Fluid flow validation
Each check compares a RunMat result with an independent reference: a closed-form solution, a manufactured solution or a published benchmark. The tests listed under a check run on every change and in the evidence report; receipts record a public study run with its build, commit and mesh. The validation methodology explains how checks are built and how tolerances are set, and Status summarizes every family.
Steady
Plane Poiseuille pressure drop
Problem. Fully developed laminar flow between parallel plates, refined in the kernel test. The public example has L/h = 30 at Re = 1.
Reference. Fully developed pressure drop ; 36 Pa for the public example. (F. M. White, Viscous Fluid Flow, 3rd ed., McGraw-Hill, 2006.)
Compared. Difference of the face-mean pressure between inlet and outlet.
Tolerance. Kernel: relative error below 5 % and decreasing with refinement. Public example: between 34.2 and 37.8 Pa (5 %), which contains the uniform inlet's entrance loss of about +3 %.
Result. Passed (2 tests) in the evidence run of 2026-11-08.
Tests: pressure_drop_of_plane_poiseuille_flow_is_twelve_mu_u_l_over_h_squared, plane_poiseuille_example_publishes_the_analytic_pressure_drop.
Couette flow
Problem. Flow between a moving and a fixed wall on canonical tetrahedra.
Reference. The linear Couette profile, its uniform vorticity and its wall shear stress . (F. M. White, Viscous Fluid Flow, 3rd ed., McGraw-Hill, 2006.)
Compared. Velocity field, wall shear on both walls, vorticity, global momentum.
Tolerance. Velocity relative L2 error below 0.25; wall shear within 0.25; global momentum below 1e-6. For the exact linear field, vorticity and wall shear within 1e-6.
Result. Passed (2 tests) in the evidence run of 2026-11-08.
Tests: affine_couette_uses_original_geometry_and_balanced_physical_fields, couette_vorticity_and_wall_shear_match_the_linear_profile.
Fully developed flow in a rectangular duct
Problem. Pressure-driven laminar flow in a rectangular duct, three refinements.
Reference. Fourier-series solution for the velocity, flow rate and wall shear; the series tail is below 1e-5 at 255 terms. (F. M. White, Viscous Fluid Flow, 3rd ed., McGraw-Hill, 2006.)
Compared. Velocity L2 error, flow rate, wall shear.
Tolerance. Each error decreases with refinement; finest velocity error below 0.35 and flow-rate error below 0.25; boundary flux below 1e-8.
Result. Passed (1 test) in the evidence run of 2026-11-08.
Tests: rectangular_duct_compares_fourier_reference_with_controlled_tail.
Uniform flow through a porous zone
Problem. Uniform flow through a porous slab.
Reference. Darcy–Forchheimer pressure gradient . (D. A. Nield and A. Bejan, Convection in Porous Media, 5th ed., Springer, 2017.)
Compared. Velocity and pressure gradient.
Tolerance. Relative 1e-6.
Result. Passed (1 test) in the evidence run of 2026-11-08.
Tests: uniform_flow_through_a_porous_zone_has_the_darcy_forchheimer_gradient.
Shear-thinning channel flow
Problem. Pressure-driven channel flow of a power-law fluid with n = 0.5, three refinements.
Reference. The closed-form power-law channel velocity profile.
Compared. Velocity profile error.
Tolerance. Error decreases with refinement and is below 5 % on the finest mesh.
Result. Passed (1 test) in the evidence run of 2026-11-08.
Tests: shear_thinning_power_law_channel_converges_to_the_closed_form_profile.
Transient
Uniformly accelerating flow
Problem. A uniform flow accelerated at a constant rate, at two reference pressures.
Reference. The exact inertia-driven pressure gradient .
Compared. Velocity and pressure gradient.
Tolerance. Velocity within 1e-8; pressure gradient within 1e-6.
Result. Passed (1 test) in the evidence run of 2026-11-08.
Tests: original_uniform_acceleration_has_exact_inertia_pressure_and_cold_binding.
Exponentially decaying translation
Problem. A uniform translation decaying as , solved with three time steps.
Reference. The exact pressure gradient of the decaying translation.
Compared. Volume-weighted pressure-gradient error.
Tolerance. Error falls below 0.6 of its previous value at each halving of the step (first order) and is below 0.02 at the finest step.
Result. Passed (1 test) in the evidence run of 2026-11-08.
Tests: exponential_translation_pressure_shows_first_order_time_convergence.