Head Loss in Pipe Systems Laminar Flow and Introduction To Turbulent Flow
Head Loss in Pipe Systems Laminar Flow and Introduction To Turbulent Flow
Gerald Recktenwald∗
∗ Associate Professor, Mechanical and Materials Engineering Department Portland State University, Portland, Oregon,
[email protected]
Typical Pipe System
ρV L VL
Re = =
µ ν
V is a characteristic velocity
L is a characteristic length
• Steady
• Incompressible
• Pipe cross-section doesn’t change with axial position
• Flow is fully-developed
dp
= constant
dx
=⇒ p(x) is linear in x
3
10
Laminar flow: (Red < 2000)
Le 2
10
≈ 0.06Red L /d = 4.4 Re1/6
d e d
Le/d = 0.06 Red
L/D
1
10
Le 1/6
≈ 4.4Red 0
10
d
See Munson, Young and Okiishi −1
10 1 2 3 4 5
§8.1.2, pp. 405–406 10 10 10
Re
10 10
In most design calculations, the flow in straight sections is assumed to be fully developed.
The entrance length correlations are used to check to see whether this is a good
assumption.
V2 V2
» – » –
p p
+ +z = + +z + hs − hL
γ 2g out γ 2g in
The pressure is assumed to decrease in the flow direction, hence the pressure on the right
hand side is p1 − ∆p.
Equation (1) shows that the pressure drop exists because of shear stress on the
circumpherential surface of the fluid element. Ultimately this shear stress is transmitted
to the wall of the pipe.
Thus, if ∆p/` is constant, i.e. if the flow is fully-developed then the shear stress varies
linearly with r . This result applies for laminar or turbulent flow.
∆p
τw = R (4)
2`
Summary so far:
• Force balance applies to laminar or turbulent flow
• For fully-developed flow, dp/dx is constant. As a consequence the shear stress profile
is linear: τ = 0 at the centerline and τ = τw at r = R.
• We need a relationship between τ and u to obtain the velocity profile.
∆p dp
= (6)
` dx
du 1 dp
−µ = r
dr 2 dx
Since dp/dx is constant (for fully developed flow), the preceding ODE can be rearranged
and integrated
du 1 dp
=− r = Kr
dr 2µ dx
where K = −(1/2µ)(dp/dx).
Integrate twice
du 1 2
= Kr =⇒ u= Kr + C1
dr 2
D 2 dp
Apply B.C. that u = 0 and r = D/2 to get C1 =
16µ dx
" «2#
2 „
D dp r
u(r) = − 1−
16µ dx R
Summary so far:
• Apply a force balance to a differential control volume to get an ODE.
• Integrate the ODE analytically to get the velocity profile.
Next: Use the velocity profile to derive formulas useful for practical engineering design.
The solution to the velocity profile enables us to compute some very important practical
quantities
D 2 dp
Vc = u(0) = −
16µ dx
" „ «2#
r
=⇒ u(r) = Vc 1 −
R
The total flow rate, and hence the average velocity, can be computed exactly because the
formula for the velocity profile is known.
R R
Q 1
Z Z
Q= u(r) dA =⇒ V = = u(r) dA
0 A A 0
Therefore
Q Vc
V = =
A 2
πR2Vc πR2 D 2 πD 4
„ « „ «
dp dp
Q= = − = −
2 2 16µ dx 128µ dx
Now, for convenience define ∆p as the pressure drop that occurs over a length of pipe L.
dp ∆p
In other words, let − ≡
dx L
Then
πD 4 ∆p
Q= (7)
128µ L
So, for laminar flow, once we know Q and L, we can easily compute ∆p and vice versa.
V2 V2
» – » –
p p
+ +z = + +z + hs − hL
γ 2g out γ 2g in
For a horizontal pipe (zout − zin) with no pump (hs = 0), and constant cross section
(Vout = in), the energy equation reduces to
pin − pout ∆p
hL = =
γ γ
128µQL 128µQL
∆p = =⇒ hL = (8)
πD 4 πγD 4
These formulas only apply to laminar flow. We need a more general approach.
128µQL
∆p = (8)
πD 4
or, in the form of a dimensional analysis
∆p = φ(V, L, D, µ)
where φ( ) is the function in Equation (8). Note that ρ does not appear. For turbulent
flow, fluid density does influence pressure drop.
For turbulent flow the dimensional form of the equation for pressure drop is
∆p = φ(V, L, D, µ, ρ, ε)
Consider the instantaneous velocity at a point in a pipe when the flow rate is increased
from zero up to a constant value such that the flow is eventually turbulent.
r 1/n
„ «
ū
= 1−
Vc R
For turbulent flow the dimensional form of the equation for pressure drop is
∆p = φ(V, L, D, µ, ρ, ε)
From practical experience we know that the pressure drop increases linearly with pipe
length, as long as the entrance effects are negligible.
The φ̃2 function is universal: it applies to all pipes. It’s called the friction factor, and
given the symbol f „ «
ρV D ε
f = φ̃2 , (10)
µ D
Combine Equation (9) and Equation (10) to get a working formula for the Darcy friction
factor
∆p D
f = 1 2L
(11)
2 ρV
128µQL
∆p = (8)
πD 4
64
flam =
ReD
Colebrook and Moody put Nikuradse’s data into a form useful for engineering calculations.
„ «
1 ε/D 2.51
√ = −2 log10 + √ (12)
f 3.7 Re f
Wholly turbulent
f
ε/D
increases
Laminar
Transition
Sm
oo
th
ReD