Multihull Structure Thoughts

Carrying on the theme of tortured plywood sided cats the Selway Fisher’s 254 catamaran has a long waterline and light weight construction using tortured ply. The pre-shaped sides are simply stitched to the bottom and epoxied at a set angle. The 6mm sides are then pulled up producing a round sectional shape. This gives a very light but stiff hull which simply uses bunk tops and shelves as longitudinal stiffening. Four berths can be fitted into the hulls along with basic cooking and toilet facilities. The headroom is only 4 foot. The boat was designed as a racing catamaran but could be a fast cruiser. LOA 26’3’’ (8.00m); Beam O.A. 16’9’’ (5.1m); Beam (Hull) 3’6’’ (1.07m); 1’9’’ (0.53m); Sail Area 375 sq.ft (34.5 sq.m.); Disp. Approx 1600 lbs. (725 kg.); Payload capacity 1600 lbs. (725 kg.). The boat is an older design but should be reasonably fast. This design is still available at a cost of 250 british pounds.

Hull Shape. U shaped with narrow flat bottom plank. Construction Method. Stitch and tape. Major plywood requirements for hull.

14 x 3mm sheets
32 x 6mm sheets
18 x 9mm sheets

 

There is a group of tri’s that have custom main hulls and use Tornado catamaran hulls for floats. Tornado rigs are often used as the motive power. The first boat shown is the original Bieker Brown variation which had a main hull built with ply sides and WRC main hull bottom. The boat was 25 x 17.5 foot weighed 950 lbs with 380 sq foot of sail. The cross arms were carbon fibre and weighed 65 lbs per arm. The boat could sleep 2. Kelsall did a version called Typhoon shown and many other variations as shown by Yet Another Trinado project which is on the following web site. Drawings - www.trinardo.com http://www.trinardo.com/y.a.t-yetanothertrinadodrawings Also look at his material section for some interesting thoughts on foam.

 

The Livewire 28 is a very modern Grainger design aimed at speed and fun. It will meet the NZ 8.5 meter racing rules and will scare the heck out of most sailors on a beam reach in strong winds. Just pure fun. Look at the hull lines and again notice the near flat bottom and sides. At lot of this boat could be built as a flat panel construction. The central spine is a rigid structure that allows you to crank on the forestay and mainsheet tension for optimum upwind performance. It serves as a secure storage pod for anchor and chain, and other items that need to readily accessible from the deck. It also provides firm mounting base for the outboard auxiliary while concentrating the weight close to the centre of buoyancy. Graingers web site gives more details.

 

Strip plank cedar construction is used in many multihulls but if falling out of favour as timber prices are increasing and cheaper sources of suitable foam are being found (China brands). But Catsketchers comment about the amount of glass he used on his 38 foot cat made me look for some testing of WRC glass structures. An article in Professional boatbuilder magazine proved informative. Basically said you can calculate a WRC glass hulls requirements a similar way as foam core glass. Basically the thinner the foam core or WRC planks require more glass which improves knock resistance. The thicker the core you can have thinner glass the less knock resistance. The hull panel tests were done for a mono called Graywolf 40 x 13 foot displacing 13500 lbs with 1100 sq ft of sail and 4500 lbs ballast also water ballast 2250 lbs.

The designer calculated the maximum hydrostatic load on the hull was going to be 4.5 PSI (pounds per square inch) and the maximum deflection should be Length of boat (or in this case test panel length) divide by 100. Test panels were 600 mm square so a maximum deflection of 6 mm in the panel.

Graywolf had 8 panels tested with a target weight of 1.9 lbs/ft. Actual weights varied by up to 0.4 lbs/ft due to fabrics not weighing as advertised and excess resin in some panels. That’s a potential 20% weight gain over a hull. Tests are at https://pbbackissues.advanced-pub.com/?issueID=69&pageID=158

Base case was 32 mm WRC with 10 oz on outside, the panel deflected 6 mm with 9 PSI The panel deflected 3 mm at 4.5 psi. The preferred panel was inside 2 x 7781 330 gsm cloth, WRC strip 25 mm, DB 120 450 gsm. Deflected 6 mm with 12.5 PSI with 2.2 mm deflection at 4.5 PSI. Best was an 18 mm duracore with 2 x 3mm timber veneers sheathed with 330 sm cloth outside that required 14.5 PSI to deflect 6 mm and defected 1.8 mm at 4.5 PSI.

Actual Gray Wolf build was outside 2 x 1576 e glass in epoxy, 25 mm wrc strip, inside 2 x 200 gsm uni s glass in epoxy . Sheer clamps 70 x 32 mm. Deck 12 mm ply with 330 gsm glass covering. Hull to deck joint external 3 x 120 DB 450 gsm. Deck beams 32 x 50 mm.

Translation of the above. Glass WRC if properly designed can exceed the requirements of designer but testing of panels will give an excellent guide as to what can be done. In the real world understand what comparable boats are built of, make a test panel, get results then build test panels to see if they match the known outcome. As Rob Denny wrote in 239 on this thread. “There are lots of test results for cored panels, but of importance to you is not the comparative stiffness so much as whether your panels are up to scratch. In particular, that resin is minimised and the core/cloth bond is adequate. You don't need a load cell to compare stiffnesses. Lay up strips 500mm long x 50mm wide (18" x 2") of the materials you want to test (fibre alignment and direction is important) and support them on a bench with 75% of them hanging over the edge. Then hang weights on the unsupported edge and measure deflections. You can also load them to breaking point, but this is not a load case that boats see. Weigh them, try and peel the laminate off, bash them with a hammer. Then saw them into pieces and see what has happened to the core and interface. If you do set up a hydraulic press for panels, make sure the load is evenly distributed or you will be testing resistance to drying out on a rock, not panel stiffness.”

 

Thanks - I can read the article's tables easily. What weight is the 1576 e glass?
I like the idea that the builder made test pieces to check the properties of actual laminates. I hope he included filler too.

 

Catsketcher At web site MatWeb - The Online Materials Information Resource http://www.matweb.com/search/datasheet.aspx?matguid=7bd56d892952403a89a13976780a7d80&ckck=1 gives the detail of 1576 e glass = 10.6 oz per square yard = approx 400 gsm. It looks like a semi unidirectional fabric. The actual web site at Hexcel Technical Data Sheets http://www.matweb.com/search/GetMatlsByManufacturer.aspx?navletter=H&manID=81&manname=Hexcel lists over 700 fabrics and other materials details.

 

A Russian catamaran developed on an open web site is looking interesting. The boat has evolved with the inputs of the web site users who range from workers to engineers to professional boatbuilders. The boat they have come to is a 6.5 meter cat. The boat suites the needs and affordability of the members of the site. The boat is plywood, 6 mm skins but no full structural details yet. The lines are realistic but some of the “crew” in the drawings are optimistic. A sample page of the Russian web site (I have no hacking problems from the site) is

Давайте нарисуем парусный многокорпусник для самостроя - Страница 96 - Проекты и чертежи - Кают-Компания "Катера и Яхты" https://forum.katera.ru/index.php?/topic/43381-davaite-narisuem-parusnyi-mnogokorpusnik-dlia/page-96

Also included is photo’s of one person’s build of a folding catamaran. The process looks OK but I do not know of the structural integrity.

 

An interesting piece on hardware placement onto boats. For those who bond eg stainless steel bolts into wooden substructures using epoxy the following article could be useful. https://pbbackissues.advanced-pub.com/?issueID=15&pageID=23 The detail of some of the failures is interesting. Rarely is it the bonding of the metal with epoxy but often it is the corrosion of the metal part that causes the failures. This happened on Newicks Rogue Wave tri when the heads of the bolts holding down the mainsheet track started to shear after 15 years. Also the bolts on a 125 foot wood wind turbine blade started to shear due to corrosion. It would appear that a salt water environment without oxygen can be a concern.

 

I have been using potted epoxy fasteners for - heck - over 20 years, I must be getting on. I have had no failures of any fasteners. There are a few things I do differently from the article.
Nowadays I try to stitch fittings on in composite rather than bond steel fittings onto a composite body. A good case in point for multis is with tramp attachments. In the bad old days you would screw on a track with heaps of fasteners. Now we glue a piece of PVC pipe on and glass over with 600gm db and cut slots where required. No fasteners. Even my genoa "track" is the same - no track just PVC and stitched on to a timber piece underneath with uni rovings. Uni rovings are great and a few turns can hold a huge amount of load - me engine nacelles are held on with about 2 sets of 8-10 wraps of the rovings that go into a chopper gun. Very strong, no point load, no fasteners, no big bucks required.
My winches get put in with nuts but still are bedded in wet epoxy. This can cause issues with removal. I had a manual anchor windlass I wanted to replace with a nice new electric one (a new toy for the boat every couple of years). I seriously thought I would have to cut out the deck or chop up the winch in place with a grinder. But I gave removing it a go first.
First I bought a small gas torch from Bunnings. Then I got a square shaft very large screwdriver that fitted the slot bolts perfectly. Then I gave the fastener about 15-20 seconds of flame, right on its head. Quickly on with the screwdriver with a spanner on the square shank and out she came - every single bolt. They were in at least 75mm of ply and cedar backing plate. I felt like I had a major win. Gotta love epoxy.

 

Wharram cats had originally solid beams then moved forward to a variety of crossbeam construction techniques. His later crossbeam offering is for his Tane or Tanenui design and is an upgrade for the original solid beams to the Tiki style crossbeams. The original solid beams often had rot problems if not well built or built of unsuitable timber like radiata pine. This beam is suitable for open bridgedeck cats with righting moments of 20,000 foot lbs. EG 4000 lbs displacement with the centreline of the hulls 10 foot apart. The Tane has 3 main crossbeams. Full plans are available from the Wharram site.

 

A generic discussion on composites. All the information is based on fact but is generalised. For a start we will compare E glass polyester resin to a high strength carbon fibre epoxy laminate (single skin laminates).

For a given compressive strength 1000 gsm E glass laminate could be replaced by a 400 gsm high strength carbon laminate. For equal laminate tensile strength 1600 gsm of e-glass laminate could be replaced by 400 gsm HS carbon laminate. For equal laminate stiffness a 2700 gsm e glass laminate could be replaced by a 400 gsm HS carbon laminate. As you can see a high strength carbon fibre laminate can be significantly lighter BUT thinner single skin carbon laminates replacing thicker E glass laminates will lose flexural stiffness unless you do a sandwich where the skins are load in pure tension or compression – provided the skins are thick enough to resist buckling.

Resin fracture toughness (peel splits) of glass fabric laminates shows that orthothalic polyester resins are the base case, isothalic polyester are 2 as tough, epoxy laminates are 3 times tougher. Special vinylesters are 6 times as tough. This is all based on the laminates have the same glass fibres used in the matrix.

Carbon fibre is flavour of the month but needs to be treated with respect. The basic issue of carbon is its low elongation (less than 2 %). When a load goes on carbon laminate, carbon is capable of handling the load to about 90% plus of its load then it goes bang. If a laminate of fiberglass or Kevlar gets a load it starts bending or cracking at about 70% of its load and you get warnings of failure. This is a problem with any high modulus product. The other issue is the selection of the resin that is to be used with carbon. Carbon is very stiff with low elongation so why would you use a resin that suits e glass that stretches 6 or 7%. The resin you choose has to be matched to the material you use. The resin must have high tensile strength and must suit the application method. EG the resin used in vacuum bagging can be more viscous than resin used in vacuum infusion.

Foams. Airex is a linear PVC foam that can be worked to about 140 F (Fahrenheit). Divinycell is a crosslink PVC foam can be worked to about 160 F. Corecell is a linear polyurethane. Airex R63 great for hull due to its flexibility but is bad for decks as it can sag in high sun temperatures. Shear strength of cores related to density, a 2 lbs/cu ft foam shears at 50 psi, 6 lbs/cu ft shears at 200 psi and 10 lbs/cu ft shears at 400 psi. Another test has shown a skin core bond strength of Airex R-82 is 12 psi and HT 70 Divinycell is 8 PSI. Corecell can outgassing in high temperature post cures unless you use foams that can be post cured at higher temperatures.

 

Sebago is a 45 x 27 foot racing catamaran displacing 7000 lbs and carrying 1000 sq foot of sail that was the “1st self-righting hull design. 4th in class in 1984 OSTAR. 1st in class, 7th overall in 1985 Round Britain” that is now converted to day charter boat. The interesting part is the self-righting capability. This was achieved by creating a “sinker” hull which had a different build structure to the “floater” hull. The boat when capsized then floods the sinker hull which sinks to 90 degrees. Then the sunk hull is pumped out and hopefully rises to the surface. Its an idea that was never tested (the thought of trying to hand pump 3000 litres of water out of a hull in windy weather and waves is beyond my comprehension). But the boat was built according to the principle and raced successfully. The structure of the starboard hull was 3(300 gsm unidirectional e glass) 450 gsm e glass double bias on both sides of 25 mm balsa. The port hull 3(300 gsm unidirectional e glass) 450 gsm e glass double bias on either side of 3 mm okume ply. The port hull also has stringers and a partial 9 mm plywood shelf to provide the rigidity lost by the thinner hull core.

 

Colt Cars was a racing trimaran raced by Rob James. The tri was 60 x 40.5 foot that displaced 11,000 lbs and carried 2000 square foot of sail. The complete hull shell and crossbeams weighed 6000 lbs. The additional weight is the 80-foot rig, engine, equipment, food and people. The hull is 5 layers of hybrid 50/50 carbon/kevlar 240 gsm unidirectional fabric either side of a 20 mm Nomex core. The deck is the same basic structure. All layups are vacuum bagged with epoxy and if possible laid in one continuous run with the 5 layers at once. That would be a major organisational issue. The boat was built in 1980 prior to resin infusion. The carbon fibre crossbeams are designed to deflect a maximum of 25 mm over the 12.5 meters. The crossbeam structure was designed for the maximum righting moment of the tri then a safety factor of 3 to 5 is applied. According to Ron Holland the designer the rig loads will be easily carried by a crossbeam designed for maximum righting moment. The 600 x 1100 mm rudder at 24 knots can put up to 300,000 ft/lbs stress on the stock, result solid stainless steel 75 mm in diameter tapered at the ends. They analysed the centreboard construction techniques and came to the following conclusions. A solid cedar board would weigh 340 lbs, a laminated birch plywood board with a central timber spar would way 190 lbs, an aluminium board about 200 lbs and Honeycomb composite about 100 lbs. They choose the birch plywood board. The boat was moderately successful in racing as the hull designs were not optimal for the times. Ron Holland is an excellent mono designer but did not do any other multi’s after Colt Cars.

 

Ocean Emu was an early Grainger trimaran design built in Tasmania for a Victorian yachtsman. The boat was 43 x 35 foot displacing 10,500 lbs and carried 1400 square foot of sail. This tri was meant to be a racer cruiser. The tri had 3 major incidents. Snapped a crossbeam suspected to be a bad initial build/design. Snapped the front of a float off when got hit by a gust in shallow water where it was suspected the float bow touched bottom. Broke another crossbeam several years later, suspected to be a bad overdate batch of carbon fibre imported into Australia. Yes folks carbon fibre has a “shelf life” when not in a laminate matrix, it degrades in light etc. The hulls and decks were strip plank duracore (balsa with thin about 2 mm hardwood faces) 18 mm thick with 450 gsm 45/45 e glass biax outside overlapped to the waterline and 220 gsm 80/20 unidrectional glass inside. The crossarms were rebuilt to a top and bottom flange 25 mm thick carbon fibre with a 10 mm solid fiberglass vertical web in between. This was in a C section then a forward face was added to C section. The forward face being a foam glass structure with 3 x 600 gsm biax on either face. Webframes every 500 mm and a D section of foam glass was placed in front of the C box and the entire beam was wrapped in 600 gsm triax. The boat was built about 1984 and is still sailing today with the rebuilt crossarms.

A Farrier tri shows another way to break crossarms and floats is to have a fight with a big mono while racing. Attached photo. Finally an updated set of picture of Grainger's new Venom 42 foot tri is at Images – Venom Sailing https://venomsailing.com/?page_id=174

 

The Crowther SP 40 cruising catamaran was done for home builders with a parallel version was previously developed for production in France, the C106, which we have discussed previously. The SP 40 comes in a few variations of structure. The boat is 40 x 21 foot weighing 14000 lbs and carrying 1080 square foot of sail. The cat was initially a WRC strip plank design but was also built in foam glass. The strip plank version had 2 layers of 576 gsm biax e glass on either side of 15 mm WRC in the hulls. The bulkheads were 576 gsm biax 15 mm WRC 576 gsm biax. The foam glass version has outside 330 gsm Kevlar, 1176 gsm e glass triax, 25 mm Klegecell, 1176 gsm triax in epoxy on the hulls with 760 gsm triax both sides of 20 mm Klegecell for the decks and cabins. Non beam bulkheads are 576 gsm biax either side of 20 mm Klegecell foam. Some boats had composite wing masts 60 foot high with a 420 mm cord and 220 mm wide. The boats sailed well but were superseded by design 226A and Duegello etc.