Deep-Sea Buoyancy: Ballast, Compressible Hull, Pressure and Depth Stability

Trim a research submersible: seawater compression, hull shrinkage, ballast tanks, added mass and drag, and why neutral buoyancy is unstable.

Switch to 2D

Presets

Playback

Parameters

Camera

Layers

Quality

Model check

The science

What is modelled

A 4 m research submersible hanging in the water column between the surface and a seabed at 500 m. Its weight is fixed; its buoyancy is set by the displaced water. You change the buoyancy by flooding or blowing the two side ballast tanks, by adding or dropping payload, by moving to fresher or denser water, and by pushing with the vertical thrusters.

Equations

  • Seawater is slightly compressible: ρ(p) = ρ0(1 + p/Kw) with Kw = 2.3 GPa. Integrating hydrostatics dp/dz = ρ g gives p(z) = Kw(eρ0 g z / Kw − 1) and ρ(z) = ρ0 eρ0 g z / Kw.
  • The pressure hull shrinks as Vp = Vp0(1 − p/Kh). Buoyancy is B = ρ g (Vp + Vb(1 − f)), where f is the flooded fraction of the ballast tanks (Vb = 1.78 m3).
  • Vertical motion: (m + ma) dv/dt = m g − B − ½ ρ Cd A v|v| − Fthr, with added mass ma = Ca ρ V (Ca = 0.7) and Cd A = 7.5 m2.
  • Terminal speed under a constant force F: vt = √(2F / ρ Cd A), approached as v = vt tanh(t/τ) with τ = (m + ma)/√(½ ρ Cd A F).
  • Depth stability: the slope d(mg − B)/dz is positive when the hull compresses more than the water does. Then neutral buoyancy is an unstable balance and a small push grows as cosh(λ t), λ = √(slope / (m + ma)). A stiff hull (Kh above about 2.3 GPa) is stable.

Checks run in the page

  • Hydrostatic pressure at 500 m: RK4 integration against the closed form.
  • Terminal velocity and the tanh law from the same stepper that drives the sub.
  • The cosh runaway growth against the analytic slope at neutral buoyancy.
  • Work-energy theorem with drag off.
  • Archimedes trim: displaced mass equals the sub mass at the design point.

Honest simplifications

  • The hull is given a single effective bulk modulus (default 1 GPa, as for a hull with syntactic-foam buoyancy); a real vehicle's compressibility is measured, and the stability result depends on it.
  • Temperature and salinity layers (thermoclines) are ignored: density changes only through pressure, so real vehicles will see extra density steps.
  • Motion is purely vertical, with fixed added-mass and drag coefficients typical of a bluff body; there is no pitch, roll or current.
  • The ballast tank is vented and free-flooding (no gas compression), pumped at a limited rate of 5 % per second.
  • Light attenuation with depth is a stylised exponential, not an optical model.

What to try

  • Press "Find neutral trim", then run at 6x: the sub still drifts away because neutral buoyancy is unstable.
  • Switch on the depth-hold autopilot: a small thruster force pins the depth.
  • Set the hull modulus to 20 GPa and the same trim becomes stable.
  • Add 500 kg of payload, then flood less ballast to recover.
Loading model...
Loading Blender model...
Depth against time
Net force against depth (slope shows stability)
Net force against ballast fill
Made with 🟠 Blender · Three.js