NS_Equations_Example_01
NS_Equations_Example_01
Consider the steady flow of an incompressible, constant viscosity Newtonian fluid between two infinitely
long, parallel plates separated by a distance, h.
g h y fluid
x
∂p
⇒ = −ρg
∂y
⇒ p ( x, y ) = f ( x ) − ρ gy (1)
where f(x) is an unknown function of x.
841
Now let’s examine the Navier-Stokes equation in the x-direction:
⎛ ∂u ∂u ∂u ⎞ ∂p ⎛ ∂ 2u ∂ 2u ⎞
ρ ⎜ x + ux x + u y x ⎟ = − + µ ⎜ 2x + 2x ⎟ + ρ f x
⎝ ∂t ∂x ∂y ⎠ ∂x ⎝ ∂x ∂y ⎠
After simplifying:
⎛ ⎞ ⎛ 2 ⎞
⎜ ∂ux ∂ux ∂ux ⎟ ∂p ⎜ ∂ ux ∂ 2 ux ⎟
ρ + ux + uy =− + µ⎜ 2 + + ρ fx
⎜! ∂t
⎜⎝ =0 (#2) !∂x ! ∂y ⎟
⎟
∂x ∂x
! ∂ y2 ⎟ !
=0 (#3)
=0 (#5) ⎠ ⎜
⎝ =0 (#3) ⎟
⎠ =0 (#4)
∂ 2 ux 1 ∂ p
⇒ =
∂ y 2 µ ∂x
Based on assumptions #1, #2, and #3 we can write:
∂ 2ux d 2ux
=
∂y 2 dy 2
so that the simplified Navier-Stokes equation in the x-direction becomes:
d 2 u x 1 ∂p
=
dy 2 µ ∂x
Integrating twice with respect to y (note that from Eq. (1) we observe that ∂p/∂x is not a function of y):
1 ∂p 2
ux = y + c1 y + c2
2µ ∂x
where c1 and c2 are unknown constants that we find using our boundary conditions. Note that this is the
equation of a parabola.
After applying these boundary conditions to determine the constants c1 and c2 we find that the fluid velocity
in the x-direction is given by:
y h2 ⎛ ∂p ⎞⎛ y ⎞⎛ y ⎞
ux = U + ⎜ − ⎟⎜ ⎟⎜1 − ⎟
h 2µ ⎝ ∂x ⎠⎝ h ⎠⎝ h ⎠
This type of flow is often referred to as a planar Couette-Poiseuille flow (pronounced “’pwäz I”)
Notes:
1. The stress acting on the fluid at any point can be found from the stress-strain rate constitutive relations
for a Newtonian fluid.
⎛ ∂u ∂u ⎞
σ ij = − pδ ij + µ ⎜ i + j ⎟
⎜ ∂x ⎟
⎝ j ∂xi ⎠
2. If we remove the pressure gradient and move the fluid using just the moving upper boundary, the
velocity profile becomes linear:
y
ux = U
h
This type of flow is referred to as a planar Couette flow.
842
3. If we hold both boundaries stationary and move the fluid using only a pressure gradient (note that flow
in the positive x-direction occurs for dp/dx < 0), the velocity profile becomes:
h2 ⎛ ∂p ⎞⎛ y ⎞⎛ y ⎞
ux = ⎜ − ⎟⎜ ⎟⎜1 − ⎟
2µ ⎝ ∂x ⎠⎝ h ⎠⎝ h ⎠
This type of flow is referred to as a planar Poiseuille flow.
4. The average flow velocity may be found by setting the volumetric flow rate using the average velocity
equal to the volumetric flow rate using the real velocity profile. For example, for planar Poiseuille
flow the average velocity is:
y=h
h 2 ⎛ ∂p ⎞⎛ y ⎞⎛ y ⎞ h3 ⎛ ∂p ⎞
Q = uh = ∫ 2µ ⎝ ∂x ⎠⎝ h ⎠⎝ h ⎠ 12µ ⎜⎝ − ∂x ⎟⎠
y =0
⎜ − ⎟⎜ ⎟⎜ 1 − ⎟ dy =
h2 ⎛ ∂p ⎞ 2
u= ⎜ − ⎟ = umax (2)
12µ ⎝ ∂x ⎠ 3
5. Recall that we assumed that these solutions only hold for laminar flows (the uy component is zero).
Experimentally we observe that planar Couette-Poiseuille flow remains laminar for:
ρ uh
Re = < 1500
µ
where Re is the Reynolds number of the flow and u is the average flow velocity. It should be noted
that the value of 1500 is only approximate and can vary considerably depending on how carefully the
experiment is conducted. Its value is given only as an engineering rule-of-thumb.
U
dp/dx > 0 Couette flow (dp/dx = 0)
y
x dp/dx < 0
843