane flow): The differential equations of streamlines are dey _ tm da _ va ds jij’ ds |i’ and for this problem dz, (at+bsinwt)z, _ ay day uy —(a+bsinwt)z2 22" The direction of the velocity field is time independent and therefore steady directional, In this case, the streamlines are also time indepen- dent. The differential equation can be solved by separating the variables dey dz / a1 / ry and after adding the integration constant we are led to Ina; =—Inz,+InC or Thus, the integral curves are hy- perbolae and represent the stag- nation point flow in the upper half plane zz > 0. For the stream- line equation through the fixed point (a19,£20), the integration constant is determined as C = 24999 Which leads to Z10 T20 rz = ——... 71 b) Pathlines: Since the flow is steady directional, the streamlines coincide with the pathlines. The pathline equation is m=. Ty We calculate the pathline again by integrating the differential equations dz; dt = Uy1.2 Kinematics 21 here dr . =e = u=(a+bsinwt) 2, a = uw =—(a+bsinwt) zr, . Separating the variables yields dz, : a iC + b sinwt) dt, daz _ . = [(a+bsinwt)de. The integration leads to b Ing; = inc + (at ~ * cosut) > ry = Cy elt Eooset) b ~(at-2 coswt) Inzg = InC;—|at——coswt => = Coe . - w The integration constants are determimed from the initial conditions Z(t = 0) = Fas C= es and Cy =fe75 . Therefore the parametric representation of the pathlines is & e(2t+3(1—coswt)) . (1) ry & en (2t+$(-coswt)) . (2) a2 If we eliminate the pathline parameter t from the first equation e(ttt+E(1-cosut)) _71 & and insert the result into the second equation, we arrive again at the equation _ fib 7 T222 °) 1 The Concept of Continuum and Kinematics Streaklines: As mentioned previously, for a steady directional flow, the streamlines coincide with the pathlines. This is also true for the streaklines. Conse- quently, the streaklines are also described by the same equation, which is in this particular case a hyperbola 242q = const = yy ya - - For y1 = yo = 0 we get 4) Tj22=0. The above equation implies that the coordinate axes (t; = 0 and zy = 0) are the wanted streaklines. From the pathline equations (1) and (2) we conclude that 2;-axis is the streakline of the fluid particles, which were located at the origin at the time t’ + —oo. For the fluid particles that would arrive at the origin at the time t’ + oo, the z2-axis is the streakline. Velocity change measured by a probe, which moves along the path ip =I2p = cot: The change of an arbitrary field quantity measured by the probe is given by the operator qd_ 9, @ di 87 9 Be; with ¢; as the absolute velocity of the probe. We apply the differential operator to the velocity vector uj and arrive at the velocity changes measured by the probe on its pathline: guj _ Ou; |, Ou dt Ot * 7 az; With ¢; = c, = co, the changes of the velocity components as seen by the probe are day Ou Out Ot de a, ey = (bw coswt)z; +e(a+bsinwt) , dug _ Ou Bug Buy m7 Ht de, + Oey = (—bw coswt) x2 +c [—(a+b sinwt)] .1.2 Kinematics 23 Along the path of the probe, we have 2; = 2, = cot, 2 = ta, = cot. This means that velocity changes expressed as functions of ¢ are “ = co(bwt coswt+a+bsinwt) , ais = eg (bwt coswt+a+6sinwt) . Problem 1.2-8 Streakline of a water jet The nozzle of a water hose is located at ¥ = h& and oscillates with the angle a = a(t), Water leaves the nozzle with a constant exit velocity U. Neglecting the air forces exerted on the water jet, determine: nozzle a) the velocity components u(t) of a fluid particle which was at the nozzle exit at the time ¢’, b) its pathline for #(0) = é, c) the equation of streaklines. d) Has this type of flow streamlines? Solution a) The velocity of the fluid particle u,(t): Neglecting the air forces exerted on the water jet, the fluid particles de- scribe trajectories in form of parabolae, where the velocity components are given by wm = CL, Cy—gt. ug The constants C, and C are determined from the condition that the fluid particle under consideration was located at the nozzle exit at time t! and had the velocity components u(t’) = U cosa(t’), u(t’) U sina(t’) .24 b) ) 1 The Concept of Continuum and Kinematics The constants are CQ; U cosa(t') , Cr U sina(t')+gt'. With the above constants, we recover the velocity components u = U cosa(t’), uz = U sina(t’)—g(t-?'). Pathline of the particle with z(t = 0) = From the pathline differential equations da, _ ’ _ ao U cosa(t') (= const) , dz. _ . , , a7 U sina(t')—g(t-#'), follows by direct integration a(t) = U cosa(t’)t+C3, a(t) = U sina(t')t- 59 (7 —2t'th +Cy. For the fluid particle under consideration the integration constants are (0) = &=Cs, m0) = &=Cy, therefore its pathline is x(t) = U cosa(t')i+é, (1) z(t) = U sina(t')t- 59 (@-tP-t) +6. (2) Streakline equation: Starting from a known pathline 2; = 2,(€;,t), we solve this equation for £; = €;(ai,t) and identify the fluid particle at time t’ and position y = hé, through the equation €; = €;(y;,t'). The resulting equation ai(t!) = ci(Es(ye,t’),t) = te(yes tt)1.2 Kinematics 25 gives for a fixed y, and ¢ the equation of a streakline. As we saw, the streakline is the connecting line of all particles having passed through the fixed location 7 at a time ¢'. This is the step-by-step procedure: We solve 2; = 2;(£;,¢) for €; = &;(2;,t) from (1): & = 2,-U cosa(t')i, from (2): f& = 22—-U sinalt’)t+ sate -t'y? i). Particle identification (¢ = t’, 2, = y; = 0,22 = y2 = h): & = —-U cosa(t’)i', ay ae & = A-U sina(t')t —a9t : Inserting the material coordinates nozzle t=const into the pathline equation (1) and (2) leads to a = U cosa(t')(t—-t'), ra h+U sinalt’)(t-t) + streakline=jet 1 -sg(t-t)*. gatt—e) This is the parametric represen- tation of the streakline at time t with ¢ as the curve parameter. d) Are there streamlines? The instantaneous picture of the water jet at time ¢ is exactly the calculated streakline. The velocity vectors are not tangential to the streaklines.26 1 The Concept of Continuum and Kinematics = To construct the curves tangential to the velocity vectors of different fluid particles at the same time t (stream- lines), adjacent particles in the direc- tion of the velocity vectors forming particle the streaklines must exist. This is not pathline the case here, the streamlines degen- * XCEC¥.U),t) erate to points, Asa result, no stream- line can be constructed for this par- ticular problem. eb a streakline Problem 1.2-9 Streamlines and Streaklines in cylindrical coordinates The velocity vector of a plane, unsteady flow field is given in cylindrical coordinates (r,) by 1 > d= (Aves + Bo(l + at) ey) with the dimensional constants (Ao, Bo, 2). Using cylindrical coordinates, calculate a) the equation of streamline through the place P(r = ra, y = 0) and b) the pathline equation of a fluid particle, which was at time t = 0 at place P. Solution a) The equation of streamline: In cylindrical coordinates a differential line element. is given by (F. M. (B.2)) dz =dré.+rdyé,+dzé, , (1) and the velocity vector (F. M. (B.2,c)) d= uy, é + tye, tus. (2) Thus, we arrive at: dee ds la’ the three components of the equation for the streamline are dru, dp uy, dz ou a ds fa’ "ds ja’ ds al’1.2 Kinematics 27 The first two equations are sufficient for describing the streamlines in the plane flow in the r,g-plane. To eliminate the curve parameter s, we divide and get drt rdp uy Introducing B(t) = Bo(1 + at), we arrive at dr Ap 7 BS? Since the streamlines are lines at fixed time, the time ¢ and also B(t) are considered as constants. The integration results for a streamline going through the point (ro,0) in 7dr Ao 7 ro 7 ~ BU) [ee r Ao => In nT BH ep or (3) r(yv) = ro enh | (4) As indicated previously, the aux- iliary function B(t) is a func- tion of t and has only a para- metric influence on the stream- line. Equations (3) and (4) rep- resent logarithmic spirals, i. e. all straight lines drawn from the origin, intersect the curve under the same angle ¥. b) Pathlines: With (1) and (2) the differential equations are: dr u dp _ dz _ Gent "ap tte Gaue For the present plane flow, only the first. two differential equations are used. For the velocity field in this problem, they assume the following1 The Concept of Continuum and Kinematics form a7 6 fe = Boll + of) . (6) Differential equation (6) is coupled with (5) via r. Since (5) is not coupled, we integrate it first: [rors [Aoat+c. Introducing the initial conditions r(f = 0) = rp, the solution found is ri=ri+2Apt. (7) Using (7), (6) can be integrated dy l+at : F l+at — = di = Bo f = at a ayaa | ? °f B+ 2Adt t t 3 In(r§ + 2401) 1 2 p(t) = Bo lax In(rp + 2Agt) +4 (sc- TA ; _ Bo Boari\ 1 2 Boat)‘ = (#- zag) 2 iro 245 |, L _ Bo Boar? 2Aq .\*? Boat - (2- AR Inyitoat + 3a" Eliminating the pathline parameter ¢ from (7) results in the following explicit representation lr) = (3e- Sey2) mn (=)+ mae -7). (8) It should be mentioned that by setting a = 0 (+8(t) = Bo) a steady flow case is generated, for which the pathline equation (8) coincides with the streamline equation (4).1.2 Kinematics 29 Problem 1.2-10 Streamlines and pathlines of standing gravity waves 0.0 -0.5 -1.0 -1.5 7 -107n 0 10m x, The velocity field u;(z;) with the components u = —U sin Mt sinka, coshk(z2+h), u2 = 4+U sinMi coska sinhk(x, +h), uy = 0 describes a standing gravity wave in an horizontal liquid layer of depth h. The velocity U, the frequency 2, the wave number k, and the depth A are constants. a) Show that the velocity field represents a potential flow and determine the velocity potential ®(2,, x2, t). b) For sufficiently small amplitude the shape of free surface can be ap- proximated by (= _l O0(x,, 29, t) g ot ny=0 with g as the gravitational field strenght. Sketch the surface at the fixed time t = 0. Compute and sketch the streamlines and pathlines. Given: U, frequency 9, wave number k, depth h Solution a) The existence of a potential flow is proved since curl i = 0, i. e. : Ou. — Juz dug _ Gus Bug — uy Oz2 Oz,” O23 Or.’ Or, Axa” The velocity potential © is obtained by integrating the differential equa- tion 00/02; = u; O(r1, 22, 1) = ‘: sin Nt cos kr, cosh k (22+ h) .30 1 The Concept of Continuum and Kinematics In the above equation the integration constant is omitted without. loss of generality. b) The surface at, time ¢ = 0: C= Ua cosh kh cos ka, . gk The unsteady flow field can be written in the form: u(z,t) = f(t) az) . The direction of the velocity vectors is thus time independent, i. e. the streamline curves coincide with the pathline’s. These curves are calculated using the streamline differential equations in the following form dz, daz nw Separation of the variables dz, _ diz “tanka, tanhk(zy+h) and subsequent integration leads to ; In (sin ka.) = zn cinh (22+ A) + Z nC 1 > 6 We obtain the integration constant C from the requirement that the streamline is to go through the point 2; = z19, t2 = 29. As a result, we have 1 ¢ = —______.. sinh k (to + A) sin key Thus, the equation of streamlines in an explicit form is sinh k (ao + h)sin kay . _ carsinh sinh k (220 + nees| -h. Xo 0.0 —0.5 -1.0 -1.5 1 SX —10n 0 1071.2 Kinematics 31 Problem 1.2-11 Change of material line elements in a Couette-flow The velocity components of a Couette-flow is given by uw=—f2, w=u3=0. k a) Determine the time rate of change of strain of material line elements dz and dz’. b) Obtain the angular velocities Dy/Dt and Dy'/Dt of the material line elements. c) Determine the material change of the right angle between dz and dz". d) Determine 1) the velocity gradient du,;/0z;, 2) the rate of deformation (rate of strain) tensor e,; and 3) the spin tensor 1,;. e) Using the tensors from d), calculate the strain rate of the material elements dz, dz’, and the material change of the right angle between them. Solution a) The rate of change of material elements dz and dz’: uj (x2+ds) dx I ds: “uXe) uz) de uy (Xe) The line elements have the directions of the coordinate axes dz = dsé&, dv = ds’&. The velocity field is only z2-dependent and only the 2,-component of dis different from zero.32 b) <) d) 1 The Concept of Continuum and Kinematics The strain rate 1/dsD(ds)/Dt corresponds to the component of the velocity difference dif between both ends of the line element in direction of the element itself, divided by the length of the element ds. Since ug = uy = 0, the velocity difference dif has also only a x,-component. It is d dy = uy(22 + ds) — uy(20) = i= eas. (1) The component in direction of the element dz and thus the strain rate of dz is zero. For the element dz’, because of u, = u;(r2) we may write dui = u(z2) — u(r.) =0, (2) i.e. the strain rate is zero. Angular velocities of the elements: The angular velocity of a line element Dy/D¢ is calculated using the component of the velocity difference normal to the line element divided by the length of the element. For dz considering (1) we obtain De dw Ug De ds a the negative sign refers to the convention that for positive du the rota- tion occurs in a mathematically negative sense. For dz’ follows from (2) Dy! _ Do” i.e, the element is neither stretched nor rotated. The material change of the right angle between the line elements dz and dz’: The material change of the right angle is the difference of both angular velocities Den Dy Dy’ U U ‘Dt Di Dt & hh Velocity gradient, rate of strain tensor, and the spin tensor: 1) The only non-zero term of the velocity gradient Ou;/Ox; is 0u,/Oz2 = Ufh = +4. 2) The rate of deformation tensor ¢;;: The rate of deformation tensor (rate of strain tensor) ¢,; is the1.2 Kinematics 33 symmetric part of the velocity gradient tensor Gu;/Oz; a (Om Oe “3-9 aa; * da:) Since e;2 = em, we have: en Ou 4 Gua Ouz\ _ 1, ” On, Oz,) 27° Again, the other components are zero. 3) Spin tensor 9,;: The spin tensor is the antisymmetric part of Ju;/Oz;, Ou; du; m= 3 (Foe) Thus a1 (Gm _ Ou) Lou 1, n= 5 (5-5) = 2 Oz, 27 On - The other components disappear. e) The strain rate: , The strain rate of a material element in direction of | = dz/ds is 1 D(ds) ds Dt The direction vectors of both elements are = ij t by * dz: T=(0,1,0); d#: ['=(1,0,0). Inserting e;;, [j, and lj in the above relation, for the element dz we obtain 1 D(ds) ds pr HN =0 d for dé” and for dz" 1 Dds!) _ bree. dy De 8a In agreement with the results in a) we obtain both strain rates as zero. The material change of the right angle is calculated from: = = -2e12 and e,2 = ay:34 1 The Concept of Continuum and Kinematics which we already obtained under c). Note: For a plane flow, the only non-zero component of the angular velocity @ is obtained from w, = dein Qj; thus 1 1 1. 1. 1. Ws = 5 (G23 Ma + €na M2) = 3 (-37- 3’) =—97- Problem 1.2-12 Change of material line elements in a three-dimensional flow For a steady state flow field, the nondimensional velocity field is given by T= 32} 0G 4 2eir3h +2, 2223. Calculate at point P = (1,1,1) a) the components of the velocity gradient. Gu;/x;, b) the components of the angular velocity of a fluid particle at P, c) the components of the rate of deformation tensor ¢;;, d) the strain rate in the r,, rz and z-direction, e) the material change of the right angle between dz, and diz of a material volume element dV = dz, dr2dz3, f) the strain rate of a fluid element in its pathline direction, g) the principal strain rates and their directions. Solution a) Velocity gradient: The components of the velocity gradients are calculated from Ouz dug dus —= = = —_—= 2=] aa, 6222 = 6 an, 0 an T2L4 a é a Je, == 8 Jn = tts 4 ia = 1 Ou, Gu Jug —-=+=0 —=277 =2 a = 2: =2. Oz Org 72 Org TF az3 b) Angular velocity of a fluid particle: From .~ Ll ~ 1 Ou, a= gona eu = 2 HD, we obtain 1 3 w= w= 5) w= O° 1 5?1.2 Kinematics 35 c) Rate of deformation tensor: é4 =6 €1 = 3/2 €13 = 1/2 1 (du; 9 Gu; cv 5 (Se +S) en =3/2 en =4 e23 = 3/2 és = 1/2 egg = 3/2 egg = 2. d) Strain rate in the coordinate directions: For the material line element dz with the direction vector [= dz/ds h we have 1 D(ds) ds Dt We successively insert for = (1,0,0), (0,1,0), (0,0,1), and obtain the strain rates in the coordinate directions = Cy A . | D(dz,) dz; Dé su=6, 1 ae = en=4, day 1 D(dz3) _ dz3 Dt faa = 2. e) The material change of the right angle between dx, and dzz is a ee == 8 | dxa | similarly, the material changes of the right angles axe between dr; and dzr3, dzz and dzy respectively ax, are D(a D(a. aE) —2ei3=—-1 and Pies) = —2e€3 = —3. f) Strain rate in path direction: The strain rate of a line element in the direction of the particle path is obtained by substituting the direction vector / in equation 1 D(ds) ds Det = eg Ul; by the particle path direction, which is given by i Is i= z36 8) 1 The Concept of Continuum and Kinematics Since @(P = (1,1,1)) = (3,2,1), ie. |@] = 14, one can obtain the components of the normalized direction vector as: 1 1 h=—=3, b= 1 Vid 2 Vid Thus, we arrive at the strain rate 1 D(ds) ds Dt 2, t3 = 1. 1 vit = entry t+ eit ts +eistits + tear ta ty + €22 te ty + eag te ts + +a ty ty + €g2 ts to + €33 ta ta « Because of e;; = ej;, the above equation can be rearranged as 1 D(ds iz (ds) _ ern ty + €2a ty + eng t + 2 (C12 ta ty + Crs tits + €29 t2 ts) - s Dt At point P = (1,1,1) the numerical value calculated is 1 D(ds)_ 9 64.1 273, 1, 3.) 99 gee Oa tig tha tig (954 39+ 92) = Principal strain rates and directions: The calculation leads to the eigenvalue problem (ei; — e6i;) 1; = 0 with e as the principal strain (= eigenvalue) and Fas unit vector in the principal direction (= eigenvector). Non-trivial solutions exist only if the determinant of the coefficient matrix (e;;—¢4j;) identically vanishes, ie. if (“ —e £12 €13 det} en e2-e € |=0. ea. €32 €33 — € This condition leads to the characteristic equation: —e? + Tyee? — Ine + Ine = 0, which allows the calculation of the three eigenvalues. The invariants of the strain tensor are he = eg =64+442=12, 1 fre = pleiieas ~ e43¢%5)1.2 Kinematics 37 with €11é11 + €rz€i2 + erséist €g1€21 + €g2€22 + €ag€23+ €31€31 + €g2€32 + €ggea3 2 2 2 2 2 2 C11 + C30 + €33 + 2 (eye + eta + C23) eseazes ()'+(6)'«()) ms 1 > Ie = 5 (12? — 65.5) = 29.25 ajay i++ il In, = det(ei;) 9 373 1/9 > te = 6(8- 7) 45 (7-8) +9 (7-2) =s125. The three roots of the third order polynomial can be calculated using appropriate solution methods such as Newton’s. e!) = 1.180, e) = 3.741 , e) = 7.079. These solutions are the principal strain rates. The components of the strain rate tensor in direction of the principal axes are: eae for tay 9) 0 for ifj. Using the above eigenvalues, the eigenvectors are calculated from the system of linear equations (ej; — e4j;)[; = 0. Since the determinant of this system of equations disappears, only two equations are linearly independent. A unique solution results from normalizing: G+b+8=1. A simpler alternative is to calculate first the vector t, which is not normalized and perform the normalization later. We delete the third equation and set instead & = 1. We obtain the first eigenvector from . 3, 1 i=l: (6-118) h+5h+5=9, 3 3 2 i=2: Sh+(4-1.18)h4+5=0,38 1 The Concept of Continuum and Kinematics using the Cramer's rule: i, = 0.07406, = ~0.5719. Thus the vector components of the non normalized eigenvector are y =0.07406, & =-05713, % =1. Normalizing the above components leads to M= 0.067, M= -0.4950, = 0.8665. The components of the second and third eigenvectors can be calculated in a similar way. For the second eigenvector we insert ¢ = e) = 3.741, which results in 79558, i@=0.702, i= 0.442 . The third eigenvector may be calculated the same way. It is worth noting that the calculation can be performed in a much simpler way by taking advantage of the fact that 0.825 R) = 1M) yf) _ | 0.515] . 0.233 Calculating F* from the above equation also determines the sign of It), The vector @) is in direction of e), 2) in direction of e), and #) in direction of e), Problem 1.2-13 Angular velocity vector and the change of material line elements in a two-dimensional flow field Given is the velocity field: w Wy = TR TTS te ae uw = +77 h3 h , uz; = 0. Determine the components of a) the velocity gradient tensor, b) rate of deformation tensor ¢;; and the spin tensor 1;;,1.2 Kinematics 39 c) the angular velocity vector @. Calculate d) the principal strain rates and the principal strain rate directions at point P = (2,2,2) and e) the pathline of the particle which at time £ = 0 was at place P = (2, 2, 2). Solution a) Velocity gradient tensor: Gt _ 9 duz _ 4 Gus 4 Oz, az, h oz, Ou w Our dug = = -23— —=-0 —_—= Or A Org Oz2 Ou we Om _ iw ous _g @t3 A 823 kh Oz, b) The components of the tensors ej; and 1,;: €1=0 en = es=—22 411 = n= BIO E = - — en =0 e2 =0 e3= > FZ ey = TY — mv =0 rr a “I= OF eg =U; w Zyw On =0 Mh = —t3 7 Ww=-FF w nw N21 = t3 — Oe. = 0 5 = — — 21 = 73 h 22 23 2h n= 2% Oya-B” ay ao, => 32 2k Nas = 0 c) The components of the angular velocity vector a: 1 wi = 5 Giik Nis with the individual components 1 mw a = 5 (ezaMse + e13283) = Mor = ae 1 tywW Wy = 3 (e2a%is + era) = Oy = "oh? rile 1 wy = 3 (eannMan + e322) = My = I H a40 1 The Concept of Continuum and Kinematics Note: Identifying the components 2;; of the spin tensor with the angular velocity components, we obtain Qn =0 Q2=-w, ha =a. Qo, = ws Ny» =0 Ng = —4y Os, = Wy Og = wy Qs = 0. The three independent components of the antisymmetric spin tensor Q;; correspond to the components of J. d) Principal strains and principal strain rate directions: The following eigenvalue problem is to be solved (ei — ey) lj =0. The rate of deformation tensor at point P = (2, 2,2) is given by w en =0 e2=0 t= —F w é7 = 0 ea =0 t3 = k w Ww ea =p em =F éx = 0. The eigenvalues are calculated (with € = eh/w) from -é 0 -1 ' det(e;; — €6;;) = det 0 -é 1] =0, -l 1 -é that is, from the characteristic polynomial: 0 q —é(é? — 1) —1(—é) =>ée(e-2) = 0. => €=0,-v2,4Vv2. Therefore the three eigenvalues are given by ell = -vi5 ; e@) =0 : eA = +95 with e = éw/h.1.2 Kinematics 41 With the above known eigenvalues, the eigenvectors (= principal strain rate direction) are determined from the following system of homoge- neous equations -é 0 -1 hy 0 0 -é 1 hl=]o]. -1 1 -é b 0 Since in the above system only two linearly independent equations exist, we may first delete one equation and set [| = 1. As a result, we obtain the inhomogeneous linear system i=2: ~€+h=0 i=3 : §-é@§-1=0, with the solution The corresponding eigenvectors are calculated subsequently as ae = 2 a2) — 9 é3) = 2 B= -1,4= V2 h = 1,%=0 B= -1,4=--v2 ji = 2 i = v2 FL = 2. The results in normalized form are: 7m) :(-] m-L{; waif 2 Me-fa}, MsazAlil, Mas] -1 |}. "V2 v2 \o *\ va To determine the sign of the vector T®) we Tequire that hy, 7), and i) should form a right-handed system, i.e. 19) = 7) x jl), a) a) a) MY) x 7) = det] 1/2 -1/2 vn) va/2 2/2 0 #9 (3) +20(}) +0(2). 1 i =~9? ip) = We obtain IP) = “IS i 5?42 1 The Concept of Continuum and Kinematics The rate of deformation tensor in principal axes system is ef) 9 0 (a)=] 0 ee of. 0 0. ¢®) The rotation matrix, which transfers the old coordinate system into the principal axes system is ay = GG eT, Thus, aj; is the i-th component of the j-th eigenvector 1/2 V@/2 -1/2 (aj)=| -1/2 vef2 1/2]. va/2 0 Va/2 The eigenvectors form the columns of the rotation matrix. Pathline through the point P = (2,2, 2): The differential equations are dz; w a= Ree da w a 7 Rt = = 0 or z3=const = &. The first. two equations along with the results of the third one are da w Ts = —_éata, dz w a7 RST The time ¢ does not appear explicitly, ie. pathlines and streamlines coincide. Thus, we divide the two equations by each other: dr zy dz, ty Separation of variables and the integration yield a? 2 OC 2573 tg 7 w+ap=C.1.2 Kinematics 43 The pathlines are circles in the plane 23 = 5. From the condition Z(t = 0) = € = (2, 2,2) then follows C' = 8 and we get Problem 1.2-14 2442 rm +22=8, ra =f =2. Rate of deformation and spin tensors of an unsteady two-dimensional flow The velocity components of an unsteady flow field are given as uw Us ug = 0, Determine the components of a) the velocity gradient Ju;/Az;, b) the rate of deformation tensor ej; and the spin tensor 2); as well as A(ziz2 - x4)e ~ Blt) A(23 — ayzg)e7 Bl-t0) . c) curld at point P = (1,0,3) and time ¢ = do. Solution a) The components of the velocity gradient tensor are: Ou 0 Buy az, Ox, ayy Bus Ory Oz, du 5 Bue Or3 Oz; = =m7,Ae -B(t-to) ty Aen Alt-to) 223 Ae Blt) Bus ny Ae Blo) Oz, Bug _ Qzz Ae Pr) Oxy dus = —2, Ae Bt) | Ozgot 1 The Concept of Continnum and Kinematics b) The rate of deformation tensor ej and the spin tensor Nj are calculated as: en =0, tae pace , t; t; 13 = -= Aen Blt") 5 en= z Ae Blt-t0) , eng = 2, Ae BU) a3 = (2y — 23) Ae~BUt~) en = —F Aerts) : ex = (2 — ty) Ae BU-4) , €9 = —ry Ae“) 5 MH =0, Ma = —B AaB), Ma = BAeWBltHte) , My = BAe BHO), Q,=0, Mey = —(a9 + 23) Aen Bt-*) , Ds, = —B Ae Bee) , Mea = (22 + #3) Aem ROH) , Ny =0. c) From Qa +2ra cul? = | Ofa | AeWie), m-0 follows the curld at point P = (1,0,3) and time t = to: 6 culd = |3]A. 01.2 Kinematics 45 Problem 1.2-15 Time change of the kinetic energy of a fluid body The velocity components of a two-dimensional flow with constant density 9 are given in cylindrical coordinates: Ua) Uys Us = 0 (A=const) . Consider a portion of the fluid at time ¢ = 0 located between the surfaces of two concentric cylinders with radius r = a and r = 6 (6 > a) where O<2z< 8. Calculate a) the pathline of the particles located on the inner and outer cylinder surfaces at t = 0. How does the material volume look like at time ¢? b) the kinetic energy e K(t) = fff ewav (v(t) Pu) = fff etav (vce) of the fluid under consideration at any arbitrarily chosen time ¢ and the substantial changes DK/Dt and DP/DE. c) DK/Dt, using Reynolds’ transport theorem. d) Describe the motion in material coordinates and calculate DA/Dt by transforming back to the volume Vo, which was occupied by the fluid at time tf = 0. and the momentum Solution a) Pathlines: The differential equations of the pathlines in cylindrical coordinates are in general (see Problem 1.2-9) dr dy _ dz a Ur y "ae te at and here dr_ A dey dz aor) at a These three differential equations are decoupled and can be solved suc- cessively as follows: frav= faa @ Pa Al+ Ch,46 1 The Concept of Continuum and Kinematics [av=o & yv=C,, fac=o & z=Ca. For a fluid particle located at time ¢ = 0 on the inner cylinder (r = a), the integration constants are calculated as Cc, = Cy= o(t=0), Cy= z(t=0), e 2° and the pathline is rdjsa4+2At , y= y(t=0), z= 2(t=0). Correspondingly, we obtain for the particle on the outer cylinder (r = 6, at time t = 0) r(tjy=04+2At, y,z=const. The boundaries of the material volume experience a displacement and the pathline equations are given as Ri(t)=a?+2At for the inner surface, Ra(t) =—P42At for the outer surface. Ax, 7 v(t) * x Ri(t} Rg “tt) The sketch shows the material volume Vo and V(t) occupied by the fluid body at time ¢ = 0 and an arbitrary time ¢ respectively.1.2. Kinematics AT b) Calculation of the kinetic energy K(), the momentum P(t), and their substantial changes DA/Dt and DP/Dt: Ki) = ff futav= f[fSw+u+ujav= | f fou av (Vie) (v(a) (v()) Qe & Rat) 1 = = 22 = K(t) = [/ / 5A rdrdedy 0 a Rit) L b+ 2At\? — 2 => Kt) = rod tin(5 ot) DK dfx #4 2At DS a [504 on(3*3)) DK => 1 1 oS = -rpAtL (——-— |. Di me (am xu) For the momentum we obtain ar L Rolt) Bit) II ozav =f | J oA grdrdzdy (¥(t)) 0 0 Ry(t) ll ae eAL(Rolt) - Ri(t)) f & ay = Pa = 0 and thus, DP/Dt = 0. c) Calculation of DA'/Dt using Reynolds’ transport theorem: The Reynolds transport theorem (see F.M. 1.96) applied to K is: Bf [lf fears [[teasas (v) (s) Since the flow is steady, the first volume integral on the right hand side vanishes.48 1 The Concept of Continuum and Kinematics 3. Di ul | 1 I = 2 Nie bo 1 2/75 Fas a & a + + => 1 2| a A = 22 ef = “pig et Rik + 5 fe 2* Rol DK 3 1 1 pe = re (7) With R; = R,(t) and Ro = Ro(t), the above result is identical to the one found under b). d) Using (p, ¢, ¢) for the material (r, y, z)-coordinates, the material de- scription of motion is r(t)=Vp?+2At, p=d, z=C. With the Jacobian determinant of this motion (rs P52) _ p O(r,¢,6) ve? +2A71.2 Kinematics 49 the integration over the time dependent volume V() can be trans- formed back to the integral over a fixed control volume Vo: DK _ D ia a a a Dt nl | gore 00 yaterat p ye 7 ms p = — = 2 ee . ml {fe aaa VE tPAt apie t do Because Vp is time independent, the operator D/Dt can be taken under the integral sign. Thus, we have DK 3 D p DK Pg Dt nhA ela pazat? _ 2 p2A wha Satie 7+ 2AtP i] LA 1_ Te? B+ 2A, fl DK 1 1 De > eM tag Bo) and obtain again the results from part b),2 Fundamental Laws of Continuum Mechanics 2.1 Conservation of Mass, Equation of Continuity Problem 2.1-1 One-dimensional unsteady flow with given density field A one-dimensional unsteady flow is given by the following velocity Ges ~yel le and density field e_fy-iri, 2)" go \ytita yt} a) Calculate the substantial change of the density. b) Check the validity of the continuity equation at Ou pit? ea =0 for this flow field! c) What is the change of density that a swimmer senses if he/she swims with the velocity ¢ = u+a or ¢ = u—a through the flow field? Use the relation 4 ae (2) - do 20 ,2.1 Conservation of Mass, Equation of Continuity 51 Solution a) The substantial change of the density is De de do Dt Ot az _ 2m (rede, 2)" (_y-121), ~ y=l\yt lta y+) +1 ag t? 2 (F ) 200 fy—l « 2 vrm} y-1 1 + ——(=-a 7 — +—_ —— y+. \t y—-l\ytlta 41 yt 1 tag 280 qyrols 2 yr" ty+1) \yFl tay 741 , b) Check the validity of the continuity equation: The second term of the continuity equation remains to be calculated. Performing the differentiation, we get gu_ (writ, 2 a aT * oe ~ yt+ltag yt+l1 ytilito We introduce the above results into the continuity equation and obtain rete Gu _ 200 grit? = °ae ~ “ty +) \yFl tay 74 200 y-le 2 \"T tn (ity) , > ee +0 ou = 0. ° ae c) Change of the density in the moving system: The swimmer feels a density change described by the equation: de _ 9e | ,9e dt at ar With ¢= uta we have dg _ 40, G0, G0 Doe, ao at ae de De ae52 2 Fundamental Laws of Continuum Mechanics Replacing the material derivative of density by the result from a), we find the expression do _ __ 260 (Soe 2 TO goiel. 2 dt t(yt+l)\y+ lta ytl “\ytita 741 2o (rated 2 YO (ya y-l\yti tao” y+) yt tao _ 200 fy-l « 2 \ ~ (y+) (Hints) * 200 y-1zrl 2 \r 4+—— | — - -— + — . éy+)) (Hist) Thus, the swimmer senses the following density changes: force =u+a:z “ = 0, ay de 400 (3 2 2 ~ fore =u—a: —=- = = 4 . di t(yt]) \ytltag | ytl Problem 2.1-2 Plane, steady flow with a given density field The density field of a plane, steady flow is given by a(2;) = kr, 22, k=const. a) Determine the velocity field, for which the flow is incompressible. b) Find the pathline equation. Solution a) For an incompressible flow the material derivative of the density must disappear. Using the equation (see F. M. (2.4)) De ao, ae De at Ba? we first obtain the following relationship for the velocity components:2.1 Conservation of Mass, Equation of Continuity 53 we (1) since the density o is not a function of time ¢. The continuity equation for an incompressible flow can be written as (see F. M. (2.5)) Ou | Our _ 0 Oxy Oz ~ Introducing (1) into the above equation results in a first order partial differential equation in terms of the component ty. Ouy _ ty Ouy _ (2) Or, 2, 02. 2 We introduce now a parameter s and write the required solution in the form of u(s) = wi(21(s), 22(s)). The derivative duy _ Our dey | Our drs ds @z, ds ° Oz, ds when compared with (2) results in a system of ordinary differential equations dey =1 3 i 1, 6) dr _ ds a’ du; uy aon (4) With equations (3) and (4) we arrive at duy dr, 2 with the solution y= C2 ’ where C' is the integration constant. Using (1), #2 is now determined: ug =—-Cx. b) From the differential equations of pathlines54 2 Fundamental Laws of Continuum Mechanics we calculate Separating the variables and integrating the above equation furnishes the pathline equation with C’ as an integration constant: @123=C=const. This is the equation for a family of hy- *z2 perbolae. Thus the flow is a stagnation point flow. The density does not change along the pathline: g=kryr=k. SS However, it changes from one pathline to another, because the constant C is differ- x, ent for each pathline. Problem 2.1-3 Velocity at the exit of a container The container (see figure) has two inlets and one out- let with circular cross sec- tions. The flow is steady and the density is constant. At stations [A] and [C] the velocities are assumed to be constant over the cross- section, whereas the veloc- ity at station [B] has a 1 parabolic distribution. The (B} radii of the inlets and out- let are given as Ra, Rp, Ac, furthermore, the velocities ua, ug = Upmaz(L —(r/ Rp). Find the velocity ug at station [C].2.1 Conservation of Mass, Equation of Continuity 55 Solution The velocity ug can be obtained from continuity equation in integral form (see F. M. (2.7) wi [fea =~ |f oa-nas. (5) The integrals have to be carried out over the fixed control volume. For the present steady state problem, the local derivative 0/0t = 0. Furthermore, the constant density can be moved outside the integral resulting in [[@aas =o. (8) i We first place the control volume inside the container with the in- Sa let control surfaces S4, Sp, and i- contrat “= 4 the outlet control surface S¢ and ' whee Ts. write the conservation equation in Lg i integral form: 7 = wT 4 ii a i Sp aoceeesseee! ii [[taas+ [faaas+4 ffa-aas+ [[taes=o. 3a Sp Se Sw = “eff as— ffvnte) 48 + uc ff aS =0. At the wall Sy, the surface integral will disappear, because the scalar product 7-7 = 0 there. We evaluate the second integral as follows Sf un(r) dS = | [vie D-(e)] rdrdy 1 2 Umer Re | ! - (z)| ie (Zz) o56 2 Fundamental Laws of Continuum Mechanics Thus, continuity equation reduces to —uat Ri Vinee 5 Ri tc # Re =0. A simple rearrangement results in wom (2) «ge (8) Problem 2.1-4 Steady flow through a circular channel Steady incompressible fluid flows through a circular channel with the inlet radius A. At the inlet cross-section A;,, the velocity u = Up is constant. r { Mout s The density g remains constant over the cross-sections A,, Az, and A3. Inside the channel, a concentric pipe with a negligible thickness divides the flow regime into an inner cylindrical core and an outer annular portion with the exit cross-sections Az and As. The velocity distribution at Az is given by u ar\? Gy Uamas R2.1 Conservation of Mass, Equation of Continuity 57 At the exit cross-section Ag, the velocity has the following distribution: wont (2) + in] Determine the mass flux mm,,,, added or carried off within the channel. Given: Up, R, ala, = alas = fos Olay = G0/2, Urmae = 2.5 U0 Solution We apply the continuity equa- tion (see F. M. (2.8)) for a steady flow to the control vol- ume shown in the figure and obtain: [fowaass [foanas+ ff ova d5+ ff ota dS trite =O. (1) Ay Sw Ag Ay At the wall, because of @-# = 0, the integration over Sy becomes zero. The evaluation of the integral at the inlet cross-section A, results in [fet #48 =~ [f cole d8 = —2olar Re. Ay At the exit cross-section A2 we find [fea-aas = J] j00Uane [-(F)] ds 1 Re ary? = eo2suarn | [:- ($=) | rdr 5 = gp 0 Vom Re and finally, at the exit cross-section Ag we arrive at: an R [fet ras = 2 Ue I] [- +5") rdrdy. © R/z58 2 Fundamental Laws of Continuum Mechanics To make the integration process easier, we substitute F = 2r/f and find g a [[ ez-aas = Feo / p-7 + om] FdF A : 3 1 3 ooYor a G (- ia) " 0.126 a9 Up rR? . Introducing the above results in (1) we get Tout = 0.718 pg Up R? . Problem 2.1-5 Squeeze film flow The gap shown in the figure has the length L, the height A(t), and is filled with a flu- id of constant density. The top wall of the gap moves down- ward with the velocity Vo. The velocity distribution at the ex- it is u(y) = 4Uo {it ~ (5) | . a) For A(t = 0) = Ao, determine the function of the gap height. b) Calculate the maximum velocity Up at the exit. Given: Vo, Ao, L, @ Solution a) The gap height A(t): To find the function of the gap height, we integrate the differential equation dh dt =—-% + Alt)=—Vot+ho.2.1 Conservation of Mass, Equation of Continuity 59 b) The maximum velocity Uo: The maximum velocity Uo at the exits is calculated using the continuity equa- tion in integral form (see F. M. (2.7)) Sif eav=— f, eit ds (¥) (s) The integration has to be carried out over the contro] volume (see figure), which coincides with the material volume at time t. Since the integration domain is fixed, the first integral can be evaluated using 9 = const: afffeo= [ff er (v) (Vv) Since the density g is constant, it can be moved outside the second integral. We now split the integral into four parts aenidS+ [f/a-adS+ [/i-idS+ [fa-adS=0. [Jonas [fo-tass [fons If Since the normal velocity component must be equal to the velocity of the top surface Sp (otherwise the fluid would penetrate 57) tei =the i= —Vo, thus, the first integral is [twas =—v ff dS =-VyBL Sr Sr with B as the plate depth. For the kinematic reasons explained above, at the bottom surface Sy: a-77 = 0. At the left and right control surfaces 5; and Sr, the normal velocity component is u- 7 = u(y). Introducing this result into the continuity equation, we find -VBL + [fowass [fuyyas =o A(t) 2 y ¥ Wwe, = 2 2-yj2 > 0 wolf (4) | a60 2 Fundamental Laws of Continuum Mechanics 2 37 AC) = 8p ly iy 2h 3F|, 4 = 7 Uoh(t). We rearrange the above equation and obtain the velocity 3.4 3 LY Uo= Ti = Tha vl: This equation indicates that the velocity Up approaches infinity for a gap height of A(t) = 0. However, in the reality, the top wall can not be made to move toward the bottom wall with a constant velocity Vo. Problem 2.1-6 Moving Piston A piston moves with the velocity Vp (see figure) inside a cylinder, which is filled with oil. The velocity distribution w(r) of the exiting oil at the top surface is measured relative to the piston and is given by: w(r) = Wo - ()} , Determine the maximum velocity Wo using a) a piston-fixed coordinate system (a coordinate moving with the piston), b) a space-fixed coordinate system. Given: ro, R, Vp, @ = const Solution a) Piston-fixed coordinate system: In a piston-fixed coordinate system the bottom cylinder wall moves with Vp upward. With 9 = const. the con- tinuity equation for a fixed control volume embedded in this relative coordinate system is written as:2.1 Conservation of Mass, Equation of Continuity 61 o [[ws-aas=o. G) We subdivide the entire control sur- face into three surfaces: => ffw-tdS+ f[e-ads+ Sw Sp + ffid-adS =0. . Sr n At the bottom wall Sg whose ve- locity is Wy = —Vpm, the kine- matic boundary conditions requires w-n=tiy: a= —-Vp (otherwise the fluid would penetrate these walls), while w-i disappears at all solid walls. This results in the equation: “We ff as+ [fw dS =0 or * Ha 2 Ver =2n | Wo (: - (=| ) rdr 0 To We substitute r\? r ra dé =(— dé =2—d: dr = 2= é (=) > de °2 CS OF and obtain Jomear= fea gr Tad _ We f-oae= "ee thus, Vea Ro = ara pT TT) 3 => Wo = 2Vp (=) To62 2 Fundamental Laws of Continuum Mechanics b) Space-fixed coordinate system: The continuity equation in this coordinate system is [fenase [Jenass [eras =o. On the fixed walls Sw, @- i dis- appears. The same condition ap- plied to Sp results in my é-a=tw-i, with &y = —Vp& €-n=—-Vpé.-n. As a consequence, the integral over Sp is [J@aas= [[ vee mas. é,-ndS is the projection of the surface element dS in z-direction, €,:nidS = +dA,. The sign of the scalar product is determined from the angle between 7 and é;. If this angle is less than 90 degrees, the sign of the product will be positive (+), otherwise negative (—). At Sp sgn (é@, +71) > 0 (enclosed angle less than 90°), i. e. the positive sign must be applied. We evaluate the integral: [ [= tas =v | [fr ardg =—Vo a(R? - 3). Sp Sp In a space-fixed coordinate system the absolute velocity at the oil exit is c= w — Vp, which is e(r) = Wo [ - (2) -Vp. The integral over Sr is then [Jeaasm2 ]{m f - (=)| - ve} rdraar (72 - ve) ,2.1 Conservation of Mass, Equation of Continuity 63 The continuity equation gives —Vp x (K-13) +23 (52 Vp) =0, or Ver? = wry 2 > W = 2vp (=) To which is identical with the result from a). Problem 2.1-7 Flow between two inclined flat plates Fluid with constant density g is located between two flat plates of length L. Both plates turn symmetric to z-axis towards each other with a constant angular velocity 1. The turning of the plates cause the fluid to move out of the plates. We assume that the plates in z-direction have an infinite width, justifying the assumption of plane flow. The velocity field in cylindrical coordinates is Ur, p) = u(r, p) & + welts e) ee whose radial component is given by u(r) = f(r)cos (9) with the unknown function f(r). a) Determine the wall velocity i(r) = uw(r)é, for both plates. b) Using the continuity equation in integral form, find the function f(r) forO cas wi = . or, R R For the same reason, we have n-é =sna <1 and m, can be neglected compared with nj. Thus, the normal component of the wall velocity can approximately be written as 7 thy ty Oy AA tye Ey FE yg Eg and with u,2 = 0h/dt it follows that [fe #dS = [f= tase [Ge wd. The expression €2+ dS is the projection of dS in é-direction, which is dz, dz3. Thus, Z1p [[t-aes & 6 [ Gan = —Wwcsnat cos des (Sa) 0 [f* #dS & ~b Rhw csinwtsin =. (Se) Inserting the above results into (2) gives V (210, t) = Rhwesinwt sin 2 o 4 Vv(0, t). R68 2 Fundamental Laws of Continuum Mechanics Problem 2.1-9 Effect of boundary layer displacement thickness Xa Incompressible fluid flows over a flat plate (width 6, length L) with constant velocity Up. The viscosity effect causes a boundary layer with the thickness 6(2,). Outside the boundary layer, the velocity is u. = Ua = const. We assume that the velocity distribution within the boundary layer follows a sine function with no-slip condition at the wall. a) Determine the mass flux through the surface BC of the sketched control volume. b) Calculate the velocity field within the boundary layer u,(x;). c) Calculate the mass flux through BC using u2(1, 22 = 6). Given: 6 = 6(2), 6, = 6(z1 = L), wn/Uo -{ sin (Sn22/6) for 0 < 22/8(x1) <1 1 for x2/8(2,) > 1 Solution a) Mass flux through the surface BC: The continuity equation in integral form for the sketched control volume gives: ffewtass ff oa-tas+ ff oa-tas+ ff ot-tas =o. (1) AB BC cD AD2.1 Conservation of Mass, Equation of Continuity 69 The mass flux through the surface BC is rngc, thus Jf ot-# a8 = sage. BC We calculate the surface integrals J[ot-tds = -ob6.00, B A 6 ffen-aas = eb [Uo sin (557) drz _ 26, Wt\)5t = ebte=* [oo (F2)], = obit vT [fewnes = and obtain from (1) o 2 —ob6, Ua + mao + = obéy Us = 0 2 + thec = 0b4,Uo (1-=). Velocity field: To calculate u2(z) with a given u,, we use the continuity equation in differential form (see F. M. (2.3a)): De Ou; De + ° an, =0. Since the flow is incompressible, for the present two dimensional case (8/823 = 0) we have: Ou, | Gug _ -f du, Bede ~9 % = [Ge te) The above integration constant may be a function of 2, but here we find f(z;) = 0 from the boundary condition up(z2 = 0,7) =0.70 & 2 Fundamental Laws of Continuum Mechanics Further we have diy —J Uo $48 cos (FB) for a2 < A(z), 0 Oey for zp > 6(x), with 6! = ag and obtain from (2) for 22 < 6(214): u2(z1,%2) = [tute cos (57) dry 0 o vtenay = toe {-2f)om(Z2)] +20 (22)} and for the edge of the boundary layer #2 = 6(1) ua(21, 2) = Uo &'(21) (1 - =) . (3) Mass flux through the surface BC with ua(r1, 22 = 6): ni We apply the continuity equation (/at = 0) to the sketched control volume: f[ot-tas=o0 (5) > f[ot-tass ffod-aas+ ffoz-tas =o. AB BC AC The first two integrals are already known from part a): + tine = eUobs, - ff ot- as. (4) AC On the surface AC is @ = Up &, + u2(21,6(21)) &2, thus: a-nads Up ey - AS + ua (a1, 6(a1)) e+ dS Up dirg dg — u2(r1, 6(21)) dry dig «2.1 Conservation of Mass, Equation of Continuity 71 With the above equations, the integral in (4) is calculated using (3) [J o-aas = et dose (1-2) sf en) ao = pUybi, —2b Uy (1-=) in, we see that (4) gives the mass flux calculated previously: mee = 9b6, Up (1 - =) . Problem 2.1-10 Flow through a diffuser with a linear velocity change in flow direction channel diffuser "88°! Two channels with the area A; and Ag are connected with each other by a diffuser of the length L. The shape of the diffuser is designed in such a way that it allows a linear change of the velocity component u in flow direction z from U; to Us, while c it remains constant over the cross-section A(z). The density 9 should also remain constant. a) Determine the distribution of the velocity component u(x) in the chan- nel, Find the change of the cross-section A(x). b) Calculate the local and convective acceleration in the diffuser for a constant inlet flow velocity Uj. c) Answer part b), for the case that U/, is time dependent with OU, /t = a, = const as given. Given: A,, Az, U,, £, 9 = const, ay Solution a) The velocity distribution is linear u(x) = ma+c. We find the constants cand m using the boundary conditions u(x =0) =U, and u(z = L) =U, Uy —-U, n=——. c=, and L72 b) 2 Fundamental Laws of Continuum Mechanics With the continuity equation in its integral form, we find Uy from Uy = U, A, /Aq and the velocity distribution wenf)Ee) We consider a control volume between z = 0 and an arbitrary fixed location 0 < x < L, The evaluation of the continuity equation ui dA = — iad flees fe aida results in u(a) A(x) = Uy Ay, and determines the diffuser cross-sectional area as a function of «: A, Ale) = (AFA, —ija/b 41 * For the case that U; is not a function of ¢, we find from (1) du/dt = 0, i.e. the local acceleration disappears. The convective acceleration is Ou UP Tf Ai \s \(4 ) w_LAf(ft_y\F2 Ary). 9 “Fe 74 Yet ylant (2) With U; = V(t) , OU, /At = ay and from (1) u = u(z,t). The local acceleration du Ay x wo ((E-) z+ 4 is not equal to zero. The velocity is a function of z only through the diffuser geometry. Since the diffuser geometry does not change, we obtain the convective acceleration again in the form of equation (2).2.1 Conservation of Mass, Equation of Continuity 73 Problem 2.1-11 Temperature boundary layer along a cold wall A gas at temperature T, and density g, is sud- denly brought in contact with a cold wall at tem- perature Ty. From the wall, a boundary layer de- velops into the gas with 6(t) = vt (v = const). The pressure is within the entire field constant. Inside the boundary layer, the gas temperature decreases linearly from T, to Tw. The gas density distribution is given by a(t) fer Soa * for 0< 2: < é(t) Bo for 22 > d(t) Relative to z1- and z-direction, the wall extension can be considered as infinite. Determine the velocity outside the boundary layer. The velocity in z,-direction is equal to zero within the entire field. Given: 9,, ew Solution We apply the continuity equation in inte- gral form to the sketched control volume (see F. M. (2.7)) [LJ sav =~ ff evmas. (1) | | | | a ww) | The left control surface is located outside F the boundary layer at an arbitrary but Xe fixed position 2 from the wall. The sur- faces with a distance a from the r2-axis can be displaced to infinity. Equation (1) per unit depth in 23-direction can be written as lim / / 28 aeydes =~ Jim / auj(d,)ndzy . (2) o a a0 The right hand side of the above equation already takes into account that no flow occurs through the cold wall and no velocity component in «1- direction exists. Outside the boundary layer all flow quantities are uniform.74 2 Fundamental Laws of Continuum Mechanics We form the local derivative of the density av oe {r= esa for 0 <2 < &(t) for z2.>6 and obtain from (2) / (ow — eat wor t2dz, = —8¢tt(Z2) . This equation is valid for arbitrary Z outside the boundary layer, where the velocity is ow v =-(*-1 . “ (* age Problem 2.1-12 Flow in a lubrication gap L 4] fa 1 2] The sketched “slide pad” with an infinite extension in z-direction has a gap height A(x) = hy — ax, with a= (hy — hy) fl €1. The slide pad, which is inclined at an angle a moves with a constant velocity U and drags the fluid with the density g = const into the gap. One would incorrectly expect that a linear velocity distribution u(x, y) would develop inside the gap. The non-slip condition at the wall is accounted for by u(z, 0) =0 and u(z, A(x)) =U cosa =U. Hint: The velocity component in y-direction at the top wall is of the order of magnitude aU and can be neglected. Furthermore, the gap pressure is only a function of x.2.1 Conservation of Mass, Equation of Continuity 75 a) Show that the volume flux in z-direction per unit depth h(x) V= f u(z,¥)dy , does not depend on 2. 0 b) The velocity distribution u(x, y) = Uy/h fulfills the required non-slip condition. Why is this velocity distribution not established? c) We introduce in the velocity distribution from part b) a correction term, which depends on the pressure gradient dp(z)/dz = —K(x) and has a quadratic term in y/h: _ yu, K(x) h? y v we =0 AOE (1) a Determine the negative pressure gradient A(z) with K(0) = Ay, such that the continuity equation is fulfilled. d) Obtain the pressure distribution in the gap by integrating K(x). The integration constants and Ky are calculated using the pressure bound- ary condition p(0) = p, = p(L) = pz = 0. e) Determine the volume flux through the gap. Given: 7, fi, Aa, L, U, p= p. =0 Solution a) V =const : For the sketched control volume, the continuity equation is hy h(x) { w0,y)dy= f ulz,y)dy. (1) o a The right hand side of (1) is equal to the volume flux V. Since the left hand side is constant, V is independent from x.76 2 Fundamental Laws of Continuum Mechanics b) The velocity u(x, y) = U y/A(z): ¢ We calculate the volume flux at z: . h(x) A(x) Ve / ue, y)dy =f Uy =5 Uh(z) = =U (hy — a2). a a (2) Thus, the volume flux is independent of x only for a = 0. For a # 0, the velocity distribution u = Uy/h does not satisfy the continuity requirement. Determine A(z): coe pure Couette flow Since, for a pure Couette flow the continuity requirement is not satis- fied, a pressure distribution p(x) must exist within the gap that causes a velocity profile at the gap inlet that is “thinner” and at the gap exit “thicker” with the result that V is now constant in the gap. The pres- sure gradient is determined by the continuity equation. We evaluate (1) for the given velocity distribution with (0) = 1 ht “er g Un + iG Be os = 5 U iz) + K(2) =V (3) or _# hh AY oy 1 Ke=-E =o (4 ae) wep ee] At the position z = «#*, with K(a*) = 0, the pressure distribution has an extremum and the velocity profile is that of a pure Couette flow. We obtain from (4) thus an equation for h(2*): Khe nter)= (14 FE) ty = ha, with the special gap height ho. We introduce this new constant into (4) and get dp ho a7 —K (x) = 6U 4 are ar : {5)2.1 Conservation of Mass, Equation of Continuity 7 d) Pressure distribution p(x): The integration of the pressure gradient (5) with respect to x results in polo polo p(t) = 6U n laep®-” | aap? : with / Vig t (Ld hap a \Alz) hy oO and rd wei(t 3 / AGP” a \iaye 0 Thus tom tse E erm} © and after a rearrangement (twice quadratic complement) Un f[to J]? [ be ; wz) = ot {[E- ~ fe The pressure is zero at the right gap boundary. From this condition and (6) we obtain the unknown special height ho: _g bik 0 hy + ha” (7) e) Volume flux through the gap: At 2 = 2" the velocity profile is a pure Couette flow, thus hy hg hy + hy” V=-Uhb=U78 2 Fundamental Laws of Continuum Mechanics 2.2. Balance of Momentum Problem 2.2-1 Principal axes of a stress tensor Given is the stress tensor in a non-dimensional form 5 v3 0 m=] V8 3 04. o 01 Calculate: a) the invariants y,, 27, and Js, of the tensor, b) its principal stresses o), o!?), and a!) c) and its principal directions. d) Determine the rotation matrix that transforms 7; to a diagonal form. Perform the transformation. Solution a) The invariants: hiz Toy Igy | . Ti = Ti + Ta + 733 (Tatjj — TT) = 20, det(ri;) 12. b) Principal stresses: The solutions of the characteristic equation -8+hro— hee + hh, =0, which is here -o° 4907-2004+12=0, are the required principal stresses eH=a1, oM=2, oM=e. c) Principal directions: The homogencous system of equations (ti _ al*} 633) nl) =0 has the solutions fork=1: nl =0 : nf) = , nasi, fork=2: n=sh , nay, nf=o, fork=3: naa, nasi, allo.2.2 Balance of Momentum 79 The solution vectors fi are already normalized and are thus unit vectors. Their direction is fixed except the sign. The sign of two vectors can be chosen arbitrarily. The direction of the third vector is determined in such a way, that #@) constitutes a right handed coordinate system. Thus, the following condition must be satisfied: HO y F@ tL ZO). Choosing mM =%, we find AS) = Principal axis transformation: 1, 3. WO = 5 1a 3, 41. qatge #4 coincide with & , 7) and #@) are in the 2,22-plane. The principal axis system is rotat- ed relative to the original coordinate system. The rotation matrix is calcu- lated from a;; = cos(d 2;, 24) and is presented in matrix form as: o f hW aj=|0 -i¥3 } . 2 2 1 0 0 The columns of the transformation matrix are the components of the eigenvectors i“) (called modal matrix). The transformation yields by , Thy = ik QjtTiy the values mM=1, thh=2, h=6, and = 0 foriZ#j, such that the tensor takes on the following matrix form: 100\ /o o 0 f=!o020]4+] 0 o 0 006 0 0 o® The stress tensor in principal axes is a diagonal matrix with the prin- cipal stresses on the diagonal.80 2 Fundamental Laws of Continuum Mechanics Problem 2.2-2 Fluid forces on a manifold The sketched pipe Ax branching is connect- wear ed with the pipe line system by three flexi- ble flanges (total spring stiffness cig:) at lo- cations [1], [2], and [3]. The pipe branch- ing can move only in 2- direction and the move- ment is considered frictionless. a) Calculate the velocity ug for the case that the flow at [1], [2], and [3] is fully uniform. b) Calculate the displacement Az of the pipe branching with respect to the position of rest (u; = ug = ug = 0), when the flexible flanges in the position of rest are not preloaded. c) Calculate the force acting on the pipe in y-direction. Given: p), p2, P3, U1, U2, Ar, Ap, 9 = const, Crot Solution a) The velocity us: To calculate us, we apply the continuity equation in integral form to the sketched control volume with Sw as the pipe walls. The flow is steady and incompressible, i. e. [[taas=o. (Ss) At the pipe walls Sy we have i-7i = 0. Since the velocities are uniform at the cross sections, we can write uA; + upA2 = usA; , or Aa ug = Uy + —Ug. Ay2.2 Balance of Momentum 81 b) Calculation of displacement Az: Using the integral form of the balance of momentum, we calculate the forces by the flow acting on the pipe branching by neglecting the body forces (see F, M. (2.43)) | ed (a- i) dS = [fies. (5) (5) For the control volume we find I oa (#7) ds+ f/f oa(a) as+ [f oa(z-a) d8+ ff ot (a) as = Ay Ay Ay Sw = [Jtas+ |ftass [frase [ftas. (1) The surface integrals are calculated term by term Jf et(t-i as = out Aé, [[ias=- [[ptas=n ae, A A A [fede %) 05 =-out nee, |[tes=- ffrtes=n ae, Ag Ag A; Jf ota) ds =o Ae, [fies [[pnas=-maz, Ag Ay As [foam aso. with uniform flow at {1], [2], and [3] resulting in #= —p#. The integra- tion of the stress vector f over the pipe wall Sy yields the force by the wall acting on the fluid. The reaction force is the required force Fret. by the flow exerting on the pipe branching. // fdS = Pein, =—Frup.. Sw Thus, we obtain from (1) ~ouiA, & — puzAaé, + uzdy & = (2) =n Aves + py Ar &e — ps Aree — Fina, « The force has only a component in z-direction: Fz = (py — ps)Ar + pA t+ o(uy — u§)A1 + ouj A82 2 Fundamental Laws of Continuum Mechanics The force equilibrium gives Fr = cot Ar, and thus, the displacement is calculated as 1 Ag = = [(p: — ps)Ar + pada + oui ~ ua)Ar + our 0 c) From (2) we obtain by a scalar multiplication with 2, F,=0. Problem 2.2-3 Calculation of drag force U. Un. -E GLEE oe ane .X,) \ Ue, Fluid with constant velocity U. and density 9 flows past an infinitely long symmetric cylindrical body. The flow direction coincides with the symmetry axis and the only force on the body is then the drag force Fp. Downstream of the body a wake flow is generated where the velocity u; is less than U/,,. With a given u;/U,. calculate the drag force Fp per unit depth acting on the body. wake Solution We choose a control volume with a control surface that encloses the body and then extends far enough from the body that all disturbances have died out sufficiently so that pressure differences to the undisturbed pres- sure vanish. Outside the wake flow we assume that the viscous stresses disappear.2.2 Balance of Momentum 83 Kg al c h D re —h a 2 ! Use II 1 Use t Hols ! q hoe I 7 {| jt x, | Z > i us 1 | we 1 I ae IK wake al 1 HH I h i contre! volume ‘lu. To determine the drag force Fp we use the integral form of the balance of momentum (steady flow, no body forces) [foma-aas = [fias. (i) (s) (8) The integral on the right hand side is decomposed into a surface integral along A, B, C, D, both sides of the slot, and one along the body surface Sg. The integrals over both sides of the slot will cancel each other, because of the opposite direction of the normal unit vectors. According to the previous assumption, on A, B, C, D is f= —pofl and we can write [fas = [[-mias + Jffas. (5) ABCD 5a The integral over A, B, C, D disappears, because no resultant force is acting on the closed surface, if fis the result of a constant pressure. The second integral is the force by the body on the fluid inside the control volume and is equal to the opposite force by the fluid on the body. We get from (1) —Futoay = ff et (@-i) dS, (s) from which we only need the 21-component: -Fo = —Fratody & = ff ou(a- it) as. (5) The surface integral of the momentum flux in z-direction can be decom- posed in part integrals, thus we write Fo = ff ow(a-a)as + ff ou(@-a) ds 4 a5 Be84 2 Fundamenta) Laws of Continuum Mechanics + ffowa-iyas + ff oui) as + CD DA + ff euta-aas + Jf una fy ds Se Satot On AB, CD, DA, uw = Uo, on BC, (i+ t) = us(x1, 22), on DA, (@-#) = —U,. and on Sg the product (i- 7) disappears leading to Us ff o(@-tas + ff owtas+ AB Be + Use ff oe) as - ff evr ds cD DA Jfevras - ff evjas+ DA BC - va ( ffewmas fe =m] ; (2) AB oD The integrals in parentheses are calculated using the continuity equation for steady flow ff oi-ndS =0: (5) [ow wast [fe Ga Hast [fees mas+ ffee-mas =0 = ffote-35 + [fete = Jf ev as ~ ffowas. AB eb DA BC The integrands are independent of x3. The force per unit depth is therefore —Fp => Fp Fp Cc Cc elk - foutar, - oU2At Uso | urd, B B > Fp = eu i (1- gt) a- Me ie2.2 Balance of Momentum 85 Since the integrand disappears outside the wake flow, its value does not depend upon h provided that h is larger than the wake width. Therefore k can approach infinity h — co and the drag force per unit depth is calculated as Uu uy Fp = 0, Sit = “(1 - a) dry. Since Fp/ol/?, is a constant, the value of the integral does not depend upon 2, , although u; = uy(x1,22) does, Thus, the integral is a measure for the momentum deficiency caused by the viscous flow. In boundary layer theory, this integral, which has the dimension of a length, is called the momentum thickness. Un. — uy = ug is called the velocity deficit uz. Thus, we may write Fp _ ae wt) Ug oU2, -/ (\-¢ Te dz. Downstream of the body is ug/Uoo 1 and the equation simplifies to Problem 2.2-4 Force on a slender nozzle A laminar flow (density p, viscos- ity 7) flows through the sketched slender nozzle: R(x) = Ry +(Ra—Ri) 5 At stations [1] and [2] the stress vector is given by f= —p7t. The pressures on both sides of the noz- zle p, and p) were measured. [1] [2] a) Calculate the velocity distribution inside the nozzle using the volume flux V and assuming a parabolic velocity profile with a mean velocity U, which is half the maximum velocity. b) Determine the force acting on the nozzle. Given: p1, po, V, R(z), |, @86 2 Fundamental Laws of Continuum Mechanics Solution a) Velocity distribution: The velocity distribution of a laminar flow through a slender nozzle is u(r,z) = Uma f - (es) | . The continuity equation is V=D(2)A(rt) = OU(2)= ES . Assuming U(x) = Umer/2, we obtain the velocity distribution Vv r \ wine) = 2 Fae) f - (ais) | — 9 Ye (arate) x (Ri + (CR: — Ri) 2/l)? Ry + (A, - M)e/l , Force on the nozzle: The momentum equation for the present case (constant density and no body forces) can be writ- ten as [feui-aas= [ftas. () (s) (Ss) b The flaw force Fy on the noz- zle has, because of flow symme- try, only a component in x- direction. To obtain this component, we multiply scalarly equation (1) with é&: [foun dAt ff oui(r)aa = [f» dA-Fy-ée+ ff —m dA. (2) Ay Ag Ay Az The integration is to be carried out over dA = r dr dy. The first integral on the left hand side is -f i} {2h [:- (z)| | + dra = eo 002.3. Balance of Angular Momentum 87 Ry “py? 2 9 2V r\? 4 ve = ~2re | (23) {1 - (x) | rdr= 3° oR (Substitution: t = 1—(r/f,)*, r dr = —Rj/2 dt) and the second integral on the left hand side Qn Ra [Je {25h 1-(£)]} rareo=$e¥, We solve (2) for the force vector: + - 4 V?/1 1 Fy-&= Fu =me melt tot (te ae) . 2.3. Balance of Angular Momentum Problem 2.3-1 Torque on pipe with slot The sketched fixed pipe has in the hori- zontal part a narrow slot with the width 5 and the depth 6 rp, where water (density ) exits horizontally. The water velocity is a linear function of r,. Viscous stresses may be neglected at all cross sections. a) Determine the torque in a3-direction as a function of Umax which is exerted by the flow on the pipe. b) For the given volume flux of V find the maximum velocity of the water jet. Given: b, to, V, 0, Po88 2 Fundamental Laws of Continuum Mechanics Solution a) The torque on the pipe: The flow is steady, the body forces Sp do not contribute to the moment of momentum in z3-direction. There- fore, we use the balance of angular momentum in the form described in (F. M. (2.54b)) and form the third component by a scalar mul- tiplication with é3 Sp 3 Se |[ea (Ex @)(@- i) dS = fe (@xidas. (1) (s) The control volume we choose, encloses the pipe and intersects it at the surfaces 5; and Sp. We split the integration into the sketched parts and consider that d+ # = 0 at solid walls Sw and Sp: / aes: (Ex a) (i-i) dS = |f& (ex as+ ff (fx ids. (2) SitSe SitSotSw To evaluate the integrals we consider individually ‘the integrands. Left hand side of (2): At the inlet 5; with @ = ug & é3- (fx wv) =—-é&-(@x 7) =-(& x d)-7=0, at the outlet S, with @ = u,@, F= 2, €, +726 +236, 1 = —-& é3-(@ xa) = éy- (21) + 2 €y + £3 €3) X U2 Ey) €3 + (w1 U2 &s — Ty uz Ey) = wy up and tw: fi = —ug. The velocity distribution at the outlet is linear: 8 to - 2) 62 For the left hand side of equation (1) we obtain Bry +b/2 8 ry —21\7 _ 2 oO 1 Vann | [ a( 625 ) dead, ug = u,(21) = —Umax for 2279 < 71 < 820. [fot (@xa(a-aas (s) 2z9 -b/2 = -TobU2..%- (3)2.3 Balance of Angular Momentum 89 Now we consider the right hand side of equation (2): Since the viscous stresses at S; and S, can be neglected (P;; = 0), the stress vector at the inlet 5; assumes the form f= —pfi and at the outlet S. we have t = —poft. At the wall Sw we also have f = —poi. Since the ambient pressure pp does not contribute to the torque, we may set Po = 0 and obtain [fw exdas+ fat (fxijdS= SitSo+Sw [fa @x (pay as+ [[e-@xijas. (4) Si Sp The first integral on the left hand side over the inlet surface S; disap- pears, because the vector product is normal to é3. The second integral is the moment in the intersection surface Sp, which is the reaction mo- ment that correspond to the moment Mr;.p, which is exerted by the fluid on the pipe // &-(@x i) dS = Ms, =—Mpiop. (5) Sp Thus, with (3), (4), and (5) we get the requested torque Mraop = Toba, U? : b) Maximum velocity Umax: With the given V, we calculate Umax using the equation V= ff z-tas So and obtain = bU nas / 8407 te, = 3bU mae to 620 =<: > Uinax = wo ~ 8 o90 2 Fundamental Laws of Continuum Mechanics Note: Frequently, it is appropriate to choose the control volume in such a way that it includes the fluid and the solid body. The concerns in connec- tion with the Gauss theorem (used in Reynolds transport theorem) which requires differentiability of the quantities in control volume, can be cir- cumvented by assuming that the transition between the fluid and the solid body is smooth but occurs with large gradients. As an alternative, one can place the control volume inside the pipe; the angular momentum on the wetted surface is the requested Mp).p. Problem 2.3-2 Moment exerted on the inlet guide vanes of a water turbine The sketched inlet of a ax, water turbine consists of t a fixed spiral casing and guide vane the inlet guide vanes. The spiral casing is designed in such a way that the fluid {9 = const) can ex- it the vanes at a constant velocity and a constant exit flow angle a. We as- sume a steady flow with fully uniform velocity profile at the inlet and exit of the vanes and we also neglect the body forces. . a) Determine for a given volume flux V, the magnitudes of the velocities &; and &. b) Find the component of &, in circumferential direction ¢,. at the vane exit. c) Calculate the torque in 23-direction exerted by the flow on the entire inlet guides. (Hint: In evaluating the surface integrals over the inlet surface .5;, the terms linear in Z, i.e. Z x Cand x fi can be set equal to the corresponding mean values and be taken out of the integrals. It can be proved that the value of the surface integral is not affected by this simplifying assumption.) Given: 9, V, e, hy b, d, Roy pj constant height h spiral casing2.3 Balance of Angular Momentum 91 Solution a) b) Velocity magnitudes c¢;, cy: From the definition of volume flux we have V=-ffedas= ffeaas. § So The following conditions are obviously valid — —¢ on 5; aai= : Cocosa@ on So thus, . Vv Ga (1) and . V = 2n Rh cosa” Components of é, in circumferential direction ¢uo: Vv ——— t. . 2 in Rah an a (2) Moment exerted on the inlet guide vanes and spiral casing: In the equation of angular momentum (steady flow, negligible body forces) [ferxe@-aas= [f#xias (3) (s) (Ss) we split the entire control surface (not shown in) in Cus = CoSINA = Cy = S=5;+5,+5w (Sw = wall surface of guide vanes) and (3) yields, since, as a consequence of the kinematic boundary con- ditions, [[e@xale-a)as=o, Sw the following results f[ezx e(é-) 45+ ff ot x &(@-it) ds i So = [fates [fextas+ [fextas. (4)92 2 Fundamental Laws of Continuum Mechanics The last integral on the right hand side is the moment, which the vanes and spiral casing exerts on the fluid. A sign change yields the reaction moment exerted by the fluid on the inlet guide vanes. The evaluation of the integrals gives the individual contributions: 1) On Sj, Fx &= (ay + dé) x (—c; &) = cd & = ffczx #2. f) dS = —pVeid & . Ss 2) On So, Fx E= (Map) X (Cro & + Cus By) = Rotus & > [fezxee-a) dS = pV Rotus & . So 3) On S;, the flow is uniform, i. e. f= —pii. Because ff = & we have therefore x T= (ei @) + dé) x (—pi&) = pide > []#*t88=ponaes, 4) On S,, the flow is also uniform, i. e. f= —pit = pé,, therefore #xT=(R,é,) x (pe) =0 > |[#xtas =o. So Thus, we obtain from (4) the moment exerted on the inlet guide vanes and spiral casing as Mpiw = oV (cid — cyoRo) Es + pibhd& , and the magnitude of the non-zero component is Mg = eV (eid — cue.) + prbhd . (5) The first term in (5) represents the change of angular momentum of the fluid and corresponds to the Euler turbine equation, the second term originates from the fact that the inlet surface is not a surface of revolution, where the stress vector (here # = —pii) generates a moment. Introducing in (5) c; and c,, from (1) respectively (2), we obtain the result 6 On 23 my = 2 (§ tana ) trond.2.3 Balance of Angular Momentum 93 Problem 2.3-3 Curvature radius of circular arc profiles of a circular cascade A swirl free fluid with density @ and the velocity c; enters the sketched fixed cascade that consists of 6 stationary blades. At the exit, the flow angle is J. a) Assume a channel height H and calculate the mass flux m through the cascade. b) Determine the torque by the flow on the cascade. c) Find the force component in circumferential direction that is exerted on one blade, if the force acts at the radius r;, on the blade. d) As we know, the lift force F; in cascade flow is perpendicular to the mean velocity = G+ wo = a Calculate the angle + between ¢, and Uno. e) Determine the lift force on one of the blades. f) For a small angle of attack, the lift force of a single circular arc profile can be calculated from the following relation: ~~ f - ou = sy = 2 (a2 | ; a=a,-7. This relation can approximately be used for one blade within the cas- cade, provided that the blade spacing is much greater than the blade chord.94 2 Fundamental Laws of Continuum Mechanics How should we choose the radius of curvature in order to determine the lift force calculated in e)? Use a circular are profile of chord length i and radial angle a. a, is the given angle between the chord and the radial direction. (Hint: For the circular arc profile the relation: P=4(2fr— f*) is valid) Given: 9, ¢1, a, 82, Ri, Re, re, Hy 1 Solution a) b) Mass flux through the cascade: m=poq27h, HH. Torque T.4s. on the cascade: The Euler turbine equation (here for swirl free inlet flow condition) for the torque on the fluid gives T = m(eu2) , where the unknown circumferential component of the exit velocity is calculated from (positive circumferential direction is anticlockwise): R tan By = —c, tan By. — o2n Ry H Ry Ca = — Cg tan By = For the torque on the cascade (Tas: = —T) we find Tease = ci Ri 2a H tan By : Blade force in circumferential direction: The relationship between the torque and the force per blade in circum- ferential direction is given Tease = 2 Fur (z = number of blades) the circumferential component of the force on a blade is Tease _ Att H Fi= uw ETE arp ontan Bo.2.3 Balance of Angular Momentum 95 d) Angle between @ and Uz»: The result can be read from the sketch: cuz _ 1 Ry / Ry tan Bp tan = = 7 Cit Ge ate Ri/ Ry t > 7 = arctan RT 1 e) Blade lift. force: . The lift force is perpendicular to U,. Thus, we have for the relationship between the lift force and the circumferential component of the blade force A fl= . cosy Introducing the circumferential com- ponent, we obtain T Zr, cosy Fy = ce? R?H omtan f, 3r,.cos (arctan nae) f) Radius of curvature: Using the relationship for the lift coefficient given in the problem defi- nition we obtain the following expression for f96 2 Fundamental Laws of Continuum Mechanics or else 2 2 7 r= e Ay A tan §, -all + 4 | 37, (U2,/c2) cos (arctan ss) 1 R? H tan By al 4 | 3r, (U2,/e2) cos (arctan As) , 2.4 Momentum and Angular Momentum in an Accelerating Frame Problem 2.4-1 Fluid sprayed on a rotating disk fa yt a Fa Ot Inviscid fluid is sprayed on a disk rotating with angular velocity &. The spray nozzle with the tip located at the center, sprays a jet with the velocity € = coéy. The body forces ok are neglected and the jet is subjected to ambient pressure. a) Determine the path of the fluid particles in the inertial system (z', y’). b) Determine the same path in the rotating system (2, y) using the coor- dinate transformation. c) Calculate the path directly by integrating the equation of motion in the rotating system. Solution a) The path in the inertial system: The fluid is inviscid, the pressure p in the spray jet is constant, and the body forces are neglected, as a result, we obtain from Cauchy's law of2.4 Momentum and Angular Momentum in an Accelerating Frame 97 b) motion e a =o R+V-T the differential equation for the pathline Dé/Dt = 0, which, with the initial condition, yields the constant vector @ ~_ di c= a =z. The components in the inertial system (7 = 2’ + y'€y) are written as da! = or x’ = cot + const ae - di a + =0 or y’ = const . With the initial conditions 2‘(0) = 0 and y’(0} = 0, the path of fluid particles in the inertial system is calculated as , r® = pt, ‘ y = 0. The path in the rotating system: For the coordinate transformation we have , j= Aji Ty. The rotation matrix a, is __ f cosMt sin Qt = \_ sin Mt cos Mt} This leads to zc = cosMizc'+sin Qty’ = cot cos Mt , y = —sinNts’ +cosMty’ = —cof sin Nt, and in cylindrical coordinate system r= yetty?=cot, y = arctan () =—-Nt, zr and after eliminating t, we obtain 2 =-—r. ° co98 2 Fundamental Laws of Continuum Mechanics c} Calculate the path directly integrating the equation of motion in the rotating system: For the present case, we take from (F. M. (2.68)) the acceleration in the rotating system Di - | wos pp TAR * E-Qx (QZ). With w = ue, +veé,, =, and u = , v = y, the components in z- and y-direction are calculated: # = 207402, § = -20240N?y. The two coupled, ordinary linear differential equations with constant coefficients can be rearranged using the following definitions _fio _ {0 20 _{- 0 m=(5t)) P= (zn G)) = (9° te) M#+Dz4+Kz=0. (qa) The solution # = Ce* leads to the eigenvalue problem (VM+AD+K)G=06. (2) Non-trivial solutions of the homogeneous systems (2) exist only if the determinant of the coefficient matrix disappears, i. e. det(A7M+AD4+K) = 0 vag? -20 > aat ( 2A en) = 0 = (VWF parr at4+ey" = 0. 4 = +i are each double eigenvalues, i. e. besides Ce*! also Cte* is a solution of (1). The eigenvectors 6 are calculated using the already known eigenvalues from (2). However, because the determinant is zero, only one equation can be used. We choose the first one (V-97)C,-249C, = 0 A? — 9? eo = AA2.4 Momentum and Angular Momentum in an Accelerating Frame 99 and obtain for A = A; = +i? the first eigenvector: —20? = of) = cM ZO = oi = 60 = a(*) with Cf =a, and for 4 = Az = —19 the second one: ) = cM = Cc?) cj => 6 = s(3,) with Cc? = ap. The general solution of (1) is z= (jetta (1; ) einen [a (;)er+a( 2 )ein| . The constants a1, 4, a2, 82 which may be complex, can be determined from the initial conditions 2(0) = 0, (0) = co é, (if complex, then four equations for four complex constants), We get #0) <0 = (7) +0 (1) > 0 = a+a,, 0 = i(a,—ay). These two equations are satisfied only if a, = a2 = 0, therefore we find t= ,(1) corsa,( 1.) eines [isin (;) ent ipa ( 1,)ering . With £(0) = (co, 0)", we finally get a(;)+a(1) =(¢) > f& = B+h, o= i(8, — B2) > A= = 3: and considering the initial conditions, we arrive at re o()ereg( Je]100 2 Fundamental Laws of Continuum Mechanics We decompose the above vector equation into its components and ob- tain the pathline in the form already know from part b): 1 i. 1 i, z(t) = cot (5 cost + 5 sine + 5 cost — = sia 2) = col cosMt , y(t) cot (5 cost - 5 sin Qt — posit — >sin 0) = cot sin. Problem 2.4-2 Velocity of a moving container with a nozzle plate A cart with a nozzle contains fluid of con- stant density. Its to- tal mass at ¢ = 0 is mo. A slow motion of a heavy plate generates a constant mass flux m through the nozzle with an exit velocity wW, relative to the con- tainer. The flow in the relative systern is steady. We assume that air drag and contact friction are negligible. Calculate the velocity v(t) of the cart. LILI LLL ELLE Given: mg, mh, w,, v(t = 0) = 0 First. Solution The equation of motion for the container is with mg as the mass of the container without fluid and /, the force com- ponent in direction of container motion.2.4 Momentum and Angular Momentum in an Accelerating Frame 101 The force F acting on- ly on the container is calculated by integrat- ing the stress vector over the entire control surface Neglecting the air drag, we set the stress vector /= —po rt on S, and obtain TdS = |] —piidS = —poi dS — ff —pids. frs- owns omnes f Since S,+ Ap is a closed surface, the first integral is equal to zero, therefore —poTi dS = po ff dS = po Age. The force component in direction of the container motion (negative in z-direction) is calculated from (1) Fy =F .(-&) =p Act ff 1 dS. (2) Si The integral on the right hand side represents the fluid force that acts on the container. The integral in (2) is car- ried out by applying the balance of momentum in integral form on the flu- id in the container. We use the sketched accelerat- ed frame. The momentum equation for the control volume is (see F. M. (2.73)) (Ff veav] + ff eztw-a) as+ix fff ozav = fff ek av+ ffias. (¥) a @&) (Y) (¥) (s)102 2 Fundamental Laws of Continuum Mechanics With . e=W+7+Nxtad+7, N=0 and S= Sw +Ap+ A, we obtain the component in z-direction (5 fife w) + ff eciti-it)as+ ff ocs(w- it) 45+ a) a Bw i + feats. - fi) aS = [Jas [nas (3) Sw+Ap At the wall Sw the product w+7 and thus the integral will disappear. The first integral on the right hand side represents the force by the con- tainer exerted on the fluid. To obtain the reaction force, the sign needs to be changed. On the surfaces Sy + Ap, the normal unit vectors show an outward orientation, whereas on the surface S; they are oriented inwardly. Therefore, we have [[uas= [- ds. SwtAp This expression corresponds to the required integral in (2). On A, the flow is uniform, i. e. ! = —pi. Furthermore, the jet pressure is equal to the ambient pressure po, consequently: ic ds = Jf vm dS =—poA, . Ao Ao We solve (3) for the required integral and obtain f-sas-nas(2 J ror] + Si v) A + ff ead-tds++ ff ocd-itas. (4) Ap Ao With c) = w,+v, = uw; — v(t), we write the first integral on the right hand © (ffeet) = fo)2.4 Momentum and Angular Momentum in an Accelerating Frame 103 The velocity w, inside the container can always be neglected w, ~ 0 and thus, also Ow,/0t = 0. At the exit nozzle the velocity w, is large but there it is steady w; = w, = const. This leads to fe feoe) Effet A with mp as the fluid mass within the fixed control velume. For the second integral in (4) we obtain Jfecd-aas= ff ow — vo(t))d-#dS = -v(¢) ff owas. Ap Ap Ap The last integral can be calculated using the continuity equation I wee [few-aas, Ise 0, (the density 9 is constant) and [[eaaas+ ffow-tas+ ffow-aas =o Sw =0 Ap Ae oT with it follows | wad = —in Ap and [feces = rho(t). Ap Finally, the third integral on the right hand side in (4) is [feaa-aas = ff ow — v(t) B+ dS = (w, ~ v(t), ‘Ao Ao such that from (4) the following equation can be generated I[- aS = po A, me + rro(t) + re(we — v(t) Si