Fanning friction factor
The Fanning friction factor, denoted as f_f, is a dimensionless quantity in fluid mechanics that characterizes the frictional resistance to fluid flow in pipes, ducts, and conduits by relating wall shear stress to the fluid's dynamic pressure.[1] It is defined as f_f = \frac{\tau_w}{\rho V^2 / 2}, where \tau_w is the wall shear stress, \rho is the fluid density, and V is the average flow velocity.[1] Equivalently, it connects to pressure drop via f_f = \frac{D_h \Delta P}{2 L \rho V^2}, with D_h as the hydraulic diameter, \Delta P as the pressure loss over length L.[2] Named after American hydraulic engineer John Thomas Fanning (1837–1911), who introduced tabulated values based on earlier experiments by Henry Darcy in his 1877 treatise A Practical Treatise on Hydraulic and Water-Supply Engineering, the factor has become a standard tool for predicting energy dissipation due to viscosity and surface roughness.[3] Unlike the Darcy-Weisbach friction factor f_d, which is four times larger (f_d = 4 f_f), the Fanning version emphasizes local shear effects and is prevalent in chemical and mechanical engineering applications for non-circular conduits using the hydraulic diameter.[3] For laminar flow in smooth pipes, it simplifies to f_f = \frac{16}{\text{Re}}, where Re is the Reynolds number, while turbulent regimes require empirical correlations like the Colebrook equation: \frac{1}{\sqrt{f_f}} = -4 \log_{10} \left( \frac{\epsilon / D_h}{3.7} + \frac{1.255}{\text{Re} \sqrt{f_f}} \right), accounting for relative roughness \epsilon / D_h.[2] This factor underpins calculations for pressure drop, pump sizing, and heat exchanger design, enabling efficient system optimization across industries like petroleum processing and HVAC.[3] Its use in the mechanical energy balance equation, \frac{\Delta P}{\rho} + \frac{\Delta V^2}{2} + g \Delta z = -2 f_f \frac{L}{D_h} V^2, integrates friction with other losses for comprehensive flow analysis.[2]Introduction
Definition and Physical Meaning
The Fanning friction factor, denoted as f, is a dimensionless quantity that characterizes the frictional losses in fluid flow through conduits, particularly in pipes and ducts. It is defined as the ratio of the wall shear stress \tau_w to the dynamic pressure of the flow, expressed mathematically as f = \frac{\tau_w}{\rho u_m^2 / 2}, where \rho is the fluid density and u_m is the mean flow velocity.[4] This definition arises in the context of internal flows where viscous effects at the solid boundary dominate the energy dissipation.[5] Physically, the Fanning friction factor represents the proportion of the flow's kinetic energy that is dissipated as heat due to friction at the wall per unit volume. The wall shear stress \tau_w quantifies the tangential force exerted by the fluid on the conduit surface, while \rho u_m^2 / 2 is the kinetic energy density (dynamic pressure) associated with the bulk motion of the fluid. Thus, f provides a measure of how effectively the flow's momentum is retarded by boundary friction, influencing the overall pressure loss and energy requirements in engineering systems such as pipelines and heat exchangers.[4] The derivation of the Fanning friction factor stems from the momentum balance for steady, fully developed flow, which is an integrated form of the Navier-Stokes equations applied to a control volume within a circular pipe. Considering a cylindrical section of length L and diameter D, the axial pressure force \pi (D/2)^2 \Delta P balances the frictional shear force \tau_w \cdot \pi D L acting on the wall, yielding the relation \frac{\Delta P}{L} = \frac{4 \tau_w}{D}. Substituting the definition of f gives the equivalent form f = \frac{\Delta P \, D}{2 \rho u_m^2 L}, which directly links the observable pressure drop to the flow parameters and friction.[4][6] This formulation assumes incompressible, Newtonian fluid behavior and neglects entrance effects or secondary flows. The applicability of the Fanning friction factor depends on the flow regime, which is determined by the Reynolds number \mathrm{Re}, a dimensionless group indicating the relative importance of inertial to viscous forces in the flow.[7]History and Development
The Fanning friction factor is named after John Thomas Fanning (1837–1911), an American hydraulic engineer who introduced it in his 1877 publication, A Practical Treatise on Water-supply Engineering, initially for applications in water supply and sanitary engineering systems. Fanning compiled empirical tables of friction coefficients from diverse experimental data sources, including those from French, American, English, and German studies, to quantify pressure losses in conduits.[8] The concept's early development traces its roots to the Darcy-Weisbach equation, formulated in 1857 by Henry Darcy and Julius Weisbach, which expressed head loss as a function of velocity, pipe length, and a friction coefficient dependent on pipe diameter and roughness.[8] Fanning adapted this framework in the late 19th century for broader engineering contexts, particularly emphasizing hydraulic radius over diameter, which resulted in his friction factor being one-fourth the value of the Darcy variant; this adaptation gained traction in chemical engineering for handling non-circular ducts and process piping.[8] A key milestone occurred with its formal adoption in Perry's Chemical Engineers' Handbook in its first edition of 1934, where it became a standard tool for fluid flow calculations in industrial processes. The distinction between the Fanning and Darcy friction factors solidified in mid-20th-century engineering texts, as chemical engineering literature consistently favored the Fanning form for momentum transfer analyses, while civil and mechanical engineering increasingly aligned with the Darcy version in hydraulic designs.[8] The evolution of the Fanning friction factor shifted post-1900 from reliance on empirical tables to dimensionless correlations, enabling predictions across scales via parameters like the Reynolds number. This transition was significantly influenced by Johann Nikuradse's 1933 experiments on artificially roughened pipes, which provided foundational data on turbulence and roughness effects, later integrated into universal friction factor charts.Relation to Other Friction Factors
Comparison with Darcy Friction Factor
The Darcy friction factor, denoted as f_D, and the Fanning friction factor, denoted as f_F, are related by the equation f_D = 4 f_F. Both factors quantify the frictional resistance in fluid flow through pipes, describing the same underlying pressure drop but normalized differently: the Fanning factor directly relates to the wall shear stress normalized by dynamic pressure, while the Darcy factor incorporates an additional factor of four to align with head loss conventions in certain applications.[9] The pressure drop \Delta P across a pipe of length L and diameter D due to friction can be expressed using either factor. In the Darcy-Weisbach form, the equation is \Delta P = f_D \frac{L}{D} \frac{\rho u_m^2}{2}, where \rho is the fluid density and u_m is the mean velocity. Equivalently, in the Fanning form, it is \Delta P = 4 f_F \frac{L}{D} \frac{\rho u_m^2}{2}. These variants ensure consistency in calculations despite the differing definitions.[9] Historically, the divergence between the two arose from disciplinary preferences in engineering practice. The Darcy friction factor emerged in civil and mechanical engineering contexts, emphasizing head loss calculations for water conveyance and large-scale systems, building on the work of Henry Darcy and Julius Weisbach in the mid-19th century. In contrast, the Fanning friction factor, introduced by John Thomas Fanning in 1877 through compilations of experimental data, gained prominence in chemical and process engineering, where focus on wall shear stress facilitated analyses of heat transfer and multiphase flows in smaller conduits.[10][11] In graphical representations like the Moody diagram, which plots friction factor against Reynolds number and relative roughness, the values correspond to the Darcy friction factor; thus, Fanning friction factor values read from such charts must be divided by four to obtain the appropriate f_F. This conversion is essential for cross-disciplinary applications to avoid errors in pressure drop predictions.[3]Usage in Engineering Contexts
The Fanning friction factor is particularly favored in chemical and mechanical engineering disciplines due to its direct linkage with heat and mass transfer analogies, such as the Chilton-Colburn analogy, where the Colburn j-factor for heat transfer is expressed as j_H = \frac{f}{2}, with f denoting the Fanning friction factor.[12] This relation facilitates the analogy between momentum transfer (friction) and thermal energy transfer in convective processes, enabling engineers to predict heat exchanger performance and reactor efficiencies without separate empirical correlations.[13] In these fields, the factor's emphasis on local wall shear stress aligns with boundary layer analyses common in process design and equipment sizing.[14] The Fanning friction factor integrates seamlessly with the Reynolds number, defined as \text{Re} = \frac{\rho u_m D}{\mu}, to delineate flow regimes and select appropriate correlations for friction losses.[14] This dimensionless pairing allows practitioners to transition between laminar and turbulent regimes systematically, with f = \frac{16}{\text{Re}} governing laminar flows and more complex functions for turbulent ones, providing a unified framework for fluid dynamic assessments in piping and duct systems.[15] In engineering standards and tools, the Fanning friction factor appears prominently in references like Perry's Chemical Engineers' Handbook, where it is employed for pressure drop estimations in process industries, and in ASME codes such as the Boiler and Pressure Vessel Code, which incorporate it for safe design of fluid-handling components.[16] For visualization, the Moody chart—originally plotted for the Darcy friction factor—can be adapted for Fanning usage by scaling the y-axis values by \frac{1}{4}, since the Fanning factor is one-quarter of the Darcy value, simplifying interpolation for relative roughness and Reynolds number effects.[17] Its advantages shine in handling multiphase flows and non-Newtonian fluids, where the direct proportionality to wall shear stress, \tau_w = f \frac{\rho u_m^2}{2}, simplifies local stress computations without additional geometric factors inherent in head loss formulations.[14] This local focus proves invaluable for modeling slurry transport or polymer processing, reducing complexity in rheological adjustments compared to global pressure drop metrics.[18]Formulas for Laminar Flow
Circular Tubes
In fully developed laminar flow through circular tubes, the Fanning friction factor is given by the formula f = \frac{16}{\mathrm{Re}} where \mathrm{Re} is the Reynolds number, defined as \mathrm{Re} = \frac{\rho V D}{\mu} with \rho as fluid density, V as average velocity, D as tube diameter, and \mu as dynamic viscosity.[19][2] This relation is derived from the Hagen-Poiseuille law, which provides an exact analytical solution for steady, incompressible flow of Newtonian fluids with constant properties.[19] The derivation begins with the simplified Navier-Stokes equations under the assumptions of laminar flow (no azimuthal or radial velocity components), axisymmetric conditions, fully developed flow (negligible entrance effects), a circular cross-section, and no-slip at the wall.[19] Balancing the axial pressure gradient \frac{dp}{dx} with the viscous shear stress gradient yields \frac{dp}{dx} = \mu \frac{1}{r} \frac{d}{dr} \left( r \frac{du}{dr} \right), where u(r) is the axial velocity.[19] Integrating twice with boundary conditions u(R) = 0 (no-slip) and \frac{du}{dr}(0) = 0 (symmetry) results in the parabolic velocity profile u(r) = \frac{\Delta P}{4 \mu L} (R^2 - r^2), where \Delta P is the pressure drop over length L, and R = D/2 is the tube radius.[19] The average velocity is V = u_{\max}/2 = \frac{\Delta P R^2}{8 \mu L}, and the wall shear stress is \tau_w = -\frac{\Delta P R}{2 L}.[19] The Fanning friction factor is defined as f = \frac{\tau_w}{(\rho V^2)/2}. Substituting the expressions for \tau_w and V gives \Delta P / L = \frac{32 \mu V}{D^2}, which, when equated to the friction factor form \Delta P / L = 2 f \frac{\rho V^2}{D}, yields f = \frac{16}{\mathrm{[Re](/page/Re)}}.[19][2] This formula applies only for Reynolds numbers \mathrm{[Re](/page/Re)} < 2000 to $2300, beyond which the flow transitions to turbulent.[19] Additionally, the flow must be fully developed, which requires a sufficient entrance length L_e \approx 0.06 \mathrm{[Re](/page/Re)} \, D to establish the parabolic profile.[20]Non-Circular Ducts
For non-circular ducts in laminar flow, the Fanning friction factor is typically approximated using the hydraulic diameter concept to extend the circular tube baseline, where the exact relation is f = 16 / \mathrm{Re}. The hydraulic diameter D_H is defined as D_H = 4A / P, with A as the cross-sectional area and P as the wetted perimeter; the modified Reynolds number is then \mathrm{Re}_H = \rho u_m D_H / \mu, where u_m is the mean velocity, \rho is the fluid density, and \mu is the dynamic viscosity. The approximation yields f \approx 16 / \mathrm{Re}_H, which provides reasonable estimates for pressure drop calculations in engineering applications.[21] This approach works well for many regular shapes but requires exact solutions or corrections for precise values. For square ducts, the exact Poiseuille solution gives f = 14.227 / \mathrm{Re}_H, derived from series expansions or point-matching methods, representing a deviation of about 11% from the circular approximation. Annular ducts, formed by concentric cylinders with inner-to-outer radius ratio r^*/r_o, exhibit f \mathrm{Re}_H values ranging from 16 (approaching circular as r^*/r_o \to 0) to 24 (as r^*/r_o \to 1, resembling parallel plates), necessitating shape-specific factors from linearization or numerical solutions for accuracy.[21] Representative examples illustrate further adaptations. For infinite parallel plates separated by distance $2h (yielding D_H = 4h), the exact relation is f = 24 / \mathrm{Re}_H, obtained analytically from the velocity profile solution. Equilateral triangular ducts require numerical corrections, with f \mathrm{Re}_H = 13.333 from least-squares fitting, about 17% below the circular value, highlighting the need for shape-dependent adjustments in highly angular geometries.[22][21] The hydraulic diameter approximation is accurate to within 5% for many common shapes like squares and rectangles but becomes less precise for highly irregular geometries, such as those with sharp corners or varying curvatures, where secondary flows or numerical simulations are recommended for refined predictions.[21]Formulas for Turbulent Flow
Hydraulically Smooth Pipes
In hydraulically smooth pipes, the wall roughness is negligible relative to the pipe diameter, such that the relative roughness ε/D approaches zero, rendering the Fanning friction factor dependent solely on the Reynolds number Re for fully developed turbulent flow of Newtonian fluids in circular cross-sections. This regime typically applies for Re > 4000, following the transition from laminar flow which occurs around Re ≈ 2300. The smooth-wall limit assumes no influence from surface protrusions on the turbulent boundary layer, allowing simplified correlations derived from experimental data and boundary-layer theory. A seminal empirical correlation for the Fanning friction factor in this regime was proposed by Blasius in 1913, given by f = 0.079 \, \mathrm{Re}^{-0.25} valid for $4 \times 10^3 < \mathrm{Re} < 10^5. This power-law relation provides an explicit approximation well-suited to moderate Reynolds numbers, capturing the decrease in friction with increasing flow speed due to reduced relative viscous effects in the turbulent core. For broader applicability across higher Reynolds numbers, the Prandtl-Karman law offers a universal implicit relation for smooth walls, derived from the logarithmic velocity profile in the wall layer: \frac{1}{\sqrt{f}} = 2.0 \log_{10} (\mathrm{Re} \sqrt{f}) - 0.8. This equation, accurate up to Re ≈ 10^7 and beyond, requires iterative solution but aligns closely with experimental measurements by integrating the universal log-law behavior near the wall with the defect law in the outer flow. An alternative explicit approximation extending the range to Re < 10^7 was introduced by Koo in 1933: f = 0.0014 + 0.125 \, \mathrm{Re}^{-0.32}. This formula incorporates a constant term to better fit data at higher Re, where the pure power-law deviates slightly, while maintaining simplicity for engineering calculations in fully developed smooth-pipe flow.Rough Pipes and General Correlations
In rough pipes, the Fanning friction factor for turbulent flow depends on both the Reynolds number (Re) and the relative roughness ε/D, where ε is the absolute roughness of the pipe inner surface and D is the pipe diameter. Typical values of absolute roughness include approximately 0.046 mm for commercial steel pipes and 0.0015 mm for drawn tubing, reflecting surface irregularities that influence flow resistance.[23] These values lead to relative roughness ε/D that quantifies the roughness effect relative to pipe size, becoming significant in the transition and fully rough regimes of turbulent flow. The Colebrook-White equation provides the standard implicit correlation for the Fanning friction factor in the transition regime between smooth and rough conditions: \frac{1}{\sqrt{f}} = -4 \log_{10} \left( \frac{\epsilon / D}{3.7} + \frac{1.255}{\mathrm{Re} \sqrt{f}} \right) This equation, developed in 1939, requires iterative solution due to the presence of f on both sides and is widely adopted for its accuracy across a broad range of Re and ε/D in turbulent pipe flow. In the fully rough regime, which occurs at high Re or large ε/D (> 0.01) where viscous effects are negligible compared to roughness, the Fanning friction factor simplifies to an explicit form independent of Re: f \approx \frac{1}{[4 \log_{10} (3.7 \, D / \epsilon)]^2} This asymptotic relation derives from the Colebrook-White equation by neglecting the Re-dependent term, capturing the dominance of form drag from surface protrusions. For practical computations avoiding iteration, the Haaland explicit approximation offers a close match to the Colebrook-White equation, particularly for 10^4 < Re < 10^8: \frac{1}{\sqrt{f}} \approx -3.6 \log_{10} \left[ \left( \frac{6.9}{\mathrm{Re}} \right)^{1.11} + \left( \frac{\epsilon / D}{3.7} \right)^{1.11} \right] Proposed in 1983, this formula provides a simple, direct solution with minimal error for engineering applications involving rough pipes. As ε/D approaches 0, these correlations reduce to the hydraulically smooth pipe limit, where roughness effects vanish.Applications
Pressure Drop Calculations
The frictional pressure drop \Delta P in straight circular pipes due to the Fanning friction factor is given by the formula \Delta P = 4 f \frac{L}{D} \frac{\rho u_m^2}{2}, where f is the Fanning friction factor, L is the pipe length, D is the pipe diameter, \rho is the fluid density, and u_m is the mean flow velocity.[24] This equation arises from a momentum balance accounting for wall shear stress and is widely used in chemical and mechanical engineering for single-phase incompressible flows.[14] For non-circular ducts, the diameter D is replaced by the hydraulic diameter D_H = 4A/P, where A is the cross-sectional area and P is the wetted perimeter, ensuring applicability to various geometries while maintaining the same form.[2] Equivalently, the frictional head loss h_f is expressed as h_f = 2 f \frac{L}{D} \frac{u_m^2}{g}, where g is the acceleration due to gravity; this form directly relates to energy requirements, such as pump power P = \rho g Q h_f / \eta, with Q as the volumetric flow rate and \eta as the pump efficiency.[24] The head loss formulation facilitates integration into the mechanical energy balance for systems involving elevation changes and other losses.[2] To compute the pressure drop, follow these steps:- Calculate the Reynolds number \operatorname{Re} = \rho u_m D / \mu, where \mu is the dynamic viscosity, to determine the flow regime (laminar if \operatorname{Re} < 2300, turbulent otherwise).[25]
- For laminar flow, use f = 16 / \operatorname{Re}; for turbulent flow, obtain f from the Moody diagram, Colebrook equation \frac{1}{\sqrt{f}} = -2 \log_{10} \left( \frac{\epsilon / D}{3.7} + \frac{1.26}{\operatorname{Re} \sqrt{f}} \right) (with \epsilon as absolute roughness), or explicit approximations.[25][2]
- Substitute f, along with fluid properties, dimensions, and velocity, into the pressure drop formula to yield \Delta P.[24]