What's so good with carbon fibre?

Any construction material has three main structural properties: strength (how much load will it bear till it breaks), stiffness (how much will it stretch for a certain load), and specific weight (how much is the weight of a given volume of this material). In our field the most used materials are stainless steel and aluminium. They have similar strength, but steel three times the stiffness of aluminium and also three times the specific weight. Both these materials are isotropous, which means that they are a homogeneous solid with the same properties in each direction.

Carbon fibres are different. As their name implies, they are very thin fibres, which must be surrounded by another material to keep them together. The other material is called matrix and is generally a polymer resin, like epoxy. The carbon fibres have really good structural properties: their strength is more than ten times that of stainless steel or aluminium, their stiffness is at least the same as stainless steel (there are high modulus fibres with much higher stiffness), and their weight is about two thirds that of aluminium. This very good properties are somewhat reduced because of the need of a matrix, so we never have pure carbon fibres, but they are nevertheless much better than most metals.

The other main difference is in their fibrous nature: a metal structure often has loads in just one direction, but their properties are present also in the other two directions, even if we don't need them. With fibres we can design the material to have the required properties only in the needed direction, with no waste of material. In this case we can have a structure which is as much lighter than a metal one as is permitted by the ratio of the structural properties. That's a lot: if stiffness is what counts, we need the same thickness of carbon fibres as of stainless steel, but the weight of carbon fibres is about five times less than stainless steel. If we compare to aluminium, the stiffness of carbon fibres is three times better and their weight is about two thirds, so we get to a total weight ratio of about 5 again.

We cannot do this on all structures: the loads are often coming from different directions, so we have to put fibres all around, to simulate an isotropic material. In this case, we come to have similar stiffness to aluminium, but the weight is nevertheless two thirds.

To summarize, carbon fibres are very strong and rigid filaments. They are also very light. Furthermore, they are very thin and they can be positioned in the most efficient way to bear the loads acting on each structure. In other words, the material itself can be engineered, as opposed to metals, where the material had fixed properties and only the shape of the structure can be engineered. This property is very important, since the loads on a generic structure are often directed in one or two directions, so we can carefully optimize the carbon fibres layup to minimize the weight.

What are the drawbacks compared to traditional materials?

The material itself is expensive, but the main drawback is the amount of very qualified manual work involved in the production of a composite structure. With metals the material itself is shaped mostly by automated processes, while with carbon fibres each layer of fabric must be manually placed in a precise sequence, following every curve of the mould. This raises production costs.

Isn't there a danger with sunlight?

Carbon fibres are insensitive to UV radiation, the dangerous component of sunlight. What can be sensitive to sunlight is the resin matrix, which could change colour after long exposure to UV light. This is an old problem and it has been addressed in various ways. The resins themselves have not the same sensitivity to UV radiation. Polyester, vinilester and the various epoxy resins have different behaviors under UV radiations. Resin producers have learned to add UV filters to the resins which are today less sensitive than several years ago. We also use high quality transparent paints which are even more able to filter UV radiations. These paints were developed for the automotive market, where UV degradation, which appears as a change of colour of the surface, was a big problem until a few decades ago, but it's now almost non existent. We can say that one shouldn't worry about UV aging of his carbon fibre product any more than one worries about UV aging of his car.

Are carbon fibres electrically conductive? Could this be a problem with lightnings?

Carbon fibres convey electricity. Much better than glass fibres or kevlar, but worse than most metals. On the other hand, the resin matrix is a very good insulator. A properly laminated carbon fibre product has all its fibres covered and insulated by resin, so it cannot convey electricity, but even a poorly laminated one, with some fibres sticking out of the resin, will be a worse electrical conductor than a similar steel or aluminium product. Concerning lightnings, there are poor electrical conductors, like trees, which can be stricken. The shape is important in this regard: points should be avoided. There are no reasons to think that similar shapes built with metals or carbon fibre composites should behave very differently with regard of lightning strikes. To our knowledge, neither are data to prove otherwise.

What can you say about galvanic corrosion?

This is a serious issue whenever there are objects built with different conductive materials immersed or wet by salt water. Since our products have metal parts close to carbon fibre parts we had to address this topic. Our answer is that, even if carbon fibres are electrically conductive, a carbon fibre prepreg is not so, because of the insulation provided by the epoxy resin. With hand wet fabrics you can have resin starved areas with poor insulation, but prepregs have a tightly controlled resin content all over their surfaces which are then totally insulated from the outside environment. Of course this is different when we drill holes in a finished part: on the inside surface of the drilled hole there are naked fibres which could in principle start galvanic corrosion with a nearby metal fitting. On our gangways we use a few screws to keep in place our metal fittings until the glue is hardened. To avoid corrosion problems, we take three steps: 1) all our screwed fittings are made with a deep anodized (0.05 mm) corrosion resistant aluminium alloy; 2) all the screws are made with stainless steel; 3) the main factor is that there is a layer of insulating glue between the fitting and the carbon fibre surface and the glue is pushed also inside the holes before the screws. Thus there is no waterway left open to start galvanic corrosion.

Are there different ways to build carbon fiber structures?

Well, carbon is now used on so many different products which often have in common only the presence of those famous fibers inside. Weight, mechanical performances, looks and cost can be hugely different.

The main difference you'll find is resin percentage. I mean how much resin you have into the laminate compared to fibers. Resin is used to keep fibers together. To do that there is an optimal proportion which is around 40%. If you have more, it's only more weight. If you have less, some fibers will be separated and will not be able to correctly transfer loads to neighboring fibers. In this case the risk of structural failure is great. It's important to notice that, even if you have too much resin, you risk structural failure if fibers are not in tight contact with each other.

One can thus make a first distinction between composites with the right amount of resin and those with too much or too less of it, and one can guess that it's not easy to reach the right proportion all over the laminate if the resin is applied by hand.

Most of the times, hand wet layups have too much resin, even if there are ways to approach the ideal proportion. You can find hand wet laminates with resin percentages varying from 50% - sometimes even 40% - up to 80%. Something to keep in mind is that the stated percentage is often relative to the whole laminate, which means that you could have areas with too much or too few resin, even if the average percentage is 40%.

How can resin percentage be controlled?

Vacuum resin injection is becoming popular in the nautical industry, since it allows a high level of control even on big laminates, such as a boat hull.

In the aerospace industry, but also for racing cars and motorbikes, the almost universal method is the use of pre-preg fabrics.

What is a pre-preg fabric?

These fabrics are made from carbon fibers, but also glass or aramid, impregnated with epoxy resin by a device which applies the same, predetermined quantity of resin to the whole fabric, in a highly controlled environment. The environmental control is important, since the resin is sensitive to humidity and impurities.

After the impregnation the resin is thickened and made less sticky at room temperature, so the pre-preg fabrics can be stocked and used later, provided that they be preserved at very low temperature, usually around minus 18° C.

Why aren't pre-pregs more widely used?

Pre-pregs must be cured at very high temperatures: usually at 120° C or more and they give the best results if a very high pressure, around 7 bars, is applied during the curing cycle.

What's the purpose of such a high pressure?

High pressure is very efficient in keeping fibers in tight contact. Composites are made of several layers of fabric and they should be able to transfer loads to their neighboring layers. This is achieved if the layers touch each other. If there is a chunk of resin or an air bubble between two layers, each one will work alone and one of them will get more load then the other, risking a failure.

How is pressure applied during the cure?

One way to apply pressure is to put the object, with its mold, inside a nylon bag, and take out the air from the bag with a pump. In this way atmospheric pressure will push the layers against each other and against the mold.

The vacuum bag will also suck most of the air trapped into the laminate and part of the excess resin, helping to reach the ideal ratio. Of course, if the ratio is already ideal, because a pre-preg was used, there are ways to prevent this.

By making vacuum, the pressure one can theoretically apply to the laminate is 1 bar. If a higher pressure is needed, one has to increase the atmospheric pressure around the vacuum bag.

How do you apply the high pressures required to reach the best structural performances?

To reach higher than atmospheric pressures, we must use an autoclave, which is basically a very strong oven which can be pressurized, somewhat like a pressure cooker. Commercial autoclaves can reach pressures of 7-10 bars. Autoclaves have very thick walls, powerful compressors and heaters and are very expensive. The pressure acting on the vacuum bag can be multiplied with an autoclave, pushing harder each layer against each other and against the mold.

Is it really necessary all this pressure from an autoclave?

Thanks to the high pressure it's possible to push out virtually all the air trapped between layers, insuring the best mechanical performances. By the way, if the laminate is hard pressed against the mold, its surface is forced to follow every detail of the mold surface: if the mold is shiny, the part will be shiny too.

For these reasons, prepregs autoclave curing is the technology of choice for both the best structural and aesthetic results.

Are there any other advantages, besides the structural ones, coming from the use of prepregs?

Dry fabrics tend to slide and move on the mold while they are being wet with resin by brush or roll, while prepregs are kept stable by their thickened resin, which gives them a leather like consistency. For this reason it's easier to laminate them and their fibers are more easily kept in the required direction, without distortions.

In other words, with prepregs it's easier to reproduce several objects with a predetermined structure and with a constant appearance across the series.

Did you have any specific problems in designing nautical equipment?

When we started to think about nautical equipment we already had a long experience in motorcycle racing, where cost is not usually such a big issue, and performances are the driving factor. Our goal was then to verify if these technologies could be applied to the nautical industry. Of course our main worry was their cost.

Of course it must be understood that this technology cannot be universally applied, but we found that the present prices of several nautical carbon fiber products are in a range compatible with ours.

From a preliminary estimate of tooling and production costs, we established that it would have been possible to produce steering wheels and gangways in autoclave cured prepregs at a competitive price.

Which were your goals when you started?

Basically we set the following goals: we should build aesthetically pleasing products, with an affordable cost and at least 20% lighter than the best competition.

Which were the alternatives already present in the market?

Talking about steering wheels, there are a few built from separate spokes and rim, glued together. This method probably allows more standardization of the tools since it is possible to build the spokes for different wheels from the same mold, adjusting only their length. Probably this is the method which requires the least tooling investment, even if the outside rim requires a mold for each size of wheel, but the production time and the weight are certainly not ideal, because of the many glueings. Mechanical performances and quality control can be problematic since the glue lines are not always on appropriate surfaces.

Then there are wheels built from two halves glued together. In this case every size of wheel requires a complete mold. If the wheel is symmetrical, one mold is enough to build both halves with fixed costs savings. This system allows a better control on fiber direction and continuity allowing a fairly good optimization of the structure.

How do you build your wheels?

First of all, we use prepreg fabrics and autoclave curing. Concerning the construction method, there was a further option which we chose to examine: to build the whole wheel in one go, with no secondary bondings. This option appeared to present big difficulties, since the complete wheel is almost completely closed, with just a small hole in its center for the coupling with the axle. The lamination process, the setting up of the vacuum bag and thus the real pressure on the wall surfaces would be much more difficult to control than with an open mold.

We had in mind a few possible methods for solving those production problems, but we couldn't be too optimistic in assuming that they would work since, as far as we knew, there were no similarly built wheels on the market.

In view of these difficulties, is it really worthwhile to build wheels in a single piece?

This method allows a big weight reduction, because of the lack of coupling flanges on the two halves to be joined. A further advantage is the surface finishing we can reach, with a perfect surface right out of the mold, with no gluing lines to worry about. Furthermore, the single piece lamination is intrinsically more consistent than secondary bondings, allowing a more efficient quality control. The only problems could come from imperfect overlapping of the fabrics or lack of airtightness in one of the vacuum bags, and both those defects would be immediately visible from the outside.

Which were your main concerns in the choice of your wheels architecture?

Between the several architectures we examine in the preliminary phase, a few are discarded because of macroscopic structural disadvantages. The survivors must be analyzed in detail to establish the proper fibers lamination. It must be remembered in fact the main property of composites construction: the mold specifies only the external shape of the part, but the thicknesses and mechanical performances of the various walls depend mainly upon the fibers layup and direction. It is possible to optimize the strength by choosing the best fibers orientation and the thickness can be increased in particular areas where, for example, the loads are higher. All this tailoring of the structure is not possible with metals.

Prepreg lamination and autoclave curing are a big bonus when performing realistic structural simulations, which wouldn't be possible with less stringent production techniques. To perform a realistic structural simulation one has to be sure the fibers are in the proper direction and the resin quantity is the specified one.

We run several finite elements simulation (FEM-FEA) with higher loads than those specified by international standards, to find a general sufficient lamination sequence. Then we start taking off layers where the local stresses are lower. In the end we modify our simulated lamination to match something which could be produced in a reasonable time by a qualified operator.

At this point we have our "virtual wheel", apparently able to pass certification tests for steering systems. We just have to check that the weight is lower than the competition. And of course that the real wheel is up to the simulations... We are pleased to say that until now we were satisfied with the results.

Which tools do you need to produce your accessories?

First of all there is the master model: the base upon which the mold is built. Our masters are machined with computer controlled tools from high density resins, to have enough stiffness and temperature resistance to bear autoclave stresses. Every detail must be right from the design phase: the parts must be able to come out easily and the couplings between separate parts of a mold must be extremely precise, otherwise the coupling lines will be visible on the parts.

To get everything right we use surface modelers, CAD software especially conceived to work with complex surfaces. They allow virtual prototyping and they can be interfaced with the NC machining tools.

On the masters we build our carbon fiber autoclave molds. From the outside they are similar to the common fiberglass molds used by any boatyard, but they must be tough enough to tolerate the stresses o hundreds of high pressure cycles and they must not have micro bubbles of air below the surface, which could break them at high temperature.

Their surface finishing is mirror like, since we absolutely want to avoid sanding and painting on our parts.

Our molds are built in prepreg carbon fiber. Very expensive but very precise and durable.

How many prototypes did you need to set up the final products?

We started from the smaller wheel, the 935mm diameter, which was built from the beginning without any major problem, demonstrating that the technology was viable. So much so that we later decide to patent it. The weight of that first wheel was 2.200 g, even if its lamination was very conservative. Our estimates were thus confirmed, since the competition ranged between 2.250 and 3.500 g.

There were a few minor problems with surface finishing. The causes were already determined and solved in the second and third prototypes. This prototypes were also lightened to 1.800 g, after testing the excellent strength of the first prototype.

The further structural testing performed on the second and third prototypes confirmed the excellence of this technology and by the end of summer 2003 we made a further 5 prototypes with a weight reduced to 1.480 g, keeping a strength well above the certification limits.

The same procedure has been followed with the bigger wheels and with the gangway.

What kind of testing did you perform on your products?

We had different target design loads for wheels and gangways. For the wheels the target load was the certification load from ISO8847, for Wire Rope and Pulley Steering Systems. We designed the wheels for 3 times this load, or about 205 kg, perpendicular to the wheel plane, applied to the outer rim. For the gangways, there is no international standard, but we chose a load of 350 kg, concentrated on the central part of the walking area, which is higher than what is normally declared by competitors and appeared to be reasonable to allow for dynamic loads. All our prototypes have been tested by Exit Engineering to their limit design loads. Furthermore, starting from the second prototype, we gave a few parts to selected boatyards who acted as field testers. A few wheels are already in use from more than one and one half years without any structural or aging problems.

What can you say about Exit Engineering carbon fiber gangways?

The first gangway design was a foldable model, which 2 parts were a totally closed, very strong shell box structure. At first this appeared to be the best solution from the strength point of view. The weight of this gangway was 9.200 g against the 15-18 kg of the competition we were able to check. We built a few prototypes which passed our tests without problems, but later we realized that with a different philosophy, more appropriate to the actual loads, it would have been possible to produce a substantially lighter gangway.

At Genoa 2003 Boat Show we brought the prototypes of a new kind of gangway whose weight, in its fixed version, was no more than 5.650 g. In 2004 we built the first foldable version of this new gangway, with a weight of 6.500 g, completing our line around 2.200 mm length. This design philosophy allowed us to introduce much more elegant shapes which favored the commercial success of this product.

What is the big deal with weight saving?

The gangway has to be lifted and stowed away quite often. This can be a complex operation with metal gangways, involving ropes and leaning out of the transom with a heavy weight in your hands. Our gangways, with their weight of a few kg, make the operation really simple and safe even for elderly people. The advantage is thus real, besides the fact that with composites structures, lightness is often an indicator of good design and workmanship.

For steering wheels, the minor weight and its concentration around the hub, where loads are bigger, translates in a smaller moment of inertia, making the helming much more sensitive and allowing quicker maneuvering.

Can you make a first review of your activity?

These first two years demonstrated that it's possible to produce nautical equipment using aerospace technologies. These technologies make available the weight savings that should be expected from composite materials at a cost which is compatible with present, less technically advanced alternatives. The higher initial investments are balanced by the easier production cycle and by better quality control, allowed by prepregs. This are qualities which could be fundamental for big shipyards and international distributors. These same considerations can be applied to any product to build even in small series of a few tens of pieces. For single or very few pieces the tooling costs could be important, but the consistent weight savings or aesthetic advantages could make this possibility interesting for those who value top quality.