See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/357435147

Plane Stagnation Flow (Hiemenz Flow)

Technical Report · December 2021

CITATIONS READS
0 789

1 author:

Brian G Higgins
University of California, Davis
176 PUBLICATIONS 2,377 CITATIONS

SEE PROFILE

All content following this page was uploaded by Brian G Higgins on 30 December 2021.

The user has requested enhancement of the downloaded file.

 BGHiggins/UCDavis/December27 _2021

Plane Stagnation Flow

(Hiemenz Flow)
Background
Plane stagnation flow ( also known as Hiemenz flow) is a classical 2-D flow in fluid mechanics that can
be solved exactly using a similarity solution. A schematic of the flow is shown in the figure below. The
development of the theory can be found in many classical fluid mechanics text books ( e.g. Boundary
Layer Theory by H. Schlichting, 1968)

Far from the plate the flow is described by a potential flow given by
U = a x, V = -ay (1)

The pressure in a potential flow is given by Bernoulli’s equation, which in the absence of a body force
can be expressed as
1 1
p0 - p = ρ U2 + V2  = ρ a2 x2 + y2  (2)
2 2
At the plate the flow must satisfy the no-slip boundary condition, giving rise to a viscous boundary
layer. For the flow in the boundary layer one postulates that the velocity field is given by:
f
u (x, y) = x , v (x, y) = -f (y) (3)
y
Note the postulate for the velocity field inspired by the potential field automatically satisfies continu-
ity, that is
2 2DStagnationFlow.nb

∂u ∂v
∇ ·v ≡ + =0 (4)
∂x ∂y
Further we assume that the pressure field derived from the potential flow is impressed on the bound-
ary layer. Thus for the pressure field we assume
1
p0 - p = ρ a2 x2 + G (y) (5)
2
where the function G(y) is to be determined.
The 2-D momentum steady momentum equations are
∂u ∂u ∂p ∂2 u ∂2 u
ρ u +v =- +μ +
∂x ∂y ∂x ∂x2 ∂y2
(6)
2 2
∂v ∂v ∂p ∂ v ∂ v
ρ u +v =- +μ +
∂x ∂y ∂y ∂x2 ∂y2
Note the effects of gravity can always be included in the pressure field by defining a scalar potential for
the body force
∇ϕ = -ρ g (7)

so that
 = p+ϕ (8)

Substituting Eqs. (3) and (5) into Eq. (6) gives the following ODES for the unknown functions f(y) and F(y)
(f')2 - ff'' = a2 + νf''' (9)
1
f f' = a2 G' - νf'' (10)
2
The no-slip boundary conditions at the surface of the plate y = 0 gives
f
f (0) = 0, (0) = 0 (11)
y
Far from the plate, the flow must merged into the potential flow. Thus
f
 a, f (y)  y, as y  ∞ (12)
y
Note that the unknown function F(y) that determines the pressure field does not appear in Eq. (8). Thus
one can solve for f(y) and then thereafter solve for G(y).
The nonlinear ODE given by Eq. (9) cannot be solved in closed form. To facilitate a numerical solution it
is best to rescale the equation so that the constants a2 and ν are removed. This can be done by writing
z = α y, and f (y) = A F (z) (13)

so that Eq. (9) becomes

F 2 F
α2 A2 F'2 - F F'' = a2 + ν A α3 F''', where F' ≡ , F'' ≡ , etc
z z2
 2DStagnationFlow.nb 3

Now if we put
α2 A2 = a2 , ν A α3 = a2 (14)

such that
A= νa, α = a/ν (15)

and
z= a / ν y, f (y) = a ν F (z) (16)

Then the differential equation for F(z) becomes

F''' + F F '' - F' 2 + 1 = 0 (17)

subject to the boundary conditions

F
z = 0 : F (0) = 0, (0) = 0
z
(18)
F
z = ∞: (∞) = 1
z
Equation (17) subject to boundary (18) is a nonlinear boundary value problem that must be solved
numerically. This is best done using a shooting method, as described in the next section

Shooting Method: Stagnation Flow in a Plane

The code for the shooting method is shown below. The parameter Ω1 is the shooting parameter that
must be determined such that BC at infinity is satisfied. Note: in a numerical solution one truncates the
infinite domain. For our numerical solution the truncated domain is 0 < z < 6. That is we guess a value
for 2 F  z2 at z=0, say Ω1 and then determine what the value for F/Z is at z=6. If it is not equal to 1,
we select a new guess for Ω1. This selection process can be done using a Newton iteration method.
In[65]:= ODE[Ω1_] := - F '[z]2 + F[z] × F ''[z] + 1 + F '''[z]  0, F[0]  0, F '[0]  0, F ''[0]  Ω1;
sol[Ω1_] := NDSolve[ODE[Ω1], F, {z, 0, 6}];
FendBC[Ω1_ ? NumericQ] := First[(F '[z] /. sol[Ω1]) /. z  6];
bc = FindRoot[FendBC[Ω1]  1, {Ω1, 1, 1.3}, MaxIterations  500]
Out[68]= {Ω1  1.23259}

Our Newton iteration determines that the correct value for Ω1 is 1.23259
Here is the plot for F(z)
4 2DStagnationFlow.nb

In[69]:= plt1 = Plot[Evaluate[F[z] /. sol[Ω1 /. bc]], {z, 0, 5},

PlotStyle  {Blue, Thick}, FrameLabel  {Style["z", 16], Style["F(z)", 16]},
Frame  True, AspectRatio  1, ImageSize  300]

3
F(z)

2
Out[69]=

0 1 2 3 4 5

z
Here is the plot for F’(z) which is the scaled u velocity in the x coordinate direction, i.e. u=x F(z)
In[71]:= plt2 = Plot[Evaluate[F '[z] /. sol[Ω1 /. bc]], {z, 0, .5},
PlotStyle  {Blue, Thick}, FrameLabel  {Style["z", 16], Style["F'(z)", 16]},
Frame  True, AspectRatio  1, ImageSize  300]
0.5

0.4

0.3
F'(z)

Out[71]= 0.2

0.1

0.0

0.0 0.1 0.2 0.3 0.4 0.5

z
The following plot shows that F’[z] does indeed approach unity as z  ∞.
 2DStagnationFlow.nb 5

In[120]:= plt2b = Plot[Evaluate[F '[z] /. sol[Ω1 /. bc]], {z, 0, 5},

PlotStyle  {Blue, Thick}, FrameLabel  {Style["z", 16], Style["F'(z)", 16]},
Frame  True, AspectRatio  1, ImageSize  300, PlotRange  All]

1.0

0.8

0.6
F'(z)

Out[120]=
0.4

0.2

0.0

0 1 2 3 4 5

z
From the above plot we can deduce that the boundary layer thickness is about z=2.0, and this value as
indicated in the figure above is independent of x.
Finally here is the plot for F’’(z).
plt2 = Plot[Evaluate[F ''[z] /. sol[Ω1 /. bc]], {z, 0, 3},
PlotStyle  {Blue, Thick}, FrameLabel  {Style["z", 16], Style["F''(z)", 16]},
Frame  True, AspectRatio  1, ImageSize  300]

1.2

1.0

0.8
F''(z)

0.6

0.4

0.2

0.0

0.0 0.5 1.0 1.5 2.0 2.5 3.0

z
The value of 2 F  z2 at z=0 is
6 2DStagnationFlow.nb

F ''[z] /. sol[Ω1 /. bc] /. z  0

{1.23259}

This value is in excellent agreement with the value determined by Howarth, see Table 5.1 in Schlicht-
ing’s Boundary-Layer Theory textbook.

NDSolve Solution
Mathematica’s NDSolve function can solve Boundary Value problems directly with a built -in shooting
method. Here is how the method can be used in the above problem:
In[111]:= ODE2 := - F '[z]2 + F[z] × F ''[z] + 1 + F '''[z]  0, F[0]  0, F '[0]  0, F '[10]  1;
sol2 = NDSolveValue[ODE2, F, z, Method  {"Shooting",
"StartingInitialConditions"  {F[0]  0, F '[0]  0, F ''[0]  1.3}}]

Domain: {{0. , 10. }}

Out[112]= InterpolatingFunction 
Output: scalar

In[60]:= First[F[z] /. sol2]

Domain: {{0. , 10. }}

Out[60]= InterpolatingFunction [z]
Output: scalar

Shown below are plots of F[z}, F’[z] and F’’[z]

In[]:= Plot[Evaluate[sol2[z]], {z, 0, 5}, PlotRange  All, Frame  True,
AspectRatio  1, FrameLabel  {Style["z", 16], Style["F(z)", 16]}]

3
F(z)

2
Out[]=

0 1 2 3 4 5

z
View publication stats

2DStagnationFlow.nb 7

In[121]:= Plot[Evaluate[D[sol2[z], {z, 1}]], {z, 0, 5}, PlotStyle  Thick, Frame  True,
AspectRatio  1, FrameLabel  {Style["z", 16], Style["F'(z)", 16]}, PlotRange  All]

1.0

0.8

0.6
F'(z)

Out[121]= 0.4

0.2

0.0

0 1 2 3 4 5

In[113]:= Plot[Evaluate[D[sol2[z], {z, 2}]], {z, 0, 3}, PlotStyle  Thick, Frame  True,
AspectRatio  1, FrameLabel  {Style["z", 16], Style["F''(z)", 16]}, PlotRange  All]

1.2

1.0

0.8
F''(z)

0.6
Out[113]=

0.4

0.2

0.0

0.0 0.5 1.0 1.5 2.0 2.5 3.0

z
Note that the solution is quite sensitive to the initial guess for F‘’[0], so some trial an error is needed for
selecting F’’[0] .

References
◼ H. Schlichting, Boundary-Layer Theory, McGraw-Hill New York, 1968