
ID	0147
Class	KNOWN-LIMITATION
Severity	2
Status	blocked

Models: —

Found: 2026-11-03, docs review of 2026-10-24 (recommendation 7b); the work ENV-0115 and ENV-0116 deferred to a design with a state per contact

Decision needed: the owner chooses the state per contact (a Dahl or LuGre bristle, or a set-up time of the soil), and whether stiction and the hold of a buried body come in one design.

Evidence

include/fhsim_environment/seabed/SeabedLaw.h:28-35: a fully buried body is pushed up by the elastic force and rises until it rests on the bed (ENV-0115, closed as a known limitation); below v_reg the friction is only a fraction of its limit, so a resting body creeps (ENV-0116, "deferred"; FlukeAnchor 1 to 4 cm/s, TorpedoAnchor 7 mm/s, measured in fhsim_base). include/fhsim_environment/seabed/SeabedContact.h:121-123 says the same. The soil law keeps no state per contact. ENV-0116 was closed as resolved only because the tracker has no deferred state; its ruling says the stiction work is to be filed anew when it is designed. This is that record.

Effect

A body put deep in the soil, or an anchor that has dug in, does not stay embedded. A resting anchor under a small steady side force slides slowly instead of holding. A fixture must start a body on the bed, and a holding anchor needs a model of its own outside the library.

Possible fix

Add a state per contact (Dahl or LuGre bristle, or a soil set-up time) so that static friction holds below the limit and a buried body is held at rest. It changes the soil law's interface (a state to integrate or step) and its results at low speed. Needs the owner's design decision.

Test that would prove it

A body at rest on the bed under a side force below the friction limit stays within a small bound for a long run; a body started fully buried stays buried. Both fail today.

Risk

The state per contact touches the ODE state and the Jacobian of every body that uses the soil law. Results of existing seabed tests move at low speed.