How Amphibious Vehicles Work: Buoyancy, Drive and Surf
An amphibious vehicle floats because its sealed hull displaces water equal to its own weight. Afloat, it moves by paddling with its tracks or wheels or, on most of the designs compared below, with a separate water drive such as propellers or waterjets. Most of the engineering is a compromise between land and water. A hull shaped for the ground is short, blunt and heavy, so it carries little spare buoyancy and cannot go much faster than 6 to 8 knots afloat. The published water speeds of the vehicles compared below range from about 4 miles per hour for a swimming Sherman tank to 40-plus knots for a hovercraft, and the type of hull, more than the engine or thrust device, explains most of that spread.
Buoyancy and displacement
The buoyant force on a floating body equals the weight of the fluid it displaces, which is Archimedes’ principle as OpenStax states it. A steel hull floats because it is mostly air: the sealed volume below the waterline displaces enough water to carry the structure, engine, crew and load. A vehicle sits at the depth where displaced weight equals total weight, so adding cargo makes it sink lower.
Three terms describe how much margin remains:
- Freeboard is the vertical distance from the waterline up to the main deck, the clearance left before the deck floods. On a vehicle the effective margin often ends lower, at the first hatch, intake or grille that is not watertight.
- Reserve buoyancy is the volume of watertight hull above the waterline. Marine Insight calls it the volume equivalent of freeboard and expresses it as the volume above the draft line divided by the total enclosed hull volume. It is the margin available if load increases or waves, spray or damage push the hull deeper.
- Trim is the fore-and-aft attitude of the hull. Weight placement and flooding at one end change it, and it governs how a vehicle meets a wave.
Land vehicles start with little reserve buoyancy. Armor, weapons and running gear add weight that the hull has to carry, so it floats deeper and keeps less watertight volume above the waterline. Openings (hatches, air intakes, exhaust grilles) are everywhere, and each one is a place where the effective freeboard ends. Most of the failures documented below involve water entering through an opening that was lower, or less sealed, than the sea conditions allowed.
The extreme case: the Sherman DD
The Sherman Duplex Drive shows displacement being added to a vehicle that was never a boat. Designer Nicholas Straussler’s own account, published by the Tank Museum, describes a rigid deck around the tank at mudguard level, tubular frames and a “very heavy rubberised canvas screen” raised by compressed-air tubes, each exerting about 600 lb of upward pressure for a total of about 12 tons. The screen turned the tank into a canvas-walled hull. HistoryNet describes the floating 34-ton tank sitting with the top of the raised screen only about a foot above the waterline, moving at “about four miles per hour”, and needing calm waters and light currents.
Freeboard that small left no margin on 6 June 1944. According to the French Ministry of Culture’s wreck survey of the Normandy landings, 29 of the 741st Tank Battalion’s 32 tanks were launched far from the Omaha coast and only two reached the shore. The U.S. Navy’s survey found the landing craft had to position further east because of a strong current, and the combination of swell and the design’s limits sank the tanks. At Utah Beach, which the page describes as fairly protected from the prevailing wind, 28 of 32 tanks reached the beach.
Why boxy hulls are slow: displacement hull speed
A boat pushing through water makes a wave that travels at the boat’s own speed. Faster waves are longer. When the wave length equals the waterline length, the hull sits between a crest at the bow and a trough under the stern, and going faster means climbing the face of its own wave. Good Old Boat’s explanation gives the rule of thumb: hull speed in knots is 1.34 times the square root of the waterline length in feet. The constant comes from the waves themselves. In deep water a wave’s speed depends on its length, as HyperPhysics sets out, and worked in knots and feet a wave L feet long travels at about 1.34 × √L knots, so the rule gives the speed at which the bow wave is exactly as long as the waterline. The stern usually starts to squat a little before that speed, and pushing past it takes more power than most displacement hulls have. Good Old Boat adds that the limit does not apply to long, slender hulls such as a multihull’s, whose wave drag is low.
| Waterline length | Hull speed (1.34 × √L) |
|---|---|
| 20 ft | 6.0 knots |
| 26 ft | 6.8 knots |
| 33 ft | 7.7 knots |
| 40 ft | 8.5 knots |
Doubling the length does not double the speed, because speed rises with the square root. A knot is one nautical mile per hour, about 1.15 statute mph, so a 26-ft hull tops out near 7.9 mph in these terms.
Worked example: the AAV
IMarEST’s account of the 2020 accident describes the vehicles involved as 26 ft (7.94 m) AAVs. BAE Systems lists the AAV7A1 with a 7-knot cruising speed at sea, driven by two 14,000 gpm waterjet pumps from a 400 hp turbocharged diesel V8. FAS’s reference page gives a maximum of 8.2 mph, about 7.1 knots, with 6 mph as the cruising figure. Overall length overstates waterline length, so the 6.8-knot hull speed from the table is a generous upper bound, yet the vehicle’s published speeds of 7 to 7.1 knots sit at or slightly above it. That is consistent with two jets and 400 hp being spent against the wave wall rather than converted into speed.
The same ceiling shows up across the displacement-type vehicles in the next table. Published speeds for the Sherman DD, LVT-4, DUKW, Amphicar, ACV and AAV all land between 3.5 and about 7 knots, regardless of whether tracks, propellers or jets provide the thrust. Escaping the limit takes a different kind of hull. A planing hull rises onto the surface, and an air cushion lifts the craft out of the water entirely.
Propulsion options compared
The table collects published water speeds from museums, manufacturers, a Navy fact file and reputable press. Speeds are as published, in the source’s unit, with a knot conversion added.
| Vehicle | Water propulsion | Published water speed | Source |
|---|---|---|---|
| LVT-4 Buffalo | Tracks with cleats that paddle | 7.5 mph (6.5 knots) | Tank Museum; paddling from the National WWII Museum |
| Sherman DD | Twin propellers driven from the idler wheels | About 4 mph (3.5 knots) | HistoryNet; drive in Tank Museum |
| DUKW | Propeller and rudder | 5 knots with 5,000 lb of cargo | National WWII Museum |
| Amphicar 770 | Two propellers; front wheels steer | 6.5 knots (7 mph) | Lane Motor Museum |
| AAV7A1 | Two waterjets | 7 knots cruise | BAE Systems |
| ACV-P | Not stated in the specification | More than 6 knots in open water | Army Recognition |
| Gibbs Quadski | Jet drive, wheels retract | 45 mph (about 39 knots), planing | NBC News |
| LCAC | Air cushion; propellers for thrust | 40-plus knots with a 60-ton payload | US Navy fact file; FAS |
| Screw-propelled amphibian | Rotating drums; hollow tubes float | Not stated in the source | The Drive |
Read down the speed column and the hull type matters more than the thrust device. Everything from the LVT to the ACV is a displacement hull. The Quadski and the LCAC are the only two that clear 10 knots, and they do it by changing how the hull meets the water. The Quadski’s wheels are retracted by a button and the vehicle reaches planing speed in as little as five seconds, according to NBC News’s review. The LCAC rides on its cushion.
Wheel or track paddling
Tracks and wheels can double as water propulsion, at an efficiency cost. Donald Roebling’s Alligator, the ancestor of the LVT family, used one drive for both land and water: the National WWII Museum describes “tracks with cleats that both gripped mud and paddled through water”. The 1935 prototype managed only 2 mph afloat, and its tracks broke after a few miles; the 1940 Navy contract version exceeded 9 mph in water. Track paddling is simple and needs no extra machinery, but it is slow, and the museum records track breakages persisting through the early redesigns. See Tracked Amphibious Vehicles for the LVT-4 and Sherman DD.
Propellers
The DUKW and Amphicar use propellers. The Lane Motor Museum records that the Amphicar’s rear-mounted 1,147 cc Triumph engine drives the rear wheels on land and two propellers in water, with the front wheels acting as rudders. The National WWII Museum describes the DUKW’s hull-shaped body with “a propeller screw and a rudder” and gives its loaded swimming speed as five knots. Wheeled Amphibious Vehicles compares the two.
Waterjets
A waterjet draws water through an intake, accelerates it with an impeller and expels it through a nozzle, and steering and reversing gear can redirect the flow. The Quadski and the AAV7A1 both use jets. How Waterjets Work explains intake, impeller and reverse.
Screws
Screw-propelled vehicles use rotating helical drums, which also provide buoyancy when hollow, to pull themselves across snow, mud and water. The Drive describes a Soviet design whose hollow tubes make it float, moving on Archimedean-screw principles. They remain specialist machines for marginal terrain. Screw-Propelled Vehicles covers the Fordson Snow Motor. Three were taken to Alaska to haul supplies for a 1926 transpolar flight attempt, and the Fountainhead Antique Auto Museum records that their revolving cylinders “didn’t work so well” in interior Alaska’s cold, dry snow and that the machines were very slow.
Air cushion
A hovercraft sidesteps displacement. The Hovercraft Museum explains that large fans blow air into a plenum chamber under the craft, a flexible skirt traps it, and propellers or additional fans provide thrust. How Hovercraft Work explains lift, thrust and skirts in more detail.
The surf zone and the beach exit
The hardest place for a swimming vehicle is the last few hundred meters. The National Park Service explains that as waves move into shallower water they slow, shorten and grow steeper until the crest “crashes forward creating surf”, and the band where this happens is the surf zone. Plunging breakers form where a sandbar, reef or rock structure abruptly changes the slope of the seabed: the crest outruns the base, rises and curls over the top. The vehicle ratings below are quoted in feet of plunging surf.
A vehicle in the surf zone is hit by water on its bow and then, as it turns, from the side or stern. It has to keep thrust and steering while its jets or propellers may be partly out of water, and it transfers from floating weight to ground contact. Rated limits are published, but they differ sharply from limits adopted after accidents:
- BAE lists the AAV7A1 as able to handle 10-foot plunging surf.
- Army Recognition’s ACV-P specification page says the ACV-P can operate in Sea State 3 and cross plunging surf up to 9 ft (2.7 m).
- On 19 July 2022 two ACVs had accidents in high surf off California, according to the Congressional Research Service’s ACV report. One tipped onto its side and another was disabled; crews returned safely. Amphibious operations were suspended. When open-ocean operations resumed on 23 September 2022, the interim limit was a significant breaker height of four feet.
The gap between a nine-foot published rating and a four-foot operating limit is the practical measure of the surf zone’s difficulty. CRS quotes Marine Corps leadership in January 2025: “There’s no vehicle that is inherently safe or unsafe. You have to operate within its parameters.” The same remarks placed the California accident in 12-foot plunging surf.
Sealing, bilge pumps and stability: four documented failures
Every watertight opening is a design decision, and pumps buy time rather than buoyancy. The cases below come from official investigations and reputable press.
| Date | Vehicle | What happened | Documented cause |
|---|---|---|---|
| 6 June 1944 | Sherman DD, Omaha Beach | 29 launched, 2 reached shore | Swell and strong current, flotation design (French Ministry of Culture) |
| 19 July 2018 | Stretch Duck 7 (DUKW) | Sank on Table Rock Lake; 17 of 31 aboard died | Operation after a severe thunderstorm warning; waves flooded a non-weathertight bow hatch; insufficient reserve buoyancy (NTSB) |
| 30 July 2020 | AAV, San Clemente Island | Sank; nine died | Maintenance failures, transmission oil loss, delayed evacuation, inadequate training (Marine Corps investigation) |
| 19 July 2022 | ACV, California surf | One tipped on its side | Operated in high surf; leadership later cited 12-foot plunging surf (CRS) |
The 2018 DUKW sinking: freeboard and reserve buoyancy
The National Transportation Safety Board’s release on its Stretch Duck 7 investigation records a 33-foot modified DUKW carrying 29 passengers and two crew when a derecho struck Table Rock Lake near Branson, Missouri, on 19 July 2018. Seventeen people died. The NTSB found waves flooded through a bow air intake hatch that was not weathertight, the vessel was built with low freeboard, an open hull and no subdivision or flotation, and the Coast Guard had not required sufficient reserve buoyancy for amphibious vessels. A fixed canopy and a closed side curtain impeded escape.
The NTSB’s case page reports 3-to-5-foot waves and a peak gust of 73 mph, and says Stretch Duck 54 entered the lake about two minutes before Stretch Duck 7 and “was able to exit the water after experiencing the severe weather”. Maritime Executive’s coverage of the findings adds that Stretch Duck 54 had more freeboard, more reserve buoyancy and a securable bow hatch. Two duck boats in the same storm had opposite outcomes, and the differences Maritime Executive lists are the margins described earlier.
The 2020 AAV sinking: bilge pumps, power and the chain behind them
On 30 July 2020, an AAV returning to USS Somerset after a training raid sank off San Clemente Island, California. Eight Marines and one Navy hospitalman died, Military Times reports. The Marine Corps’ command investigation, summarized by IMarEST, found:
- sea conditions rougher than forecast and beyond the mission’s no-go standards;
- water entering through the hatches, an improperly installed headlight and the intake and exhaust grilles;
- a transmission failure after oil ran low, which stopped the vehicle in the water. Military Times traces the oil loss to a loose transmission drain line;
- rising water reaching the generator, which knocked out the bilge pumps, radio and emergency lighting;
- no safety boats deployed, and a vehicle that had been sinking for about 45 minutes before help arrived;
- a final swamping through an open hatch after the two AAVs collided, the vehicle rising nose-high and sinking stern first.
The investigation’s main causes were maintenance failures from disregarding procedures, crewmen not evacuating personnel when they should have, and improper training of embarked personnel on AAV safety. Military Times adds that a second investigation, led by Lt. Gen. Carl Mundy, found maintenance in 3rd Assault Amphibian Battalion had declined from 2017 to 2019. Maritime Executive reported the Marine Corps would reinstate waterborne egress breathing devices that had been discontinued in 2015.
The mechanism is instructive. A bilge pump removes water, but on this AAV the pumps depended on electrical power from the generator. When the generator flooded, the pumps stopped, the water kept rising, and nothing remained between the vehicle and the sea except freeboard and time.
What the principles mean when reading a specification
- Rated surf is a ceiling. A published plunging-surf figure describes the vehicle’s design capability. Operating limits set after accidents were far lower in the ACV case.
- Water speed is a hull-length question. At 6 to 7 knots, a vehicle of 26 to 33 ft is already close to its displacement hull speed; extra engine power buys little.
- Freeboard predicts survival in weather. The Stretch Duck pair shows the difference between a vehicle that exits a storm and one that sinks.
- Support systems fail in sequence. Pumps, radios and emergency lighting depend on power that rising water removes.
Sources
- OpenStax: Archimedes’ principle
- Marine Insight: What is reserve buoyancy in ships
- Tank Museum: How to invent a DD tank (Straussler)
- Tank Museum: LVT-4 Buffalo
- HistoryNet: Duplex drive tank
- French Ministry of Culture: Floating tanks at the Normandy landings
- Good Old Boat: Displacement hull speed 101
- HyperPhysics (Georgia State University): Water waves and wave velocity
- NOAA: Nautical miles and knots
- BAE Systems: Assault Amphibious Vehicle
- FAS: AAVP7A1 specifications
- Army Recognition: ACV-P
- National WWII Museum: From hurricane rescue to armored landing (LVT)
- National WWII Museum: Over-the-shore logistics of D-Day (DUKW)
- Lane Motor Museum: 1964 Amphicar 770
- NBC News: Gibbs Quadski review
- US Navy fact file: LCAC
- FAS: Landing Craft, Air Cushion (LCAC)
- The Drive: Screw-driven amphibious truck
- Fountainhead Antique Auto Museum: Snow Devils (Fordson Snow Motors in Alaska)
- Hovercraft Museum: How hovercraft work
- National Park Service: Coastal processes, waves
- Congressional Research Service: The Marine Corps’ Amphibious Combat Vehicle
- NTSB: Stretch Duck 7 release (April 2020)
- NTSB: Stretch Duck 7 investigation page
- Maritime Executive: NTSB findings on the DUKW sinking
- Military Times: New investigation reveals more details on fatal AAV accident
- IMarEST: Deadly failings that caused US Marine Corps tragedy
- Maritime Executive: Marine Corps promises safety changes after AAV accident