Fluid Mechanics — Engineering Reference

1. At a glance

Fluid mechanics is the engineering discipline that describes how fluids — liquids and gases — move, and how they exchange momentum, energy, and mass with the solid surfaces that contain or are immersed in them. The field splits into a handful of working regimes that share a common mathematical core (the Navier–Stokes equations) but use different specialisations in practice:

• Hydrostatics — fluids at rest; buoyancy, pressure on submerged surfaces, manometry.
• Incompressible internal flow — pipes, ducts, manifolds, plumbing, hydraulics.
• Incompressible external flow — drag on vehicles, wind on buildings, lift on wings at low Mach.
• Compressible flow — gas dynamics, nozzles, diffusers, shocks, aircraft cruise, turbomachinery.
• Boundary layers — the thin near-wall region where viscosity matters even at very high Reynolds number.
• Turbomachinery flow — pumps, fans, blowers, compressors, turbines.
• Multiphase / non-Newtonian flow — slurries, cavitating pumps, two-phase pipe flow, polymer extrusion, blood.

Every fluid system an engineer sizes — a building’s chilled-water loop, an aircraft’s hydraulic system, a refinery’s transfer line, an HVAC duct run, a wastewater lift station, an automotive cooling jacket, a refrigeration suction line — comes back to flow rate, pressure drop, pump/fan curve matched to a system curve, cavitation margin, and flow-induced vibration. Get these wrong and pumps cavitate, valves fail, ducts whistle, pipes hammer, vessels collapse from sub-atmospheric pressure. Place in the design stack: thermodynamics → fluid mechanics → heat transfer → turbomachinery / HVAC / aerodynamics / hydraulics → CFD → commissioning testing.

2. Why it matters

A working engineer touches fluid mechanics within hours of finishing school. Three classes of error are universal: (1) under-estimated pressure drop, so the pump or fan can’t deliver design flow and the system never reaches steady state; (2) ignored cavitation margin, so the pump destroys itself in months; (3) ignored compressibility or water-hammer, so the system hits a transient pressure five or ten times design and bursts. A correctly executed fluid analysis is short — Reynolds number, friction factor, Bernoulli with head loss, NPSH check — but it has to be done. The same calculations sit under the certification of every aircraft (drag polar, lift curve, intake mass flow), the heat-rate of every power plant (steam-turbine stage efficiency, condenser pressure drop), the comfort of every building (ASHRAE 90.1 duct sizing), and the safety of every chemical process (relief-valve sizing, two-phase flashing flow per API 520).

3. First principles

Continuum hypothesis

Fluid mechanics treats the fluid as a continuous medium with smoothly varying density, velocity, pressure, and temperature. This is valid when the Knudsen number Kn = λ / L < 0.01, where λ is the molecular mean free path and L is the smallest geometric scale. For air at STP, λ ≈ 68 nm, so the continuum approximation is excellent down to L ≈ 10 µm. It breaks at MEMS scales, at very low pressure (rarefied flow, satellite re-entry), and in microchannels — there a kinetic-theory (Boltzmann) description is required.

Fluid properties

• Density ρ [kg/m³]. Water at 20 °C: 998. Air at 20 °C, 1 atm: 1.204.
• Dynamic viscosity µ [Pa·s = N·s/m²]. Water: 1.00 × 10⁻³. Air: 1.81 × 10⁻⁵.
• Kinematic viscosity ν = µ/ρ [m²/s]. Water: 1.00 × 10⁻⁶. Air: 1.51 × 10⁻⁵.
• Surface tension σ [N/m]. Water/air at 20 °C: 0.0728.
• Bulk modulus K [Pa]. Water: 2.2 × 10⁹ (so liquid water compresses by ~0.5 % at 100 bar — usually negligible).
• Vapor pressure P_vap [Pa]. Water at 20 °C: 2.34 kPa. Sets cavitation onset.

Newtonian and non-Newtonian fluids

For a Newtonian fluid, shear stress is linear in the velocity-gradient (rate of strain):

τ = µ · du/dy            (Newton 1687)

Water, air, oils (in their normal range), low-molecular-weight gases — all Newtonian. Non-Newtonian fluids appear as soon as you handle polymers, slurries, suspensions, biological fluids, and foodstuffs:

• Shear-thinning (pseudoplastic) — viscosity drops with shear rate. Paint, blood, polymer melts. Modelled as τ = K·γ̇ⁿ with n < 1 (power-law / Ostwald–de Waele).
• Shear-thickening (dilatant) — viscosity rises with shear rate. Cornstarch–water, dense suspensions. Power-law with n > 1.
• Bingham plastic — no flow below yield stress τ_y, then Newtonian above. Toothpaste, mud, fresh concrete: τ = τ_y + µ_p·γ̇.
• Viscoelastic — combines elastic and viscous response. Polymer solutions, biological tissue. Maxwell and Kelvin–Voigt models.

Compressibility regime

Liquids are treated as incompressible except in water-hammer analysis. Gases are treated as incompressible when the Mach number M = V/c stays below 0.3 — at M = 0.3 the density change is about 4.5 %, which is the conventional engineering threshold. Above M = 0.3 use compressible-flow relations (Section 8p).

Boundary conditions

• No-slip at solid walls: the fluid velocity equals the wall velocity. Universal for continuum-scale viscous flow. Holds even at very high Re — it is what makes a boundary layer exist.
• Free surface: pressure equal to ambient (gauge zero), and kinematic condition that fluid particles on the surface remain on the surface.
• Symmetry / inlet / outlet conditions specified for CFD.

4. Fluid statics

For a fluid at rest, momentum balance reduces to:

dP/dz = − ρ g                 ⇒    P(h) = P_atm + ρ g h

This is the hydrostatic equation: pressure rises linearly with depth in an incompressible fluid. For a 10 m water column: ΔP = 998 · 9.81 · 10 = 97.9 kPa, very nearly 1 atm — the origin of “1 atm ≈ 10 m of water” rule.

Manometry. A U-tube manometer reads pressure difference ΔP = (ρ_m − ρ_f) g h, with ρ_m the manometer fluid (mercury, ρ ≈ 13 550 kg/m³, or red oil, ρ ≈ 827 kg/m³) and ρ_f the working fluid. Inclined-tube and micromanometer designs amplify h by 1/sin θ for low-ΔP gas work.

Forces on submerged surfaces. Resultant force on a plane surface of area A submerged in liquid: F = ρ g h_c · A, applied at the centre of pressure h_cp = h_c + I_c / (h_c A), which sits below the centroid because pressure rises with depth.

Buoyancy (Archimedes, 250 BCE). The buoyant force equals the weight of displaced fluid: F_B = ρ_fluid · V_displaced · g, acting at the centre of buoyancy (the centroid of the displaced volume). A floating body is stable when its metacentric height GM > 0, i.e. the metacentre M sits above the centre of gravity G. For a rectangular ship hull, GM = I_waterplane / V_displaced − BG. Hull design is a balance: high GM = stiff, snappy roll (uncomfortable, hard on cargo); low GM = sluggish, more comfortable, but a roll-stability risk.

Pressure measurement. Bourdon-tube gauge (0–700 bar, ±1 %), piezoresistive transducer (0–1000 bar, ±0.1 %), capacitance diaphragm (low ranges, 0–10 kPa, ±0.05 % for absolute / vacuum work), MEMS pressure sensors (consumer / automotive), strain-gauge transducers.

5p. Conservation equations

The three governing equations of fluid mechanics, in their general (compressible, viscous, time-dependent) form:

Continuity (mass conservation):

∂ρ/∂t + ∇·(ρ v) = 0

For incompressible flow this reduces to ∇·v = 0.

Momentum (Navier–Stokes, Navier 1822 + Stokes 1845):

ρ (∂v/∂t + v·∇v) = − ∇P + µ ∇²v + ρ g

Per unit volume: inertia (LHS) = pressure gradient + viscous diffusion + body force. The non-linear convective term v·∇v is the source of turbulence and the reason the Navier–Stokes equations have no general closed-form solution (the Clay Millennium “existence and smoothness” prize remains open as of 2025).

Energy (incompressible, thermal):

ρ c_p (∂T/∂t + v·∇T) = k ∇²T + Φ + Q̇

with Φ = viscous dissipation and Q̇ = volumetric heat source. Couples to heat-transfer for thermal–fluid problems.

Bernoulli’s equation (Bernoulli 1738) — the special case of steady, inviscid, incompressible flow along a streamline:

P + ½ ρ v² + ρ g z = constant

Or, dividing by ρg, in head form:

P/(ρg) + v²/(2g) + z = H = constant

Each term has units of length: pressure head, velocity head, elevation head, total head. The engineering “modified Bernoulli” between sections 1 and 2 adds machinery work and losses:

P₁/(ρg) + v₁²/(2g) + z₁ + h_pump − h_turbine − h_loss = P₂/(ρg) + v₂²/(2g) + z₂

This is the working equation for pipe-system analysis: the system curve is the right-hand side as a function of flow, and the pump curve supplies h_pump(Q).

Control-volume momentum for steady flow:

ΣF = ∮_CS v (ρ v·n) dA

Gives the reaction force on a nozzle, jet impacting a plate, pipe bend, or rocket engine. A 90° elbow with water entering at 5 m/s and Q = 0.1 m³/s sees a reaction R = ρQ·v · √2 = 998 · 0.1 · 5 · √2 = 706 N on the supports — small but it’s why long pipe runs need thrust-blocks.

6p. Internal flow (pipes and ducts)

Reynolds number Re = ρVD/µ = VD/ν (Reynolds 1883) classifies the flow regime:

Re range	Regime in a circular pipe
< 2 300	Laminar (Hagen 1839, Poiseuille 1840)
2 300 – 4 000	Transitional (correlations diverge — design for worst case)
> 4 000	Fully turbulent

For non-circular ducts use the hydraulic diameter D_h = 4 A / P_wet, where A is cross-sectional area and P_wet is the wetted perimeter. For a square duct of side a: D_h = a. For an annulus: D_h = D_o − D_i.

Laminar fully-developed flow

The exact Hagen–Poiseuille solution for a circular pipe:

v(r) = (ΔP / (4 µ L)) · (R² − r²)         (parabolic profile)
v_max = ΔP R² / (4 µ L) = 2 V_avg
Q = π R⁴ ΔP / (8 µ L)                      (Hagen–Poiseuille flow)
ΔP_laminar = 32 µ L V / D²
f_laminar = 64 / Re                        (Darcy friction factor)

Turbulent fully-developed flow

ΔP = f · (L/D) · (ρ V² / 2)               (Darcy–Weisbach equation)

where f is the Darcy friction factor, read from the Moody diagram (Moody 1944) or computed from correlations. The two relevant inputs are Re and the relative roughness ε/D.

Friction-factor correlations:

• Smooth-pipe turbulent, Re < 1 × 10⁵: f = 0.316 / Re^0.25 (Blasius 1913).

• General turbulent, implicit (most accurate): Colebrook–White (1939)

  1/√f = − 2 log₁₀ [ ε/(3.7 D) + 2.51 / (Re √f) ]
  

  Iterate (3–4 fixed-point passes converge from f₀ = 0.02).

• Explicit equivalent: Swamee–Jain (1976), accurate within 1 % for 10⁻⁶ < ε/D < 10⁻² and 5 000 < Re < 10⁸:

  f = 0.25 / { log₁₀ [ ε/(3.7 D) + 5.74 / Re^0.9 ] }²
  

• Alternative explicit: Haaland (1983):

  1/√f = − 1.8 log₁₀ [ (ε/(3.7 D))^1.11 + 6.9/Re ]
  

Minor losses (fittings)

Add a K-factor head loss h_L = K · v² / (2g), or use an equivalent length L_eq/D.

Fitting	K (typical)	L_eq/D
Sharp-edged entrance	0.50	25
Rounded / bell-mouth entrance	0.04	2
Pipe exit (to large reservoir)	1.00	n/a
90° standard elbow (threaded)	0.75	30
90° long-radius elbow (flanged)	0.30	15
45° elbow	0.35	16
Tee, flow-through (run)	0.20	10
Tee, branch flow	1.0	50
Gate valve, fully open	0.15	7
Gate valve, ½ open	2.1	—
Globe valve, fully open	6 – 10	300
Swing check valve	2.0	100
Ball valve, fully open	0.05	—
Butterfly valve, fully open	0.45	20
Sudden contraction (D₂/D₁ = 0.5)	0.30	—
Sudden expansion (D₁/D₂ = 0.5)	0.56	—

Sources: Crane TP-410, Idelchik, Cengel/Cimbala. Use Idelchik for non-standard geometries — it remains the definitive K-factor compilation.

Surface roughness (commercial pipe)

Material	ε [mm]	ε [in.]
Drawn tubing (Cu, brass, glass)	0.0015	6 × 10⁻⁵
Commercial steel / wrought iron	0.046	1.8 × 10⁻³
Galvanized iron	0.15	6 × 10⁻³
Cast iron	0.26	1.0 × 10⁻²
Asphalted cast iron	0.12	4.8 × 10⁻³
Concrete	0.3 – 3.0	0.01–0.12
Riveted steel	0.9 – 9.0	0.04–0.4
HDPE / smooth plastic	0.0015	6 × 10⁻⁵

Note: ε grows in service from biofilm, scale, and corrosion — after 20 years a 0.046 mm steel pipe can effectively be 0.5–2 mm. Always design with allowance.

Multi-pipe and open-channel

• Series pipes add head losses: h_loss,total = Σ h_loss,i at common Q.
• Parallel pipes add flows at common head loss: Q_total = Σ Q_i with Δh equal in each branch.
• Pipe networks (water distribution): Hardy-Cross method (1936), or modern solvers like EPANET (US EPA, open-source) and Bentley WaterGEMS.
• Open-channel flow: Manning’s equation V = (1/n) R_h^(2/3) S^(1/2) (Manning 1889) with n the roughness coefficient (n = 0.013 for concrete, 0.025 for natural stream). Flow regime classified by Froude number Fr = V/√(gL): Fr < 1 subcritical, Fr > 1 supercritical, Fr = 1 critical (hydraulic-jump location).

7p. External flow

External flow problems — drag on a vehicle, wind on a building, lift on a wing — are dominated by what happens in the boundary layer and wake.

Boundary-layer theory

Prandtl (1904) recognised that even at very high Re the viscous no-slip condition forces a thin region near the wall where velocity climbs from 0 to U_∞. The boundary-layer thickness δ grows along the flow:

• Laminar (Blasius 1908): δ/x = 5.0 / √Re_x, C_f,local = 0.664 / √Re_x.
• Turbulent (Prandtl 1921, power-law): δ/x = 0.37 / Re_x^0.2, C_f,local = 0.0592 / Re_x^0.2.

Transition on a smooth flat plate occurs around Re_x ≈ 5 × 10⁵, but free-stream turbulence, surface roughness, and pressure gradients shift it. Boundary-layer separation occurs where dP/dx > 0 (adverse pressure gradient) outpaces the wall shear’s ability to keep flow attached — flow detaches, forms a wake, dramatically raises drag, and can stall a wing.

Drag and lift

F_D = ½ ρ V² C_D A         F_L = ½ ρ V² C_L A

Reference area convention: frontal area for blunt bodies (cars, cylinders, spheres), planform area for wings.

Geometry	C_D (typical)
Smooth sphere, Re = 10⁴ (laminar BL)	0.47
Smooth sphere, Re = 5 × 10⁵ (post-transition)	0.07
Golf ball (dimpled, Re ≈ 10⁵)	≈ 0.25 (transition forced)
Circular cylinder, Re = 10⁴ (laminar)	1.2
Circular cylinder, Re = 10⁶ (turbulent)	0.3
Flat plate normal to flow	1.17
Flat plate parallel (skin friction only, Re = 10⁶)	≈ 0.003 (C_f)
Streamlined teardrop (L/D ≈ 5)	0.04
Modern passenger car	0.25 – 0.35
Pickup truck / SUV	0.35 – 0.45
Tractor-trailer (highway)	0.6 – 0.9
Cyclist (upright / racing tuck)	1.1 / 0.7
Parachute (round canopy)	1.3

Vortex shedding. Behind a bluff body, alternating vortices form a Karman vortex street at frequency f_s. The Strouhal number St = f_s D / V ≈ 0.21 for a circular cylinder in the range 300 < Re < 2 × 10⁵. When f_s matches a structural natural frequency, flow-induced vibration can be catastrophic — Tacoma Narrows Bridge (1940) is the famous case, but every heat-exchanger U-tube manufacturer fights the same problem, and offshore-platform riser design is wholly governed by it.

Drag-reduction strategies. Streamlining (turn pressure drag into skin friction by closing the wake), turbulators (dimples, trip wires — paradoxically reduce sphere drag by forcing earlier transition that delays separation), riblets (longitudinal grooves, ~5 % skin-friction reduction in turbulent BL, used on aircraft), polymer drag reduction (Toms effect — parts-per-million polyacrylamide in water cuts turbulent friction by up to 80 % in pipelines).

8p. Compressible flow

Above M = 0.3, density variations matter. Define M = V/c with the sound speed c = √(γ R T) (ideal gas; γ = 1.4 for air, R = 287 J/(kg·K)). At T = 288 K (ISA sea level), c = 340 m/s.

Isentropic-flow relations (steady, adiabatic, reversible, ideal gas):

T₀ / T = 1 + (γ−1)/2 · M²
P₀ / P = [1 + (γ−1)/2 · M²]^(γ/(γ−1))
ρ₀ / ρ = [1 + (γ−1)/2 · M²]^(1/(γ−1))
A / A* = (1/M) [ (2/(γ+1)) (1 + (γ−1)/2 · M²) ]^((γ+1)/(2(γ−1)))

Subscript 0 = stagnation. A* = sonic-throat area.

Normal shock — a discontinuity from M₁ > 1 (supersonic) to M₂ < 1 (subsonic) across which mass, momentum, and energy are conserved but entropy rises. Rankine–Hugoniot relations give the jumps in P, ρ, T and M downstream as algebraic functions of M₁ and γ.

Nozzle flow. A converging nozzle accelerates subsonic flow toward M = 1 at the throat. A converging–diverging (de Laval) nozzle continues acceleration to supersonic in the diverging section provided the back-pressure is low enough. The flow is choked when M = 1 at the throat — mass flow becomes independent of back-pressure:

ṁ_max = P₀ A* · √(γ / (R T₀)) · [2/(γ+1)]^((γ+1)/(2(γ−1)))

Critical pressure ratio for choking: P* / P₀ = [2/(γ+1)]^(γ/(γ−1)) = 0.528 for air. So an air line at P₀ = 200 kPa exhausting to atmosphere (P_back ≈ 101 kPa) is choked, since 101/200 = 0.505 < 0.528.

Fanno flow (constant-area duct with friction) and Rayleigh flow (constant-area duct with heat addition) are the two canonical 1-D variable-area-substitute models — friction or heating both drive subsonic flow toward M = 1 and supersonic flow toward M = 1, with sonic conditions setting the limiting length / heating.

9p. Turbulence

Turbulence is the rotational, three-dimensional, broadband chaotic state of high-Re flow. The energy cascade (Kolmogorov 1941) spans from the integral scale (geometric, set by L) down to the Kolmogorov microscale η = (ν³/ε)^(1/4) where viscous dissipation balances energy transfer. The ratio L/η scales as Re^(3/4) — at Re = 10⁶ a fully resolved simulation needs ~Re^(9/4) ≈ 3 × 10¹³ degrees of freedom, which is the basic reason DNS is infeasible for engineering geometries.

RANS (Reynolds-Averaged Navier–Stokes). Decompose v = V̄ + v′, average; the non-linear term yields the Reynolds stress tensor −ρ v′v′ that has to be modelled (closure problem). Common closures:

Model	Best for	Weaknesses
Mixing length	Boundary layers, pipe flow (1-D)	No history, fails in recirculating flow
k-ε	Free shear flow, jets, far field	Inaccurate near walls, weak in separation
k-ω (Wilcox)	Near-wall, adverse pressure gradient	Sensitive to free-stream ω
k-ω SST	All-purpose default in modern CFD	Some swirling-flow inaccuracy
RSM (Reynolds Stress)	Strongly swirling, secondary flows	Costly, harder to converge
Spalart–Allmaras	Aerospace external aero	Tuned for attached BL, weak in massively-separated

LES (Large-Eddy Simulation) — directly resolves large eddies, models the sub-grid scales (Smagorinsky 1963; dynamic-Smagorinsky, WALE). Required for jet noise, unsteady aero, combustion stability, certain HVAC mixing problems. Cost: 50–1000× RANS for the same geometry.

y+ for first cell off the wall: y+ = u_τ · y / ν, where u_τ = √(τ_wall/ρ). Engineering practice:

• y+ < 1 for resolved viscous sublayer (low-Re wall treatment, k-ω SST).
• 30 < y+ < 300 for wall-function treatment (high-Re wall law, standard k-ε).
• Anything in between is the trouble zone — many commercial codes provide automatic wall-treatment switching, but the user still has to mesh accordingly.

Engineering judgment: RANS handles roughly 90 % of design problems. Reach for LES only when the design is governed by unsteady physics (cabin acoustics, flutter, combustion stability, vortex shedding off bluff bodies near resonance).

10p. Cavitation, two-phase, special phenomena

Cavitation. When local liquid pressure falls below P_vap, vapor pockets form; downstream they collapse implosively, generating shock waves that pit metal. Pump suction is the canonical case: design NPSH_available ≥ NPSH_required + margin (typically 0.6 m of safety, per ANSI/HI 9.6.1). NPSH_a = (P_atm − P_vap)/(ρg) + z_s − h_loss_suction (with z_s positive if liquid level is above pump centerline). First symptoms in the field are noise (“gravel in the pump”), drop in delivered head, and pitting on impeller leading edges discovered at the next overhaul. Untreated cavitation destroys an impeller in weeks.

Water hammer (Joukowsky surge, 1898). Sudden valve closure converts kinetic energy of the column into a pressure transient propagating at the wave speed c_wave = √(K/ρ) / √(1 + (K D)/(E t)) (≈ 1200 m/s in steel water pipe). The pressure surge:

ΔP_surge = ρ · c_wave · ΔV

For water at 2 m/s suddenly stopped: ΔP = 998 · 1200 · 2 = 2.4 MPa ≈ 24 bar added to the static pressure — enough to burst many distribution pipes. Mitigation: slow valve closure (closure time > 2L/c_wave), surge tanks, accumulators, air chambers. Software: Bentley HAMMER, AFT Impulse.

Sloshing. Free-surface dynamics inside partially-filled tanks (rail cars, ship cargo holds, aircraft fuel tanks, launch-vehicle propellant tanks). Resonant frequencies low (often 0.1–1 Hz) and easily coupled into vehicle dynamics. Baffles suppress.

Two-phase pipe flow. Gas–liquid mixtures take on flow regimes — bubble, plug, slug, stratified, wavy, annular, mist — that depend on the mass-flux ratio, pipe orientation, and fluid properties. Two-phase pressure drop is not the sum of single-phase contributions and can be 10–50× higher. Lockhart–Martinelli (1949), Friedel, Chisholm correlations; for industrial sizing use simulators like OLGA (SLB), LedaFlow (Kongsberg), PIPESIM (SLB).

Non-Newtonian flow. Polymer extrusion, blood, fresh concrete, drilling mud, slurries. Use power-law / Bingham / Herschel–Bulkley constitutive models in the Navier–Stokes momentum balance.

11p. Worked examples

Example A — Pipe pressure drop, water through commercial steel

Problem. Water at 20 °C flows at Q = 50 L/s through 200 m of 150-mm Schedule 40 commercial-steel pipe (ID = 154.1 mm, ε = 0.046 mm). The line has 2 fully-open gate valves, 4 long-radius 90° elbows, 1 sharp-edged entrance, and 1 pipe exit. Find ΔP.

Step 1 — Properties and velocity. ρ = 998 kg/m³, µ = 1.00 × 10⁻³ Pa·s. A = π · (0.1541)² / 4 = 1.865 × 10⁻² m². V = Q / A = 0.050 / 0.01865 = 2.682 m/s.

Step 2 — Reynolds number. Re = ρVD/µ = 998 · 2.682 · 0.1541 / 1.00 × 10⁻³ = 4.124 × 10⁵ → fully turbulent.

Step 3 — Friction factor (Swamee–Jain). ε/D = 0.046 / 154.1 = 2.985 × 10⁻⁴. Inside the log: ε/(3.7 D) + 5.74 / Re^0.9 = 8.07 × 10⁻⁵ + 5.74 / 1.30 × 10⁵ = 8.07 × 10⁻⁵ + 4.42 × 10⁻⁵ = 1.249 × 10⁻⁴. log₁₀(1.249 × 10⁻⁴) = −3.904. Squared: 15.24. f = 0.25 / 15.24 = 0.0164.

(Colebrook iteration gives f = 0.0166 — agreement within 1 %.)

Step 4 — Major (friction) head loss. h_f = f · (L/D) · V²/(2g) = 0.0164 · (200 / 0.1541) · (2.682)² / (2 · 9.81) h_f = 0.0164 · 1297.9 · 0.3666 = 7.81 m of water.

Step 5 — Minor losses. ΣK = 2(0.15) + 4(0.30) + 1(0.50) + 1(1.00) = 0.30 + 1.20 + 0.50 + 1.00 = 3.00. h_m = ΣK · V²/(2g) = 3.00 · 0.3666 = 1.10 m.

Step 6 — Total. h_total = 7.81 + 1.10 = 8.91 m of water. ΔP = ρ g h_total = 998 · 9.81 · 8.91 = 87.2 kPa ≈ 12.6 psi.

Comment. Friction dominates (88 %) — typical for long runs. For short skid-mounted piping (< 20 m) minor losses can easily exceed friction. The system designer now sizes the pump to deliver 87 kPa at 50 L/s plus the elevation difference plus a margin for pipe-roughness growth.

Example B — Drag on a passenger car

Problem. A sedan with C_D = 0.30, frontal area A = 2.2 m², mass m = 1500 kg cruises at 100 km/h = 27.78 m/s on level ground. Air at 20 °C, ρ = 1.204 kg/m³. Rolling-resistance coefficient C_rr = 0.012. Find drag force, rolling resistance, total tractive power.

Step 1 — Aerodynamic drag. F_D = ½ · ρ · V² · C_D · A = 0.5 · 1.204 · (27.78)² · 0.30 · 2.2 F_D = 0.5 · 1.204 · 771.7 · 0.66 = 306.5 N.

Step 2 — Rolling resistance. F_rr = C_rr · m · g = 0.012 · 1500 · 9.81 = 176.6 N.

Step 3 — Tractive power. F_total = 306.5 + 176.6 = 483.1 N. P = F · V = 483.1 · 27.78 = 13.42 kW ≈ 18.0 hp.

Aero share: 306.5 / 483.1 = 63 % of cruise tractive power. At 130 km/h the V² scaling pushes drag to 518 N and aero share above 70 %. This is why a 10 % C_D reduction on a highway sedan gives 4–5 % real-world fuel economy gain.

Example C — Choked converging nozzle, air

Problem. A converging nozzle is fed from a reservoir at P₀ = 500 kPa, T₀ = 300 K, exhausting through a throat of area A* = 1.0 cm² to atmospheric back-pressure P_b = 100 kPa. Air: γ = 1.4, R = 287 J/(kg·K). Find throat conditions and mass flow.

Step 1 — Check for choking. P_b / P₀ = 100 / 500 = 0.200. Critical ratio = (2/2.4)^(1.4/0.4) = (0.8333)^3.5 = 0.528. Since 0.200 < 0.528 → choked, M = 1 at the throat.

Step 2 — Throat conditions. T*= T₀ · 2 / (γ+1) = 300 · 2 / 2.4 = 250.0 K. P* = P₀ · 0.528 = 264 kPa. ρ*= P*/(R T*) = 264 000 / (287 · 250) = 3.68 kg/m³. c* = √(γ R T*) = √(1.4 · 287 · 250) = 316.9 m/s = V* (since M = 1).

Step 3 — Mass flow. ṁ = P₀ A* · √(γ / (R T₀)) · [2/(γ+1)]^((γ+1)/(2(γ−1))) = 500 000 · 1.0 × 10⁻⁴ · √(1.4 / (287 · 300)) · (0.8333)^(2.4/0.8) = 50 · √(1.626 × 10⁻⁵) · (0.8333)^3.0 = 50 · 4.033 × 10⁻³ · 0.5787 = 0.1167 kg/s ≈ 0.12 kg/s.

Comment. Further reduction in P_b below the critical does not increase ṁ — that’s the practical definition of choked flow. To get more mass flow the engineer must raise P₀ or increase A*, not lower P_b. This is why pneumatic actuators and pressure relief valves are sized by the upstream condition only.

12p. Edge cases and gotchas

1. Transition Re (2 300 – 4 000). Correlations diverge here; design for the worst case, which is usually the laminar formula extended into the transitional range (it predicts higher ΔP per unit length).
2. Cold-start of viscous fluid. Heavy oil in a chilled pipeline can have µ 100× warm-design value; ΔP and pump load go up proportionally. Design for cold-start, not just steady-state.
3. Cavitation is silent at first. Earliest symptoms are noise and a small head drop. By the time vibration is obvious, the impeller is already pitted. Take NPSH margin seriously.
4. Air entrainment at pump suction. Even 2 % air by volume can drop NPSH-effective by half. Vortex breakers, submerged inlets, and minimum-submergence rules (per ANSI/HI 9.8) exist for this reason.
5. Flow-induced vibration. Strouhal-shed frequency matching a structural mode (heat-exchanger U-bend, riser, chimney, sign post) is a classic failure mode. Check the lock-in range V/(fD) ≈ 4–8.
6. Pulsating flow from positive-displacement pumps. Plunger and gear pumps generate harmonics in the flow; downstream measurement is corrupted unless a pulsation dampener is installed.
7. Pipe roughness grows in service. Design fresh-pipe ε with a service multiplier (×5–10 over 20 years for untreated water).
8. Two-phase pressure drop is not additive. Use a recognized two-phase correlation (Lockhart-Martinelli, Friedel) and a proper multiphase simulator.
9. The “incompressible” gas assumption breaks at M = 0.3. Density changes 4.5 %; pressure-drop predictions become non-conservative.
10. Boundary-layer separation ≠ flow stoppage. Flow reattaches downstream, but with significant pressure-recovery loss and possible local heating.
11. Hydraulic-grade line below pipe centerline = vacuum. Air comes out of solution; pipe may collapse if thin-walled. Profile the pipeline so HGL stays above the pipe.
12. Turbulence-model choice matters. k-ε is fine for free jets and ducted flows but fails in adverse pressure gradients (over-predicts attachment). k-ω SST is the safer default for general engineering CFD; RSM only when swirl or strong secondary motion governs.
13. Mesh y+ mismatch. A first-cell y+ between 5 and 30 is the worst place to be — buffer-layer wall treatment is least accurate there. Either resolve (y+ < 1) or use wall functions (y+ > 30).
14. Compressible “throat” can be a long radius. In real nozzles M = 1 occurs at the geometric minimum only if the flow is one-dimensional and ideal; for real nozzles, the sonic surface curves and CFD is needed for thrust-coefficient prediction.

13p. Tools and software

Pipe-network and process flow.

• AFT Fathom — incompressible liquid networks, pump-curve matching.
• AFT Arrow — compressible gas networks.
• AFT Impulse — water-hammer / surge transients.
• PIPE-FLO — utility-grade pipe sizing.
• EPANET (US EPA, open-source) — municipal water distribution.
• Bentley WaterGEMS / HAMMER — engineering-grade water and surge.
• Pipe Flow Expert — UK SME tool, popular for HVAC and process.

Multiphase / oil-and-gas pipe. OLGA (SLB), LedaFlow (Kongsberg/SINTEF), PIPESIM (SLB).

General-purpose CFD.

• ANSYS Fluent — commercial leader; finite-volume, broad physics.
• ANSYS CFX — historically turbomachinery-strong, now a Fluent sibling.
• Siemens Star-CCM+ — automotive and external-aero strong, polyhedral meshing.
• OpenFOAM (OpenFOAM Foundation / ESI-OpenCFD) — open-source, broad solver library, requires Linux/CLI fluency.
• Autodesk CFD — Designer-friendly, automotive electronics and electronics-cooling.
• COMSOL Multiphysics CFD Module — strong on coupled multi-physics (FSI, electro-magneto-fluid).
• SimScale — cloud-hosted OpenFOAM/CalculiX front-end.
• Numeca FINE/Open (Cadence) — turbomachinery-specialised.
• CONVERGE (Convergent Science) — automatic meshing, internal-combustion-engine flagship.

Mesh. Pointwise, ANSYS Meshing, Cubit / Trelis (Sandia / Coreform), Fidelity Pointwise, snappyHexMesh (OpenFOAM utility), cfMesh (open-source).

Hand-calc and scripting.

• Excel + Colebrook iteration is still the workhorse of process engineering.
• MATLAB for parametric studies and 1-D modeling.
• Python: scipy.optimize.fsolve for Colebrook, CoolProp for fluid properties, NIST REFPROP (paid) for higher accuracy.
• Crane TP-410 (“Flow of Fluids through Valves, Fittings and Pipe”, 2024 edition) — the reference for K-factor / equivalent-length pipe sizing.
• Idelchik, “Handbook of Hydraulic Resistance”, 4th ed — the definitive K-factor compilation for non-standard fittings.

14. Cross-references

• thermodynamics — energy framing, gas properties, enthalpy / entropy used in compressible-flow.
• heat-transfer — convective heat transfer rests on the same velocity field; Re, Pr, Nu correlations.
• vibration-dynamics — flow-induced vibration, Strouhal lock-in, water-hammer transients in pipes.
• electric-motors — pump and fan drives; service factor and starting torque depend on fluid load.
• mechanics-of-materials — pipe stress under internal pressure, water-hammer surge stress, thrust-block reactions.
• pumps-turbomachinery — planned; pump-curve / system-curve matching, NPSH, affinity laws.
• hvac-fundamentals — planned; duct sizing per ASHRAE, psychrometrics, fan laws.
• aerodynamics — planned; lift, drag polars, compressibility corrections.
• propulsion — planned; nozzles, ramjets, turbojets, rocket motors.
• scientific — planned; mesh-description and case-setup file formats.

15. Citations

1. White, F. M. Fluid Mechanics, 9th ed. McGraw-Hill, 2021. ISBN 978-1260575545. The canonical undergraduate text.
2. Munson, B. R.; Okiishi, T. H.; Huebsch, W. W.; Rothmayer, A. P. Fundamentals of Fluid Mechanics, 9th ed. Wiley, 2020. ISBN 978-1119613237.
3. Çengel, Y. A.; Cimbala, J. M. Fluid Mechanics: Fundamentals and Applications, 5th ed. McGraw-Hill, 2023. ISBN 978-1266022388.
4. Crowe, C. T.; Elger, D. F.; Williams, B. C.; Roberson, J. A. Engineering Fluid Mechanics, 12th ed. Wiley, 2020. ISBN 978-1119598466.
5. Fox, R. W.; McDonald, A. T.; Pritchard, P. J.; Mitchell, J. W. Introduction to Fluid Mechanics, 11th ed. Wiley, 2024. ISBN 978-1119723547.
6. Schlichting, H.; Gersten, K. Boundary-Layer Theory, 9th ed. Springer, 2017. ISBN 978-3662529171. The reference treatise on boundary layers.
7. Pope, S. B. Turbulent Flows, Cambridge University Press, 2000. ISBN 978-0521598866. The canonical graduate-level turbulence text.
8. Wilcox, D. C. Turbulence Modeling for CFD, 3rd ed. DCW Industries, 2006. ISBN 978-1928729082. Definitive on RANS closure.
9. Anderson, J. D. Modern Compressible Flow with Historical Perspective, 4th ed. McGraw-Hill, 2020. ISBN 978-1260570823. Compressible-flow standard.
10. Idelchik, I. E. Handbook of Hydraulic Resistance, 4th ed. Begell House, 2007. ISBN 978-1567002515. The K-factor reference.
11. Crane Technical Paper No. 410 (TP-410), “Flow of Fluids through Valves, Fittings and Pipe”, Crane Co., 2024 edition.
12. Moody, L. F. “Friction Factors for Pipe Flow.” Transactions of the ASME, vol. 66, 1944, pp. 671–684. The Moody diagram source.
13. Colebrook, C. F. “Turbulent Flow in Pipes, with Particular Reference to the Transition Region between the Smooth and Rough Pipe Laws.” Journal of the ICE, vol. 11(4), 1939, pp. 133–156.
14. Swamee, P. K.; Jain, A. K. “Explicit Equations for Pipe-Flow Problems.” Journal of the Hydraulics Division (ASCE), vol. 102(HY5), 1976, pp. 657–664.
15. Haaland, S. E. “Simple and Explicit Formulas for the Friction Factor in Turbulent Pipe Flow.” J. Fluids Engineering (ASME), vol. 105(1), 1983, pp. 89–90.
16. Blasius, H. “Grenzschichten in Flüssigkeiten mit kleiner Reibung.” Zeitschrift für Math. und Phys., vol. 56, 1908, pp. 1–37. The flat-plate boundary-layer solution.
17. ISO 4006:1991 — Measurement of fluid flow in closed conduits — Vocabulary and symbols.
18. ASME PTC 19.5-2004 (R2018) — Flow Measurement Performance Test Codes.
19. ANSI/HI 9.6.1-2017 — Rotodynamic Pumps Guideline for NPSH Margin.
20. ANSI/HI 9.8-2018 — Rotodynamic Pumps for Pump Intake Design.
21. AGA Report No. 3 / API MPMS Ch. 14.3 (2013) — Orifice Metering of Natural Gas and Other Related Hydrocarbon Fluids.
22. ASHRAE Handbook — Fundamentals (2021), Chapter 21 “Duct Design.”