
ID	0086
Class	KNOWN-LIMITATION
Severity	3
Status	blocked
Models	Net/Sphere
Found	2026-09-25, review of the MARE-0084 fix
Decision needed	triage

Evidence

src/net/Sphere.cpp:80, Sphere::OdeFcn:

XDot[m_VelIdx + i] = (DragForce[i] + Buoyancy[i] + ExtForce[i]) / m_sphereMass;

The inertia is SphereMass alone, and the only fluid forces are the drag on the velocity relative to the water and the buoyancy.

Effect

A sphere accelerating in still water has no added mass (0.5 ρ V for a sphere), and a sphere in waves or a varying current gets no Froude–Krylov force (ρ V times the water acceleration), so a light sphere responds faster and follows the water motion less than it should.

Possible fix

Add the added mass to the inertia and the Froude–Krylov and added-mass forces from the water acceleration, or document that SphereMass must include the added mass and that wave excitation is left out.

Test that would prove it

A sphere released from rest in still water under a constant force: its acceleration must be F / (m + 0.5 ρ V).

Risk

Changes every model with a sphere; the analytical Jacobian must follow.