Viscous Flow and Internal Pipe Systems

Introduction

Most engineering fluid systems transport liquids or gases through pipes, ducts, valves, bends, pumps, and fittings. In such systems, viscosity converts mechanical energy into thermal energy and produces pressure loss. Correct prediction of velocity distribution, wall shear, frictional loss, and local losses is therefore essential for sizing pipes, selecting pumps, estimating operating cost, and ensuring reliable flow.

This chapter develops the fundamentals of viscous internal flow. It begins with Reynolds’ distinction between laminar and turbulent motion, derives the principal laminar-flow results, and then introduces engineering methods for turbulent pipe-flow analysis. The Darcy–Weisbach equation, friction-factor correlations, Moody charts, minor losses, equivalent pipes, and pipe networks are treated as connected design tools.

Learning Objectives

By the end of this chapter, the reader should be able to:

1. Distinguish laminar, transitional, and turbulent pipe flow using Reynolds number.
2. Explain the physical meaning of viscosity, wall shear stress, and no-slip conditions.
3. Derive and apply the Hagen–Poiseuille relation for fully developed laminar flow.
4. Determine velocity profiles and pressure losses in circular pipes.
5. Calculate major and minor head losses in practical pipe systems.
6. Use friction-factor equations and Moody-chart information appropriately.
7. Analyze pipes in series, parallel branches, and simple networks.
8. Recognize the assumptions, limitations, and uncertainty associated with pipe-flow calculations.

1. Viscosity and the Origin of Flow Resistance

Viscosity is a fluid’s resistance to deformation by shear. For a Newtonian fluid, shear stress is proportional to the velocity gradient:

Here, τ is shear stress in Pa, μ is dynamic viscosity in Pa·s, u is velocity parallel to the wall in m/s, and y is the coordinate normal to the wall in m. The relation applies to Newtonian fluids such as water, air, and many light oils under ordinary conditions. Non-Newtonian fluids require a different constitutive model.

The kinematic viscosity is

where ν is in m²/s and ρ is density in kg/m³. Viscosity generally decreases with temperature for liquids and increases with temperature for gases. Because viscosity affects Reynolds number and friction factor, fluid temperature must be considered in engineering calculations.

At a solid boundary, the no-slip condition is normally applied: the fluid velocity relative to the wall is zero. Momentum is transferred from faster fluid layers to slower layers through viscous action. In a pipe, this produces a velocity gradient and a pressure drop.

2. Reynolds Number and Flow Regimes

Reynolds’ experiment showed that flow can be smooth and orderly or irregular and fluctuating. The governing dimensionless group is the Reynolds number:

where V is mean velocity, D is pipe diameter, and the other symbols have their definitions above. Reynolds number represents the ratio of inertial effects to viscous effects.

For flow in a circular pipe, the usual engineering classifications are:

These limits are not absolute. Entrance disturbances, roughness, vibration, bends, and flow pulsation can cause transition earlier or later. In the transitional range, results are inherently uncertain and should not be treated as fully developed laminar or fully developed turbulent flow.

The Reynolds experiment can be interpreted physically. At low velocity, dye injected into a pipe remains in a thin filament, indicating limited transverse mixing. At high velocity, the dye rapidly disperses because turbulent eddies transport momentum across the pipe.

3. Fully Developed Flow in a Circular Pipe

A flow is fully developed when the velocity profile no longer changes in the axial direction. For steady, incompressible, fully developed flow in a straight, constant-diameter pipe:

• the mean velocity is constant along the pipe;
• pressure decreases linearly with axial distance;
• axial velocity varies only with radius;
• radial velocity is zero;
• the wall satisfies the no-slip condition.

These assumptions exclude entrance regions, fittings, sudden expansions, and other locations where the profile is changing.

3.1 Laminar velocity profile

Consider a cylindrical fluid element of radius r and length L. A pressure force is balanced by viscous shear. The resulting shear-stress distribution is

where p is pressure and x is the pipe-axis coordinate. Since pressure falls in the flow direction, dp/dx < 0, so shear stress is positive in magnitude.

Using τ = −μ du/dr and integrating with the condition u = 0 at r = R, where R = D/2, gives the parabolic velocity profile:

The maximum velocity occurs at the centreline:

The cross-sectional mean velocity is

Therefore,

This centreline-to-mean velocity ratio of 2 is a distinctive result for fully developed laminar flow in a circular pipe.

3.2 Hagen–Poiseuille equation

Since −dp/dx = ∆p/L, the mean-flow relation becomes

Equivalently,

where Q is volumetric flow rate in m³/s and ∆p is pressure drop in Pa. This is the Hagen–Poiseuille equation. It applies only to steady, incompressible, Newtonian, fully developed laminar flow in a straight circular pipe of constant diameter. The strong dependence on D⁴ explains why small reductions in diameter greatly increase resistance.

The wall shear stress is

The friction factor for fully developed laminar flow is

Here f is the Darcy friction factor. It must not be confused with the Fanning friction factor, which is one quarter of the Darcy value.

Worked Example 1: Laminar pressure drop

Water at approximately 20 °C has μ = 1.00×10^(−3) Pa·s and ρ = 998 kg/m³. It flows through a horizontal tube of diameter 10 mm and length 5 m at Q = 1.00×10^(−6) m³/s. Determine the pressure drop.

The area is

Thus,

The Reynolds number is

The flow is laminar. Applying Hagen–Poiseuille:

The corresponding head loss is h_L = ∆p/(ρg) ≈ 0.00208 m of water. The very low Reynolds number confirms that the laminar model is appropriate.

4. Turbulent Pipe Flow

Turbulent flow contains velocity fluctuations superimposed on a time-averaged motion. Turbulent mixing produces a flatter mean velocity profile than laminar flow, but a steep velocity gradient remains near the wall. The near-wall region is commonly divided into:

• viscous sublayer, where molecular viscosity dominates;
• buffer layer, where viscous and turbulent effects are comparable;
• logarithmic or turbulent region, where turbulent transport is important.

The velocity profile depends on Reynolds number and relative roughness, ε/D, where ε is the average roughness height. Roughness may be hydraulically smooth when it is buried within the viscous sublayer, or hydraulically rough when it protrudes through that layer.

Unlike laminar flow, there is no single elementary velocity-profile equation valid for all turbulent pipe conditions. Log-law and power-law approximations are useful, but friction-factor methods are generally preferred for system calculations.

5. Darcy–Weisbach Equation and Major Losses

For steady flow through a constant-diameter pipe, the frictional head loss is

The associated pressure loss is

Here L is pipe length, D is diameter, V is mean velocity, g is gravitational acceleration, and f is the Darcy friction factor. The equation applies to both laminar and turbulent flow when the correct friction factor is used.

For laminar flow, f = 64/Re. For turbulent flow, f depends on Reynolds number and relative roughness. The implicit Colebrook–White equation is

It applies to fully developed turbulent flow in commercial pipes. Since f appears on both sides, iteration or an explicit approximation is required. A useful explicit approximation is the Swamee–Jain relation:

It is an engineering approximation, not a replacement for careful computation near sensitive design limits. The Moody chart graphically represents the relationship among Re, ε/D, and f.

Worked Example 2: Turbulent pipe loss

Water flows through a commercial steel pipe with D = 0.10 m, L = 50 m, Q = 0.010 m³/s, ρ = 998 kg/m³, μ = 1.00×10^(−3) Pa·s, and ε = 0.045 mm. Find the major head loss using Swamee–Jain.

First,

The relative roughness is

Therefore,

Finally,

The major pressure loss is approximately ρgh_f ≈ 7.8 kPa. Minor losses must be added if fittings and valves are present.

6. Minor Losses

Local components disturb the flow and cause additional energy loss. The standard representation is

where K is the dimensionless loss coefficient based on the selected reference velocity. Typical components include entrances, exits, elbows, tees, contractions, expansions, valves, screens, and strainers.

For an abrupt enlargement, an idealized expression based on the Borda–Carnot model is

where V₁ and V₂ are upstream and downstream mean velocities. Since downstream area is larger, V₂ < V₁. The corresponding loss coefficient based on V₁ is K = (1 − V₂/V₁)².

For a gradual or well-designed fitting, K is smaller than for an abrupt change. Values should be taken from reliable manufacturer data, recognized handbooks, or validated tables. Loss coefficients can vary with Reynolds number, geometry, valve position, and installation arrangement.

At a pipe entrance, the loss may be expressed as

At a large reservoir exit, the mean kinetic energy is dissipated and the usual exit coefficient is approximately K_exit = 1 when based on pipe velocity.

The total head loss in a system is commonly written

The first term is the major loss and the summation represents minor losses. The adjective “minor” refers to the cause, not necessarily the magnitude: a valve or poorly designed entrance can dominate the total loss in a short system.

7. Energy Equation for Pipe Systems

Between two points in a flowing system, the steady-flow energy equation is

Here z is elevation, p/(ρg) is pressure head, αV²/(2g) is kinetic-energy head, h_p is pump head added, h_t is turbine head removed, and h_L is total loss. The kinetic-energy correction factor α is approximately 2 for fully developed laminar flow and close to 1 for many turbulent flows. It should be included when velocity profiles are significantly nonuniform.

This equation provides the system-level connection between pressure, elevation, velocity, pumps, turbines, and losses. It is valid when the selected control points and loss models are consistent.

8. Equivalent Pipes

Equivalent-pipe methods replace a complicated pipe arrangement with a simpler pipe producing the same discharge and head loss under specified conditions. They are useful for preliminary design but should not obscure the actual geometry.

For pipes in series, the discharge is the same through every pipe, while head losses add:

For turbulent pipes with fixed friction factors and negligible minor losses,

For pipes in parallel between the same two junctions, the head loss is the same in every branch and discharges add:

Because V_i = Q_i/A_i, each branch must be solved using its own diameter, length, roughness, and friction factor. Assuming equal flow division is generally incorrect unless branch resistances are equal.

For laminar flow, resistance can be written as

where

For laminar pipes in series, hydraulic resistances add. For parallel laminar branches,

Turbulent resistance is not generally constant because friction factor changes with Reynolds number; consequently, turbulent networks require iteration.

9. Pipe Networks

A pipe network consists of pipes connected at junctions and loops. Network analysis must satisfy two fundamental conditions:

1. Continuity at each junction: total inflow equals total outflow.
2. Energy consistency around each closed loop: algebraic head changes sum to zero.

For a junction,

A practical solution procedure is:

1. Specify known reservoir heads, pipe properties, and boundary conditions.
2. Assume flow directions and initial branch discharges.
3. Apply continuity at all internal nodes.
4. Calculate Reynolds numbers, friction factors, and losses.
5. Correct discharges until loop energy residuals approach zero.
6. Check that final flow directions and pressures are physically reasonable.

Modern software uses nonlinear solvers, such as Newton–Raphson methods, for network equations. The engineer remains responsible for selecting valid component data, checking convergence, identifying unreasonable pressures, and verifying that the numerical model represents the actual system. A converged solution is not necessarily a correct solution if the inputs or topology are wrong.

10. Entrance Length and Applicability Limits

Fully developed formulas should not be applied immediately after an inlet. Approximate entrance lengths for a circular pipe are:

for laminar flow, and commonly

for turbulent flow, depending on disturbance level and geometry. The laminar estimate is an order-of-magnitude relation; actual development depends on inlet conditions.

Important limitations include:

• compressibility effects in gases at high Mach number or large pressure changes;
• temperature-dependent viscosity and density;
• non-Newtonian rheology;
• unsteady or pulsating flow;
• multiphase flow;
• noncircular ducts, where hydraulic diameter may be used cautiously;
• severe curvature, rotating flow, or rapidly changing cross-section.

For a noncircular duct, the hydraulic diameter is

where A is flow area and P_w is wetted perimeter. The use of D_h is most reliable for ducts with established internal-flow correlations and should not be assumed universally valid.

11. Applied Case: Cooling-Water Circuit for a Machine Tool

Background

An illustrative manufacturing cell uses a pump to circulate water from a storage tank through a heat exchanger and back to the tank. The main pipe is 40 mm in diameter and 30 m long. The circuit also contains four standard elbows, a fully open valve, an entrance, and an outlet. The required flow rate is 0.004 m³/s.

Problem

The designer must determine whether the selected pump can provide the required flow while maintaining reasonable energy consumption. Ignoring fitting losses would underestimate the required pump head.

Analysis

The mean velocity is

For water at approximately 20 °C, Re ≈ 1.27×10⁵, so the flow is turbulent. Suppose the pipe has relative roughness ε/D = 0.001 and the estimated Darcy factor is f = 0.022. The major loss is

Assume representative coefficients: entrance K = 0.5, four elbows each K = 0.3, valve K = 0.2, and exit K = 1.0. Thus,

If the reservoirs have equal free-surface elevation and negligible surface velocities, the pump must supply approximately

If the required hydraulic power is desired,

For water, this gives approximately P_h = 998×9.81×0.004×10.0 ≈ 391 W. With a pump efficiency of 70%, the shaft power would be approximately 391/0.70 ≈ 559 W.

Outcome and learning points

The minor losses contribute about 15% of the total loss in this illustrative system. A smaller-diameter pipe would increase velocity and losses sharply, whereas a larger pipe would reduce operating cost but increase installation cost. A responsible design would obtain actual valve and elbow data, include heat-exchanger loss, examine pump-system intersection, evaluate transient behavior, and verify that the pressure remains above vapor pressure to avoid cavitation.

12. Design Practice, Safety, and Contemporary Tools

Good pipe-system practice includes:

• use consistent units and distinguish gauge from absolute pressure;
• evaluate fluid properties at the actual operating temperature;
• use Darcy rather than Fanning friction factors unless explicitly specified;
• include fittings, entrances, exits, filters, heat exchangers, and valves;
• check both normal and maximum-flow conditions;
• avoid operation in the transitional Reynolds-number range when possible;
• verify pressure ratings, corrosion allowance, leakage control, and maintenance access;
• check for cavitation using available net positive suction head;
• validate computational results with hand estimates and field measurements.

Computer-aided network models and digital twins can rapidly evaluate many operating conditions. AI-based anomaly detection may identify unusual pressure-loss patterns or pump degradation from sensor data. However, such tools do not replace conservation laws, calibrated measurements, engineering judgement, or safe design margins. Data quality, model transparency, cybersecurity, and responsibility for decisions remain important professional issues.

Chapter Summary

Viscosity produces shear stress and mechanical-energy loss in internal flows. Reynolds number compares inertial and viscous effects and classifies pipe flow as laminar, transitional, or turbulent. Fully developed laminar flow has a parabolic velocity profile, a centreline velocity twice the mean velocity, and is governed by the Hagen–Poiseuille equation. Its Darcy friction factor is f = 64/Re.

For both laminar and turbulent flow, the Darcy–Weisbach equation relates frictional head loss to pipe length, diameter, velocity, and friction factor. Turbulent friction factors depend on Reynolds number and relative roughness and may be obtained from the Moody chart, Colebrook equation, or suitable explicit approximations. Local fittings produce minor losses represented by h_m = K V²/(2g).

Pipes in series carry the same flow and have additive losses. Parallel branches have the same head loss and share the total flow according to their hydraulic resistances. Pipe networks must satisfy continuity at junctions and energy balance around loops. Correct engineering analysis requires awareness of entrance effects, property variation, roughness, component data, uncertainty, and the limitations of simplified models.

Key Terms

Viscosity: Resistance of a fluid to shear deformation.

Dynamic viscosity: Proportionality constant between shear stress and velocity gradient.

Kinematic viscosity: Dynamic viscosity divided by density.

No-slip condition: Fluid velocity equals the solid-wall velocity at a stationary boundary.

Reynolds number: Ratio of inertial effects to viscous effects.

Laminar flow: Smooth, predominantly ordered internal flow.

Transitional flow: Unstable regime between laminar and turbulent flow.

Turbulent flow: Flow containing irregular velocity fluctuations and eddies.

Fully developed flow: Flow whose velocity profile does not vary along the pipe.

Hagen–Poiseuille equation: Pressure-flow relation for fully developed laminar flow in a circular pipe.

Wall shear stress: Shear stress exerted by the fluid at the pipe wall.

Darcy friction factor: Dimensionless factor used in the Darcy–Weisbach equation.

Relative roughness: Ratio of roughness height to pipe diameter.

Moody chart: Graph relating Reynolds number, relative roughness, and Darcy friction factor.

Major loss: Head loss caused by wall friction along a pipe length.

Minor loss: Local head loss caused by fittings, entrances, exits, and geometry changes.

Loss coefficient: Dimensionless coefficient relating local loss to velocity head.

Equivalent pipe: Simplified pipe representation with the same specified hydraulic behavior.

Hydraulic network: Connected set of pipes requiring continuity and energy balances.

Hydraulic diameter: Four times flow area divided by wetted perimeter.

Review Questions

5 Short Answer Questions

1. What physical effects are compared by Reynolds number?
2. State the no-slip boundary condition at a stationary pipe wall.
3. Why is the centreline velocity twice the mean velocity in fully developed laminar circular-pipe flow?
4. Distinguish major losses from minor losses.
5. State the continuity conditions for pipes in series and pipes in parallel.

5 Long Answer Questions

1. Derive the parabolic velocity profile for steady, fully developed laminar flow in a circular pipe and obtain the Hagen–Poiseuille equation.
2. Explain the physical meaning of the Moody chart and describe how relative roughness affects turbulent friction factor.
3. Derive the Darcy–Weisbach expression for pressure loss in terms of head loss and explain every variable.
4. Discuss the assumptions and limitations of applying fully developed pipe-flow equations to real systems.
5. Explain how a pump-system energy balance is constructed for a pipe circuit containing elevation changes, fittings, and a valve.

3 Analytical / Application Questions

1. Water flows through a 25 mm pipe at 0.0015 m³/s. Determine the mean velocity, Reynolds number using ρ = 998 kg/m³ and μ = 1.00×10^(−3) Pa·s, and the likely flow regime.
2. Two parallel branches connect the same upstream and downstream junctions. Explain why equal flow cannot be assumed when the branches have different diameters and lengths, and outline an iterative solution method.
3. A designer proposes reducing a pipe diameter by 20% while maintaining the same laminar flow rate. Using the Hagen–Poiseuille relation, determine qualitatively—and, if desired, quantitatively—how the pressure drop changes and discuss the design consequence.

Multiple Choice Questions with Answers

Question 1

For fully developed laminar flow in a circular pipe, the Darcy friction factor is:

A. 16/Re

B. 32/Re

C. 64/Re

D. 128/Re

Correct Answer: C. 64/Re

Question 2

In fully developed laminar flow through a circular pipe, the centreline velocity is:

A. Equal to the mean velocity

B. Half the mean velocity

C. Twice the mean velocity

D. Four times the mean velocity

Correct Answer: C. Twice the mean velocity

Question 3

A pipe has diameter 0.05 m and mean velocity 1.0 m/s. For water with ν = 1.0×10^(−6) m²/s, its Reynolds number is:

A. 500

B. 5,000

C. 50,000

D. 500,000

Correct Answer: C. 50,000

Question 4

Which equation is used to calculate frictional head loss along a straight pipe?

A. Bernoulli’s equation without losses

B. Darcy–Weisbach equation

C. Continuity equation only

D. Hydrostatic pressure equation

Correct Answer: B. Darcy–Weisbach equation

Question 5

For pipes connected in parallel, which quantity is the same for every branch between common junctions?

A. Discharge

B. Mean velocity

C. Head loss

D. Pipe diameter

Correct Answer: C. Head loss

Question 6

The usual expression for a local fitting loss is:

A. h_m = KV²/(2g)

B. h_m = 2g/(KV²)

C. h_m = fD/(LV²)

D. h_m = ρgV

Correct Answer: A. h_m = KV²/(2g)

Question 7

If pipe diameter is reduced while flow rate remains constant, mean velocity:

A. Decreases

B. Increases

C. Remains zero

D. Is independent of area

Correct Answer: B. Increases

Question 8

The relative roughness used in turbulent pipe-flow calculations is:

A. D/ε

B. εD

C. ε/D

D. ε + D

Correct Answer: C. ε/D

Question 9

For pipes in series, the discharge under steady incompressible conditions is:

A. Different in every pipe because losses occur

B. The same through every pipe

C. Zero in the smallest pipe

D. Determined only by roughness

Correct Answer: B. The same through every pipe

Question 10

The Hagen–Poiseuille equation is directly applicable when the flow is:

A. Steady, incompressible, Newtonian, fully developed, and laminar

B. Always turbulent and compressible

C. Unsteady and multiphase

D. Only flow through open channels

Correct Answer: A. Steady, incompressible, Newtonian, fully developed, and laminar