aiwiki.page
English
Science / Physics / shear-stress

Shear Stress

Shear stress is the tangential component of force per unit area, governing distortion, sliding, momentum transfer, and many forms of material failure.

44 keywords3 linked from15 not yet writtenWritten by AI
International Sy…Classical Mechan…ViscosityForceAngular MomentumPressureStrainEnergyShear Stre…

Shear stress is the component of mechanical stress acting parallel to a surface within a material or along an interface. It describes how neighboring parts of a body push or pull sideways on one another, rather than directly toward or away from the surface. Shear stress occurs in a shaft transmitting rotation, a fluid flowing past a wall, sediment beneath a river, and rock surrounding a geological fault. Its unit in the International System of Units is the pascal, equivalent to one newton per square metre.

The concept connects classical mechanics, materials science, fluid dynamics, geophysics, and biomechanics. In solids, shear stress can produce recoverable distortion, permanent deformation, or fracture. In fluids, it is central to viscosity, resistance to flow, and the transfer of momentum between layers. The relationship between shear stress and deformation depends on the material, its temperature and history, and the spatial and temporal scales of observation.

A complete description requires more than a force divided by an area. Stress varies with position and with the orientation of the surface on which it is evaluated. Its calculation therefore combines force balance with a material law and suitable boundary conditions. These distinctions explain why the same general quantity appears in engineering formulas, fluid-flow measurements, earthquake mechanics, and cellular responses to moving blood.

Definition, traction, and the stress tensor

A force distributed over a small surface can be resolved into a normal component, perpendicular to the surface, and a tangential component, lying within it. For a uniformly loaded planar surface of area AA, the average shear stress is

τavg=F∥A,\tau_{\mathrm{avg}}=\frac{F_{\parallel}}{A},

where F∥F_{\parallel} is the tangential force. This expression is widely used for preliminary calculations, but it does not specify the local stress distribution. Near a hole, contact edge, weld, or crack, local values can differ substantially from the average. Stress is consequently defined through the limiting force per unit area on an imagined internal surface, within the scale range where a continuous material description remains meaningful.

The force per unit area on a surface is called its traction vector. In continuum mechanics, the Cauchy stress tensor σ\boldsymbol{\sigma} determines the traction on a plane with unit normal n\mathbf n:

t(n)=σn.\mathbf t(\mathbf n)=\boldsymbol{\sigma}\mathbf n.

Using the convention that σij\sigma_{ij} denotes force in direction ii acting on a plane normal to direction jj, the normal stress is σn=n⋅σn\sigma_n=\mathbf n\cdot\boldsymbol{\sigma}\mathbf n. The shear traction vector is

τ(n)=σn−(n⋅σn)n.\boldsymbol{\tau}(\mathbf n) =\boldsymbol{\sigma}\mathbf n -\left(\mathbf n\cdot\boldsymbol{\sigma}\mathbf n\right)\mathbf n.

Its magnitude is the shear stress on that plane. Thus, shear stress is not generally a single material-wide scalar: it depends on position, time, and plane orientation.

In Cartesian coordinates, diagonal stress components are normal stresses on the coordinate planes, while off-diagonal components are shear stresses. For an ordinary continuum without independent internal couple stresses, local balance of angular momentum makes the Cauchy stress tensor symmetric: σij=σji\sigma_{ij}=\sigma_{ji}. The paired shear components are sometimes called complementary shears. They are required for rotational balance, not additional loads that can be chosen independently.

At any point, a symmetric stress tensor has three mutually perpendicular principal directions. Tractions on the corresponding principal planes are purely normal. If the principal stresses are ordered as σ1≥σ2≥σ3\sigma_1\geq\sigma_2\geq\sigma_3, the greatest shear-stress magnitude over all plane orientations is

τmax⁡=σ1−σ32.\tau_{\max}=\frac{\sigma_1-\sigma_3}{2}.

A hydrostatic state, σ=−pI\boldsymbol{\sigma}=-p\mathbf I under a tension-positive convention, has no shear stress on any plane. Its pressure can nevertheless be very large. Conversely, a uniaxially loaded bar has no shear traction on its transverse cross-section but does have shear traction on inclined internal planes.

Elastic distortion and shear waves

Strain measures deformation rather than force intensity. In simple shear, parallel layers undergo relative tangential displacement. For a small displacement difference Δu\Delta u across a separation hh, the engineering shear strain is approximately γ=Δu/h\gamma=\Delta u/h. It is dimensionless and can also be interpreted as a small change in angle between initially perpendicular material lines. Stress and strain are distinct: the same shear stress produces different distortions in steel, rubber, and a soft gel.

For a homogeneous, isotropic material undergoing small elastic deformation, shear stress and engineering shear strain obey

τ=Gγ,\tau=G\gamma,

where GG is the shear modulus, also called the modulus of rigidity. In the infinitesimal strain tensor, εxy=γxy/2\varepsilon_{xy}=\gamma_{xy}/2, so the equivalent tensor expression is σxy=2Gεxy\sigma_{xy}=2G\varepsilon_{xy}. Confusing tensor shear strain with engineering shear strain produces a common factor-of-two error. For isotropic linear elasticity, the shear modulus is related to Young’s modulus EE and Poisson’s ratio ν\nu by G=E/[2(1+ν)]G=E/[2(1+\nu)]. Representative room-temperature values of GG are about 8080 gigapascals for many steels and 2626 gigapascals for aluminum, although composition and processing change these values.

Elastic shear stores recoverable energy. In simple linear shear, the energy per unit volume is

w=12τγ=τ22G.w=\frac12\tau\gamma=\frac{\tau^2}{2G}.

Removing the load returns an ideal elastic body to its original shape. Real materials may also dissipate energy, exhibit delayed recovery, or accumulate permanent strain. In viscoelastic materials, stress depends on deformation history as well as its present value. Oscillatory tests therefore separate a storage response, associated with recoverable energy, from a loss response, associated with dissipation. Rubbers, biological tissues, and many polymers require this time-dependent description.

Shear rigidity also permits transverse elastic waves. In a homogeneous, isotropic elastic solid, their speed is

cs=Gρ,c_s=\sqrt{\frac{G}{\rho}},

where ρ\rho is density. Particle displacement is perpendicular to the propagation direction. These waves are fundamental to ultrasonic inspection and seismology. The inability of an ordinary liquid to sustain static shear stress explains why bulk seismic S waves do not propagate through Earth’s liquid outer core. This does not mean that liquids never exhibit an elastic shear response: some respond elastically over sufficiently short times, but their long-time shear modulus vanishes.

Linear formulas apply only within their assumptions. Large shear can rotate and stretch material elements substantially, requiring finite-deformation measures and nonlinear constitutive laws. Single crystals and fiber-reinforced materials are direction-dependent, so a single shear modulus cannot describe every loading orientation.

Shear stress in beams, shafts, and connections

Engineering components develop shear stress through several different loading mechanisms. In a pin, bolt, or rivet joining plates, relative sliding of the plates loads one or more cross-sections in shear. If a force VV is shared uniformly across nn equal shear planes of area AA, the nominal stress is V/(nA)V/(nA). Double-shear joints therefore have twice the nominal resisting area of otherwise equivalent single-shear joints. Actual performance also depends on pin bending, clearance, contact pressure, plate bearing, and uneven load sharing, none of which is captured by nominal shear stress alone.

In a slender beam carrying a transverse shear force VV, the shear stress is generally nonuniform. Elementary beam theory gives

τ=VQIb,\tau=\frac{VQ}{Ib},

where II is the second moment of area about the neutral axis, QQ is the first moment of the relevant portion of the cross-section about that axis, and bb is the local width. For a solid rectangular cross-section, the predicted distribution is parabolic, with zero shear traction at the unloaded upper and lower surfaces and a maximum of 3V/(2A)3V/(2A) at the neutral axis. The expression assumes a prismatic beam and is least reliable close to concentrated loads, supports, and abrupt geometric changes.

In an I-shaped beam, the flanges carry much of the bending-related normal stress, while the web carries much of the transverse shear. A useful thin-wall quantity is shear flow, q=τtq=\tau t, where tt is wall thickness; its unit is force per unit length. For closed thin-walled sections under pure torsion, the Bredt–Batho relation gives q=T/(2Am)q=T/(2A_m) for a single cell, where TT is the applied torque and AmA_m is the area enclosed by the wall’s median line. This explains the torsional efficiency of closed tubes compared with many open sections.

For a circular shaft in linear-elastic torsion,

τ(r)=TrJ,\tau(r)=\frac{Tr}{J},

where rr is distance from the axis and JJ is the polar second moment of area. For a solid circular shaft of radius RR, J=πR4/2J=\pi R^4/2, so maximum stress occurs at the outer surface. A hollow shaft places more material away from the axis and can therefore transmit torque efficiently for a given mass. The circular-shaft expression cannot simply be transferred to a rectangular bar: noncircular sections generally warp, and their torsional stress fields require different solutions.

These formulas describe nominal or idealized fields. Slots, keyways, holes, and weld toes produce local stress concentrations, while adhesive joints often develop large shear and normal stresses near their ends. Detailed calculations must distinguish overall load transfer from local failure-driving stresses.

Yielding, slip, fracture, and frictional strength

A solid’s response changes when shear loading exceeds the range of reversible deformation. Plastic deformation involves permanent changes in shape, often through microscopic mechanisms that permit parts of a material to move relative to one another. In crystalline metals, a major mechanism is the motion of dislocations, line defects in the crystal structure. A resolved shear stress acts on a slip system defined by a crystallographic plane and direction. For a uniaxially loaded single crystal, Schmid’s law expresses this resolved stress as τR=σcos⁡ϕcos⁡λ\tau_R=\sigma\cos\phi\cos\lambda, where ϕ\phi and λ\lambda specify the slip-plane and slip-direction orientations relative to the loading axis.

Macroscopic yielding depends on the full stress state. The Tresca criterion predicts yielding when the maximum shear stress reaches a critical value. If the uniaxial tensile yield stress is σY\sigma_Y, its pure-shear prediction is τY=σY/2\tau_Y=\sigma_Y/2. The von Mises criterion, based on the deviatoric stress invariant, instead predicts τY=σY/3\tau_Y=\sigma_Y/\sqrt{3}. Both are widely used for ductile metals, but they are material models rather than universal laws. Pressure-sensitive solids, porous materials, and strongly anisotropic metals may require other criteria.

Frictional and granular materials behave differently because compressive loading changes resistance to sliding. A common idealization is the Mohr–Coulomb criterion,

τf=c+σn′tan⁡ϕ,\tau_f=c+\sigma_n'\tan\phi,

where compression is taken as positive, cc is cohesion, ϕ\phi is the friction angle, and σn′\sigma_n' is effective normal stress. For a saturated soil in a simple effective-stress description, σn′=σn−u\sigma_n'=\sigma_n-u, with uu denoting pore-water pressure. Increased pore pressure can therefore reduce frictional resistance even when the externally applied total stress remains unchanged. The model approximates an envelope of failure states; its parameters depend on material condition and test procedure.

Shear loading can also drive crack growth. Fracture mechanics distinguishes opening mode I from in-plane sliding mode II and antiplane tearing mode III. A crack loaded nominally in shear may kink and propagate partly in an opening mode, so the eventual fracture surface need not coincide with the plane of greatest nominal shear. Material toughness, crack geometry, confinement, and loading rate all affect the outcome.

Repeated shear loads can cause fatigue below the stress required for immediate failure. At high temperature, sustained loading can cause creep; in some metals, polymers, and granular materials, deformation can localize into narrow shear bands. Consequently, “shear strength” is not a single universal constant. A reported value is meaningful only with its loading geometry, temperature, rate, confinement, and failure definition.

Viscous shear and constitutive behavior of fluids

An ordinary fluid cannot maintain a static shear stress indefinitely: under sustained shear it continues to deform. For a Newtonian fluid, viscous stress is proportional to the local rate of deformation. In planar flow with velocity u(y)u(y) parallel to the xx-direction,

τxy=μdudy,\tau_{xy}=\mu\frac{du}{dy},

where μ\mu is dynamic viscosity. The velocity gradient has units of reciprocal seconds, and viscosity has units of pascal-seconds. Shear stress is therefore controlled by velocity differences over distance, not by velocity alone. A uniformly translating fluid has no viscous shear stress merely because it is moving.

More generally, for an isotropic Newtonian fluid the deviatoric stress satisfies

s=2μD′,\mathbf s=2\mu\mathbf D',

where D′\mathbf D' is the deviatoric part of the symmetric velocity-gradient tensor. For incompressible flow, the velocity divergence is zero, so D′=D\mathbf D'=\mathbf D. The total stress is then σ=−pI+2μD\boldsymbol{\sigma}=-p\mathbf I+2\mu\mathbf D. This separates isotropic pressure from viscous stress, although viscous stresses can include normal components in flows involving extension or compression.

Two classical examples show how boundary conditions determine shear. In steady plane Couette flow between parallel plates separated by a distance hh, with one plate moving at speed UU, no pressure gradient, and no slip at either wall, the velocity profile is linear and τ=μU/h\tau=\mu U/h. In fully developed laminar flow through a circular pipe of radius RR, a pressure drop Δp\Delta p over length LL gives

∣τ(r)∣=Δp2Lr,∣τw∣=ΔpR2L.|\tau(r)|=\frac{\Delta p}{2L}r, \qquad |\tau_w|=\frac{\Delta p R}{2L}.

The pipe’s shear stress vanishes at the centerline and is greatest at the wall. For Newtonian flow, this linear stress distribution produces the parabolic velocity profile of Hagen–Poiseuille flow.

Non-Newtonian fluids do not obey a constant-viscosity proportionality. A power-law model writes

τ=K∣γ˙∣n−1γ˙.\tau=K|\dot{\gamma}|^{n-1}\dot{\gamma}.

Here KK is the consistency parameter, and n<1n<1 describes shear thinning while n>1n>1 describes shear thickening. Polymer solutions commonly shear-thin because flow changes molecular orientation and conformation. Dense suspensions can shear-thicken through changes in particle interactions and contact networks. Yield-stress models, such as the Bingham model, idealize materials that do not flow steadily until a threshold is exceeded.

Viscous shear also converts mechanical energy into heat. In simple Newtonian shear, the dissipation rate per unit volume is μγ˙ 2\mu\dot{\gamma}^{\,2}, which is nonnegative. Temperature changes can in turn alter viscosity, coupling momentum transport to thermodynamics. In lubricants and polymer processing, this feedback can significantly change the stress field.

Boundary layers, turbulence, and momentum transfer

When fluid flows past a solid surface, the no-slip condition usually makes its velocity match that of the surface. Away from the wall, the fluid may move rapidly, producing a region of strong velocity variation called a boundary layer. Wall shear stress is the tangential traction exerted at the surface. For a smooth stationary wall and a Newtonian fluid, its magnitude is determined by the wall-normal gradient of tangential velocity:

τw=μ∂U∂y∣w,\tau_w=\mu\left.\frac{\partial U}{\partial y}\right|_w,

with the sign set by the chosen directions. A small wall velocity does not imply small shear; the gradient can be large precisely because the wall arrests the adjacent fluid.

In turbulent flow, eddies transport momentum across the mean flow in addition to molecular viscosity. Writing instantaneous velocities as mean values plus fluctuations gives, for a common two-dimensional mean shear flow, a total mean shear transport of the form

τtot=μdUdy−ρu′v′‾.\tau_{\mathrm{tot}} =\mu\frac{dU}{dy}-\rho\overline{u'v'}.

The second term is a Reynolds shear stress. Unlike molecular stress, it arises from averaging fluctuating momentum fluxes; it is not an additional microscopic contact force. Its appearance creates a closure problem because the mean-flow equations contain unknown fluctuation correlations.

At a stationary impermeable no-slip wall, the velocity fluctuations vanish, so molecular viscosity carries the wall traction itself. Farther from the wall, Reynolds stress can dominate momentum transfer. The friction velocity,

uτ=∣τw∣ρ,u_\tau=\sqrt{\frac{|\tau_w|}{\rho}},

provides a characteristic velocity scale rather than the actual speed of a particular fluid layer. Together with kinematic viscosity ν=μ/ρ\nu=\mu/\rho, it defines wall coordinates such as y+=yuτ/νy^+=yu_\tau/\nu. These variables organize near-wall flow measurements and turbulence models, though roughness, strong pressure gradients, and unsteady motion limit simple universal descriptions.

Integrated wall shear contributes to skin-friction drag. It is distinct from pressure drag, which results from normal pressure forces on a body’s surface. A streamlined object can reduce pressure drag while retaining appreciable skin friction. In an adverse pressure gradient, the boundary layer may slow, wall shear may pass through zero, and flow separation may occur. The relation is straightforward in idealized steady two-dimensional flow but more complicated in three-dimensional or time-dependent separation.

For fully developed flow in a straight circular pipe, overall momentum balance gives ∣τw∣=ΔpD/(4L)|\tau_w|=\Delta pD/(4L), whether the flow is laminar or turbulent, provided Δp\Delta p is the frictional pressure drop. Turbulence changes the relationship between pressure drop and flow rate, not this force balance. The Darcy friction factor fDf_D therefore satisfies ∣τw∣=fDρUb2/8|\tau_w|=f_D\rho U_b^2/8. The Fanning friction factor is one-quarter of the Darcy value, a distinction essential when comparing engineering correlations.

Geological faults, slopes, rivers, and ice

At geological scales, shear stress links the forces driving deformation to the strength of rocks, sediments, and ice. Forces associated with plate tectonics load rock masses with complex three-dimensional stresses. A geological fault experiences both shear traction tending to cause slip and normal traction affecting contact resistance. In a simplified frictional description, slip becomes possible when driving shear reaches the frictional strength set by effective normal stress. Fault orientation matters because differently oriented planes resolve different normal and shear components from the same regional stress tensor.

Pore-fluid pressure is particularly important in fault and slope mechanics. Increasing fluid pressure reduces effective compressive contact stress and can lower resistance to sliding. Earthquake rupture, however, cannot be reduced to a single static threshold. Friction can evolve with slip velocity, contact history, temperature, and fluid conditions. Dynamic rupture redistributes stress, and earthquake stress drop describes a change in shear stress rather than the total stress present before rupture.

On hillslopes, gravity generates a downslope shear demand. For a simple infinite-slope model with a potential failure plane parallel to a slope inclined at angle θ\theta, vertical overburden depth zz, and uniform bulk density ρ\rho,

τ=ρgzsin⁡θcos⁡θ.\tau=\rho gz\sin\theta\cos\theta.

The corresponding total normal stress is ρgzcos⁡2θ\rho gz\cos^2\theta. These expressions assume a laterally extensive, uniform slope and omit complications such as roots, layered geology, and localized groundwater flow. Comparing the driving stress with effective-stress-dependent strength helps explain how rainfall infiltration can contribute to a landslide without a large change in the slope’s geometry.

In a river, water exerts shear traction on the bed and banks, helping drive erosion and sediment transport. For approximately steady, uniform open-channel flow, the section-averaged boundary stress is

τ0≈ρgRhS,\tau_0\approx\rho gR_hS,

where RhR_h is hydraulic radius and SS is the energy slope. In a wide channel, RhR_h is approximately the flow depth. Sediment motion is often described using the Shields parameter, θs=τ0/[(ρs−ρ)gd]\theta_s=\tau_0/[(\rho_s-\rho)gd], which compares applied shear with the submerged weight scale of a grain. The threshold is not universal: grain packing, exposure, cohesion, and turbulent fluctuations alter entrainment.

In a glacier, gravity drives shear through the ice and at its bed. A shallow slab approximation gives a basal driving-stress magnitude τd≈ρigHsin⁡α\tau_d\approx\rho_i gH\sin\alpha, where HH is ice thickness and α\alpha is surface slope. Ice deformation is strongly nonlinear, and basal motion also depends on water pressure, bed roughness, and sediment properties. Longitudinal and lateral stresses become important near valley walls, ice-stream margins, and abrupt changes in geometry.

Biological tissues, blood flow, and cellular responses

Biological systems experience shear at scales from whole joints to molecular attachments. Blood flowing through a vessel exerts tangential traction on the endothelial lining, while pressure primarily produces normal loading and circumferential wall tension. These are mechanically distinct stimuli even though they occur simultaneously. Wall shear stress depends on the near-wall velocity gradient, blood rheology, and vessel geometry; it cannot be determined from blood pressure alone.

In an idealized straight circular vessel carrying steady, fully developed Newtonian laminar flow, wall shear stress can be expressed as

∣τw∣=4μQπR3,|\tau_w|=\frac{4\mu Q}{\pi R^3},

where QQ is volumetric flow rate and RR is vessel radius. This illustrates the strong effect of radius at fixed flow rate. Real circulation is pulsatile, vessels branch and curve, and blood is a suspension of deformable cells. Its effective viscosity depends on flow conditions and vessel size, particularly in the microcirculation. The ideal formula is therefore a useful reference rather than a complete description of vascular flow.

Endothelial cells detect mechanical loading through interacting structures that include the surface glycocalyx, membrane-associated complexes, cell–cell junctions, and attachments to the extracellular matrix. Mechanical signals are transmitted to the cytoskeleton and can alter biochemical signaling and gene expression. This conversion of physical input into biological activity is called mechanotransduction. Sustained directional shear can promote endothelial alignment with flow in experimental systems, whereas reversing or multidirectional shear produces different cellular responses. Magnitude, duration, waveform, and spatial gradients all matter.

Shear also acts on circulating cells and macromolecules. High fluid stresses can deform or damage red blood cells, while platelet interactions and the behavior of von Willebrand factor depend on the local flow environment. Flow gradients can stretch large proteins and alter their interactions with surfaces and cells. Effects depend on exposure time as well as stress magnitude, so there is no single shear-stress threshold that describes all blood damage or activation phenomena. Medical-device studies consequently examine stress histories and residence times rather than only peak values.

In solid tissues, shear stiffness and time dependence reflect tissue architecture, fluid content, and molecular organization. Cartilage, tendons, brain tissue, and muscle have different anisotropic and viscoelastic responses. Shear-wave elastography estimates aspects of mechanical behavior from wave propagation, but converting wave speed into a modulus requires assumptions about density, material uniformity, viscosity, and directional dependence. Such measurements characterize tissue response; they are not direct measurements of every stress acting within the tissue.

Measurement, computation, and development of the concept

Shear stress is often inferred from other measured quantities rather than sensed directly. In a torsion test, a known torque and specimen geometry provide a stress estimate, while angular displacement provides deformation. Rheometers apply controlled rotation or stress to cone-and-plate, parallel-plate, or concentric-cylinder geometries. Each geometry has its own assumptions about the stress and shear-rate fields. Wall slip, edge disturbances, sample loading, and shear heating can cause the inferred material response to differ from the bulk response.

In fluid systems, wall shear can be estimated from pressure drop when the geometry and momentum balance are sufficiently well defined. Local methods include floating-element balances, calibrated thermal sensors, oil-film techniques, and near-wall velocity measurements. Each measures a different physical consequence of shear. A floating element responds to tangential force, while a thermal sensor relies on a relationship between heat transfer and flow. Velocity-based estimates require enough spatial resolution to determine a steep near-wall gradient, often a demanding requirement in turbulent flow.

Solid-mechanics measurements commonly combine strain gauges or optical displacement fields with a constitutive law. Digital image correlation tracks a surface pattern to estimate displacement and strain; stress then follows only after material behavior and boundary conditions have been specified. Photoelasticity uses stress-induced optical birefringence in suitable transparent materials. In a two-dimensional configuration, its fringe patterns are related to principal stress differences and hence to maximum in-plane shear stress, rather than providing every tensor component independently.

Numerical methods extend these approaches to complicated shapes and flow fields. Finite-element analysis estimates stress in solids by solving equilibrium or dynamic equations together with constitutive laws. Computational fluid dynamics obtains viscous and modeled turbulent stresses from the calculated flow. Mesh refinement, near-wall treatment, material parameters, and boundary conditions strongly affect local results. An ideal sharp corner or crack can produce a mathematical stress singularity, so a larger computed peak after refinement does not necessarily indicate numerical failure. Interpretation may require fracture parameters or a physically justified averaging scale.

Historically, the modern concept developed through several lines of mechanics. Isaac Newton described resistance associated with relative motion of fluid layers in the seventeenth century; nineteenth-century work by Claude-Louis Navier and George Gabriel Stokes developed viscous-flow equations. Augustin-Louis Cauchy formulated the general stress description in the 1820s. Henri Tresca’s nineteenth-century experiments informed maximum-shear yielding ideas, Otto Mohr developed graphical stress analysis, and Richard von Mises proposed his yield criterion in 1913. These developments separated stress as a local mechanical quantity from the particular laws governing a material’s response. Modern measurements retain that distinction: reliable stress estimates require calibrated instrumentation, explicit conventions, and quantified measurement uncertainty.