Advanced Transport Phenomena Module 4 - Lecture 14: Momentum Transport: Flow Over A Solid Wall

Advanced Transport Phenomena
Module 4 - Lecture 14
Momentum Transport: Flow over a Solid Wall
Dr. R. Nagarajan
Professor
Dept of Chemical Engineering
IIT Madras
FLOW OVER A SOLID WALL: SURFACE
MOMENTUM-TRANSFER COEFFICIENTS
 Applications:
 Design of automobiles
 Design of aircraft, etc.
 Property of interest:
 Momentum exchange between surface & surrounding
fluid
 Associated net force
 “drag” in streamwise direction

 ‘lift” in direction perpendicular to motion
 Obtained by solving relevant conservation equations,
subject to relevant boundary conditions, or
 By experiments on full-scale or small-scale models
Momentum exchange between the moving fluid and a representative

segment of a solid surface (confining wall or immersed body)
 X  approach stream direction
 x  distance along surface
 n  distance normal to surface
 p (x,0)  local pressure
 vx (x,n)  velocity field

 tnx (x,0) = tw(x)  local wall shear stress
 Associated momentum exchange:
 Force on fluid
 Equal & opposite force on solid

 Solid surface motionless => vn (x,0) = 0
 vx (x,0) ≠ 0 => nonzero “slip” velocity

 However, experimentally: local tangential velocity of
fluid = that of solid, i.e., 0, under continuum conditions
 Wall shear stress depends on local fluid-deformation rate:
 vx 
t w ( x)   w  
 n  n 0
 Can be determined if local normal gradient of
tangential fluid velocity can be measured
 Dimensionless local momentum transfer coefficients:
 Pressure coefficient:
p( x,0)  p
C p ( x) 
1
U 2
and 2
 Skin-friction coefficient
t w  x
C f ( x) 
1
U 2
2
 Measured or predicted
 Alternative definition of skin-friction coefficient:
 In terms of properties at the edge of momentum transfer
boundary layer t w ( x)
C f ( x) 
1
e vx2,e  x 
2
 For an incompressible fluid (   e ), in the absence of
gravitational body-force effects, Bernoulli’s equation
yields: 1 1 2
p0  p  U  pe ( x)   vx ,e ,
2
2 2
 Reflects negligibility of viscous dissipation far from

surface
This equation implies that:
 1  C p 
v x ,e ( x ) 1/2
U
and hence:
C f ( x)
c f ( x) 
1  C p ( x)
Experimentally determined angular dependence of the skin-friction-

and pressure- coefficients around a circular cylinder in a cross-flow
at Re= 1.7x105
 Total (net) drag force D’ per unit length of cylinder:
 Reference Force: 1
Dref  U 2 Aproj
' '
,
2
where projected area of cylinder per unit length
 proj 
A'
cylinder
 dw
and D'
 CD cylinder  1
U 2 d w
2
 Calculated from Cp, Cf data
 By projecting pressure & shear forces in direction of
approach flow:
 dw 
D '  2    p   cos   t w   sin   d 
0 2 
and

 CD cylinder   C p   cos   C f ( )  sin  d ,
0
(polar angle expressed in radians)

 cos  term  from (locally normal) pressure force (“form”
drag)
 sin  term  from (locally tangential) aerodynamic shear

force (“friction” drag)
 Thus, drag coefficient may be split into:
 CD cylinder   CD  form   CD  friction

Experimental values for the overall drag coefficient (dimensionless

total drag) for a cylinder (in cross-flow), over the Reynolds’ number
ranger 101  Re  106
Experimental values for the overall drag coefficient (dimensionless

total drag) for a sphere over the Reynolds’ number range 101  Re  106
 Asymptotic theories: Re >> 1, Re << 1
 Re >> 1 case is of greatest engineering interest
 e.g., flow past flat plate at zero incidence

 Momentum Diffusion Boundary Layer Theory: Laminar
Flow Past Flat Plate at Zero Incidence
 1904: L Prandtl
 large but finite Reynolds number

 vn and vx vanish at solid surface
 Thin transition layer near surface across which vx
abruptly drops to zero
 Inside this “boundary layer”, velocity gradients
vx / n large enough to make momentum diffusion
important (though  is small)
 Exterior: inviscid region
Division of flow field at Re1/2 >>1 into an inviscid “outer” region and a thin
tangential momentum diffusion boundary layer (BL)(after L. Prandtl).
 1904: L Prandtl
 For Re >> 1, within the BL:

 vn << vx
 Momentum diffusion important, but only in normal
direction (tnx >> txx)
 Pressure at any streamwise location x is nearly
constant– i.e., p ≈ pe(x)
 BL equations therefore simplified, solutions to match
inner behavior of external inviscid flow
 e.g., 2D steady flow of incompressible constant-
property Newtonian fluid past a semi-infinite flat
plate at zero incidence (Blasius, 1908)
D'
 CD cylinder 
 D cylinder  CD  form1  C2D  friction
C 
U d w
2
Newtonian incompressible fluid flow past a flat plate; configuration, nome-

nclature, and coordinate system
 Laminar BL on a flat plate:
 For thin flat plate, pressure constant everywhere => no
need for y-momentum equation
 2 scalar PDE’s governing vx ≡ u(x,y), vy ≡ v(x,y)
u v
 0 (mass),
x y
u u  2u
u v v 2 ( x  momentum),
x y y
Subject to boundary conditions:
u  , y   U ,
u  x,    U ,
u ( x,0)  0,
v( x,0)  0,
 Solved by Blasius using “combination of variables”

 Blasius’ solution:
u  1 y  Ux 1/2 
 fct1  .     fct1  
U  2 x  v  
 
and
1/2
v  Ux   1 y  Ux 1/2 
 fct2   fct2  
   2 x  v  
.
U v   
 Blasius derived & numerically solved nonlinear ODE
governing

u ( )
f     d ,
0
U
and constructed tangential fluid-velocity profiles
Laminar BL on a flat plate:
  ( x) = local BL thickness = y-location at which u/U =
0.99  occurs at   5
 Therefore:
1/2
 Ux 
  5x  
 v 
(grows as square root of distance x from LE of plate)
 Wall shear stress: 1/2
 Ux 
t w ( x)  0.332 U . 
2
 v 
 Local dimensionless skin-friction coefficient cf given
1/2
by: tw  Ux 
 c f ( x)  0.664  
1
U 2  v 
2
 Total friction drag coefficient:
D(both sides) 1.328
cf  
 
1/2
1
U Aw
2 Ux / v
2
(for plate of finite length L, set x = L)
 Effect of “blowing” or “suction” through porous solid wall:
 cf values are modified

 Blowing can reduce skin-friction drag
c f ( x / L, Re,...)  c f ( x / L, Re,...)  . F (blowing ),
0
where (cf)0  no-blowing momentum-transfer coefficient, and

F(blowing) function of dimensionless variable
 w vw
eue c f ,0
(  w vw m ''
w= local mass injection rate)
CONSERVATION EQUATION GOVERNING
VELOCITY AND PRESURE FIELDS
 Navier-Stokes (linear momentum conservation) law:
 v   2 
   v.grad v   grad p  div  2 Def v    div v  I    g
 t   3 
 Nonlinear vector PDE
 Equivalent to 3 independent, scalar 2nd order PDEs
 Includes “Stokes’ Postulate”: bulk viscosity  can be

neglected
 Total mass conservation (“continuity”):

 v.grad     div v
t
CONSERVATION EQUATION GOVERNING
VELOCITY AND PRESURE FIELDS
 Conservation equations provide 4 PDEs for 5 fields: v (3
scalar fields), p, 
 Hence, necessary to specify an EOS for closure
 Unless  is constant (incompressible flow; div v  0)
 “Caloric” EOS: h as a function of T, p
 In addition to usual p (  , T ),  as a function of local

state variables
 Turbulent flows:
 Conservation equations are time-averaged
  replaced by   t
TYPICAL BOUNDARY CONDITOINS
By applying a “pillbox” control volume to straddle a

moving interface, we can write:
Gn  normal component of mass flux
t  tangential plane
Mass balance:
Gn   0
Momentum Balance:
Gn vn     p  t nn 
Tangential linear momentum:
Gn vt   t nt 
TYPICAL BOUNDARY CONDITOINS
 These conservation equations allow:

 Discontinuity in normal component of velocity,
 Discontinuity in pressure across interface,
 Discontinuity in tangential velocity (“slip”) across

interface
 Thus, the “classical” boundary conditions:
vn   0, vt   0,  p  0, t nt   0
are only sometimes true.

TYPICAL INITIAL CONDITOINS
 State of independent field variables at t = 0
 Start-up of a chemical reactor, separator, etc.
 The present, if we want to predict future (e.g., weather,

climate)
 Governing conservation equations are first-order in time
 Invariant wrt shift in origin (zero point) chosen for time
 Principal of “local” action in time (determinacy)
 Future cannot influence present!
 Only applies in time-domain, not space

SOLUTION METHODS
 Coupled PDEs + bc’s + ic’s need not always be solved to

extract valuable information
 e.g., similitude analysis
 Only relatively simple fluid-dynamic problems need to be

solved to interpret instrument readings
 e.g., flowmeters
 Mathematical solutions have become possible with advent

of powerful digital computers
 Computational fluid mechanics, CFD
 Discretizing by finite-difference, finite-element methods

SOLUTION METHODS
 Modularization:
 In sub-regions, explicit results may be possible in
terms of well-known special functions
 e.g., Bessel functions, Legendre polynomials
 Numerical:
 Reduce problem to solution of one lor more nonlinear
ODEs
 Then solve numerically
SOLUTION METHODS
“Road map” of common methods of solution to problems in transport (convection

/diffusion ) theory
SOLUTION METHODS
 Results should be independent of method chosen

 But effort should be minimized!
 Idealizations of complex problems serve a purpose

 Capture concepts
 Bring out qualitative features
 Sanity check on more complex predictions

Thank You

Advanced Transport Phenomena Module 4 - Lecture 14: Momentum Transport: Flow Over A Solid Wall

Uploaded by

Copyright:

Available Formats

Advanced Transport Phenomena Module 4 - Lecture 14: Momentum Transport: Flow Over A Solid Wall

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Advanced Transport Phenomena Module 4 - Lecture 14: Momentum Transport: Flow Over A Solid Wall

Uploaded by

Copyright:

Available Formats

Advanced Transport Phenomena

Dept of Chemical Engineering

 Design of aircraft, etc.

 “drag” in streamwise direction

Momentum exchange between the moving fluid and a representative

 vx (x,n)  velocity field

 Equal & opposite force on solid

 vx (x,0) ≠ 0 => nonzero “slip” velocity

 Reflects negligibility of viscous dissipation far from

Experimentally determined angular dependence of the skin-friction-

(polar angle expressed in radians)

 sin  term  from (locally tangential) aerodynamic shear

 Thus, drag coefficient may be split into:

 CD cylinder   CD  form   CD  friction

Experimental values for the overall drag coefficient (dimensionless

Experimental values for the overall drag coefficient (dimensionless

 Re >> 1 case is of greatest engineering interest

 e.g., flow past flat plate at zero incidence

 large but finite Reynolds number

 For Re >> 1, within the BL:

Newtonian incompressible fluid flow past a flat plate; configuration, nome-

 Solved by Blasius using “combination of variables”

 cf values are modified

where (cf)0  no-blowing momentum-transfer coefficient, and

 Nonlinear vector PDE

 Equivalent to 3 independent, scalar 2nd order PDEs

 Includes “Stokes’ Postulate”: bulk viscosity  can be

 In addition to usual p (  , T ),  as a function of local

By applying a “pillbox” control volume to straddle a

 These conservation equations allow:

 Discontinuity in pressure across interface,

 Discontinuity in tangential velocity (“slip”) across

are only sometimes true.

 State of independent field variables at t = 0

 Start-up of a chemical reactor, separator, etc.

 The present, if we want to predict future (e.g., weather,

 Invariant wrt shift in origin (zero point) chosen for time

 Principal of “local” action in time (determinacy)

 Future cannot influence present!

 Only applies in time-domain, not space

 Coupled PDEs + bc’s + ic’s need not always be solved to

 Only relatively simple fluid-dynamic problems need to be

 Mathematical solutions have become possible with advent

 Discretizing by finite-difference, finite-element methods

“Road map” of common methods of solution to problems in transport (convection

 Results should be independent of method chosen

 Idealizations of complex problems serve a purpose

 Bring out qualitative features

 Sanity check on more complex predictions

You might also like