Open Channel Hydraulics: Manning, Energy & Gradually Varied Flow (2026)

1. Introduction to Open Channel Hydraulics and Free-Surface Mechanics

Civil engineers designing flood control corridors, irrigation aqueducts, culverts, and urban drainage networks rely on open channel hydraulics. Unlike pressurized closed conduit flow where pipe geometry confines fluid boundaries, open channel flow features a free liquid surface exposed to ambient atmospheric pressure.

Because the liquid boundary remains unrestrained, flow depth, cross-sectional flow area, and local velocity adjust dynamically in response to bed slope, boundary roughness, and channel transitions. The driving mechanism in open channel hydraulics is gravity acting along the longitudinal invert slope $S_0 = \sin\theta \approx \tan\theta$, resisted by boundary shear stress $\tau_0$ along the wetted perimeter.

Understanding open channel hydraulics enables practicing hydrologists and hydraulic engineers to predict water surface profiles, prevent bank overtopping, avoid destructive bed scour, and design resilient conveyance infrastructure.

2. Fundamental Classifications of Open Channel Flow

Flow regimes in open channels display diverse behavior depending on temporal variation, spatial acceleration, and turbulence intensity.

2.1 Steady vs Unsteady Flow Fields

A flow field is steady if velocity $\mathbf{v}$ and depth $y$ at any given cross-section remain invariant with respect to time $t$:

$$\frac{\partial \mathbf{v}}{\partial t} = 0, \quad \frac{\partial y}{\partial t} = 0, \quad \frac{\partial Q}{\partial t} = 0$$

When flow parameters fluctuate over time—such as during storm runoff routing, dam-break wave propagation, or tidal fluctuations—the flow regime is unsteady, requiring Saint-Venant shallow water equations for numerical resolution.

2.2 Uniform vs Non-Uniform Flow States

Flow is uniform if water depth, flow area, and mean velocity remain identical at every successive longitudinal station along the channel:

$$\frac{\partial y}{\partial x} = 0, \quad \frac{\partial v}{\partial x} = 0 \implies S_0 = S_w = S_f$$

Under uniform flow, the channel invert slope $S_0$, the free-surface slope $S_w$, and the energy grade line slope $S_f$ run parallel. When boundary geometry, cross-sectional dimensions, or bed slopes change, flow becomes non-uniform (varied flow), subdivided into:
1. Gradually Varied Flow (GVF): Flow depth changes smoothly over long channel reaches ($dy/dx \ll 1$), with hydrostatic pressure distribution remaining valid.
2. Rapidly Varied Flow (RVF): Flow depth changes abruptly over short distances ($dy/dx \sim 1$), generating severe streamline curvature, non-hydrostatic pressure distributions, and intense turbulent eddies (e.g., hydraulic jumps, drop structures, and overflow weirs).

2.3 Laminar, Transitional, and Turbulent Regimes

Open channel turbulence is governed by the Reynolds number $Re$, defined using hydraulic radius $R_h = A / P$:

$$Re = \frac{\rho v R_h}{\mu} = \frac{v R_h}{\nu}$$

• Laminar Flow: $Re \le 500$ (rare in civil engineering practice, observed only in thin sheet flow or high-viscosity slurry conveyance).

• Transitional Flow: $500 < Re < 2000$.

• Turbulent Flow: $Re \ge 2000$ (predominant regime in rivers, canals, culverts, and engineered conduits).

3. Uniform Flow Physics: Boundary Shear, Chezy and Manning Formulations

Uniform flow represents a state of dynamic equilibrium where downstream gravitational driving forces balance boundary frictional resistance forces.

3.1 Force Balance Equilibrium and Hydraulic Radius

Consider a control volume of length $L$ in a prismatic channel. Hydrostatic forces at inlet and outlet balance ($P_1 = P_2$). The gravitational force component along the bed equals:

$$F_g = W \sin\theta = (\gamma A L) S_0$$

The resisting boundary shear force acting over wetted perimeter $P$ is:

$$F_r = \tau_0 P L$$

Equating driving and resisting forces yields the average boundary shear stress $\tau_0$:

$$\tau_0 = \gamma \left( \frac{A}{P} \right) S_0 = \gamma R_h S_0$$

Where the hydraulic radius $R_h$ is:

$$R_h = \frac{\text{Cross-Sectional Flow Area } A}{\text{Wetted Perimeter } P}$$

3.2 Chezy Resistance Equation Derivation

Antoine de Chezy (1775) expressed boundary shear stress as proportional to kinetic energy per unit volume: $\tau_0 = K_f \rho v^2 / 2$. Substituting into the boundary shear equation gives:

$$\frac{1}{2} K_f \rho v^2 = \rho g R_h S_0 \implies v = \sqrt{\frac{2g}{K_f}} \sqrt{R_h S_0} = C \sqrt{R_h S_0}$$

Where $C = \sqrt{2g / K_f}$ is the Chezy resistance coefficient ($\text{ m}^{1/2}/\text{ s}$).

3.3 Manning Semi-Empirical Equation and Roughness Coefficients

Robert Manning (1889) analyzed experimental flume data, establishing that Chezy’s $C$ scales with hydraulic radius to the $1/6$ power: $C = \frac{1}{n} R_h^{1/6}$. Substituting this yields the universal Manning equation:

$$v = \frac{k_n}{n} R_h^{2/3} S_0^{1/2} \quad \text{ and} \quad Q = \frac{k_n}{n} A R_h^{2/3} S_0^{1/2}$$

Where:

• $n$ is Manning’s roughness coefficient ($\text{ s}/\text{ m}^{1/3}$).

• $k_n = 1.0$ for SI metric units ($v$ in $\text{m/s}$, $A$ in $\text{ m}^2$, $R_h$ in $\text{ m}$).

• $k_n = 1.486$ for US Customary units ($v$ in $\text{ft/s}$, $A$ in $\text{ ft}^2$, $R_h$ in $\text{ ft}$).

3.4 Hydraulically Optimal Cross-Section Geometry

A cross-section is hydraulically optimal (most efficient) when it delivers maximum discharge $Q$ for a given cross-sectional area $A$, which corresponds to minimizing wetted perimeter $P$.

Cross-Section Type  | Optimal Geometry Criteria                          | Hydraulic Radius $R_h$
--------------------+----------------------------------------------------+-------------------------
Rectangular         | Aspect ratio: $b = 2y$                             | $R_h = y / 2$
Trapezoidal         | Side slope $z = 1/\sqrt{3}$ ($\theta = 60^\circ$), $b = \frac{2}{\sqrt{3}}y$ | $R_h = y / 2$
Triangular          | Apex angle: $2\theta = 90^\circ$ (Side slope $z = 1$) | $R_h = \frac{y}{2\sqrt{2}}$
Circular (Pipe)     | Flow depth for max discharge: $y = 0.938 D_0$      | $R_h = 0.29 D_0$

4. Energy Conservation and the Specific Energy Concept

Boris Bakhmeteff introduced specific energy analysis to evaluate transitions, contractions, steps, and choking in open conduits.

4.1 Total Head and Specific Energy Formulation

Total energy head $H$ referenced to an arbitrary vertical datum $Z = 0$ is:

$$H = Z + y \cos\theta + \alpha \frac{v^2}{2g}$$

For channels with small bed slope ($\cos\theta \approx 1$) and uniform velocity distribution ($\alpha \approx 1.0$), specific energy $E$ measures energy head referenced directly to the channel invert ($Z = 0$):

$$E = y + \frac{v^2}{2g} = y + \frac{Q^2}{2g A^2}$$

For a rectangular channel of breadth $b$ with unit discharge $q = Q/b$:

$$E = y + \frac{q^2}{2g y^2}$$

4.2 Specific Energy Curve and Alternate Depths

Plotting depth $y$ against specific energy $E$ for a constant discharge $Q$ reveals two distinct branches:
1. As $y \to \infty$, $E \to y$ (the asymptote is the $E = y$ line of zero kinetic energy).
2. As $y \to 0$, $E \to \infty$ (kinetic energy head dominates).

For every energy level $E > E_{min}$, there exist two physical depths called alternate depths ($y_1$ and $y_2$) that convey the identical flow rate $Q$. The upper depth corresponds to subcritical flow, while the lower depth represents supercritical flow.

4.3 Critical Flow Criteria and Froude Number Mechanics

Differentiating specific energy with respect to depth at constant discharge:

$$\frac{dE}{dy} = 1 – \frac{Q^2}{g A^3} \frac{dA}{dy}$$

Since top water surface width $T = \frac{dA}{dy}$, setting $\frac{dE}{dy} = 0$ yields the universal critical flow condition:

$$\frac{Q^2 T}{g A_c^3} = 1 \implies \frac{v_c^2}{g (A_c / T)} = \frac{v_c^2}{g D_h} = 1$$

Where hydraulic depth is $D_h = A / T$. The non-dimensional Froude number $Fr$ is:

$$Fr = \frac{v}{\sqrt{g D_h}} = \frac{Q / A}{\sqrt{g (A / T)}}$$

• $Fr < 1$: Subcritical Flow (tranquil flow; gravity waves travel faster than flow velocity, allowing downstream disturbances to propagate upstream).

• $Fr = 1$: Critical Flow (minimum energy state for given discharge; maximum discharge for given specific energy).

• $Fr > 1$: Supercritical Flow (rapid shooting flow; disturbances cannot propagate upstream).

For rectangular cross-sections ($A = b y$, $T = b$):

$$y_c = \sqrt[3]{\frac{q^2}{g}}, \quad E_{min} = \frac{3}{2} y_c, \quad v_c = \sqrt{g y_c}$$

5. Momentum Conservation and Specific Force

While energy conservation fails across zones of severe turbulent dissipation, the momentum equation tracks hydrodynamic forces accurately. The specific force $F_s$ (or momentum function $M$) is defined as:

$$F_s = \frac{Q^2}{g A} + \bar{y} A$$

Where $\bar{y}$ represents the depth of the cross-sectional centroid below the free surface. Across a hydraulic jump, neglecting external boundary shear over a short distance:

$$F_{s1} = F_{s2} \implies \frac{Q^2}{g A_1} + \bar{y}_1 A_1 = \frac{Q^2}{g A_2} + \bar{y}_2 A_2$$

Depths $y_1$ and $y_2$ sharing the same specific force are termed sequent (conjugate) depths.

6. Gradually Varied Flow (GVF) Profile Analysis

Gradually varied flow describes longitudinal depth variations in non-uniform channels where boundary streamlines remain nearly parallel.

6.1 Governing Differential Equation of GVF

The governing first-order ordinary differential equation relates water surface slope $dy/dx$ to bed slope $S_0$, friction slope $S_f$, and Froude number $Fr$:

$$\frac{dy}{dx} = \frac{S_0 – S_f}{1 – Fr^2}$$

From Manning’s equation, the friction slope at any intermediate depth $y$ is:

$$S_f = \frac{n^2 v^2}{R_h^{4/3}} = \frac{n^2 Q^2}{A^2 R_h^{4/3}}$$

6.2 Classification of the Twelve Water Surface Profiles

Water surface profiles depend on the relationship between bed slope $S_0$, normal depth $y_n$ ($S_f = S_0$), and critical depth $y_c$ ($Fr = 1$):

Mild Slope (M1, M2, M3 Curves):
     ────────────────────────────────── Free Surface (M1 Curve: dy/dx > 0)
     - - - - - - - - - - - - - - - - -  Normal Depth Line (y = y_n)
     .................................. Critical Depth Line (y = y_c)
     ────────────────────────────────── Channel Invert Bed (Slope S_0 < S_c)

6.3 Numerical Integration via Direct Step Method

For prismatic channels, the Direct Step Method solves for reach length $\Delta x$ between two specified flow depths $y_1$ and $y_2$:

$$\Delta x = x_2 – x_1 = \frac{E_2 – E_1}{S_0 – \bar{S}_f}$$

Where $\bar{S}_f = \frac{S_{f1} + S_{f2}}{2}$ is the arithmetic mean friction slope across the reach.

7. Comprehensive Worked Engineering Calculation: GVF Backwater Profile

7.1 Channel Geometry and Design Conditions

A trapezoidal flood conveyance canal requires a backwater profile computation caused by a downstream diversion weir. The engineering parameters are:

• Design Discharge: $Q = 35.0\text{ m}^3/\text{ s}$

• Bed Width: $b = 8.0\text{ m}$

• Side Slopes: $z = 1.5$ (Horizontal : Vertical, $1.5\text{ H}:1\text{ V}$)

• Longitudinal Invert Bed Slope: $S_0 = 0.00060$ ($0.60\text{ m/km}$)

• Manning Roughness: $n = 0.020$ (Finished cast-in-place concrete lining)

• Downstream Control Depth at Weir: $y_1 = 3.60\text{ m}$

7.2 Determination of Normal and Critical Depths

Step 1: Compute Critical Depth $y_c$ by solving $\frac{Q^2 T}{g A^3} = 1$:

$$A(y) = (b + z y) y = (8.0 + 1.5 y) y$$

$$T(y) = b + 2 z y = 8.0 + 3.0 y$$

$$f(y) = \frac{(35.0)^2 (8.0 + 3.0 y)}{9.81 [(8.0 + 1.5 y)y]^3} – 1 = 0$$

Solving iteratively via Newton-Raphson yields:

$$y_c = 1.182\text{ m}$$

Step 2: Compute Normal Depth $y_n$ by solving Manning’s uniform flow equation:

$$Q = \frac{1.0}{n} A R_h^{2/3} S_0^{1/2} = \frac{1.0}{0.020} \frac{[(8.0 + 1.5 y)y]^{5/3}}{[8.0 + 2 y \sqrt{1 + (1.5)^2}]^{2/3}} (0.0006)^{1/2} = 35.0$$

Solving iteratively:

$$y_n = 1.945\text{ m}$$

Since $y_n (1.945\text{ m}) > y_c (1.182\text{ m})$, the channel has a Mild (M) slope. Because initial backwater depth $y = 3.60\text{ m} > y_n$, the resulting water surface profile is an $M_1$ backwater curve ($dy/dx > 0$).

7.3 Step-by-Step Direct Step Computation of M1 Profile Length

We compute the upstream reach distance over two depth steps: from $y_1 = 3.60\text{ m}$ down to $y_2 = 3.10\text{ m}$, and from $y_2 = 3.10\text{ m}$ down to $y_3 = 2.60\text{ m}$.

Step 3: Interval 1 Analysis ($y = 3.60\text{ m} \to 3.10\text{ m}$):

At $y_1 = 3.60\text{ m}$:

• $A_1 = (8.0 + 1.5 \times 3.60) \times 3.60 = 48.240\text{ m}^2$

• $P_1 = 8.0 + 2 \times 3.60 \times \sqrt{1 + 1.5^2} = 20.981\text{ m}$

• $R_{h1} = 48.240 / 20.981 = 2.298\text{ m}$

• $v_1 = 35.0 / 48.240 = 0.7255\text{ m/s}$

• $E_1 = 3.60 + \frac{(0.7255)^2}{2 \times 9.81} = 3.60 + 0.0268 = 3.6268\text{ m}$

• $S_{f1} = \frac{(0.020)^2 (0.7255)^2}{(2.298)^{4/3}} = \frac{0.0004 \times 0.5264}{3.003} = 0.0000701$

At $y_2 = 3.10\text{ m}$:

• $A_2 = (8.0 + 1.5 \times 3.10) \times 3.10 = 39.215\text{ m}^2$

• $P_2 = 8.0 + 2 \times 3.10 \times 1.8028 = 19.177\text{ m}$

• $R_{h2} = 39.215 / 19.177 = 2.045\text{ m}$

• $v_2 = 35.0 / 39.215 = 0.8925\text{ m/s}$

• $E_2 = 3.10 + \frac{(0.8925)^2}{2 \times 9.81} = 3.10 + 0.0406 = 3.1406\text{ m}$

• $S_{f2} = \frac{(0.020)^2 (0.8925)^2}{(2.045)^{4/3}} = \frac{0.0004 \times 0.7966}{2.607} = 0.0001222$

Step 4: Reach 1 Length Calculation ($\Delta x_1$):

$$\bar{S}_{f,1-2} = \frac{0.0000701 + 0.0001222}{2} = 0.00009615$$

$$\Delta x_1 = \frac{E_2 – E_1}{S_0 – \bar{S}_{f,1-2}} = \frac{3.1406 – 3.6268}{0.00060 – 0.00009615} = \frac{-0.4862}{0.00050385} = -964.97\text{ m}$$

The negative sign indicates integration proceeds in the upstream direction. The water surface drops from $3.60\text{ m}$ to $3.10\text{ m}$ over a distance of $965.0\text{ m}$ upstream of the weir.

8. Practical Engineering Applications in River and Canal Design

Practical application of free-surface hydraulics spans multiple critical infrastructure domains:

1. Flood Inundation Delineation: Hydrodynamic models (such as HEC-RAS and Telemac) solve full Saint-Venant equations across complex natural river bathymetry to generate floodplain hazard maps.
2. Irrigation Canal Optimization: Channels designed to maintain non-silting ($v > 0.6\text{ m/s}$) and non-scouring ($v < 1.8\text{ m/s}$) flow velocities preserve unlined canal integrity while minimizing sediment maintenance.
3. Culvert and Bridge Pier Backwater: Sizing bridge openings to restrict backwater afflux $\Delta h$ prevents upstream roadway overtopping during 100-year flood events.

9. Synthesis and Hydrodynamic Equilibrium Wrap-Up

Mastering open channel hydraulics equips water resource engineers with the analytical tools necessary to balance gravity driving potentials against boundary turbulence resistance across all conveyance networks. Rigorous evaluation of specific energy, critical thresholds, and backwater curves ensures hydraulic structures deliver long-term operational resilience.

References & Standards Cited

1. Chow, V.T. (1959): Open-Channel Hydraulics. McGraw-Hill Book Company, New York.
2. USACE (1994): Hydraulic Design of Flood Control Channels. Engineering Manual EM 1110-2-1601, U.S. Army Corps of Engineers, Washington, D.C.
3. French, R.H. (1985): Open-Channel Hydraulics. McGraw-Hill, New York.
4. Henderson, F.M. (1966): Open Channel Flow. Macmillan Publishing Co., New York.
5. ASCE Task Committee (1963): Friction Factors in Open Channels. Journal of the Hydraulics Division, ASCE, Vol. 89, No. HY2.
6. Brunner, G.W. (2016): HEC-RAS River Analysis System: Hydraulic Reference Manual. US Army Corps of Engineers Hydrologic Engineering Center, Davis, CA.

Frequently Asked Questions (FAQ)

What is the fundamental difference between pipe flow and open channel flow?

Pipe flow is driven by longitudinal piezometric pressure gradients inside a closed conduit where the fluid fills the entire cross-section without an air boundary. Open channel flow has a free water surface exposed to ambient atmospheric pressure, where flow is driven by the component of gravity acting parallel to the bed slope.

Why is critical flow inherently unstable in physical flumes and channels?

At critical flow ($Fr = 1.0$), specific energy reaches its mathematical minimum for a given discharge ($dE/dy = 0$). Consequently, small variations in specific energy cause large, undulating oscillations in water surface depth, leading to standing waves, instability, and surface air entrainment.

How does channel bed slope classification affect backwater profiles?

When normal depth exceeds critical depth ($y_n > y_c$), the bed slope is classified as Mild (M), and downstream backwater disturbances propagate upstream as $M_1$ or $M_2$ profiles. When $y_n < y_c$, the slope is Steep (S), and upstream control conditions govern supercritical flow profiles ($S_2, S_3$).

What causes the transition between subcritical and supercritical flow?

Transitions from subcritical to supercritical flow occur smoothly at channel steepening transitions, drop structures, or under sluice gates without energy dissipation. Conversely, transitions from supercritical to subcritical flow cannot occur smoothly; they force a turbulent, energy-dissipating hydraulic jump.

How is Manning's roughness coefficient $n$ adjusted for composite or vegetated channels?

For composite channels with varying roughness along the wetted perimeter (e.g., concrete bed with riprap banks), equivalent roughness $n_e$ is computed using the Horton-Einstein or Cowan method:
$$n_e = left( frac{sum P_i n_i^{1.5}}{P} right)^{2/3}$$