CS-310 is the Water family's persistent survey-and-carrier AUV: a ring-framed polymer pressure hull in a free-flooded moulded fairing, a 20 kg / 20 L wet bay at the centre of buoyancy, and a screw sized for range. This page is the hull, and why it is the shape it is.
Two bodies, one dry. CS-310 is a torpedo-form body of revolution, 2.60 m overall over a Ø345 mm fairing at a fineness ratio of 7.5, rated to 100 m.
A glass-filled nylon cylinder with 8 mm hemispherical closures and four ring frames at 300 mm pitch, holding the 2.0 kWh battery on its trim rail, the motor, the avionics and the ballast machinery: the only volume aboard at one atmosphere.
Nose, mid and tail shells flood through slots and see no pressure differential, so they are light, inexpensive and quietly shaped; an unloaded part is eligible for recycled PET.
A flooded bay about 700 mm long, nominal 20 kg / 20 L, behind one define-once interface. A drop from the centre of buoyancy disturbs trim minimally.
Only the inner cylinder is a pressure vessel; everything the water touches first is moulded polymer off the fleet's own tooling, so the hard engineering concentrates in one simple, testable part, a tube with two end caps. Its discipline is ovality control and proof testing; the fairing is cut on the fleet's own 5-axis die shop, on the manufacturing page. An all-plastic vehicle is inexpensive to build in numbers, and the same hull carries no steel, so there is no magnetic anomaly to read and only a weak sonar return.
The screw is audible, and that trade is accepted for the survey role: range and payload sit at this end of the family, and the last, quietest kilometre belongs to CS-320, which rides here as a cartridge. The two hulls sit side by side on the Water family; the buyer's view is who it is for.
External pressure at 100 m is 1.006 MPa. Hoop stress on a 10 mm wall is about 12.5 MPa, far below the material's strength, so hoop is the wrong check: a polymer tube under external pressure buckles first.
Two knock-downs govern a polymer pressure hull and neither is optional. A polymer under sustained external pressure loses stiffness with time, and buckling resistance scales with stiffness rather than strength, so the design modulus runs from 9 GPa dry to 6 GPa wet and to 3.0 GPa after a creep retention of 0.50. Real tubes collapse below the perfect-geometry prediction, so an ovality factor of 0.70 comes off the pressure.
| Plain wall, Ø250 mm | Short-term collapse | After creep and ovality | SF at 100 m | Verdict |
|---|---|---|---|---|
| 10 mm | 0.88 MPa, 87 m | 0.31 MPa | 0.30 | Fails outright |
| 15 mm | 2.95 MPa, 294 m | 1.03 MPa | 1.03 | Collapses at nearly its design depth |
| 19 mm | 6.00 MPa, 597 m | 2.10 MPa | 2.09 | Passes, on mass: the wall eats the payload |
A Ø250 mm plain cylinder 2.0 m long has a critical length of 1,425 mm and sits in the long-cylinder regime.
Every plain-wall option fails a safety factor of 2 at 100 m once creep and ovality are applied, and the one that passes does so at 19 mm, where the wall is eating the payload. The way out of a long-cylinder problem is to stop having a long cylinder.
Ring frames divide the shell into short bays, and short cylinders resist external pressure far better because the end restraint carries load. Stiffening a shell also opens a second failure mode.
With frames at 300 mm pitch on a Ø240 mm hull the critical length is 1,224 mm, so each bay sits deep in the short-cylinder regime. What makes “just add rings” a trap is general instability: the shell and its frames can buckle together as one assembly, at a pressure that has nothing to do with the bay length. A modest frame looks like sensible engineering and buys almost nothing, because it moves the failure from one mode to another at essentially the same pressure.
| Ring frame | Inter-bay buckling | General instability | Governing mode | SF at 100 m | Verdict |
|---|---|---|---|---|---|
| 20 × 15 mm, the intuitive frame | 2.73 MPa | 0.74 MPa | General instability | 0.73 | Fails |
| 40 × 18 mm | 2.73 MPa | 1.52 MPa | General instability | 1.51 | Fails |
| 40 × 25 mm, selected | 2.73 MPa | 2.84 MPa | Inter-bay | 2.71 | Passes |
Four ring frames of 40 × 25 mm section at 300 mm pitch on the Ø240 × 12 mm shell.
Frames are what make the depth rating exist at all. The same shell unstiffened reaches 0.60 MPa after knock-downs, a safety factor of 0.60 and a hull that collapses at 60 m. Every bit of the 100 m rating is bought by four rings weighing 3.0 kg between them, and they cost bore: Ø216 mm clear between frames and Ø166 mm at each frame, which the trays and the ballast tank thread through.
Spherical shells are far more imperfection-sensitive than cylinders. The empirical knock-down on a dome is about 0.2 to 0.3, not the 0.70 used on the cylinder, and applying it moves the governing mode back to the cylinder.
A 6 mm hemispherical closure sits at 9.25 MPa classical and 2.31 MPa after a 0.25 sphere knock-down, below the 2.73 MPa cylinder it closes, so at 6 mm the closure would set the hull's rating rather than the cylinder. Going to 8 mm costs 0.98 kg over the 6 mm caps, sits at 16.44 MPa classical and clears at 4.11 MPa. That is the cheapest kilogram in the design.
Shell buckling between frames on the Windenburg–Trilling short-cylinder relation. The mode the hull is sized to.
The ring-shell assembly buckling as one, on the Bryant n = 2 relation. Clear because the frames are 40 × 25 mm.
The 8 mm caps on the classical sphere relation with a 0.25 knock-down. Clear by a wide margin, for 0.98 kg.
The governing 2.73 MPa is a nominal collapse depth of 271 m. At the 100 m design depth the safety factor is 2.71 after creep at 0.50 on modulus and ovality at 0.70 on pressure, and 5.42 at 50 m. The closure is written at a wet modulus of 6 GPa and scales linearly with it, falling to 1.81 at a saturated 4 GPa, so the modulus is carried as the design basis, and every hull is proof-tested hydrostatically before its first wet use.
Collapse pressure scales as roughly (t/D)2.5, so strength rewards a small hull; displacement scales as D², so buoyancy rewards a large one. The battery decided where they balance.
Cell-level energy density is about 200 Wh/kg; a packaged marine pack with its cells, management, enclosure, interconnect and potting is 120 to 150 Wh/kg. At 140 Wh/kg the 2 kWh pack is 14.3 kg rather than 10.0 kg, and the extra 4.3 kg lands on the reserve. Sweeping the diameter at a 12 mm wall with 40 × 25 mm frames at 300 mm:
| Hull | SF at 100 m | Hull | Total dry | Buoyancy | Reserve | Wet | Verdict |
|---|---|---|---|---|---|---|---|
| Ø200 × 1,400 | 3.54 | 18.7 kg | 73.4 kg | 64.3 kg | −9.1 kg | 84.3 kg | Sinks |
| Ø225 × 1,500 | 2.98 | 22.7 kg | 78.6 kg | 83.1 kg | 4.4 kg | 103.1 kg | Closes on a thin reserve |
| Ø225 × 1,500 + 8 L buoyancy | 2.98 | 22.7 kg | 80.6 kg | 91.3 kg | 10.6 kg | 111.3 kg | Closes, buying in a material class |
| Ø240 × 1,500, selected | 2.71 | 24.6 kg | 80.9 kg | 93.1 kg | 12.2 kg | 113.1 kg | Closes on geometry alone |
| Ø250 × 1,800 | 2.51 | 30.3 kg | 88.8 kg | 116.7 kg | 27.9 kg | 136.7 kg | Over-buoyant, carries dead lead |
Every row at a 12 mm wall with four 40 × 25 mm frames, the battery at 140 Wh/kg packaged.
A Ø200 hull is the strongest option on the board and cannot float its own contents, at 9.1 kg negative. A Ø250 × 1,800 hull floats easily and then carries nearly 28 kg of trim lead back to neutral, which is payload thrown away. Ø225 closes on a 4.4 kg reserve, about 5% of displacement; 8 L of closed-cell buoyancy closes it properly at 111.3 kg wet with 9% less frontal area, but the two closing options differ by 1.9 kg wet and 9% of frontal area, a near wash, and the foam buys a material class and a qualification item. Ø240 closes on geometry alone with nothing bought in, and that is the hull.
A submarine's structure is bought twice, once in material and once in the buoyancy the material consumes. Every line carries a mass and a displaced volume, so displacement is a sum rather than an estimate.
| Item | Mass | Displaced |
|---|---|---|
| Pressure hull, 8 mm domes, four ring frames | 24.55 kg | 75.10 L |
| Fairing, planes and duct, free-flooded | 15.89 kg | 11.77 L |
| Battery, 2.0 kWh at 140 Wh/kg packaged | 14.29 kg | 0 |
| Thruster motor and drive electronics | 3.00 kg | 0 |
| Propeller, shaft and coupling rotor, wet | 1.00 kg | 0.40 L |
| Four plane actuators and linkages, wet | 2.50 kg | 1.50 L |
| Avionics, INS and compute | 4.00 kg | 0 |
| DVL and modem transducers, wet | 2.00 kg | 1.20 L |
| Variable-ballast machinery, dry | 4.00 kg | 0 |
| Ballast water at half fill, neutral datum | 2.50 kg | 0 |
| Moving-mass trim rail | 1.20 kg | 0 |
| Penetrators, harness, connectors | 3.00 kg | 0.50 L |
| Drop-weight abort, cast iron, wet | 3.00 kg | 0.38 L |
| Total | 80.93 kg | 90.85 L |
| Displacement in seawater | 93.12 kg | — |
| Fixed trim and reserve, 13% of displacement | 12.19 kg | — |
| Payload cartridge, neutral by rule | 20.00 kg | — |
| Vehicle wet displacement | 113.12 kg | — |
The selected baseline's ledger; items inside the pressure hull displace nothing of their own.
Carrying the same 20 kg payload to 100 m in polymer lands at about 113 kg, and that is what the requirement costs once the parts have masses. Two-person handling survives on a dolly or a davit; an unaided lift does not. The fresh-to-salt swing on this displacement is 2.27 kg, the floor for variable-ballast authority; freshwater and salt water sit on opposite sides of it, so the ±3 kg authority is a requirement.
The non-metallic posture has a cost, and it is about 2% of displacement. An unstiffened 6061-T6 aluminium hull sized to the same safety factor of 2 at 100 m over the full 1,500 mm unsupported length needs a 6.0 mm wall and weighs 22.3 kg against the polymer's 24.6 kg: about 2.3 kg, roughly 2% of displacement, for a hull with no steel in it.
One low-RPM shrouded screw behind four cruciform stern planes, driven by a brushless motor that never leaves the dry core. Every wire that crosses the pressure boundary is a leak path, a pressure-test item and a cost line.
Propulsion is a single shrouded stern thruster, a low-RPM propeller in a duct that doubles as guard and snag protection, at low tip speed and low blade loading. The motor sits inside the pressure hull and drives the screw through a magnetic coupling, torque only, with no dynamic seal and no penetrator. At the 0.91 m/s best-range point the electrical propulsion draw is 16.8 W and shaft power about 14.3 W, so an 8% coupling loss is about 1.1 W against a 47 W budget, roughly 0.8% of range, set against a flooded hull being a total loss.
The stronger argument shows up when the harness is drawn. A naive layout puts ten penetrations through the hull. The coupling deletes the motor penetration, and one tail multi-way carries all four control-plane actuators with their power and position feedback. Five remain.
| Penetration | Carries |
|---|---|
| P1 · Shore umbilical | Charge and data, blanked in the water |
| P2 · Tail multi-way | Four control-plane actuators, power and position feedback |
| P3 · DVL transducer | The Doppler velocity log |
| P4 · Acoustic-modem transducer | The sparse check-in link |
| P5 · Payload-bay wet-mate | Power, data and the release command |
| — · Thruster drive | Magnetic coupling, torque only: no wire, no seal, not a penetration |
Ten leak paths become five, 50% fewer. The circuit and penetrator schedules and the test plan are a Wired Industries deliverable.
Trim is done by moving mass: the battery rides a lead-screw trim rail for static pitch, and a small pumped variable-ballast tank inside the hull handles density change and payload-drop compensation. The stern planes are effective only with way on, so station-keeping is bought as a cartridge rather than built into the hull.
The one device aboard that has to work is a drop weight on a preloaded spring. On command, on timer lapse or on loss of power, the weight goes and the vehicle rises.
The drop weight is held by a normally-energized coil, and cutting power releases it. A flat battery, a hung computer, a flooded bus or a severed harness all end the same way: the weight drops and the vehicle surfaces. Every other circuit fails safe by being switched off, and so does this one, wired inverted relative to everything else aboard. It is the one subsystem that must work when nothing else does.
The weight is a 3.00 kg cast-iron assembly on the ventral keel, cast iron rather than lead because every abort leaves it on the seabed. Recoverability is an environmental requirement and an economic one: a fault surfaces the vehicle and brings the hull, the battery and the data home. Nominal recovery is a planned rendezvous at the surface; off-nominal recovery is the weight. Around it the autonomy stack is disciplined behaviour design: a pre-planned mission in geofenced corridors, loiter or bottom-sit then surface on loss of navigation confidence, return to recovery when a communications timer lapses, and the weight on any critical fault. Human authority lives in the mission plan and the abort criteria, at mission grain rather than joystick grain.
Drag is computed on an ITTC-57 friction line with a 1.25 form factor and a 1.30 appendage factor over 2.94 m² of wetted surface at a 2.60 m reference length, because the fairing, not the pressure hull, is the widest section.
The equivalent frontal drag coefficient is 0.193. Hotel load is the other half of the budget: with an inertial navigator and a Doppler log running, Tier 3 draws about 30 W; Tier 1, on preset dead-reckoning with neither, draws 6 W. On 2,000 Wh at a propulsive efficiency of 0.45, in still water:
| Speed | Drag | Tier 1 | Tier 1 range | Tier 3 | Tier 3 range |
|---|---|---|---|---|---|
| 0.50 m/s | 2.90 N | 217.0 h | 391 km | 60.2 h | 108 km |
| 1.00 m/s | 10.01 N | 70.8 h | 255 km | 38.3 h | 138 km |
| 1.50 m/s, passage | 20.78 N | 26.6 h | 144 km | 20.1 h | 109 km |
Still water. Tier 1 is preset dead-reckoning at a 6 W hotel load; Tier 3 is INS and DVL live at 30 W. Each row carries the speed and the current it assumes.
Tier 3's best-range speed is 0.91 m/s, delivering 139 km over 42.7 h; at that optimum the 16.8 W propulsion draw is about half the 30 W hotel draw, which is what an optimum looks like on a hotel-dominated vehicle. So 0.91 m/s is the norm, and 1.5 m/s is a passage speed for when the clock matters, at 20.1 h and 109 km. Tier 1 optimizes at 0.51 m/s for 391 km over 212.8 h.
Power is set by speed through the water and range by speed over the ground; the gap is the current. Re-optimizing against the set, the vehicle holds 1.06 m/s into a 0.25 m/s current for 0.81 m/s over the ground and 104 km, 75% of still water; into 0.50 m/s it holds 1.26 m/s for 0.76 m/s and 75 km, 54%. The right response to an adverse current is to speed up, because time in the water costs hotel energy. A representative emplacement, 15 km out and 15 km back in still water, is 9.2 h at 0.91 m/s and 22% of the battery.
One vehicle envelope, one bay, many cartridges. The bay is free-flooded, ventral, about 700 mm long, at the centre of buoyancy, nominal 20 kg / 20 L, and every cartridge is a sensor, a relay, a sampler or an inspection tool.
The interface is defined once: mount rails, a wet-mate connector for power and data, retention and release, and the buoyancy bookkeeping for each cartridge, so a new cartridge is tooling rather than redesign, and cartridge internals are the customer's or the partner's scope. One rule pays for itself: every cartridge is neutrally buoyant, ±1 kg in seawater. The vehicle trims the same whichever cartridge is fitted, releasing one is trim-neutral at the moment the vehicle is trying to hold position, and it is what makes carrying the creeper cheap instead of a stability problem. The cost is that every cartridge is ballasted at build, a fixture and a scale.
Passive acoustic or environmental packages for persistent harbour and approach monitoring; the node is the customer's or the partner's payload. A relay node is carried out and moored the same way.
Hull, pier, pipeline and intake inspection, and water-column and sediment sampling: the civil work, on the same hull as the survey work.
Low-signature delivery or recovery of tags, beacons and instrumentation where a surface presence is unwanted. Treated as export-sensitive in its own right.
The family's creeper carried in the bay and released at a surveyed point for the final low-signature leg, below.
Station-keeping is bought the same way. Releasing a node from a vehicle moving at 0.91 m/s over a 30 m water column, at a sink rate of order 0.5 m/s, puts it roughly 50 m downstream: fine for seabed acoustic monitoring, not fine for a precision tag or for holding against a set at a pier. So tunnel thrusters are in, as a bolt-in hover module that occupies about 6 L of the 20 L bay, fitted only on the missions that need it; the base vehicle keeps its clean hull, its five penetrations and its low drag.
The list ends there, by rule rather than by omission. No warhead, no fuzing, no energetic material and no terminal-effect payload, in any cartridge, in any variant, for any customer. Any proposal to arm the platform is a counsel-first corporate question and never a product decision.
The small hull is a cartridge in the large one. CS-320 closes at 16.34 kg and 15.95 L wet, inside the 20 kg / 20 L bay with 18% margin on mass and 20% on volume.
From the bay's side the whole mechanism is a retention cradle and a release command. The creeper rides in the cradle and leaves on a signal the bay interface already carries, and it does not come back: the mothership carries and releases, and the creeper transits and surfaces at a planned point for its own recovery. That is all the hull has to build, and because the creeper is a neutral cartridge by rule, the release transient is zero.
The engineering reason belongs to the small hull. At creeper scale the vehicle cannot make ground against its own design current, so its terminal leg has to start close to the target and inside a tide window. Delivery by CS-310 to a surveyed release point is what makes the creeper's delivered-precision tier load-bearing rather than decorative, and for much of the mission set it is the only thing that makes that tier work at all. The expensive navigation stays on the mothership, and a buyer who needs both buys one relationship. The creeper's own page is CS-320.
Under water the fleet's no-radio, no-GPS doctrine is not a design choice. RF and GPS do not penetrate seawater, and every underwater vehicle already lives the way Canadian Shield platforms are designed to live.
Pre-programmed legs on magnetic heading, a depth cell, a water-speed estimate and time. Jam-proof by construction at a 6 W hotel load; position error grows with distance, so it is the tier for short legs.
A human in the loop over optical fibre from a payout spool, direct heritage from the fibre-tether work on CS-410, and kinder here with no launch shock and low speeds. The natural tier for inspection work.
Inertial navigation corrected by a Doppler velocity log with bottom lock. The capable tier and the export-sensitive one: navigation-grade INS and DVL are bought rather than built, and export-gated in their own right.
The acoustic modem aboard is a sparse check-in, abort and retask link at kilobit-class rates over kilometre-class ranges under good conditions, and nothing like a control loop. The vehicle is autonomous between contacts and the human holds mission authority at contacts. Acoustic emissions cost signature too, so silent-running segments with zero transmissions are a planned mode.
Nothing in the fleet transmits by radio and nothing guides on a link, and under water the rule costs nothing to hold: no receiver to jam, no satellite to lose, and one low-rate acoustic modem that can take a check-in or an abort at kilometre-class range and can never fly the vehicle. Human safety runs first: launch and recovery are a two-person shore-crew task on a dolly or a davit, the abort is mechanical, and a fault surfaces the vehicle. The four-domain doctrine is on the doctrine page.
The hull suits a buyer with water to know and a job at the far end of the leg. It suits nobody looking for a weapon.
If you plan route survey, harbour awareness, seabed monitoring or sovereign underwater capability, the conversation starts with the hull that carries everything else.
A conversation starts with a screened enquiry rather than a quotation. Bring the approach you have to know, the depth band, the current and the job at the far end of the leg; we bring the hull, the bay and the interface between them. Enquiries are screened, counsel comes first on any international transfer, and permits are taken per shipment. The Tier 3 navigator and the emplace-and-retrieve cartridge each carry export treatment of their own. Every design in the family is patent pending.
The Water family is the two hulls side by side, with KMM-01, the uncrewed clearance workboat, on the surface above them. CS-320 is the creeper this hull carries. Who it is for is the buyer's view. Government and allied procurement has its own door.
All Water family designs are patent pending; international transfer is export-controlled and subject to Canadian government permits taken per shipment, and nothing on this page is an offer.
