FM-06-02 Chapter 06 · Pipe Flow

Turbulent Pipe Flow, Darcy-Weisbach Equation, Moody Diagram & Pipe Roughness

In this chapter (4 topics)
On this page

1. Module Overview & Pedagogical Scope

Turbulent flow represents the predominant fluid dynamic regime in municipal water supply networks, hydroelectric penstocks, industrial transport conduits, and environmental pipelines. When the conduit Reynolds number exceeds the critical threshold (Re≈2000–4000Re \approx 2000\text{--}4000), infinitesimal disturbances undergo exponential amplification, destroying laminar laminae and establishing a chaotic, three-dimensional, rotational flow field characterized by intense continuous velocity fluctuations:

u=uˉ+u′,v=vˉ+v′,w=wˉ+w′u = \bar{u} + u', \quad v = \bar{v} + v', \quad w = \bar{w} + w'

In turbulent pipe flow, momentum transfer is dominated by macroscopic eddy mixing rather than molecular diffusion. This produces a much flatter, fuller velocity profile across the conduit core and exceptionally steep velocity gradients confined to a microscopic near-wall boundary layer.

This module delivers a comprehensive mathematical and engineering treatment of:

  1. Turbulence Physics & Reynolds Stresses: Reynolds averaging, the apparent turbulent shear stress tensor (−ρu′v′‾-\rho \overline{u'v'}), and Prandtl's mixing length hypothesis.
  2. Universal Law of the Wall: Asymptotic three-zone boundary layer structure (viscous sublayer, buffer zone, and logarithmic overlap layer), and Prandtl's velocity defect law.
  3. Hydrodynamic Roughness Regimes (Nikuradse Criteria): The physical interplay between equivalent sand roughness height ksk_s and viscous sublayer thickness δ′\delta', establishing the exact boundaries between hydrodynamically smooth, transitional, and completely rough regimes.
  4. Friction Factor Formulations & The Moody Diagram: Exact analytical equations (Blasius, von Kármán-Prandtl smooth, Nikuradse rough), the implicit Colebrook-White equation, and explicit engineering approximations (Swamee-Jain, Haaland) for municipal conduit design.

2. Official Examination Syllabus Mapping

Website note: this mapping is from the author’s notes. It has not been checked against the official notification for each examination; confirm the current syllabus with the examining body.

Examination Body Syllabus Clause / Core Topic Reference Focus Regimes & Weightage
GATE CE (IITs / IISc) Fluid Mechanics: Turbulent flow in pipes; Darcy-Weisbach equation; Moody chart; boundary layer thickness; velocity profiles in smooth and rough pipes; friction factor. Critical Core Subject (1 to 2 questions annually; friction factor calculations, smooth vs rough classifications, Colebrook iteration).
UPSC ESE (Prelims & Mains) Paper-II Fluid Mechanics: Turbulent flow through pipes; Reynolds stress; Prandtl mixing length; velocity defect; hydrodynamically smooth and rough boundaries; age effect on pipes. Extremely High (20–25 marks in Mains subjective; 3–5 MCQs in Prelims).
SSC JE & State AE / JE Concept of turbulent flow, Darcy-Weisbach formula, Chezy formula, loss of head due to friction, hydraulic gradient line, Moody chart. High Frequency (Direct formula questions: hf=fLV2/2gDh_f = f L V^2 / 2g D, smooth vs rough ks/δ′k_s / \delta' values).
RRB JE (CBT-2) Pipe flow, head loss, friction factor variation with Reynolds number and roughness. Standard MCQs on friction factor trends.

3. Fundamental Physical Principles & Governing Assumptions

  1. High Reynolds Number Flow: The flow occurs well above the critical transition regime (Re=ρVDμ>4000Re = \frac{\rho V D}{\mu} > 4000).
  2. Reynolds Time-Averaging (RANS Formulation): Instantaneous flow variables are decomposed into steady time-mean components (uˉ\bar{u}) and turbulent fluctuating components (u′u'), where by definition: u′‾=1T∫0Tu′dt=0,while u′2‾≠0, u′v′‾≠0\overline{u'} = \frac{1}{T} \int_0^T u' dt = 0, \quad \text{while } \overline{u'^2} \ne 0, \ \overline{u'v'} \ne 0
  3. Fully Developed Core Flow: The pipe is of uniform diameter DD and sufficient length (L/D>40–50L / D > 40\text{--}50) such that entrance effects have decayed and the time-averaged velocity profile uˉ(r)\bar{u}(r) is invariant with axial distance xx.
  4. Isotropic Homogeneous Fluid: The fluid is continuous, homogeneous, and Newtonian (ρ=const,μ=const\rho = \text{const}, \mu = \text{const}).
  5. No-Slip Boundary Condition at Wall: At the solid conduit boundary (y=0y = 0), both mean and fluctuating velocities vanish identically: uˉ=0,u′=v′=w′=0\bar{u} = 0, u' = v' = w' = 0.

4. Mathematical Derivations & Rigorous Proofs

Sub-derivation A: Reynolds Decomposition, Reynolds Stresses & Prandtl Mixing Length

Consider the instantaneous Navier-Stokes momentum equation along the longitudinal flow coordinate xx. Decomposing velocities into mean and fluctuating parts (u=uˉ+u′u = \bar{u} + u', v=vˉ+v′v = \bar{v} + v') and time-averaging over a characteristic turbulent timescale TT:

ρ(uˉ∂uˉ∂x+vˉ∂uˉ∂y)=−∂pˉ∂x+∂∂y[μ∂uˉ∂y−ρu′v′‾]\rho \left( \bar{u}\frac{\partial \bar{u}}{\partial x} + \bar{v}\frac{\partial \bar{u}}{\partial y} \right) = -\frac{\partial \bar{p}}{\partial x} + \frac{\partial}{\partial y}\left[ \mu \frac{\partial \bar{u}}{\partial y} - \rho \overline{u'v'} \right]

The bracketed term represents the total effective shear stress:

τtotal=τlam+τturb=μ∂uˉ∂y−ρu′v′‾\tau_{\text{total}} = \tau_{\text{lam}} + \tau_{\text{turb}} = \mu \frac{\partial \bar{u}}{\partial y} - \rho \overline{u'v'}

The extra convective momentum transfer term:

τt=−ρu′v′‾\tau_t = -\rho \overline{u'v'}

is known as the Reynolds Apparent Turbulent Shear Stress.

Physical Origin of −ρu′v′‾>0-\rho \overline{u'v'} > 0:

When a fluid lump moves upward from a lower layer (v′>0v' > 0), it arrives at a faster layer with a velocity deficit (u′<0u' < 0). Thus, the product u′v′<0u'v' < 0. Conversely, when a lump moves downward (v′<0v' < 0), it brings excess momentum (u′>0u' > 0), again yielding u′v′<0u'v' < 0. Therefore, the time-averaged correlation u′v′‾\overline{u'v'} is negative, rendering −ρu′v′‾-\rho \overline{u'v'} strictly positive.

Prandtl's Mixing Length Hypothesis:

In 1925, Ludwig Prandtl postulated that a turbulent fluid lump retains its original momentum over a characteristic transverse distance ll (the mixing length) before coalescing with surrounding fluid:

u′≈lduˉdy,v′≈−lduˉdyu' \approx l \frac{d\bar{u}}{dy}, \quad v' \approx - l \frac{d\bar{u}}{dy} u′v′‾≈−l2(duˉdy)2\overline{u'v'} \approx - l^2 \left(\frac{d\bar{u}}{dy}\right)^2

Substituting this into the Reynolds stress expression:

τt=ρl2(duˉdy)2\tau_t = \rho l^2 \left(\frac{d\bar{u}}{dy}\right)^2

Near a solid boundary, Prandtl reasoned on dimensional grounds that the eddy scale is constrained by the distance to the wall:

l=κyl = \kappa y

where κ≈0.40\kappa \approx 0.40 is the dimensionless von Kármán constant.


Sub-derivation B: The Universal Law of the Wall

Near the pipe wall (y≪Ry \ll R), the shear stress is approximately constant and equal to the wall shear stress: τ(y)≈τ0\tau(y) \approx \tau_0.

Define the characteristic scaling parameters:

  • Shear Velocity (Friction Velocity): uτ=τ0ρu_\tau = \sqrt{\frac{\tau_0}{\rho}}
  • Dimensionless Distance from Wall: y+=yuτνy^+ = \frac{y u_\tau}{\nu}
  • Dimensionless Velocity: u+=uˉuτu^+ = \frac{\bar{u}}{u_\tau}

The turbulent boundary layer resolves into three distinct asymptotic sublayers:

1. Viscous Sublayer (0≤y+≤50 \le y^+ \le 5):

Adjacent to the wall, turbulent fluctuations are damped out by wall proximity. Viscous shear stress completely dominates (τtotal≈μduˉdy=τ0\tau_{\text{total}} \approx \mu \frac{d\bar{u}}{dy} = \tau_0). Integrating:

uˉ=τ0μy  ⟹  uˉuτ=τ0μuτy=ρuτ2μuτy=uτyν\bar{u} = \frac{\tau_0}{\mu} y \implies \frac{\bar{u}}{u_\tau} = \frac{\tau_0}{\mu u_\tau} y = \frac{\rho u_\tau^2}{\mu u_\tau} y = \frac{u_\tau y}{\nu} u+=y+u^+ = y^+

The velocity varies linearly with distance from the wall.

2. Buffer Layer (5<y+<305 < y^+ < 30):

In this transition zone, molecular viscous stress and Reynolds turbulent stress are of the same order of magnitude. No simple analytical equation exists.

3. Logarithmic Overlap Layer (y+>30y^+ > 30):

Turbulent Reynolds shear dominates completely (τ0≈ρl2(duˉ/dy)2\tau_0 \approx \rho l^2 (d\bar{u}/dy)^2 with l=κyl = \kappa y):

τ0=ρ(κy)2(duˉdy)2  ⟹  duˉdy=1κyτ0ρ=uτκy\tau_0 = \rho (\kappa y)^2 \left(\frac{d\bar{u}}{dy}\right)^2 \implies \frac{d\bar{u}}{dy} = \frac{1}{\kappa y} \sqrt{\frac{\tau_0}{\rho}} = \frac{u_\tau}{\kappa y}

Integrating with respect to yy:

uˉ=uτκln⁡y+C\bar{u} = \frac{u_\tau}{\kappa} \ln y + C

Dividing by uτu_\tau:

uˉuτ=1κln⁡(yuτν)+B\frac{\bar{u}}{u_\tau} = \frac{1}{\kappa} \ln\left(\frac{y u_\tau}{\nu}\right) + B u+=1κln⁡(y+)+Bu^+ = \frac{1}{\kappa} \ln(y^+) + B

Substituting experimental constants κ≈0.40\kappa \approx 0.40 and B≈5.5B \approx 5.5 (Nikuradse, 1933):

u+=2.5ln⁡(y+)+5.5=5.75log⁡10(y+)+5.5u^+ = 2.5 \ln(y^+) + 5.5 = 5.75 \log_{10}(y^+) + 5.5

This is the celebrated Prandtl-von Kármán Universal Law of the Wall.


Sub-derivation C: Velocity Defect Law & Mean Velocity

At the pipe centerline (y=Ry = R), the velocity reaches its maximum value umax⁡u_{\max}:

umax⁡uτ=1κln⁡(Ruτν)+B\frac{u_{\max}}{u_\tau} = \frac{1}{\kappa} \ln\left(\frac{R u_\tau}{\nu}\right) + B

Subtracting the local velocity equation from the centerline equation eliminates the wall roughness constant BB:

umax⁡−uˉ(y)uτ=1κln⁡(Ry)=2.5ln⁡(Ry)=5.75log⁡10(Ry)\frac{u_{\max} - \bar{u}(y)}{u_\tau} = \frac{1}{\kappa} \ln\left(\frac{R}{y}\right) = 2.5 \ln\left(\frac{R}{y}\right) = 5.75 \log_{10}\left(\frac{R}{y}\right)

This is von Kármán's Velocity Defect Law. Remarkably, this defect profile is universal—it holds identically for both hydrodynamically smooth and completely rough conduits!

Average Flow Velocity (VavgV_{\text{avg}}):

Integrating the defect profile over the pipe cross-section (dA=2π(R−y)dydA = 2\pi (R - y) dy):

Vavg=1πR2∫0Ruˉ(y)⋅2π(R−y)dy=umax⁡−3.75uτV_{\text{avg}} = \frac{1}{\pi R^2} \int_0^R \bar{u}(y) \cdot 2\pi (R - y) dy = u_{\max} - 3.75 u_\tau Vavg=umax⁡−3.75uτV_{\text{avg}} = u_{\max} - 3.75 u_\tau

Radial Position of Mean Velocity:

Equating local velocity to mean velocity:

umax⁡−Vavguτ=3.75=5.75log⁡10(Ryavg)\frac{u_{\max} - V_{\text{avg}}}{u_\tau} = 3.75 = 5.75 \log_{10}\left(\frac{R}{y_{\text{avg}}}\right) log⁡10(Ryavg)=3.755.75≈0.65217  ⟹  Ryavg=100.65217=4.489\log_{10}\left(\frac{R}{y_{\text{avg}}}\right) = \frac{3.75}{5.75} \approx 0.65217 \implies \frac{R}{y_{\text{avg}}} = 10^{0.65217} = 4.489 yavg=R4.489≈0.2228Ry_{\text{avg}} = \frac{R}{4.489} \approx 0.2228 R

Radial distance from centerline:

ravg=R−yavg=R−0.2228R=0.7772Rr_{\text{avg}} = R - y_{\text{avg}} = R - 0.2228 R = 0.7772 R

Key Comparison for Civil Engineering Exams:

  • In Laminar Flow (Hagen-Poiseuille): u=Vavgu = V_{\text{avg}} occurs at r=R2≈0.7071Rr = \frac{R}{\sqrt{2}} \approx 0.7071 R.
  • In Turbulent Flow (Logarithmic): u=Vavgu = V_{\text{avg}} occurs at r≈0.7772Rr \approx 0.7772 R.

Sub-derivation D: Hydrodynamic Roughness Regimes (Nikuradse Criteria)

The thickness of the laminar/viscous sublayer is formally defined as the distance where the linear sublayer matches the logarithmic profile (y+≈11.6y^+ \approx 11.6):

δ′=11.6νuτ\delta' = \frac{11.6 \nu}{u_\tau}

J. Nikuradse (1933) systematically coated pipe interiors with graded uniform sand grains of diameter ksk_s and established three hydrodynamic roughness regimes based on the ratio of sand roughness height to sublayer thickness:

Hydrodynamic Regime Relative Sublayer Criteria Roughness Reynolds Number Re∗=uτksνRe_* = \frac{u_\tau k_s}{\nu} Physical Flow Mechanism & Governing Friction Factor
Hydrodynamically Smooth Boundary ksδ′<0.25\frac{k_s}{\delta'} < 0.25 Re∗<4Re_* < 4 Sand grains are completely submerged inside the quiescent viscous sublayer. Boundary behaves as if hydraulically polished. Friction factor depends only on Reynolds number: f=fn(Re)f = \text{fn}(Re).
Transition Regime 0.25≤ksδ′≤6.00.25 \le \frac{k_s}{\delta'} \le 6.0 4≤Re∗≤1004 \le Re_* \le 100 Roughness peaks emerge partially into the buffer layer. Form drag on asperities and viscous wall shear coexist. Friction factor depends on both Reynolds number and relative roughness: f=fn(Re,ks/D)f = \text{fn}(Re, k_s/D).
Completely Rough Boundary ksδ′>6.0\frac{k_s}{\delta'} > 6.0 Re∗>100Re_* > 100 Roughness peaks completely project beyond the sublayer, destroying it. Flow resistance is governed entirely by pressure form drag (wake shedding behind asperities). Friction factor is independent of Reynolds number: f=fn(ks/D)f = \text{fn}(k_s/D) only.

Sub-derivation E: Friction Factor Formulations & The Moody Chart

The Darcy-Weisbach head loss equation is defined as:

hf=fLDV22gh_f = f \frac{L}{D} \frac{V^2}{2g}

where ff is the dimensionless Darcy friction factor. (Note: Fanning friction factor f′=cf=f/4f' = c_f = f / 4, wall shear stress τ0=f8ρV2\tau_0 = \frac{f}{8}\rho V^2).

1. Blasius Formula (Smooth Pipes, Re≤105Re \le 10^5):

Empirical formula derived by Heinrich Blasius (1913):

f=0.3164Re0.25=0.3164(ρVDμ)0.25f = \frac{0.3164}{Re^{0.25}} = \frac{0.3164}{\left(\frac{\rho V D}{\mu}\right)^{0.25}}

2. Kármán-Prandtl Resistance Equation for Smooth Pipes:

Derived from the logarithmic law of the wall:

1f=2.0log⁡10(Ref)−0.80\frac{1}{\sqrt{f}} = 2.0 \log_{10}(Re \sqrt{f}) - 0.80

3. Nikuradse Resistance Equation for Completely Rough Pipes:

When the viscous sublayer is completely destroyed (Re∗>100Re_* > 100), viscosity drops out of the problem:

1f=2.0log⁡10(D2ks)+1.74=1.14−2.0log⁡10(ksD)=2.0log⁡10(3.7Dks)\frac{1}{\sqrt{f}} = 2.0 \log_{10}\left(\frac{D}{2 k_s}\right) + 1.74 = 1.14 - 2.0 \log_{10}\left(\frac{k_s}{D}\right) = 2.0 \log_{10}\left(\frac{3.7 D}{k_s}\right)

4. Colebrook-White Transition Formula (1939):

C. F. Colebrook and C. M. White devised an ingenious transcendental equation that smoothly bridges the Kármán-Prandtl smooth equation and the Nikuradse rough equation across the entire commercial transition regime:

1f=−2.0log⁡10(ks/D3.7+2.51Ref)\frac{1}{\sqrt{f}} = -2.0 \log_{10}\left( \frac{k_s / D}{3.7} + \frac{2.51}{Re \sqrt{f}} \right)

This equation is the mathematical foundation of Lewis F. Moody's celebrated Moody Diagram (1944).

5. Explicit Engineering Approximations:

Because the Colebrook-White equation is implicit in ff, engineers utilize accurate explicit approximations:

  • Swamee-Jain Equation (1976) (Accurate within ±1%\pm 1\% for 10−6≤ksD≤10−210^{-6} \le \frac{k_s}{D} \le 10^{-2} and 5000≤Re≤1085000 \le Re \le 10^8): f=0.25[log⁡10(ks/D3.7+5.74Re0.9)]2f = \frac{0.25}{\left[ \log_{10}\left( \frac{k_s / D}{3.7} + \frac{5.74}{Re^{0.9}} \right) \right]^2}
  • Haaland Equation (1983): 1f≈−1.8log⁡10[(ks/D3.7)1.11+6.9Re]\frac{1}{\sqrt{f}} \approx -1.8 \log_{10}\left[ \left(\frac{k_s / D}{3.7}\right)^{1.11} + \frac{6.9}{Re} \right]

5. Technical Diagrams & Interactive Architectural Schematics

Figure 1: Near-Wall Sublayer Structure & Nikuradse Roughness Regimes

The diagram below illustrates the three-zone boundary layer structure (y+≤5y^+ \le 5, buffer, and core), and the physical mechanism distinguishing hydrodynamically smooth, transition, and completely rough boundaries.

Near-Wall Sublayers and Roughness Regimes

Figure 2: The Moody Diagram Architecture

The comprehensive Moody chart below displays the log-log mapping of Darcy friction factor ff across Reynolds numbers from 10310^3 to 10810^8, depicting laminar flow (f=64/Ref = 64/Re), the critical transition zone, smooth conduit limits, and the complete turbulence asymptote.

Moody Diagram Architecture

6. Comparative Engineering Analysis & Parameter Regimes

Flow Property / Feature Laminar Pipe Flow (Re≤2000Re \le 2000) Turbulent Pipe Flow (Re>4000Re > 4000)
Momentum Transport Mechanism Molecular diffusion (τ=μdu/dy\tau = \mu du/dy) Macroscopic eddy mixing (τt=−ρu′v′‾\tau_t = -\rho \overline{u'v'})
Velocity Profile Geometry Quadratic Parabola: u(r)=umax⁡[1−(r/R)2]u(r) = u_{\max}[1 - (r/R)^2] Logarithmic / 1/7th Power Law: u(y)∝y1/7u(y) \propto y^{1/7}
Centerline-to-Mean Ratio (umax⁡/Vavgu_{\max} / V_{\text{avg}}) Exactly 2.002.00 1.18–1.251.18\text{--}1.25 (Typically ≈1.20\approx 1.20)
Mean Velocity Radial Location (ravgr_{\text{avg}}) r=R2≈0.7071Rr = \frac{R}{\sqrt{2}} \approx 0.7071 R r≈0.7772Rr \approx 0.7772 R
Kinetic Energy Factor (α\alpha) α=2.00\alpha = 2.00 α≈1.058–1.06\alpha \approx 1.058\text{--}1.06 (Taken as ≈1.0\approx 1.0)
Momentum Correction Factor (β\beta) β=1.333\beta = 1.333 (4/34/3) β≈1.020\beta \approx 1.020 (50/4950/49)
Friction Factor Dependence f=64Ref = \frac{64}{Re} (Strictly independent of roughness ksk_s) f=fn(Re,ks/D)f = \text{fn}(Re, k_s/D) (Becomes independent of ReRe in rough regime)
Head Loss Dependence on Velocity hf∝V1h_f \propto V^1 (Linear with velocity) hf∝V1.75h_f \propto V^{1.75} (Smooth) to V2.00V^{2.00} (Completely rough)
Shear Stress Variation Across Pipe Linear from 0 at CL to τ0\tau_0 at wall Linear from 0 at CL to τ0\tau_0 at wall (Fundamental momentum identity!)

7. Worked Numerical Examples (Pre-Calculated & Verified)

Problem 1: Boundary Classification & Viscous Sublayer Thickness

Statement:
Clean water at 20∘C20^\circ\text{C} (ρ=1000 kg/m3,ν=1.00×10−6 m2/s\rho = 1000\text{ kg/m}^3, \nu = 1.00 \times 10^{-6}\text{ m}^2/\text{s}) flows through a horizontal municipal pipe of diameter D=200 mmD = 200\text{ mm} (0.20 m0.20\text{ m}) at an average flow velocity V=2.50 m/sV = 2.50\text{ m/s}. The measured Darcy friction factor is f=0.020f = 0.020.

  1. Compute the wall shear stress τ0\tau_0 and the friction velocity (shear velocity) uτu_\tau.
  2. Calculate the nominal thickness of the viscous sublayer δ′\delta'.
  3. Determine the hydrodynamic boundary classification if the pipeline is constructed of:
    • Case (a): Drawn smooth brass tubing (ks=0.020 mmk_s = 0.020\text{ mm}).
    • Case (b): Commercial galvanized iron pipe (ks=0.250 mmk_s = 0.250\text{ mm}).
    • Case (c): Severely tuberculated cast iron main (ks=1.500 mmk_s = 1.500\text{ mm}).

Solution:

Step 1: Wall Shear Stress and Shear Velocity

τ0=f8ρV2=0.0208×1000 kg/m3×(2.50 m/s)2=0.0025×1000×6.25=15.6250 Pa\tau_0 = \frac{f}{8} \rho V^2 = \frac{0.020}{8} \times 1000\text{ kg/m}^3 \times (2.50\text{ m/s})^2 = 0.0025 \times 1000 \times 6.25 = 15.6250\text{ Pa} uτ=τ0ρ=15.62501000=0.015625=0.1250 m/su_\tau = \sqrt{\frac{\tau_0}{\rho}} = \sqrt{\frac{15.6250}{1000}} = \sqrt{0.015625} = 0.1250\text{ m/s}

(Check: uτ=Vf/8=2.50×0.0025=2.50×0.05=0.1250 m/su_\tau = V \sqrt{f/8} = 2.50 \times \sqrt{0.0025} = 2.50 \times 0.05 = 0.1250\text{ m/s}. Exact match).

Step 2: Viscous Sublayer Thickness

δ′=11.6νuτ=11.6×1.00×10−6 m2/s0.1250 m/s=9.280×10−5 m=0.0928 mm=92.8 μm\delta' = \frac{11.6 \nu}{u_\tau} = \frac{11.6 \times 1.00 \times 10^{-6}\text{ m}^2/\text{s}}{0.1250\text{ m/s}} = 9.280 \times 10^{-5}\text{ m} = 0.0928\text{ mm} = 92.8\ \mu\text{m}

Step 3: Hydrodynamic Classification for Each Roughness Case

  • Case (a): Drawn Smooth Tubing (ks=0.020 mm=2.0×10−5 mk_s = 0.020\text{ mm} = 2.0 \times 10^{-5}\text{ m}):

    ksδ′=0.020 mm0.0928 mm=0.2155\frac{k_s}{\delta'} = \frac{0.020\text{ mm}}{0.0928\text{ mm}} = 0.2155 Re∗=uτksν=0.1250×2.0×10−51.00×10−6=2.50Re_* = \frac{u_\tau k_s}{\nu} = \frac{0.1250 \times 2.0 \times 10^{-5}}{1.00 \times 10^{-6}} = 2.50

    Since ksδ′=0.2155<0.25\frac{k_s}{\delta'} = 0.2155 < 0.25 and Re∗=2.50<4.0Re_* = 2.50 < 4.0, the pipe behaves as HYDRODYNAMICALLY SMOOTH. The roughness peaks are fully submerged within the 92.8 μm92.8\ \mu\text{m} viscous sublayer.

  • Case (b): Galvanized Iron (ks=0.250 mm=2.5×10−4 mk_s = 0.250\text{ mm} = 2.5 \times 10^{-4}\text{ m}):

    ksδ′=0.250 mm0.0928 mm=2.6940\frac{k_s}{\delta'} = \frac{0.250\text{ mm}}{0.0928\text{ mm}} = 2.6940 Re∗=uτksν=0.1250×2.5×10−41.00×10−6=31.25Re_* = \frac{u_\tau k_s}{\nu} = \frac{0.1250 \times 2.5 \times 10^{-4}}{1.00 \times 10^{-6}} = 31.25

    Since 0.25≤2.6940≤6.00.25 \le 2.6940 \le 6.0 and 4.0≤31.25≤100.04.0 \le 31.25 \le 100.0, the pipe operates in the TRANSITION REGIME. Friction depends simultaneously on ReRe and ks/Dk_s/D.

  • Case (c): Tuberculated Cast Iron (ks=1.500 mm=1.5×10−3 mk_s = 1.500\text{ mm} = 1.5 \times 10^{-3}\text{ m}):

    ksδ′=1.500 mm0.0928 mm=16.1638\frac{k_s}{\delta'} = \frac{1.500\text{ mm}}{0.0928\text{ mm}} = 16.1638 Re∗=uτksν=0.1250×1.5×10−31.00×10−6=187.50Re_* = \frac{u_\tau k_s}{\nu} = \frac{0.1250 \times 1.5 \times 10^{-3}}{1.00 \times 10^{-6}} = 187.50

    Since ksδ′=16.16>6.0\frac{k_s}{\delta'} = 16.16 > 6.0 and Re∗=187.50>100.0Re_* = 187.50 > 100.0, the pipe is COMPLETELY ROUGH. The viscous sublayer is totally annihilated by eddy shedding; ff is independent of ReRe.


Problem 2: Logarithmic Velocity Profile & Velocity Defect

Statement:
In a turbulent pipe flow of air through a smooth duct of internal radius R=150 mmR = 150\text{ mm} (0.150 m0.150\text{ m}), the measured centerline maximum velocity is umax⁡=3.60 m/su_{\max} = 3.60\text{ m/s} and the shear velocity is uτ=0.150 m/su_\tau = 0.150\text{ m/s}. Using von Kármán's universal velocity defect law:

umax⁡−uˉ(y)uτ=5.75log⁡10(Ry)\frac{u_{\max} - \bar{u}(y)}{u_\tau} = 5.75 \log_{10}\left(\frac{R}{y}\right)
  1. Compute the local time-averaged flow velocity at distance y=15.0 mmy = 15.0\text{ mm} and y=75.0 mmy = 75.0\text{ mm} from the duct boundary.
  2. Calculate the cross-sectional mean velocity VavgV_{\text{avg}} and the ratio Vavg/umax⁡V_{\text{avg}} / u_{\max}.
  3. Determine the radial coordinate where local velocity equals the mean flow velocity.

Solution:

Step 1: Local Velocities Across Conduit

  • At y=15.0 mm=0.015 my = 15.0\text{ mm} = 0.015\text{ m} (radial position r=R−y=0.150−0.015=0.135 m=135 mmr = R - y = 0.150 - 0.015 = 0.135\text{ m} = 135\text{ mm}):

    Ry=150 mm15 mm=10.0  ⟹  log⁡10(10.0)=1.0000\frac{R}{y} = \frac{150\text{ mm}}{15\text{ mm}} = 10.0 \implies \log_{10}(10.0) = 1.0000 umax⁡−uˉ=5.75×uτ×1.0000=5.75×0.150 m/s=0.8625 m/su_{\max} - \bar{u} = 5.75 \times u_\tau \times 1.0000 = 5.75 \times 0.150\text{ m/s} = 0.8625\text{ m/s} uˉ(15 mm)=3.60 m/s−0.8625 m/s=2.7375 m/s\bar{u}(15\text{ mm}) = 3.60\text{ m/s} - 0.8625\text{ m/s} = 2.7375\text{ m/s}
  • At y=75.0 mm=0.075 my = 75.0\text{ mm} = 0.075\text{ m} (mid-radius position r=75 mmr = 75\text{ mm}):

    Ry=150 mm75 mm=2.000  ⟹  log⁡10(2.000)=0.30103\frac{R}{y} = \frac{150\text{ mm}}{75\text{ mm}} = 2.000 \implies \log_{10}(2.000) = 0.30103 umax⁡−uˉ=5.75×0.150 m/s×0.30103=0.8625×0.30103=0.2596 m/su_{\max} - \bar{u} = 5.75 \times 0.150\text{ m/s} \times 0.30103 = 0.8625 \times 0.30103 = 0.2596\text{ m/s} uˉ(75 mm)=3.60 m/s−0.2596 m/s=3.3404 m/s\bar{u}(75\text{ mm}) = 3.60\text{ m/s} - 0.2596\text{ m/s} = 3.3404\text{ m/s}

Step 2: Cross-Sectional Mean Flow Velocity

Vavg=umax⁡−3.75uτ=3.60 m/s−3.75×(0.150 m/s)=3.60−0.5625=3.0375 m/sV_{\text{avg}} = u_{\max} - 3.75 u_\tau = 3.60\text{ m/s} - 3.75 \times (0.150\text{ m/s}) = 3.60 - 0.5625 = 3.0375\text{ m/s} Vavgumax⁡=3.03753.60=0.84375≈84.38%\frac{V_{\text{avg}}}{u_{\max}} = \frac{3.0375}{3.60} = 0.84375 \approx 84.38\%

(Compare with laminar flow where Vavg/umax⁡=50.0%V_{\text{avg}} / u_{\max} = 50.0\%; the turbulent profile is significantly more uniform).

Step 3: Radial Position of Mean Velocity

5.75log⁡10(Ryavg)=3.75  ⟹  log⁡10(Ryavg)=3.755.75=0.6521745.75 \log_{10}\left(\frac{R}{y_{\text{avg}}}\right) = 3.75 \implies \log_{10}\left(\frac{R}{y_{\text{avg}}}\right) = \frac{3.75}{5.75} = 0.652174 Ryavg=100.652174=4.4893  ⟹  yavg=150.0 mm4.4893=33.41 mm\frac{R}{y_{\text{avg}}} = 10^{0.652174} = 4.4893 \implies y_{\text{avg}} = \frac{150.0\text{ mm}}{4.4893} = 33.41\text{ mm}

Radial distance from centerline:

ravg=R−yavg=150.0 mm−33.41 mm=116.59 mmr_{\text{avg}} = R - y_{\text{avg}} = 150.0\text{ mm} - 33.41\text{ mm} = 116.59\text{ mm} ravgR=116.59150.0=0.7772≈0.777R\frac{r_{\text{avg}}}{R} = \frac{116.59}{150.0} = 0.7772 \approx 0.777 R

Problem 3: Commercial Penstock Sizing (Colebrook-White vs Swamee-Jain)

Statement:
A hydroelectric penstock constructed of commercial welded steel has diameter D=600 mmD = 600\text{ mm} (0.60 m0.60\text{ m}) and length L=1200 mL = 1200\text{ m}. It conveys water (ρ=1000 kg/m3,μ=0.0010 Pa⋅s\rho = 1000\text{ kg/m}^3, \mu = 0.0010\text{ Pa}\cdot\text{s}) at design discharge Q=0.850 m3/sQ = 0.850\text{ m}^3\text{/s}. The absolute roughness of the steel conduit is ks=0.045 mmk_s = 0.045\text{ mm} (4.5×10−5 m4.5 \times 10^{-5}\text{ m}).

  1. Determine the relative pipe roughness ks/Dk_s / D and the flow Reynolds number ReRe.
  2. Calculate the Darcy friction factor ff using the explicit Swamee-Jain formula, and verify convergence using the implicit Colebrook-White equation.
  3. Compute the friction head loss hfh_f, hydraulic gradient slope SfS_f, and the pumping power lost to viscous friction.

Solution:

Step 1: Flow Kinematics and Reynolds Number

A=πD24=π(0.60)24=0.282743 m2A = \frac{\pi D^2}{4} = \frac{\pi (0.60)^2}{4} = 0.282743\text{ m}^2 V=QA=0.850 m3/s0.282743 m2=3.00626 m/sV = \frac{Q}{A} = \frac{0.850\text{ m}^3\text{/s}}{0.282743\text{ m}^2} = 3.00626\text{ m/s} Re=ρVDμ=1000×3.00626×0.600.0010=1,803,756≈1.804×106Re = \frac{\rho V D}{\mu} = \frac{1000 \times 3.00626 \times 0.60}{0.0010} = 1,803,756 \approx 1.804 \times 10^6

Since Re≫4000Re \gg 4000, flow is highly turbulent. Relative roughness:

ksD=4.5×10−5 m0.60 m=0.000075=7.50×10−5\frac{k_s}{D} = \frac{4.5 \times 10^{-5}\text{ m}}{0.60\text{ m}} = 0.000075 = 7.50 \times 10^{-5}

Step 2: Explicit Swamee-Jain Friction Factor Calculation

fSJ=0.25[log⁡10(ks/D3.7+5.74Re0.9)]2f_{\text{SJ}} = \frac{0.25}{\left[ \log_{10}\left( \frac{k_s / D}{3.7} + \frac{5.74}{Re^{0.9}} \right) \right]^2} ks/D3.7=7.50×10−53.7=2.0270×10−5\frac{k_s / D}{3.7} = \frac{7.50 \times 10^{-5}}{3.7} = 2.0270 \times 10^{-5} Re0.9=(1,803,756)0.9=431,686.8  ⟹  5.74Re0.9=5.74431,686.8=1.3297×10−5Re^{0.9} = (1,803,756)^{0.9} = 431,686.8 \implies \frac{5.74}{Re^{0.9}} = \frac{5.74}{431,686.8} = 1.3297 \times 10^{-5} Sum=2.0270×10−5+1.3297×10−5=3.3567×10−5\text{Sum} = 2.0270 \times 10^{-5} + 1.3297 \times 10^{-5} = 3.3567 \times 10^{-5} log⁡10(3.3567×10−5)=−4.47409\log_{10}(3.3567 \times 10^{-5}) = -4.47409 fSJ=0.25(−4.47409)2=0.2520.0175=0.012489≈0.012499f_{\text{SJ}} = \frac{0.25}{(-4.47409)^2} = \frac{0.25}{20.0175} = 0.012489 \approx 0.012499

Step 3: Colebrook-White Implicit Convergence

1f=−2.0log⁡10(7.50×10−53.7+2.511,803,756f)\frac{1}{\sqrt{f}} = -2.0 \log_{10}\left( \frac{7.50 \times 10^{-5}}{3.7} + \frac{2.51}{1,803,756 \sqrt{f}} \right)

Using f0=0.012499  ⟹  f0=0.111799f_0 = 0.012499 \implies \sqrt{f_0} = 0.111799:

2.511,803,756×0.111799=1.2447×10−5\frac{2.51}{1,803,756 \times 0.111799} = 1.2447 \times 10^{-5} Bracket=2.0270×10−5+1.2447×10−5=3.2717×10−5\text{Bracket} = 2.0270 \times 10^{-5} + 1.2447 \times 10^{-5} = 3.2717 \times 10^{-5} −2.0log⁡10(3.2717×10−5)=−2.0×(−4.48523)=8.97046-2.0 \log_{10}(3.2717 \times 10^{-5}) = -2.0 \times (-4.48523) = 8.97046 f1=(18.97046)2=(0.111477)2=0.012427f_1 = \left(\frac{1}{8.97046}\right)^2 = (0.111477)^2 = 0.012427

Second iteration yields exact convergence to 5 figures:

fconverged=0.012430f_{\text{converged}} = 0.012430

The Swamee-Jain formula provides an outstanding agreement within 0.56%0.56\% error!

Step 4: Friction Head Loss, Hydraulic Slope & Dissipated Power

hf=fLDV22g=0.012430×(1200 m0.60 m)×(3.00626 m/s)22×9.81 m/s2h_f = f \frac{L}{D} \frac{V^2}{2g} = 0.012430 \times \left(\frac{1200\text{ m}}{0.60\text{ m}}\right) \times \frac{(3.00626\text{ m/s})^2}{2 \times 9.81\text{ m/s}^2} hf=0.012430×2000×9.037619.62=24.860×0.46063=11.4510 mh_f = 0.012430 \times 2000 \times \frac{9.0376}{19.62} = 24.860 \times 0.46063 = 11.4510\text{ m} Sf=hfL=11.4510 m1200 m=0.0095425=9.543 m/kmS_f = \frac{h_f}{L} = \frac{11.4510\text{ m}}{1200\text{ m}} = 0.0095425 = 9.543\text{ m/km}

Pumping power lost to friction:

Plost=ρgQhf=1000 kg/m3×9.81 m/s2×0.850 m3/s×11.4510 m=95,484.4 W=95.48 kWP_{\text{lost}} = \rho g Q h_f = 1000\text{ kg/m}^3 \times 9.81\text{ m/s}^2 \times 0.850\text{ m}^3\text{/s} \times 11.4510\text{ m} = 95,484.4\text{ W} = 95.48\text{ kW}

8. High-Yield Examination Traps & Ambiguities

S.No. Examination Trap / Confusion Rigorous Engineering Distinction Primary Exam
1 Smooth vs Rough Pipe Definition: Believing that a pipe made of "rough" material is always rough. A pipe is hydrodynamically smooth if ks<δ′/4k_s < \delta'/4, regardless of the material! At low velocities, δ′\delta' is large, submerging the roughness. At high velocities, uτu_\tau increases, δ′\delta' shrinks, and the exact same pipe becomes hydrodynamically rough! GATE CE, UPSC ESE
2 Friction Factor Independence on ReRe: Stating ff always decreases with ReRe. In the completely rough regime, the Moody curves are completely horizontal; ff is strictly independent of ReRe and depends solely on ks/Dk_s/D. SSC JE, State AE
3 Linear vs Parabolic Shear Stress: Assuming turbulent shear stress varies non-linearly across the pipe. The total shear stress τtotal(r)=−r2dpdx\tau_{\text{total}}(r) = -\frac{r}{2}\frac{dp}{dx} is strictly linear across the conduit in both laminar AND turbulent flows (dictated purely by static equilibrium ∑Fx=0\sum F_x = 0). Only the split between laminar and turbulent components changes. GATE CE, ESE Mains
4 Velocity Ratio (umax⁡/Vavgu_{\max} / V_{\text{avg}}): Using 2.02.0 for turbulent flow. In laminar flow, umax⁡/Vavg=2.0u_{\max} / V_{\text{avg}} = 2.0. In turbulent pipe flow, umax⁡/Vavg≈1.20–1.25u_{\max} / V_{\text{avg}} \approx 1.20\text{--}1.25. SSC JE, RRB JE
5 Fanning vs Darcy Friction Factor: Writing hf=4f′LV2/2gDh_f = 4 f' L V^2 / 2g D. In British/American literature, fDarcy=4fFanningf_{\text{Darcy}} = 4 f_{\text{Fanning}}. Ensure whether the exam question provides Darcy friction factor (f≈0.015–0.04f \approx 0.015\text{--}0.04) or coefficient of friction (f′≈0.004–0.01f' \approx 0.004\text{--}0.01). GATE CE, ESE Prelims

9. Real-World Engineering Applications & Case Studies

  1. Water Distribution Networks & Conduit Aging (Tuberculation):
    Over 20–50 years of municipal service, unlined cast iron water mains suffer internal corrosion and bio-chemical encrustation (tuberculation). Absolute roughness ksk_s increases from 0.12 mm0.12\text{ mm} (new) to 1.5–3.0 mm1.5\text{--}3.0\text{ mm} (aged), causing friction factor ff to double and carrying capacity to drop by over 40%40\%.
  2. Hydraulic Penstocks in Hydroelectric Stations:
    In high-head hydroelectric plants, water velocities exceed 4–6 m/s4\text{--}6\text{ m/s}. Under these high shear velocities (uτu_\tau), the sublayer δ′\delta' shrinks below 30 μm30\ \mu\text{m}, forcing the penstock into the completely rough regime. Smooth epoxy or polyurethane internal linings are applied to suppress roughness and conserve millions of kilowatt-hours annually.
  3. Slurry Pipelines & Dredging Conduits:
    Mining slurry pipelines transporting fly ash, copper tailings, or dredged silt require turbulent suspension velocities to prevent sedimentation. Operating at Re>105Re > 10^5, friction factor calculations directly dictate booster pump station spacing along hundreds of kilometers.

10. Cross-Disciplinary & Downstream Dependencies

[Fluid Mechanics: Ch 06 Turbulent Pipe Flow]
        │
        ├──► [Subject 09: Environmental Engineering]
        │         ├─► Water Supply Networks (Hardy Cross Analysis, Hazen-Williams vs Darcy)
        │         └─► Sewer Line Self-Cleansing Velocities & Friction Slopes
        │
        ├──► [Subject 04: Water Resources & Hydrology]
        │         └─► Penstock Hydraulics, Surge Tank Dynamics & Hydraulic Turbines
        │
        ├──► [Fluid Mechanics: Ch 06 Topic 03]
        │         └─► Minor Losses, Parallel Pipes & Equivalent Pipe Formulations
        │
        └──► [Fluid Mechanics: Ch 07 Boundary Layer Theory]
                  └─► Turbulent Boundary Layer on Flat Plates & Form Drag

11. Summary Formula Cheat Sheet & Memory Anchors

  • Reynolds Stress & Mixing Length: τt=−ρu′v′‾=ρl2(dudy)2,l=κy(κ≈0.40)\tau_t = -\rho \overline{u'v'} = \rho l^2 \left(\frac{du}{dy}\right)^2, \quad l = \kappa y \quad (\kappa \approx 0.40)
  • Friction Velocity & Sublayer: uτ=τ0ρ=Vf8,δ′=11.6νuτu_\tau = \sqrt{\frac{\tau_0}{\rho}} = V \sqrt{\frac{f}{8}}, \quad \delta' = \frac{11.6 \nu}{u_\tau}
  • Nikuradse Roughness Boundaries:
    • Smooth: ks/δ′<0.25  ⟺  Re∗<4k_s / \delta' < 0.25 \iff Re_* < 4
    • Transition: 0.25≤ks/δ′≤6.0  ⟺  4≤Re∗≤1000.25 \le k_s / \delta' \le 6.0 \iff 4 \le Re_* \le 100
    • Rough: ks/δ′>6.0  ⟺  Re∗>100k_s / \delta' > 6.0 \iff Re_* > 100
  • Universal Velocity Defect: umax⁡−u(y)uτ=5.75log⁡10(Ry),Vavg=umax⁡−3.75uτ\frac{u_{\max} - u(y)}{u_\tau} = 5.75 \log_{10}\left(\frac{R}{y}\right), \quad V_{\text{avg}} = u_{\max} - 3.75 u_\tau u=Vavg at r=0.777Ru = V_{\text{avg}} \text{ at } r = 0.777 R
  • Friction Factors:
    • Blasius (Smooth, Re≤105Re \le 10^5): f=0.3164Re0.25f = \frac{0.3164}{Re^{0.25}}
    • Nikuradse (Rough): 1f=1.14−2log⁡10(ksD)=2log⁡10(3.7Dks)\frac{1}{\sqrt{f}} = 1.14 - 2 \log_{10}\left(\frac{k_s}{D}\right) = 2 \log_{10}\left(\frac{3.7 D}{k_s}\right)
    • Colebrook-White: 1f=−2.0log⁡10(ks/D3.7+2.51Ref)\frac{1}{\sqrt{f}} = -2.0 \log_{10}\left(\frac{k_s/D}{3.7} + \frac{2.51}{Re \sqrt{f}}\right)
    • Swamee-Jain Explicit: f=0.25[log⁡10(ks/D3.7+5.74Re0.9)]2f = \frac{0.25}{\left[\log_{10}\left(\frac{k_s/D}{3.7} + \frac{5.74}{Re^{0.9}}\right)\right]^2}

12. Comprehensive References & Standard Codes

  1. Moody, L. F. (1944). Friction Factors for Pipe Flow, Transactions of the ASME, Vol. 66, pp. 671–684.
  2. Colebrook, C. F. (1939). Turbulent Flow in Pipes, with Particular Reference to the Transition Region Between the Smooth and Rough Pipe Laws, Journal of the ICE, Vol. 11, pp. 133–156.
  3. Nikuradse, J. (1933). Laws of Flow in Rough Pipes, VDI-Forschungsheft 361. (NACA TM 1292).
  4. White, F. M. (2016). Fluid Mechanics, 8th Edition, McGraw-Hill Education.
  5. IS 7784 & IS 4880 (Bureau of Indian Standards: Codes of Practice for Design of Tunnels, Shafts and Water Conduits).