Multihull Structure Thoughts

David DeVillier’s who designed the 62 foot aluminum cat featured a few days ago was also asked to design a smaller version for offshore cruising. The core theme for this boat was to be a safe, capable and speedy offshore cruising catamaran which can be easily handled by a couple. A simple blue-water passagemaker. The Purecat 52 was designed. The 52 is 52.5 x 25.6 foot and displaces at DWL 32,000 lbs. The ketch rig version 1539 square foot on a 60 foot main mast and 43 foot mizzen mast. The fractional sloop rig has a 72 foot mast that carries 1663 square foot of sail. There is an option of daggerboards or low aspect ratio keels.

The real difference between these 2 rigs is ease of handling and stability. The ketch rig has smaller sails with more options for sail reduction in stronger winds. Yes, a mizzen mast costs extra money and is an additional hassle but the ketch rig centre of effort of the rig is 20% lower than the fractional sloop rig. This means the ketch rig will be able to carry its sail area in stronger winds, but the real advantage of the ketch is you can eg just drop the mainsail and still have enough “balanced” sail area up to drive the boat in up to gale force conditions. Reefing mains can be a painful experience in stronger winds, it’s a lot easier just to fully drop the main.

The forward cockpit is the safest place to be when working the sails. All headsail sheets, furling lines, mainmast halyards etc. are led to the area around the base of the mainmast and at shelf height it’s very comfortable working the winches and jammers. Control of the mizzen is in the aft cockpit around the base of the mizzen mast. The PureCat 52 is capable of 250 nautical mile days according to the designer.

The interior has 3 double berth cabins and loo’s in the hulls. The main bridgedeck cabin has the galley, dinette and inside steering position. The 50 HP engines are in separate engine rooms aft.

The cat is mainly built from 5083 aluminum with frames and T section stringers covered with 4 or 5 mm aluminon sheet. Only the bottom and a few other panels need to be shaped. This boat is designed for serious cruising with such features as 20 mm thick aluminum skegs for prop protection and 90 mm solid aluminum rudder shafts on the rudders. The rudder shafts exceed all standards societies requirements. The rugged alloy cruising cat has an inherent safety feature of 6 watertight compartments and high underwing clearance. The designer claims an aluminum cat has lower build costs than an equivalent custom composite boat (as long as you’re not comparing costs to pop-out production boats).

The jpegs give the idea. Good design for those who have the money and time (10,000 hours).

 

A small discussion about composite boats and things to watch for. This is a few people’s thoughts combined into one item. Also a scientific advance in aluminium fatigue.

The majority of fiberglass boats are done in female moulds. In basic shops with little or no vacuum bagging or resin infusion the gelcoat is laid then the glass skin with a wet CSM layer then followed by any core material such as scrim kerfed foam or balsa. This is often allowed to cure then the following day the interior skin is then laid. There is no or little compaction or filling of any kerf gaps. Slightly more advanced builders first let the outer skin cure, then apply bonding paste, and push the foam into that. Again, issues with bonding can develop, kerfs still not completely filled, etc.

Either of these approaches can lead to delamination as there is an unknown or incomplete bond in the fiberglass core layer which can dramatically decrease strength in that bond. The real issue here is it is easy to stick glass to a core, it’s hard to stick foam or balsa cores to glass.

Technical bulletins from resin foam companies ask for foam to be hot coated and vacuum bagged onto the glass skin. A guide to vacuum bagging is at https://www.westsystem.com/wp-content/uploads/VacuumBag-7th-Ed.pdf The best solution is resin infusion of the glass foam (balsa) interface. A guide and web site devoted to resin infusion of a Farrier F 39 is Using Controlled Vacuum Resin Infusion technique for my project http://www.fram.nl/infusion.html

A well done polyester resin foam core interface is better than a badly done epoxy core interface. If you have a good vinylester foam/balsa composite vacuum bagged or resin infused you will be getting a good build solution.

Also, when you have chosen a design please advise your designer of the materials that are available to you. Some fabrics or resins may not be available in your location without exorbitant shipping or acquisition costs. I know in some countries you cannot import some materials. The designer often can modify layups to suit your circumstances. In one case of a build I was involved in it was going to take 4 months to get some fabrics from the US. The build time line did not allow for that delay. A slight redesign of the layup resulted in an extra 40 lbs of weight but the build time line was meet.

Also build a hull in the right temperature with minimal humidity or postcure the part(s) so that you get a complete cure of the resin glass matrix. If you do not cure your resin correctly it will weaken the structure. If you cannot postcure say an entire hull, roll the hull out into the sunlight on a warm day for 8 hours making sure both sides get a dose of sunlight. It has a similar effect to post curing. Do not layup an epoxy glass layup on a core in cooling evening moist air then not do some post curing. I have seen bond failures with epoxy months after construction because the epoxy was never fully cured.

And do not apply eg a radiant heater directly to a newly glassed part in hope of “speeding up the cure”. I have seen a bulkhead ruined by a builder as the resin got so hot it partially evaporated away just leaving white glass patches on the bulkhead.

Get resin ratio’s correct for eg part A and B in epoxy resins. Epoxy is especially important in this regard. Also, when you are mixing bog materials mix the resin first then add the microballons etc.

Now a final thought from a person with a lot of resin infusion experience. If you are a home builder, build the male or female strip plank mould. Lay your foam into or on that mould and hand lay or vacuum bag the glass onto foam. Take away the mould and then resin infuse the other side of the foam. Reason why? Because resin infusion requires a very good non porous surface and associated bag to hold the vacuum required to pull the resin through the fabric layup. Therefore, it requires an excellent mould or a flat table surface to get a good infused part. Not all builders can afford an expensive mould. Rob Denny “Intelligent Infusion” uses relatively cheap moulds to achieve a good base for his infusions but his hull shapes are also simpler than most designers (the hull shapes work well in his boats).

Aluminium scientific advances. Not a current product but this will have a real impact when developed. Aluminium alloys are used because they are light and have great corrosion resistance. But they fatigue badly. Australian engineers may have solved that problem, after creating aluminium alloy microstructures that can heal themselves while in operation. A report in Nature Communications, has demonstrated a 25-fold improvement in the life of high-strength alloys.

The team led by Christopher Hutchinson from Monash University showed that poor fatigue performance is due to weak links called precipitate free zones (PFZs). Hutchison says 80% of all engineering alloy failures are due to fatigue – an alternating stress. The failure occurs in stages. The alternative stress leads to microplasticity (undergoing permanent change due to stress) and the accumulation of damage in the form of a localisation of plasticity at the weak links in the material. This causes a crack, which grows and leads to final fracture. Working with three commercially available aluminium alloys, the researchers used the mechanical energy imparted into the materials during the early cycles of fatigue to heal the weak points in the PFZs. This significantly delayed the localization of plasticity and the initiation of cracks. Translation less masts falling down and longer life boats, but the real advantage will be lighter boats as designers will need smaller safety factors and can use thinner parts.

 

The second picture shown With the black bottom paint is not a core to glass fail, it’s a repair caused by a lightning strike. Quite a few pinholes in the hull below waterline.
On edit: after looking at the thermal imaging report there was a core to glass failure due to the lightning strike not due to bad building practices.

 

Smj1. Thanks for the input, I am often surprised by the updates we get. You can still see a clean failure between the green foam and the outer glass layer and the inner glass layer above the foam has little foam adhesion. Yes it may have been caused by lightning but it still did not have a good initial bond.

 

That’s interesting as the green foam was a repair done after a previous lightning strike!

 

Smj1. I bow to your knowledge of this cat and its repair.

 

Secondary bonding of fiberglass to fiberglass is a tricky subject. Some people assume that a quick surface grind and clean is enough, others think an acetone wash is ok but we need to understand the different approaches have on differing resin types. Adhesion and secondary bonding are 2 different things.

Polyester has relatively poor adhesion but can have reasonable secondary bonding (eg bonding a bulkhead to a cured hull) with the sanding of the hull/bulkhead surfaces. The best secondary bonding occurs within 18 hours of the creation of the original part. There can be at least some chemical bonding if you have eg unwaxed polyester resin. After about 18 hours you are depending on a mechanical bond which generally means a sanded surface to increase the surface area of the join location.

Vinylester has high adhesion but also has cured surface that is more resistant to secondary chemical bonds due to its cell structure (the reason its more “waterproof”). Again, after about 18 hours the main strength in a secondary bond is a mechanical which generally means a sanded surface to increase the surface area of the join location.

Epoxy has very high adhesion but to Joe Parker of Gougeon “Once polyester resins have thoroughly cured there is no opportunity for a chemical linkage to the cured polyester laminate and the item being bonded regardless of the type of resin used” “It is critical to grind, clean, and dry the laminate to be bonded to”.

Other approaches to doing secondary bonding that are tried include acetone or MEK wipes etc. These products “plasticize” the surface (soften) but they do not promote chemical linkage. In tests surface grinding results in better secondary bonds than trying to “glue” to a fully cured surface.

Peel ply is often touted to make it easier to do secondary bonds. Peel ply function is to provide a barrier between the curing glass layer and the external excess glass and contaminates. This peel ply does this job well but it also seals the underlying resin surface very well, which allows the underlying surface to more effectively cure and reduces the ability of a chemical bond. Result you again have to do a lighter sand of the surface to get a better secondary bond.

Sanding the surface of a resin glass layup is a risky business. You want the resin roughed up but you do not want to cut the woven roving or biaxial fabrics. Conversely, you do not want a resin rich surface layer which weakens and adds weight to the total structure. The higher the fiberglass fabric content the better. Sanding the full joining surface is important, there can be no shiny spots. Depending on what builder you talk to they claim from 16 grit to 120 grit sanding surface is the “ideal” for secondary bonding. Also, make sure you clean the surfaces after your sanding as dust and dirt can affect the secondary bond strength.

Again choice of fabrics in the initial hull layup help. Unidirectionals, biax and triax fabrics tend to lay flat and need minimal resin on the surface allowing you to get closer with your secondary bonding tapes. Woven rovings and CSM generally have resin rich layups to fill the gaps between the layers of the glass filaments. Sanding a smooth surface with woven rovings or CSM either leaves a resin rich surface or damage to the glass filaments.

Doing secondary bonding between wood or plywood and a resin glass (eg hull) be sure to “precoat” the timber with the bonding resin prior to doing the bonding. Timber can soak up a lot of surface resin which may starve the bonding glass resin which will weaken the joint again.

One of the better tricks to help secondary bonding is to lay an additional strip of biaxial or CSM on the inside of the hull, when laying up the internal hull glass skin, at any point of a bulkhead. This allow you to sand the area later for the secondary bonding of a bulkhead into a hull without cutting into any structural biaxial fabric. The US Navy specifies secondary bond locations should be sanded if the original part has been curing for longer than 18 hours.

Most home built foam glass multi’s will require secondary bonding throughout the build process. The important factor is to prepare and optimise the surface on both parts for the best chance of getting a good bond. Finally try and minimise the need for secondary bonds. Built parts as large as you can so they have continuity of glass fabric lines to minimise joints. Use vacuum bagging or resin infusion where possible to ensure solids bonds and it helps to build, eg half a hull side at a time. And do not think a dose of epoxy will solve all, an ill prepared joint will cause you a lose of strength and could cost a lot more.

For a general information document on repairs and secondary bonding please look at http://atlcomposites.com.au/icart/products/50/images/main/Fibreglass Boat Repair and Maintenance Manual.pdf

 

Today a few trimaran designs around 20 foot. The first 3 jpegs are of “home designs” that are intended to be built, the last jpegs are of the “professional” Seaclipper 20 and Richard Woods Strike 20 tri. The hull shapes in all these designs are simple and similar. I do not suggest the amateur designers hulls will be as good as the professionals but they will probably work well enough.

The floats can be from a beach cat or built from scratch to suit the design. The rigs are normally from bigger beach cats like Hobie 18 or 20. For an amateur designer this size of trimaran is often a first build and a bit of an experiment. You often can cheaply optimise the boat as required EG more rake to the rig, move the float slightly forward or aft etc. The professional designs are very likely to work as designed and if built to plan work sail very well (Seaclipper 20 and Strike 20 both sail very well). The small cabins on the amateur designs reflect the desire to build a “fast cruiser”.

The designs are all intended to be built in plywood with a variety of crossbeams. Wooden box beams in the amateur designs or aluminum tubes. Flat solid timber beams in the Seaclipper 20. The plywood hulls range from 4 mm ply on the sides of the lighter designs to 9 mm ply on the main hull bottom of the heavier design. The following is the general material list for the Strike 20 trimaran to give a feel for this size of trimaran.

Strike 20 Basic Materials List

5 sheets 4mm plywood
6 sheets 6mm plywood
2 sheets 6 or 9mm plywood

1in x 1in 15m
2in x 1in 50m
11/2in x 1in 7m
11/2in x 11/2in 10m
4in x 1in 12m
4in x 2in 2m

15kgs epoxy
500g wood flour or similar filler
4kgs 200g glass cloth (biaxial +/-45deg recommended)
1000 stainless steel countersunk screws 3/4in x 6
filler/paint as required

 

Todays trimaran is partially home and professionally built. The “Alien” trimaran is 22 x 16 foot weighs between 600 to 800 lbs and carries a 28.75 foot mast from a Prindle 19. The original main was from a Hobie but a new main is a custom from Whirlwind Sails, a black pentex mylar square top. The headsails are a GM self- tacking jib with a Doyle screecher from a F18 with a Harken roller furler and cleating snatch blocks from the F18 cat. Mark Schriebman the original owner claims the tri can reach 25 knots (optimistic, I suspect 20 knots is more realistic). The tri is wet above 10 knots but it is mainly sailed in Florida which means its an advantage not a problem.

The basic main hull was built from foam glass by Chislett boat builders of Rhode Island then had some modifications and painted white by Mike. A class floats that weighed 68 lbs were added with 70 mm aluminum cross beam tubes that have wire waterstays for additional support strength. The full length tubes go through fiberglass tubes attached to the main hull. The original A class aluminum chainplates were replaced by stainless steel chain plates because of the greater rig loads of the tri. There is a 5 foot long Intercat daggerboard that draws 3 foot and a Prindle 19 rudder on the main hull.

It takes about 2 hours to set the tri up from the trailer with the trampolines taking a lot of time, and under an hour to break down to put on the trailer. The engine power is a Honda 2.3 HP outboard.

The jpegs give an idea of a very nice fun boat.

 

Nice boat, and I agree with you : " Mark Schriebman the original owner claims the tri can reach 25 knots (optimistic, I suspect 20 knots is more realistic)"
Classic A cat ultimate top speed (right boards) is about 19 kt, I doubt about exceeding this with same hull and much higher displacement !

 

This is a little history which is providing us all with a current education. One warning this is about structures of the boats NOT about the type of boat. The boats we will discuss are HarryProa’s. Rob Denny had a history of innovative products and materials before he decided to develop a freestanding rig and a small 16 foot cross between Atlantic and Pacific proa. Result was a HarryProa. Rob’s test proa worked OK so he decided to build a cruiser version. He built the first Harryproa (he asked his wife what he should call the boat, she replied Harry). The Harryproa was 40 x 23 foot weighing 1500 lbs and displacing 2500 lbs. The sail area was 375 square foot on a freestanding rotating rig. Jpeg attached.

The original 40 foot proa was built from 9 mm “bendi ply” (ply that had about 80% plies in one direction). Result you could torture the ply into round bilge hull shapes. Then a layer of glass is laid over. The main structural strength of a Harryproa is between the beams, the hull ends can be constructed very lightly as there are minimal forces on the ends. The proa sailed well but was a test bed for developing the concept further. EG rudders needed optimisation etc. But the concept worked sailing at near wind speed with a minimal rig.

Next development was strip plank cedar variants of the concept with the 35 foot “trailable” Harrigarmi being the most public example. Again, the structure was very light but about this time Rob was developing how home builders could build their own carbon fibre masts with some innovative construction techniques. Rob promoted the use of carbon tow instead of more expensive unidirectionals etc.

The next phase of the development was the use of foam glass in proas such as the 25 foot Elementary. The initial ones were vacuum bagged and had “round” bilge hull shapes. They were very light using aluminium beams.

The next phase came after Rob attended a Kelsall resin infusion hull shaping workshop. Rob saw the potential in this approach and developed the resin infusion approach by 2 aspects. One he simplified his hull shapes to allow them to be built in simpler molds (these shapes are not a problem in proas but do not work as well in EG cats). Two he developed the resin infuse process to build large parts with many “add ons” built in. He also found cheaper/simpler ways to do the actual infusion. EG less plumbing, simpler breather materials and pumps etc. All of these boats were mainly foam glass with some in epoxy others used vinylester. A jpeg of the latest 25 foot Elementary is attached.

Up until now Rob had put his money where his mouth was funding and building quite a few of these boats to verify the design and build concepts.

The next phase is being developed now. The 80 foot Harryproa cargo vessel. You can view some of the discussion on the following Cargo Ferry – HARRYPROA http://harryproa.com/?p=2561 and 80 foot cargo harryproa https://www.boatdesign.net/threads/80-foot-cargo-harryproa.64736/ and Harryproa https://www.facebook.com/Harryproa/

The 80 foot proa build and associated 28 foot tender build is being documented on the Facebook site. The main point of interest is that both boats are being built in solid glass with stringers frames and reinforcing ribs being used to develop panel stiffness. This allows faster construction at less cost according to Rob. There is a slight weight increase in the 80 footer being in solid glass but the rib /stringer/frame skin is still relatively light in relation to the total displacement. Again, Rob is putting his money where his mouth is.

Now a short discussion on panel construction. You require 3 things from a panel: structural strength, stiffness and puncture resistance. Hull strength is often obtained long before the panel is stiff enough for a hull. Stiffness is generally obtained by “thickness” which is the reason for foam cores with the structural strength faces separated. The other way to obtain panel stiffness is to reduce the panel size by the installation of ribs, frames, stringers etc. Puncture resistance is generally the ‘hardness” of the material. Solid glass, aluminium ply all can achieve this.

A short detour into timber/ply. A 26 foot cat can be built from 6 mm ply on stringers and frames. A 26 foot tri in constant camber is 9 mm thick without much framing. A 28 foot tortured ply cat can be built from 4 mm ply with a minimum of framing. All the panels have similar stiffness. The tortured ply cat gains stiffness from its round shape. A flat ply panel needs reinforcement to gain stiffness. Constant camber needs thickness with minimal framing.

Rob Denny has helped us learn a lot about light weight construction. This is not a discussion about the virtues of Harryproa’s as boats, it is about structural development. The jpeg give a clue.

 

Captain Bones is a fun camper cruising trimaran that can be trailed. The tri is 18.5 x 13.5 foot that can be compressed to under 8 foot for trailing. The displacement is unknown. The 19.5 foot roller furling mast carries a 97 square foot mainsail and 70 square foot jib. The floats are 12.25 foot long. The lee board on the main hull provides lateral resistance. The kickup rudder is on the main hull.

The idea is to have a tri that you can sleep on the side panels of the main cockpit with general storage in the footwell and various compartments fore and aft. The motive power is an oar.

The tri is a “kit” boat from B&B (I think). The kit contains all the plywood components for the design which is basically a stich and glue design from the precut plywood panels. The cross arms are aluminium tubes on the main hull which have float aluminium attachment tubes that slide in and out for trailering and sailing. Guessing here but I suspect 6 mm ply skin and bulkheads with 16 x 16 mm inwales. The panels are assembled in female frames with plastic ties holding the panels into shape. The chines are filleted together then glassed.

This is more a visual approach as to the build. Jpegs give the idea. The build is at July 2020, Bones & Miss Lynn's Excellent Adventure: http://captbones.com/2020/07/

 

Thanks Old Multi. Not a bad description of the journey.

Rob Denney

 

Coach, i just want to say that this is my third time reading through this epic treatise on multihull design methods and considerations. It may just be the pre-eminent condensed source of multihull history and design in the world. At least that's what i have found.

2 quick questions:
1) any considerations/methods on design of 'escape hatches' and 'survival rooms'? Capsize recovery is one thing, plans for getting out of a capsized boat and surviving is another.

2)Post #54 describes a patent and article on stringer-less plywood lapstrake construction. Have you found the patent and article? Sounds like a great idea, stringers construction is such a pain these days...

post #54 - One of the simplest plywood build methods I know of was invented and patented in 1970 (the patent has run out). It was initially used to build high speed round bilge power boats but could be adapted to any hull shape. The method is basically 24 inch (600 mm) wide plywood strips glued together to form a clinker hull. Now the best part get your 24 inch strips and scarf them together lengthwise to the length of hull. Use eg 3 mm ply for say a 23 foot boat. Put a series of round bilge hull shaped former's on a strongback about 18 inches apart. Place the first 24 inch strip from the keel line outward. Lay the next 24 inch strip about 10 inch from the keel outward covering just over half the first 24 inch strip. Lay the next 24 inch strip 22 inches from the keel line over lapping the 2nd strip and 2 inches of the first strip. Repeat until you get to the gunnel line. Result you have a 6 mm thick hull with an 9 mm thick hull every for 2 inches every 10 inches. The 9 mm thick part act as a default stringer every 10 inches. This techniques build a monocoque hull structure that only requires BH's to support the hull shape. A stem, keel and gunnels still needs to be inserted. Also at the keel and gunnel you will have double any 3 mm ply to form a 6 mm ply hull. This technique can be used for 4 mm and 6 mm ply which means it will work on hulls up to 50 feet. Yes it looks like a clinker ply boat (as Peter Spronk cats were) but it is very fast and easy to do with a minimal need for timber stringers etc. If this needs further explanation I will find the original article which had pictures and diagrams.

 

Tom2x4. I will try and find the original article as requested on the "molded ply' approach and upload it. give me a day or so. Escape hatches will also take a day to do. Hopefully in the next couple of days for both.