
ID	0117
Class	KNOWN-LIMITATION
Severity	2
Status	blocked

Models: —

Found: 2026-10-19, plan C1 task 5 (seafloor sanity criteria, feature old-vs-new), default SoilModel and default SoilState

Decision needed: Is a friction ratio above mu intended (adhesion and passive resistance are part of the sliding resistance), or should the sanity criterion compare against mu + adhesion? Physics: owner.

Evidence

include/fhsim_environment/seabed/SeabedContact.h:135-137 (SoilForce): the sliding resistance is mu F_n + alpha s_u A_s plus the passive term N_p s_u A_f, so friction over normal force is not mu.
Failing test: tests/SeabedContact_Sanity_Test.cpp, SeabedSanity_VALIDATION.SlowDragFrictionOverNormalEqualsMu (v = 0.5 m/s, body resting at its static penetration, dragged along x, default soil mu = 0.2, alpha = 0.35, s_u0 = 5 kPa). Criterion: |ratio - mu| < 5 % of mu.
Friction / normal: sphere D 0.1 m (35 N) 0.314; sphere D 1 m (5 kN) 0.325; cylinder D 0.1 m (100 N/m) 0.350; cylinder D 1 m (2 kN/m) 0.323.
SlowDragWithoutAdhesionIsNearMu (alpha = 0): cylinders 0.2001 and 0.2031 (pass the criterion in value; the test fails on the spheres), spheres 0.249 and 0.259. So adhesion adds 0.11 to 0.15, and the passive face of a sphere adds 0.05 to 0.06.

Effect

A slowly dragged body on the default soil feels 1.6 to 1.75 times the Coulomb friction of mu F_n. Cable pull-in or drag tests that expect mu = 0.2 see a larger force.

Possible fix

Either accept (adhesion is real for clay) and change the criterion to mu + alpha s_u A_s / F_n + passive; or change the soil defaults (alpha = 0 for sand). Owner decision; the law is not changed by this issue.

Test that would prove it

The two SlowDrag* tests, once the owner has ruled which ratio is right.

Risk

Low: a validation test only.