What is it, where is everything, and could you live in it?
The ship
The arrangement and the interior: a cutaway, an inboard profile, four sections and a plan of every room with what is in it. All drawn from the same stations, footprints and masses the budget integrated.
The ship
Not a concept render. Every box below is placed and sized from the same station, extent, width and height the mass statement integrated to get its volume, and every one of those volumes went into the lift figure and the habitability check. The fins are the planform the yaw stability was computed from. The gas cells occupy exactly the volume the buoyancy came from, minus the keel corridor they give up.
The near half of the hull and cover removed. Cells, keel corridor, gondola and every compartment.
Drawing this made the ship 25 metres longer
The baseline was 90 m for as long as the mass budget was a fraction. Giving the compartments, the machinery, the tanks and the array real positions and real masses turned it into a statement, and the statement was that 90 m comes out 5,495 kg heavy at the fill fraction that gives it pressure height. It closes at 104.8 m and needs 112.3 m to carry the 15 percent growth that every preliminary mass estimate suffers between concept and first flight.
An aeroplane that comes out heavy loses range and still flies. An airship has no such trade: the buoyancy is fixed by the envelope. A design that closes exactly is a design that will not close.
Where everything is
The drawing an airship is actually designed on. Every habitable space is below the gas cells, because a leak rises: the gondola hangs under the hull and the keel corridor runs along its bottom, and nothing a person occupies is inside the cell volume. The engine is aft and low because the exhaust must leave below and downstream of the whole envelope, which costs trim and is worth it.
| Compartment | Deck | Station | Volume | Mass |
|---|---|---|---|---|
| Nav station and helm | gondola | 0.23 | 22 m³ | 210 kg |
| Saloon | gondola | 0.26 | 37 m³ | 240 kg |
| Galley | gondola | 0.30 | 19 m³ | 260 kg |
| Head and washroom | gondola | 0.32 | 11 m³ | 180 kg |
| Sleeping cabin | gondola | 0.35 | 26 m³ | 220 kg |
| Crew and personal effects | gondola | 0.29 | 4 m³ | 220 kg |
| Mooring cone, anchor winch, drogue and sea anchor | keel | 0.05 | 43 m³ | 320 kg |
| Food stores | keel | 0.20 | 19 m³ | 584 kg |
| Consumables and spares | keel | 0.26 | 12 m³ | 356 kg |
| Forward water tank | keel | 0.22 | 10 m³ | 1,250 kg |
| Systems bay: fuel cell, electrolyzer, battery | keel | 0.44 | 45 m³ | 1,150 kg |
| Aft water tank | keel | 0.50 | 10 m³ | 1,250 kg |
| Hydrocarbon reserve | keel | 0.35 | 7 m³ | 1,200 kg |
| Hydrogen COPV storage | keel | 0.67 | 21 m³ | 400 kg |
| Seawater ballast bladder and pump | keel | 0.44 | 3 m³ | 110 kg |
| Exercise and wet space | keel | 0.62 | 18 m³ | 180 kg |
| Workshop | keel | 0.73 | 35 m³ | 300 kg |
| Engine and generator bay | keel | 0.81 | 30 m³ | 270 kg |
Sections
Looking forward, at the four stations where the arrangement changes character. The gas cell fills the section above the keel; nothing a person occupies is inside it.
Mass by group
| structure | 9,361 kg | 36.9% |
| consumable | 4,640 kg | 18.3% |
| energy | 4,058 kg | 16.0% |
| gas | 3,640 kg | 14.3% |
| machinery | 1,890 kg | 7.4% |
| habitat | 1,590 kg | 6.3% |
| crew | 220 kg | 0.9% |
| Gross weight | 25,398 kg | |
| Gross lift, design altitude | 32,093 kg |
Lift is computed at both ends of the operating band and the binding one is used. At sea level the cells are at 85% fill on dense air; at the design altitude they have expanded to fill completely on thin air, which is what pressure height means.
What the arrangement has to obey
PASSNo enclosed or habitable volume above or adjacent to a gas cell.
Every habitable space is in the gondola below the hull or in the keel corridor below every cell. Hydrogen that escapes a cell rises away from all of them.
PASSHabitable spaces are continuously ventilated, never sealed.
No habitable compartment is sealed. Ventilation is what keeps a slow leak below a quarter of the lower flammability limit.
PASSConfined runs are narrower than the 150 mm critical passage width, shorter than the run-up distance, or open at both ends.
1100 mm by 104 m keel corridor. Open at both ends to the free stream, so hydrogen cannot accumulate to a flammable concentration.
PASSEngine exhaust leaves below and downstream of the entire gas envelope.
Exhaust exits at station 0.94, aft of the cell block at 0.92, and 13.0 m below the axis. This constraint is what pins the machinery aft, and it costs real trim to obey.
PASSCentre of gravity at least 2.4 m below the centre of buoyancy.
3.10 m of pendulum lever, 26 percent of hull radius. This is the entire static stability of the vehicle: there is no other restoring moment in pitch or roll.
PASSCentre of gravity within 2 percent of length of the centre of buoyancy.
Centre of gravity at 52.3 m, centre of buoyancy at 51.7 m: 0.52 percent of length aft. Corrected by moving water between keel tanks, which is why the water is distributed rather than in one drum.
PASSGross lift at the design fill fraction exceeds gross weight.
32093 kg of lift against 25398 kg of ship: 6695 kg spare.
PASSLift margin at least 15 percent of gross weight, to absorb preliminary-estimate growth.
6695 kg of margin on 25398 kg, 26.4 percent. Preliminary mass estimates grow 10 to 20 percent between concept and first flight, every time, and an airship has no way to trade payload for the difference: it either lifts or it does not. A design that closes exactly is a design that will not close.
WARNEmpty weight per cubic metre of gas at or below the Hindenburg's 0.59 kg/m3.
0.599 kg/m3 empty, against a historical fleet band of 0.505 to 0.79, all of it duralumin. Compared on EMPTY weight because that is what the fleet table records; consumables and crew are excluded from both sides. This ship also carries a photovoltaic array and a habitat that no ship in that table did, so beating the band is a harder claim than it looks.
PASSEvery keel bay fits inside the keel corridor envelope, longitudinally and by volume.
286 m3 of bays inside a 1345 m3 corridor running station 0.015 to 0.9. That corridor is subtracted from the gas volume, so the space you live in is paid for in lift rather than assumed free.
PASSBallast transfer between the two water tanks can correct the standing trim offset.
468 kg has to move 33 m to bring the centre of gravity onto the centre of buoyancy, and 1250 kg can. The margin is what absorbs stores burning off over the year, which is a nose-up trend of its own.
PASSEvery keel bay fits inside the hull section at both of its ends.
Every bay clears the hull skin over its full length. The hull tapers and the bays do not, so this is a real constraint rather than a formality: it is what stops the corridor running out past the cover near the tail.
PASSFin area at least 1.3 times the minimum that balances the Munk moment.
351 m2 of YAW-EFFECTIVE fin, summed as cos-squared over the four surfaces of the 702 m2 tail at its 45 degree roll offset, against a 226 m2 minimum on a 54.5 m arm: a static margin of 1.55. The lift slope is 3.38 per radian, from an exposed aspect ratio of 0.87 doubled by the hull acting as an end plate, knocked down 15 percent for the boundary layer the tail sits in, and multiplied by 1.68 for the load the fin induces on the hull itself, which restores at the same arm. The Munk moment is certain and the fin effectiveness is not, so the margin is the honest part of this number.
WARNThe propulsors lift the landing trim with one of them stopped, without pitching the ship.
4 ducted propulsors at an effective 6.0 m make 8.3 kN on 72 kW, which is a 842 kg thrust budget. THAT IS NOT WHAT IT LIFTS. Every unit is aft of the centre of gravity, the aft pair by 42 m, so tilting them up pitches the ship as much as it lifts it: at full authority the fleet makes 153 kN m of nose-down moment and settles 9.0 degrees down on the pendulum, which is an attitude and not a hover. Held LEVEL on the pendulum alone the fleet lifts 413 kg, 53 percent of the budget, and 207 kg after the worst single loss, which is port-mid rather than any unit: the mid pair does the lifting and the aft pair is a trim tab. Against a 550 kg trim that DOES NOT HOLD. The moment can be bought back with ballast: 474 kg moved 33 m aft against 405 kN m of transfer authority, and then the budget is available again: 575 kg after the worst loss, which holds. SO THE VECTORED LANDING IS A BALLAST-TRIMMED MANOEUVRE and the trim has to be staged aft before the descent begins, not called for during it. THE LANDING TRIM IS SET BY THIS CASE and not by the sea state: a trim the vehicle can only leave with every propulsor running turns one failure into a vehicle that cannot take off again. Thrust goes as the TWO-THIRDS power of diameter and the two-thirds power of power, equally, and the duct is worth 1.18 at equal power rather than the factor of two the folklore quotes, which is the area effect at fixed induced velocity and above the ideal-flow ceiling.
PASSAt least 19 m3 of habitable volume per person for a mission past the Celentano asymptote.
169 m3 across 2 crew, 85 m3 each. Tolerable is 5, the performance limit is 10, and 19 is where more volume stops helping. For a year, designing to tolerable is how you get a crew that stops maintaining the ship.
PASSThe daily superheat lift excursion is answered, by a trim that swallows it or by a ballast loop that tracks it.
17.1 K of diurnal swing, 10.0 K of superheat at 100 percent cloud and 7.0 K of supercooling under a clear night sky, moves lift by 2112 kg, which is 3.8 times the 550 kg the vehicle rests on water at, so NO PASSIVE WATER-CONTACT DEVICE CAN BE SIZED FOR IT: a relief valve set for the trim is bypassed at the night load and useless at the day load. The arrangement answers it with 2.6 m3 of seawater bladder against the 2.1 m3 the swing needs, pumped at 70 kg a minute on 115 W. THE OCEAN IS THE BALLAST and moving water costs about a three-thousandth of what compressing lifting gas does. It works only afloat, which is where the problem is.
PASSThe lowest point of the tail clears the crest of a sea state 3 wave, so the vehicle can float without immersing a fin.
The lower fin tip is 12.3 m below the hull axis against a 14.7 m keel, clearing still water by 2.5 m and the 0.81 m crest of a sea state 3 wave by 1.6 m.
PASSAt least one laterally separated pair with full 90 degree vectoring.
Differential thrust across the mid pair gives yaw authority at zero airspeed, which is the entire control system during mooring, during a water landing, and any time the fins have no flow over them.
PASSPropeller discs clear the hull surface.
The tightest unit's INBOARD tip clears the hull at its own station by 0.7 m. The outrigger has to carry that offset, and its bending moment is what sets the mount mass.
The four parts that are not an airship's
A conventional rigid airship is a hull, a keel, a gondola, an engine and a tail. This one has four things that are not on that list, and each of them exists because a number said so rather than because it looked right.
Wings, 40 m span
NOT for efficiency. There is no speed at which it does. The trade a wing normally wins is taking weight off something that pays induced drag to carry it, and buoyancy does not: the gas carries the whole weight at zero speed and for free. So the wing adds profile and induced drag at every airspeed and takes nothing away, and no crossover exists at any speed the vehicle could reach. It is for carrying: 1,147 kg of extra weight at 14 m/s on the power the vehicle already has, for 342 kg of structure.
Outboard rather than a fatter hull, because induced drag goes as span squared and not as area, so area at the extremities is worth about ten times area in the envelope. At the centre of buoyancy, so the lift split can change without a trim excursion. And it costs 1.7 percent of the station-keeping power every hour it is not carrying anything, which is the argument for folding it.
Centreboard, 18 m²
The part that decides whether boat mode exists. Holding a heading and travelling along it are different things: at an angle to the wind the envelope makes an enormous side force and a hull sitting centimetres into the water resists almost none of it.
On bare hulls the usable cone from dead upwind is five degrees, because the vehicle points where the fins say and goes where the wind says. At this area it is the whole compass, and no amount of thrust substitutes: the speed through the water is identical either way.
A 2.1 m³ seawater bladder and a 115 W pump
17.1 kelvin of diurnal swing moves the lift by 2,112 kg against a 550 kg landing trim, so the vehicle takes itself off by mid-afternoon and presses that onto its gear before dawn. No passive device can be sized for a load that swings by 3.8 times the trim, twice a day.
That swing is computed rather than assumed, and it is not the number this project used to carry. It was graded against a flat 20 K of superheat, described as the standard figure for a dark envelope. This envelope is not dark: the cover is reflective by design and only the array is optically black, so the clear-sky peak is nearer 10.0 K. The half that was missing entirely is supercooling. On a clear night the sky radiates as though it were 25 K colder than the air, the hull follows it down, and the gas goes 7.0 K BELOW ambient. That is the excursion that puts the ship on its float, and it adds to the superheat rather than cancelling it.
The worst superheat case is also not the obvious one. It arrives at 100 percent cloud, not a clear sky. Cloud blocks the night-time radiative loss for twenty-four hours a day while only cutting the solar gain for twelve, so the daily mean temperature of the envelope rises even as its peak falls, and a gas volume with a twenty-minute time constant follows the mean rather than the peak.
That result is bounded by something the model does not do: ambient air temperature is an input and does not respond to cloud, when in reality the same blanket that keeps the hull warm keeps the air warm. So the honest reading is that cloud does not relieve the thermal problem, rather than that it makes it much worse. The swing, which is what the ballast loop actually has to track, moves less than a tenth across the whole range of skies.
A vehicle afloat is sitting on unlimited ballast, and moving water costs about a three-thousandth of what compressing lifting gas does. The pump moves 70 kg a minute for 0.27% of what the habitat draws in a day, and it tracks the superheat instead of fighting it. It works only afloat: in the air there is nothing to pump from.
Four ducted propulsors at 6.0 m
Sized to lift the vehicle’s residual heaviness rather than its weight, which on a buoyant ship is a few percent and two orders of magnitude less thrust than a helicopter of the same mass would need. They make a 842 kg thrust budget on 72 kW.
They do not lift that. Where they sit decides it: all four are aft of the centre of gravity, the aft pair by forty-two metres, so tilting them up pitches the ship as hard as it lifts it. Held level the fleet lifts 413 kg, 53 percent of the budget, and 207 kg after losing port-mid, which is the loss that costs most because the mid pair does the lifting and the aft pair is a trim tab.
Ducted because the shroud is worth 1.18 at equal power, not the factor of two the folklore quotes: two is the area effect at fixed induced velocity, and it is above the ideal-flow ceiling. 6.0 m because thrust goes as the TWO-thirds power of diameter and the two-thirds power of installed power, equally, so a bigger disc and a bigger motor buy the same percent and which one to fit is a mass and drag question.
Living in it
The arrangement gives each room a size and a mass. That is enough to check whether the vehicle flies and not nearly enough to check whether a person can live in it. A galley with 18 cubic metres and 260 kilograms is a number; a galley with a two-zone induction hob, a 120 litre fridge and 1.4 metres of worktop is a room.
What the volume figure does not tell you
8.4 m3 of lockers in the accommodation against 1.0 m3 of stores in daily use. The rest lives in the keel, which is a walk rather than a problem.
The fitout adds up to 866 kg against the 1110 kg the arrangement carries for the same rooms, a ratio of 0.78. Two routes to the same number and they agree.
Sleeps 3 for a crew of 2. The spare is the settee, and a vehicle carrying two people for a year needs somewhere for one of them to sleep when the other is ill, working, or simply awake at the wrong time.
2 independent ways out of the accommodation: the passage forward into the keel, and hatches from the saloon and the sleeping cabin.