Skip to content

ashfordeOU/aero-agent-skills / skills/aerodynamics/boundary-layer/stokes-creeping-flow-drag

stokes-creeping-flow-drag

Verified

Use when you must compute the steady low-Reynolds-number viscous drag on a sphere in creeping flow, the Stokes solution for the slow motion of a sphere through a viscous fluid: evaluate the stokes streamfunction and the velocity field about the sphere, the surface pressure and wall-shear distributions with their high-pressure-facing-the-stream signature, the total stokes drag F = 6*pi*mu*a*U split one third pressure drag to two thirds friction drag, the drag coefficient Cd = 24/Re_D at the diameter Reynolds number, the Oseen correction factor 1 + (3/8)*Re_a on the radius-based Reynolds number, and the terminal settling velocity (2/9)*(rho_p - rho_f)*g*a^2/mu of a small dense sphere in still air. Produces the creeping-flow drag, the field values and the settling speed in SI units that anchor low-Reynolds-number body-drag estimates and viscous-flow checks. Trigger: stokes-creeping-flow-drag, creeping-flow, stokes-drag, stokes-streamfunction, oseen-correction, terminal-velocity.

Stokes Creeping-Flow Drag (aerodynamics/boundary-layer/stokes-creeping-flow-drag)

Use when you must compute the steady creeping (Stokes, 1851) flow of a viscous fluid past a sphere at Reynolds number well below one: the slow-motion solution of the full Navier-Stokes equations with the inertia terms dropped, in the form Schlichting Boundary-Layer Theory section 4 presents it. This leaf implements the Stokes streamfunction and the velocity field about the sphere, the surface pressure and wall shear, the total drag F = 6pimuaU with its exact one-third pressure (form) to two-thirds friction split, the drag coefficient Cd = 24/Re_D, the Oseen first-order drag correction and the terminal settling velocity, in pure Python stdlib, closed form, no iteration. It is the steady low-Reynolds-number member of the exact-viscous family, complementing the time-dependent plate Stokes layers of unsteady-laminar-stokes-layers, and it anchors low-Reynolds-number body-drag estimates for particle drift and droplet settling checks. It does not do time-dependent plate Stokes layers, high-Mach compressible sphere drag, flat-plate boundary layers or parachute descent balances: incompressible constant-property laminar flow only, uniform mu and nu, Reynolds numbers in the creeping range.

Domain quick reference

Spherical polar coordinates (r, theta) with theta measured from the downstream pole: the uniform stream U runs along +z toward theta = 0 and the windward stagnation point sits at theta = pi. Module constants: NU_AIR = 1.46e-5 m2/s, RHO_AIR = 1.225 kg/m3, G = 9.81 m/s2, RHO_WATER = 1000.0 kg/m3 (particle density for the settling example). Dynamic viscosity is always derived MU_AIR = RHO_AIR * NU_AIR = 1.7885e-05 Pa s, never an input.

  • Stokes streamfunction: psi(r, theta) = 0.5Ur^2sin(theta)^2 * (1 - 1.5a/r + 0.5*(a/r)^3), m3/s. psi(a, theta) = 0 identically (the sphere surface is the psi = 0 streamline) and psi -> 0.5Ur^2*sin(theta)^2 as r -> inf (uniform stream value).
  • Radial velocity: u_r = Ucos(theta)(1 - 1.5a/r + 0.5(a/r)^3), m/s. Negative on the axis upstream of the windward point (theta = pi, flow approaching), positive downstream (theta = 0); u_r(a, theta) = 0 (no penetration).
  • Tangential velocity: u_theta = -Usin(theta)(1 - 0.75a/r - 0.25(a/r)^3), m/s. u_theta(a, theta) = 0 (no slip). Fore-aft symmetry at zero Reynolds number: u_r(2a, 0) = -u_r(2a, pi) = +-0.3125*U, no wake and no separation.
  • Surface pressure: p - p_inf = -1.5*(muU/a)cos(theta), Pa: HIGH by 1.5muU/a at the windward stagnation point (theta = pi), LOW by the same amount at the downstream pole (theta = 0), zero at the equator. The reversed-pressure signature of creeping flow: no dynamic-pressure head, the pressure does not rise on the lee side.
  • Wall shear: tau_w = 1.5*(muU/a)sin(theta), Pa: zero at the stagnation points theta = 0 and pi, peak 1.5muU/a at the equator theta = pi/2.
  • Stokes drag: F = 6pimuaU, N. Pressure (form) drag F_p = 2pimuaU from the surface pressure integral, friction drag F_f = 4pimuaU from the wall-shear integral: F_p + F_f = F identically and F_p/F_f = 1/2 exactly (one third form drag, two thirds friction drag).
  • Drag coefficient: Cd = F/(0.5rhoU^2pia^2) = 24/Re_D = 12mu/(rhoUa), dimensionless, with Re_D = U2a/nu the diameter-based Reynolds number.
  • Oseen first-order correction: F_oseen = 6pimuaU*(1 + (3/8)Re_a), N, with Re_a = Ua/nu the radius-based Reynolds number, identical to the factor 1 + (3/16)*Re_D on the diameter-based number. The factor is 1.1875 at Re_a = 0.5. The Oseen term is a first-order drag correction toward Re ~ 1, kept inside this leaf, never claimed as a separate flow theory.
  • Terminal settling velocity: weight minus buoyancy (4/3)pia^3*(rho_p - rho_f)g = 6pimuaU gives U_t = (2/9)(rho_p - rho_f)ga^2/mu, m/s, a closed-form balance valid while Re_D(U_t) stays below about 0.1. The Oseen-corrected terminal velocity is the closed root of U*(1 + (3/8)Ua/nu) = U_t: U = (-1 + sqrt(1 + 4cU_t))/(2c) with c = (3/8)(a/nu).
  • SI units throughout; Stokes 1851 and Schlichting section 4 material cited through NACA TR-824, summary-only.

Workflow

  1. Fix the creeping-flow state: the sphere radius a, the stream speed U, the fluid density rho and kinematic viscosity nu (mu = rho*nu derived, never an input), then confirm the regime with the radius-based and diameter-based Reynolds numbers (radius_reynolds, diameter_reynolds): Re_D below about one keeps the pure Stokes law operative.
  2. Traverse the streamfunction and velocity field about the sphere (stokes_streamfunction, radial_velocity, tangential_velocity) over r >= a and the polar angle: the no-slip surface values psi(a, theta) = u_r(a, theta) = u_theta(a, theta) = 0 and the fore-aft symmetric far field u_r(2a, 0) = -u_r(2a, pi) = 0.3125*U are the creeping-flow fingerprints.
  3. Traverse the surface pressure difference with surface_pressure_delta over the polar angle: read the high-pressure windward stagnation point (theta = pi) and the equal suction at the downstream pole (theta = 0), the reversed-pressure signature.
  4. Traverse the wall shear with wall_shear_stress over the polar angle: zero at both stagnation points, peak 1.5muU/a at the equator.
  5. Assemble the drag: the total stokes_drag with the pressure_drag (one third) and friction_drag (two thirds) split; the closed-form identity pressure_drag + friction_drag = stokes_drag is exact.
  6. Reduce the drag with drag_coefficient against the freestream dynamic pressure and frontal area; cross-check Cd = 24/Re_D on the diameter-based Reynolds number.
  7. Apply the Oseen first-order correction for Reynolds numbers approaching one: oseen_correction(Re_a) multiplies the drag in oseen_drag, a 2.6 percent rise at Re_a = 0.068 and 18.75 percent at Re_a = 0.5.
  8. Balance settling: terminal_velocity from weight minus buoyancy against the Stokes drag, and run the creeping Reynolds check Re_D(U_t) below about 0.1 at the settling speed.
  9. Correct the terminal velocity with oseen_terminal_velocity, the closed root of the corrected balance, always below the Stokes value.
  10. Confirm the deterministic checks: rerun the closed-form identities and reject non-physical inputs (the input-rejection traverse) with the contract test scripts/test_stokes_creeping_flow_drag.py.

Worked example

Air at standard conditions nu = 1.46e-5 m2/s, rho = 1.225 kg/m3, mu = rho*nu = 1.7885e-05 Pa s, g = 9.81 m/s2. All values below are real outputs of the module.

Worked sphere: radius a = 1.0e-4 m (0.1 mm) at U = 0.01 m/s in air:

  • Reynolds numbers: Re_a = Ua/nu = 0.068493151 and Re_D = U2a/nu = 0.1369863, the low end of the creeping range where the Oseen correction still matters.
  • Stokes drag F = 6pimuaU = 3.3712431e-10 N (0.337 nN). Pressure (form) drag F_p = 2pimuaU = 1.1237477e-10 N, friction drag F_f = 4pimuaU = 2.2474954e-10 N: the split is one third of the total to pressure and two thirds to friction, an exact 1:2 ratio, and the sum check F_p + F_f - F leaves a zero residual (module value 0.0 N). Creeping linearity: F(2U)/F(U) = 2.0 exactly.
  • Drag coefficient Cd = 12mu/(rhoUa) = 175.2, identical to 24/Re_D = 175.2 and to F/(0.5rhoU^2pi*a^2) = 175.2, three independent routes agree.
  • Surface pressure p - p_inf = -1.5*(mu*U/a)*cos(theta): +2.68275e-03 Pa at the windward stagnation point (theta = pi), -2.68275e-03 Pa at the downstream pole (theta = 0), and zero at the equator (module residual -1.64271060e-19 Pa): high pressure faces the oncoming stream, an equal suction trails it.
  • Wall shear tau_w = 1.5*(mu*U/a)sin(theta): 1.89699072e-03 Pa at theta = pi/4, peak 2.68275e-03 Pa at theta = pi/2, symmetric 1.89699072e-03 Pa at theta = 3pi/4, and 0.0 Pa at theta = 0 with a 3.285e-19 Pa float residue at theta = pi.
  • Velocity field at r = 2a: u_r = +3.12500000e-03 m/s downstream on the axis (theta = 0), zero at the equator (module residual 1.91351062e-19 m/s), and -3.12500000e-03 m/s upstream (theta = pi): the flow is fore-aft symmetric with no wake at this Reynolds number. Tangential component u_theta = -4.19844651e-03 m/s at theta = pi/4 and -5.93750000e-03 m/s at theta = pi/2.
  • No slip on the surface: u_r(a, pi/3) = 0.0 m/s and u_theta(a, pi/3) = -0.0 m/s; the streamfunction is psi(a, theta) = 0 on the surface, psi(2a, pi/2) = 6.25000000e-11 m3/s (0.625Ua^2) and psi(2a, pi/4) = 3.12500000e-11 m3/s.

Oseen correction:

  • At the worked sphere Re_a = 0.068493151 the factor 1 + (3/8)Re_a = 1.025684932, a 2.56849 percent drag increase, giving F_oseen = 6pimuaU(1 + 3*Re_a/8) = 3.45783322e-10 N.
  • Factor values: 1.1875 at Re_a = 0.5, identical to 1 + (3/16)*Re_D at Re_D = 1.0 (the two Reynolds conventions give the same correction).

Terminal velocity: a water droplet (rho_p = 1000 kg/m3) of radius a = 1.0e-5 m (10 microns) settling in still air:

  • Stokes terminal velocity U_t = (2/9)*(rho_p - rho_f)ga^2/mu = 1.21740537e-02 m/s (1.217 cm/s).
  • Balance check: stokes_drag at U_t over the buoyancy-adjusted weight (4/3)pia^3*(rho_p - rho_f)*g = 1.0 exactly (module value).
  • Droplet Reynolds numbers at U_t: Re_a = 0.0083383929 and Re_D = 0.016676786, deep inside the creeping regime (Re_D well below 0.1), so the pure Stokes balance is the operative one.
  • Oseen factor at U_t: 1 + (3/8)*Re_a = 1.003126897; the Oseen-corrected terminal velocity (closed-form quadratic root) is 1.21362229e-02 m/s, 0.310749 percent below the Stokes value.

Read-off: a 0.1 mm sphere drifting at 1 cm/s through air carries only 0.337 nN of drag, one third of it pressure and two thirds friction, with a drag coefficient of 175, while a 10 micron water droplet settles at 1.22 cm/s with a Reynolds number of 0.017; the reversed surface pressure (high facing the stream, suction behind) and the fore-aft symmetric field are the fingerprints of creeping flow, and the Oseen factor adds 2.6 percent drag at Re_a = 0.068 and 18.75 percent at Re_a = 0.5.

Verification

  • Confirm radius_reynolds(0.01, 1.0e-4, 1.46e-5) returns 0.068493151 and diameter_reynolds returns 0.1369863, exactly twice the radius-based value.
  • Confirm stokes_drag(1.7885e-5, 1.0e-4, 0.01) returns 3.3712431e-10 N (between 3.0e-10 and 4.0e-10 N) and that doubling the speed exactly doubles the drag (creeping linearity).
  • Confirm the split: pressure_drag 1.1237477e-10 N, friction_drag 2.2474954e-10 N, friction/pressure exactly 2, and the sum equals stokes_drag to float noise.
  • Confirm drag_coefficient(1.225, 1.7885e-5, 0.01, 1.0e-4) returns 175.2, equal to 24/Re_D and to F/(0.5rhoU^2pia^2).
  • Confirm surface_pressure_delta is +2.68275e-03 Pa at theta = pi and -2.68275e-03 Pa at theta = 0 (antisymmetric, equator zero) and wall_shear_stress peaks at 2.68275e-03 Pa at theta = pi/2.
  • Confirm the no-slip surface values, the fore-aft symmetric field at r = 2a and the oseen_drag 3.45783322e-10 N at the worked sphere.
  • Confirm terminal_velocity(1000.0, 1.225, 1.7885e-5, 1.0e-5) returns 1.21740537e-02 m/s (between 1.1e-2 and 1.3e-2 m/s) with the drag-weight balance at 1.0 and Re_D(U_t) = 0.016676786 below 0.1.
  • Confirm every non-positive mu, a, U, rho, nu, every r below a, a negative Re_a and every rho_particle at or below rho_fluid raises ValueError.
  • Run the contract test offline: python3 scripts/test_stokes_creeping_flow_drag.py (35 tests, deterministic).

Related leaves

  • aerodynamics/boundary-layer/unsteady-laminar-stokes-layers: the unsteady member of the exact-viscous family, time-dependent Stokes layers of an infinite plate in a quiescent fluid; this leaf is the steady low-Reynolds-number body-drag complement.
  • aerodynamics/boundary-layer/boundary-layer-theory: steady flat-plate layers (Blasius and 1/7-power skin friction), flat plate only, no sphere and no body drag.
  • aerodynamics/high-speed/hypersonic-flow: the compressible high-Mach sphere drag of impact theory at Mach well above 5, the regime fence that never meets the incompressible creeping limit of the same geometry.
  • space-systems/mission-design/entry-descent-landing: the high-Reynolds parachute terminal-velocity balance under the parachute drag coefficient, the bluff-body descent fence against the creeping Stokes balance.

Pitfalls

  • Applying the Stokes law above the creeping range: once Re_D climbs past about one, inertia, wake and separation break the linear drag and Cd = 24/Re_D under-predicts; run the Reynolds check and take the Oseen correction only toward Re ~ 1.
  • Mixing the Reynolds conventions: Cd = 24/Re_D uses the diameter while the Oseen factor (3/8)*Re_a uses the radius, so 1 + (3/16)*Re_D is the diameter-based twin of the factor; swapping conventions halves or doubles the correction.
  • Reading the 1:2 split backwards: creeping sphere drag is one third pressure and two thirds friction, the reverse of a high-Reynolds bluff body, and the surface pressure is high at the windward stagnation point with suction at the lee pole, opposite to Bernoulli intuition, because viscosity dominates at Re << 1.
  • Feeding mu as an input: dynamic viscosity is always derived mu = rhonu (MU_AIR = RHO_AIRNU_AIR = 1.7885e-05 Pa s in the module); an inconsistent rho-nu pair silently shifts every drag and settling value.
  • Skipping the creeping check on a settling sphere: the Stokes balance holds while Re_D(U_t) stays below about 0.1; larger or denser droplets need the Oseen-corrected root or a higher-Reynolds drag law.
  • Forgetting the buoyancy: U_t uses rho_p - rho_f, and a sphere with rho_p at or below rho_f does not settle (the module raises ValueError).
  • Reversing the polar-angle convention: theta is measured from the downstream pole with the windward stagnation point at theta = pi; swapping the convention flips the pressure sign and mirrors the streamfunction field.

Behavior contract (gate 3)

Run the deterministic contract test (stdlib unittest, offline):

python3 scripts/test_stokes_creeping_flow_drag.py

The test covers the worked-example contract (Reynolds numbers, drag with magnitude bounds, the 1:2 pressure-to-friction split, Cd = 175.2 by three routes), the surface pressure and wall-shear traverses with their anchors, the no-slip surface and fore-aft symmetric far-field values, the streamfunction anchors, the Oseen correction and drag, the terminal-velocity balance with its creeping Reynolds check, the Oseen-corrected settling root, deterministic repeat calls, and the input-rejection traverse of non-physical mu, a, U, rho, nu, r, Re_a and buoyancy arguments.

Compliance

  • Standards referenced, not reproduced: NACA TR-824 is the cited reference for the viscous-layer family; the classical creeping-flow treatment follows Stokes 1851 and Schlichting Boundary-Layer Theory section 4, whose material is cited through the report. Summary-only methodology per standards-map.yaml.
  • compliance: STANDARDS-REF, gated: false.
slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack."Stars: 198kUpdated Claude CodeCodexVerifiedInstall
ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt.Stars: 194kUpdated Claude CodeCodexVerifiedInstall