Multihull Capsize Recovery

The Problem

Multihulls can be light and fast but cruising ones must never capsize: They stay inverted.

Single hulled yachts can be self-righting and seaworthy, some large and small ones can survive storms and work to windward in gales but to be quick they need deep weighted keels and a wide hull. Built in buoyancy can avoid sinking but they are outperformed by multihulls.

Can a multihull be seaworthy? Could one work to windward in a gale or sail on when mid Atlantic breakers flatten her? After that happened in my Contessa, I got up (from safe mode: In a bunk with a lee cloth) looked round, nothing broken, self-steering working well, no need to heave to, back to the bunk. Another year we picked up an empty life raft south of Bermuda after hearing of an inverted catamaran. We found no name on the life raft but it was last serviced in Southampton.

The trimaran I had long ago was quick but never seaworthy: It filled floats at night and taught a nervous skipper to swear as gusts came through. Cruising catamarans I have sailed on since could not go to windward in bad conditions, wind and wave forces were more than the drive from a safe amount of sail. Running with trailing warps was the experienced owner’s alternative.

Could a proa be seaworthy? ”Cheers” long ago showed speed and knock-down survival. What about an automatic system for capsize recovery? A proa should be able to sink and then re-inflate the float to allow self-righting. With those two problems solved could a proa heave to in heavy weather and work slowly to windward like a Contessa? Probably not, but perhaps it would be as seaworthy as the offshore racing fleet.

Project Description

A cruising proa is being built with folding float beams and small accommodation but also low weight and cost. Performance and comfort should be high with a shallow draft allowing inshore work.

Pods help proas right themselves from a knock down. A slender hull with the cabin extended each side lets Dragonfly trimarans win races. A proa is longer for the same accommodation but has lower beam forces: A simpler structure.

Proas have been built in various configurations but my preference is for the main hull with accommodation and rig to leeward and the float carrying heavy items to windward.

Inversion following capsize or pitchpole would need air release from the float to sink it then air or CO2 to right the proa. This all works on paper but presents problems with different solutions from normal yachts.

Design Features

Knockdown recovery may be possible with the full sail up. If the position of the centre of gravity needs adjustment then the sail and topmast can be lowered. A high cabin structure raises the centre of buoyancy and a buoyant topmast prevents heeling beyond 90 degrees.

After a knockdown the sail will drag in the water and the wind push the hull to leeward. Wind on the float should then help to unstick the sail from the water.

Inversion is a risk in a serious capsize or pitchpole. Recovery should be possible with air released through the ventilation system of an inverted float. It must carry enough weight, such as anchor and engine, to sink it. An air replacement or CO2 system when the float is below the main hull should then right it. Lowering the rig can again correct the centre of gravity position.



Dragonfly trimarans demonstrate performance with folding wings.

The high cabin is beneficial for knockdown righting. It will be built as a 50mm box structure of thin plywood. Filled with light weight open cell foam this should have strength and give insulation as well as buoyancy if inverted

A short proa suffers, like wave penetrating multihulls, from low buoyancy at the ends. This is good for performance but bad for stability.

A proa with a heavy float can trail it and lift the bow with little loss of righting moment. This also adjustments the position of the float mounted centreboard for sailing balance.

A mast mounted to leeward makes knockdown recovery possible but on a proa gives a problem rigging a backstay. An unstayed mast avoids this but is heavy. A telescopic topmast will reduce both the weight aloft and windage when lowered. This will help performance in strong winds.

Normal masts are aluminium or carbon fibre but an unstayed mast needs strength not stiffness: If the top bends it eases gust loads. Stayed masts on the other hand depend on stiffness to resist rigging tension. Carbon fibre has many times the stiffness of GRP but less additional strength. Timber lies between them but is compatible with glass in strength and stiffness.

A simple conical layout with 1:100 taper makes a two section telescopic mast possible. A thin wall timber stave design with unidirectional epoxy glass coating has been built. Timber has a lower density than resin giving good stiffness for the weight and allows composite construction without a mould.

Unstayed masts are larger diameter and heavier but the absence of rigging allows a wrap round “Freedom” type rig which performs well except to windward.

External battens from jib luff to main leach allow a fully battened wrap-round mainsail but with a jib to give windward performance. The jib will be set inside a parrel system threaded round the mast to give a soft wing rig. The tension in the parrel sets the mainsail shape.

The rig can allow tacking when shunting would be too slow.

The full bending moment of the unstayed mast is taken on the beams so a ladder structure has been designed: The main beams will have structural handrails above them. Conventional netting

between the beams will give access to the float at sea but allow float folding in harbour. The ladder layout allows simple hinges at the beam ends.

Contribution to Nautical Science

This project is hoped to demonstrate feasibility of both knockdown and inversion recovery in a practical multihull. Safer cruising and harder racing should be possible if multihulls can survive squalls and breakers as well as monohulls.

Thin wall timber / GRP masts may have been researched long ago to give the right mix of timber/glass. Without that answer the large cost saving makes a composite timber stave/GRP telescopic mast attractive but risky. We will have to test ours.

The fully battened wrap round main and jib arrangement may give a useable soft wing rig.

Progress so far

A 9.5m proa is being built to test these ideas. It is based on an American (Madness) design from

Fyne Boat Kits in Cumbria. This is large enough to give realistic weight and buoyancy distribution but with cabin, mast and beam details all entirely different from the original design.

Many versions of the drawings and calculations were followed by building the rudders, centreboard, hulls, decks and mast in the last year. The cabin, beams, sails and all the fitting out and finishing may take another year.

Project plan/budgeting

I am retired, with my time spent designing and building the many details that are different on a proa from any normal sailing craft. The following items are still to be done:

The cabin structure with the cockpit, its bulkheads, and beam connections are underway. The float needs outside glass/epoxy.

The beams and float connections are barely started.

A masthead unit is needed to set the sails shown and a ghoster to the bow. It then all needs paint and antifouling.

Sails and battens to be made

Interior fitting out will be left until sailing trials have been done.

The original budget of £15,000 has risen to £17,700 including £3,600 in overheads but now excluding sails and outboard. The wrap round mainsail can be made from an old heavy weather spinnaker and our yuloh can provide calm weather propulsion.

Original To Current
Budget Spent Spend Budget
Plywood kit £3,750 £3,750 £3,750
Epoxy package £2,400 £2,400 £2,400
Timber £1,000 £964 £964
Glass fibre £2,000 £1,724 £1,724
Mast £1,000 £1,000 £1,000
Sails £1,000 £0 £0 Old spinnaker
Ropes £250 £164 £250 £414
Fittings £500 £250 £250
Furling £250 £0 £0 Omit
Outboard £850 £0 £0 Yuloh
Trailer £1,000 £1,000 £1,000
Towbar £1,000 £400 £400
s/s rudder tubes £678 £678
Fixings & brushes £470 £470
Gloves & overalls £397 £397
Paint £427 £200 £627
Machinery £820 £820
Hand Tools £540 £540
Shed £242 £242
Electrics £375 £375
Barn Use £500 £500
Labour £1,162 £1,162
£15,000 £15,613 £2,100 £17,713

I am funding this from my own savings. I am happy to let any of the details submitted be freely available and have little interest in commercial exploitation.

Contribution Requested

If the unstayed mast is satisfactory then properly made sails and battens would be valuable. If it is too heavy or not useable a conventional stayed mast from a Windermere One Design yacht (there is a pile of old ones at the club) would be purchased with conventional sails.

An outboard would also allow use at sea, especially getting out of our Cumbrian marinas.

Minimising Risks

Amateur boat design and construction are the major risks with error rectification a possible long delay. As a water industry design engineer, working to yacht codes (Principles of Yacht Design by Larssen) is not too different. Construction delay is not a cost problem with my own time and a borrowed barn. My own boats started with a 24’ trimaran rebuilt long ago and I have cruised a Contessa for the last 10 years.

We are proposing several novelties: Knockdown and inversion recovery, Unstayed telescopic mast,

Fully battened wrap-round main with a pocketed jib,

Float trailing for parking and to adjust both centreboard position and buoyancy

If these are identified as new features, each should work, with perhaps a 25% risk of failure. This gives a 30% chance of total success. The ones that do not work will need replacement:

Knockdown and inversion recovery both depend on weight and buoyancy distribution. Items are being weighed as work proceeds: The hull and the mast are both overweight. The cabin, beams and rig weights are still estimates; it is only when these are built that the calculations can be rerun and finally verified afloat.

Initial calculations indicated knockdown recovery was possible with the topmast up. Current weights show lowering it may be necessary.

Inversion may allow the buoyant topmast and battens to be lowered with a single rope release. If the float ventilation system traps water the float bilge pump pipe may need to be used for air evacuation. Once the float has sunk full righting only needs a CO2 system to bring it the right way up.

The unstayed telescopic mast needs a secure fit between the halves. Both have been built as 1:100 conical sections. A 1.5m overlap with 1:5 mating tapers of graphited epoxy at each end should allow full bending moment transfer. Only testing will show if these will pull up tight enough to prevent mast movement but allow lowering.

The lifting system has an 8:2 pulley system. Only testing will show the best arrangement. The mast has a 1:50 reverse taper where it sits into the support. Graphite epoxy mating surfaces again should allow the mast to be lifted out, but sit firmly without movement.

If the lifting system proves unreliable the mast could be rebuilt as a single spar. If it proves too heavy it may need replacing with a conventional stayed mast but with running lee shrouds set up on each tack. These would be set forward like inner forestays to support the rig if taken aback; it would not allow tacking in restricted water. The wide shroud angles will reduce mast loads but float

trailing would need careful rigging design.

The split jib with wrap round main combines two elements that work individually: The wrap round “Freedom” mainsail has been used for many years. The added external battens from jib luff to main leech and multiple sheets are for reefing and sail shape control. Slieve McGalliard’s split junk jib has careful sail shape design between the battens. Setting between the parrels may allow a simpler jib for this proa.

Mainsail shape is intended to be set by the battens and parrels with mast stiffness having less effect. Comparison of it in use with junk rig sails and their parrels may be instructive.

The main risk may be snagging of the parrels, especially when it is lowered over the mainmast head. Several years of development are expected to make it into a useable rig with wingsail efficiency but junk rig reefing ability.

Float folding allows the beam to be reduced from 5.5 to 3 metres for marina parking. Diagonal ropes between the beams can let the float trail to windward and lead downwind. Only when sailing can the effect on centre of effort of the rig and bow buoyancy be assessed. Low bow buoyance and pitching resistance may be a problem that trailing the float a small amount does not alleviate. A longer proa may be needed.

Design Intentions

Proa for serious


Different and fun:


Fun, fast, light, simple & cheap?

cruising Fast, light, simple & cheap
Knockdown recovery

& inversion recovery

Safety over wider operating window Righting from sails in water

Righting from inversion

Cabin & cockpit to leeward Crew survival when cruising

Sit on float when racing

Safe in cockpit for knockdown

Cabin escape route after pitchpole

Other weight to windward Inversion recovery and performance Weight to sink inverted float without slowing sailing
Unstayed telescopic mast Proa backstay layout

Lowering in gales

Lower without sticking Sail to windward in gale Enable righting
Folding beams Marina berthing

Float trailing for buoyancy and sailing balance

Robustness & pontoon height

Bow buoyancy if trailed

Sailing balance

Battened wrap round main with jib Performance

Ease of use

Racing success

Extended use in varied conditions

Rudders rotate 360o

& liftable

One used for fast sailing

Both for manoeuvring,

Control together or separately

Lift from cockpit with block system.

Foam filled cabin structure Insulation

Inversion recovery

Cabin condition

Buoyant volume to support boat

Fold-under centreboard Poor navigation Folding tests for grounding and rope snagging

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.