PRIORITY

The instant appllicatin claims priority to US Provisional Application Number 61/519,520 filed 24 May 2011.

TECHNICAL FIELD

The instant invention relates to spars or beams and methods for making spars or beams. More specifically, the instant invention relates to spar caps and especially wind blade spar caps comprising pultruded components and methods for making spar caps and especially wind blade spar caps comprising pultruded components.

BACKGROUND ART US Patent 5,324,563 is art related to the instant invention. US Patent 5,324,563 discloses a carbon fiber pultruded rod composite structure for reinforcing a substrate wherein the rods are bonded together into a preform for subsequent lay-up on the substrate with cross plied tape or fabric to form a wing or fuselage structure wherein the rods become longerons or stringers carrying axial load while the cross plied skins carry shear. Although the structure disclosed in US Patent 5,324,563 provided improved compressive strength, the structure disclosed in US Patent 5,324,563 is not amenable to forming shapes having compound curvature because the rods are bonded together before lay-up on the substrate. In addition, US Patent 5,324,563 did not disclose improved fatigue resistance for its structure.

US Patent 7,625, 185 is art related to the instant invention. US Patent 7,625, 185 discloses a composite structure wherein fiber rovings are formed into preformed fiexurally stiff strips which strips are then bonded together to form the final structure. Although the structure disclosed in US Patent 7,625,185 reduced fiber undulation and reduced problems of heat management during resin curing, the structure disclosed in US Patent 7,625,185 uses conventional resin impregnation of the fiber rovings which results in uneven fiber-resin distribution which results in reduced physical properties for the final structure. The imperative for alternative energy is clear throughout the world. Of the many alternative energy options large scale wind turbines are among the most viable both because they are able to produce electricity on a large scale and because they are one of the most cost effective alternatives. However significant further improvement is possible and each improvement in aerodynamic efficiency, structural efficiency and manufacturing cost increases their viability.

Wind turbine design requires multiple compromises between aerodynamics, structure, manufacturing cost and operating life. Thin blade sections are more efficient aerodynamically, while thicker blades are more efficient structurally. Thinner blades require more material (or more exotic material) in order to achieve adequate structure.

Larger turbines produce more power but as blades become larger, structural issues increase, in particular compression fatigue. Initial cost of a wind turbine and the very high cost of a failure dictate that blades have a long performance life. Inherent structural flaws make this difficult to achieve without significant safety factors which add to both weight and cost. Development of higher performance materials, with improved consistency without cost penalty can contribute significantly to increasing the cost effectiveness of wind power.

Large scale wind power has already proven to be one of the more cost effective forms of alternative energy. However given all of the costs and other factors associated with manufacturing, siting and installing large wind turbines, it is clearly beneficial to all levels of society that their performance and longevity continues to be improved. The level of improvement required is unlikely to be achieved by incremental improvements. LM Wind Power's VP of R&D, Frank V. Nielsen notes, "The industry needs a quantum leap in materials and production technology."

From a wind turbine system view, improved durability, longevity and materials efficiency are all success measures important to the industry. More efficient use of materials based on improvement in fatigue life and reduction in design knockdowns will allow either larger blades resulting in more energy capture for a given installation, and/or lighter blades for a given diameter resulting in reduced loads and more efficient

downstream material usage on all turbine mechanical systems. Either track will provide greater material efficiency and energy production for a given system. The instant invention addresses the above-mentioned problems and provides a significant technical advance in the art. SUMMARY OF THE INVENTION

The instant invention is a substantial improvement in the art and specifically upon the art of US Patents 5,324,563 and 7,625,185. The instant invention overcomes many of the most significant structural and manufacturing issues facing large blade producers today namely; near perfect placement, spacing and collimation of fibers, elimination of weal- spots resulting from resin rich or resin poor areas, ability to incorporate tapered sections and reinforcement levels and the ability to produce smooth curves for curved spar caps all while maintaining excellent short term & fatigue properties. In particular the multiple rod technique of the instant invention has been demonstrated to significantly improve compression fatigue of a wind blade spar cap. Additionaly the multiple rod technique of the instant invention has been shown to yield higher modulus and higher strength.

More specifically, the instant invention in one embodiment is a spar cap comprising a plurality of separate composite rods bonded together in a mold to form the spar cap. In another embodiment, the instant invention is a process for manufacturing a spar cap by bonding together separate composite rods in a mold, with a bonding agent to form the spar cap. Applications for the spar cap of the instant invention include wind blades, aircraft wings, aircraft rotors, marine spars, marine masts, and civil engineering applications such as bridge beams.

In yet another embodiment, the instant invention is a system for generating electricity, the system comprising a wind turbine installation configured to generate electricity, the wind turbine including at least one wind blade, the wind blade comprising a spar comprising spar caps, each spar cap comprising a plurality of plutruded composite rods bonded together in a mold to form each spar cap.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1 shows data points for a reference material and materials of the instant invention in terms of compressive strength v. fiber volume fraction;

Fig. 2 shows data points for a reference material and materials of the instant invention in terms of compressive modulus v. fiber volume fraction;

Fig. 3 shows data points for a reference material and materials of the instant invention in terms of compressive modulus v. compressive strength;

Fig. 4 shows compression fatigue data curves for a reference material and materials of the instant invention in terms of stress v. cycles; Fig. 5 shows a compression fatigue data curve for a material of the instant invention in terms of stress v. cycles;

Fig. 6a depicts a section of woven fabric composed of strands of fibers;

Fig. 6b depicts an enlargement of the strands of fibers shown in Fig. 6a as well as a resin surrounding the fibers;

Fig. 6c depicts an enlargement of the strands of fibers shown in Fig. 6b as well as a resin surrounding the fibers;

Fig. 7 depicts a cross sectional view of a pultruded composite rod showing fibers surrounded by a resin;

Fig. 8 depicts a number of pultruded rods laid in a curved mold;

Fig. 9 depicts a wind blade spar cap produced by flowing a bonding agent around the rods shown in Fig. 8 followed by curing the bonding agent to finish the spar cap;

Fig. 10 depicts a wind blade spar cap that tapers in both directions from one end to the other end and which is curved in both directions from one end to the other end;

Fig. 11 depicts a joint between a stud and a bundle of pultruded rods at one end of a wind blade spar cap of the instant invention;

Fig. 12 shows a cross sectional view of the stud of Fig. 11;

Fig. 13 shows a cross sectional view of the joint of Fig. 11 at a location where the rods surround the stud;

Fig. 14 shows a cross sectional view of the joint of Fig. 11 at a location before the rods surround the stud; and

Fig. 15 is a simplified cross sectional view of a wind blade showing its spar caps, spar web and aerodynamic skin.

DETAILED DESCRIPTION OF THE INVENTION

The advantages of using pre-manufactured pultruded rods, which rods are then bonded together to produce a final spar cap are significant. The pultruded rods are currently readily producible by a multitude of commercial pultruders globally. A particular advantage of pultrusion over resin infusion or pre-pregs is that the fibers have a more even distribution and more complete wet-out and are nearly perfectly straight and aligned. The inherent desire of the rods to remain straight ensures that as multiple rods are placed in a mold the near perfect fiber alignment is continued in the final spar cap. The diameter of the pultruded rods of the instant invention is not critical as long as the final spar cap comprises a plurality of rods. As a starting point, rods having a diameter in the range of from 1.5mm to 12.5mm are suggested.

With even moderate automation, the proposed method is capable of substantially increasing manufacturing throughput due to faster reinforcement placement and increased infusion speed which allows faster curing resins to be utilized. Pre-manufactured pultruded rods can be bonded together to produce a final spar cap with mechanical properties superior to current products.

The structure of the instant invention is in effect a composite within a composite. The first composite is the pultruded rod consisting of multiple highly aligned fibers nearly equally spaced within a matrix resin which is cured during the pultrusion process (the above-referenced US Patent 5,324,563 disclosed the benefits of fiber straightness in a pultruded rod structure but does not disclose the benefits of equally spaced fibers in a such a structure). The second composite consists of multiple rods bonded together by an infused resin matrix. This dual scale structure has been named C-Squared (meaning a composite within a composite). We refer to the fibers and resin in the rod as the primary

reinforcement and primary matrix and the rods and resin between them as the secondary reinforcement and secondary matrix respectively.

Samples of the instant invention are manufactured using 2 to 6mm diameter pultruded rods. The samples are manufactured by grouping together multiple rods to form a larger hexagonal or rectangular sections. These bundles of pultrusions are held in shape within a vacuum bag and infused with epoxy resin. The rods infuse easily although some voids are observed. The voids are not considered serious enough to prevent testing and their prevention was deferred to later in the project. The rods are cut into short sections and mounted for compression testing. The purpose of the mounting is to mimic as closely as possible the end restraints used in ASTM D6641 and D695 testing to ensure true

compressive failure without 'brooming' of the ends that occurs in unsupported specimens. Some problems are encountered in achieving this. Initially end restraint was attempted by fiber wrapping the specimen ends but this proved insufficient. Subsequent attempts progressed through potting in copper tube and potting in 32mm wall steel tube. All of these methods prove to be insufficient, with the restraint bursting before specimen compressive failure occurred. As the restraints were strengthened the measured compressive properties increased so finally a thick walled solid steel restraint was machined. This prevented the brooming and resulted in the highest properties.

The short term compressive strength and modulus are measured using this restraint (described as Pultruded Rods 1 or 2) and are shown in Figs 1-3, plotted along with reference data for prior art materials (the NREL/Sandia data from Sandia National Laboratories Report SAND 97-3002). Although the data sampling is relatively small, a number of key points are immediately apparent. The material modulus for the samples of the instant invention and the conventional samples varies linearly and consistently with fiber content. This is to be expected, the data follows a consistent 'rule of mixtures'. However for strength the Sandia data for conventional materials shows a maximum around 50% fiber loading. Above this value, while modulus continues to rise, strength drops off. This is conventionally attributed to 'fiber crowding', in other words the fibers begin to be so close to each other that it becomes impossible for them to be fully aligned and perfectly wet-out so defects such as dry fibers and kinks occur. The C-Squared samples of the instant invention show higher strength values even at higher fiber loadings. The significance of this can be seen in Fig 3 where strength is plotted against modulus. The conventional prior art samples require a trade-off between strength and modulus - strength of 800Mpa is possible at a modulus of 34,500Mpa but if higher modulus is required lower strength is obtained. Since both strength and modulus are important in a blade spar, the ability of the instant invention to bring both higher modulus and higher strength is a significant attribute.

Short term data is measured on specimens with cross sections between 2.3 and 3.2cm . Long term data is measured on specimens with a maximum cross sectional area of 1. lcm . For many composites, but more especially for the structure of the C-Squared material of the instant invention there is a noticeable size effect on samples. Repeating the same single point data with the smaller samples give a slightly lower ultimate compressive failure stress. This is attributed largely to the effect of having fewer rods in the specimen, increasing the relative importance of surface effects. This data is used as the starting point for fatigue testing.

A total of four sets of fatigue samples were produced (Series 1, 2, 3 and 4). Each set consists of a single infusion of a bundle of rods about 60cm long. Ten samples are cut from each infused specimen and mounted for testing. Initial infusions show the same void problem that appeared in the short term specimens. Significant effort is expended to understand the cause of the voids. Test series 1 through 4 show progressive improvements in void content. The final procedure in order to fully eliminate voids consisted of : Degas both resin and hardener overnight under high vacuum and moderate temperature (<5Torr, 40C), mix resin and hardener by metering through an enclosed static mixer in order to avoid air entrainment, degas again under high vacuum prior to infusion, degas mold under high vacuum to draw off any volatiles or surface moisture on the composite, infuse under relatively low vacuum (500Torr) in order to prevent any further volatiles being pulled from the resin during infusion. As this procedure is developed, the void content in the samples continues to drop. The last sample (Series 4) has no visible voids.

The resulting samples are squared off, mounted and end polished for testing.

Initially ultimate compression strength is measured on 2 or 3 samples to establish a baseline for testing. Fatigue testing is carried out with R=10 i.e. the load is cycled between a set maximum compression and 10% of the maximum. Initial frequency is set 1 Hz, however as many specimens ran to unexpectedly long times, the frequency is progressively increased to 2, 5 and finally 10 Hz as testing times increase. Fig 4 shows the data from the fatigue testing. A progressive improvement in performance can be seen in subsequent test series. This is attributed in part to eliminating voids through improved manufacturing, but also to improvements in specimen preparation techniques as the effect of even tiny errors became increasingly apparent.

It can be seen in Fig. 4 that all curves related to the instant invention (Series 1, 3 and

4) are flatter than prior art Mandell materials (the Mandell data from the Department of Energy/Michigan State University Composite Material Fatigue Database) but because of the increased performance as specimen preparation techniques improved, Series 4 is considered the most accurate representation of the true performance of the material of the instant invention.

The full data for Series 4 is shown in Fig 5. It should be noted that while all data points are included in the least squares curve fit calculations, multiple specimens had not failed and were either 'parked' due to lack of further time for testing or are ongoing at the time of the preparation of this application. Thus the true curve is better than shown in the curve fit. Based on the flatness of the series 4 curve and the un-failed data points that were included in the curve-fit, it is not an exaggeration to claim that the instant invention results in a truly fatigue resistant composite structure. The enhanced performance of the instant invention is attributed to the significantly improved fiber alignment and elimination of manufacturing defects. In prior art methods of spar manufacture the predominant reinforcement format is fabric. Fabrics have multiple inherent defects which are detrimental to the performance of a structure whose primary performance requirement is high uni-axial properties. First the weave of the fabric inevitably creates marcelling (waviness): as the warp and weft cross over each other fiber bends must be induced. Secondly there is a regular variation in the glass content due to the spacing of the fiber strands as a result of the weave and finally during the infusion process the pressure of the resin against the fibers inevitably causes a degree of crowding in some areas resulting in a fiber rich region and spacing in others resulting in resin rich areas. Figs 6a, 6b and 6c depict these problems. Fig. 6a depicts the fiber strands 11. Fig. 6b depicts an enlargement of the fiber strands showing individual fibers 12 surrounded by resin 13. Fig. 6c depicts a further enlargement of the individual fibers 12 surrounded by resin 13. In addition to all of these built in defects which are a consequence of the reinforcement choice, there is the further possibility of localized folds, creases and pulls in the fabric as a result of the cutting, handling and laying process. In contrast Fig. 7 depicts the arrangement of fibers 14 and resin 15 in a pultruded rod structure, wherein the fiber spacing is much more even and although exaggerated in Fig. 7, there is a thin layer of resin 15 between the fibers 14 even at their closest points.

All of the defects of the prior art woven structure, although seemingly minor, reduce the ultimate performance of the final composite. In tension, minor misalignments result in a commensurate de-rating of properties according to the degree of angular misalignment. In compression the effect of every minor misalignment is exaggerated, allowing compression to be translated to localized out of plane buckling. The use of stitched fabrics helps reduce these effects by eliminating the 'under-over' effect of the weave but the stitching itself still causes a degree of marcelling. Elimination of these defects and inconsistencies using the instant invention reduces required safety or knock-down factors and significantly improves blade cost and performance.

The use of pre-pultruded rods of the instant invention eliminates all of these defects. The rods, once pultruded, are rigid and can be stacked in a mold in perfect (or near perfect) alignment, completely eliminating any marcelling from the weave or stitching.

Additionally the pultrusion process inherently creates almost perfect fiber alignment and spacing with the rod. The fibers are continuously in tension as the resin is applied and cured which helps to create near perfect axial alignment and unlike resin infusion there is never any sideways force on the fiber to cause bunching. Rather the process by which resin and fibers are pulled into the die in the pultrusion process automatically causes the fibers to space equally through the cross section of the rod. This is especially true for small diameter rods. The combination of these two factors results in reinforcement fibers which are near perfectly axially aligned and near perfectly spaced from each other which are both very desirable attributes in reaching the maximum possible longitudinal properties. In particular the near perfect alignment and near perfect spacing is beneficial in fatigue where failures are almost always the result of a crack propagating from some minor defect.

Because the fatigue results are so unexpectedly good in the instant invention the eventual failure mechanism is obviously of interest. The limitations on the fatigue tester prevent determination of failure mechanism on samples that have been taken all the way to failure as it is not possible to stop the machine quickly enough to prevent complete destruction of the sample. Instead we inferred the failure mechanism by looking at samples for which the testing is stopped prior to catastrophic failure. Two forms of cumulative damage are apparent. First on both ends of the specimen it is apparent that 'brooming' has begun at the outside of the specimen and is gradually working its way towards the center. This localized brooming is in some ways an 'unfair' failure mechanism as it is a

consequence of the way the mounting method transmits a disproportionate load to the outer portions of the specimen. It is not known which mechanism is ultimately responsible for the sample failure. Also importantly, there is no failure in either the epoxy between the rods or at the interface of the epoxy and the rods. This is particularly encouraging as prior to this work it was felt that this interface might be a weak link in the composite.

In addition to the near perfect fiber alignment and spacing, a further advantage of the pultrusions of the instant invention is their inherent consistency. The process is highly automated and extremely consistent. The degree of variability in pultrusions is extremely small compared to resin infusion. Inherent variations in the strength of uni-axial pultruded rods are generally accepted to be +/- 3% while vacuum infused structures of the prior art have significantly higher variation. Further the pultrusions of the instant invention are much easier to test for guaranteed properties, either batch-wise offline or even using online stress rating techniques. A further advantage of the proposed multiple rod composite of the instant invention is that it makes the production of curved and or tapered spar caps less problematic. Curved blades (and thus curved spar caps) are increasingly considered as a means to increase blade performance while shedding loads from destructively high wind gusts. The 'STAR' (Sweep Twist Adaptive Rotor) proposed by Zuteck et al is one such example. However, even though the required spar cap curvature is relatively small, any degree of curvature increases the likelihood of kinks or creases in the reinforcing fabric of prior art structures as the fabric is laid into a mold. The STAR blade partly overcame this difficulty by laying down 'ropes' of wet impregnated fibers, however even those involved with the development suggest that this is a highly impractical (and messy) process that they would not repeat. The proposed multiple rod system of the instant invention allows a spar cap with considerable curvature to be produced without problems due to the ability of the small diameter rods to slide over each other and create curvature without the mismatch or wrinkling issues caused in fabric. Fig. 8 depicts pultruded rods 16 laid in a mold to have a curved shape. Fig. 9 depicts a curved spar cap 19 according to the instant invention, wherein a resin 18 has been flowed around the pultruded rods 16, which resin 18 has then been cured to finish the curved spar cap 19.

A sample of the instant invention is produced with approximately 12mm of curve over a one meter length. The flexibility of the rods allowed them to easily take the curvature of the mold and be locked into this curvature by the secondary matrix resin. This level of curvature is extreme compared with that required on a full size blade but is not even close to the degree of curvature that is possible using this technique.

Although the curvature on today's blades is relatively slight it is quite possible that with more advanced aerodynamic analysis & and construction methods, the shape of the blade may evolve in the same way as jet engine fan blades have evolved from the straight blades of 20 years ago to significantly more complex geometries and improved performance of today's blades.

All large wind blades (and wind blade spar caps) are tapered both in blade profile for aerodynamic reasons and in construction to reduce outboard weight and match strength and stiffness to applied loads. The creation of this taper is problematic with prior art construction because the relatively thick layers of reinforcement fabric used result in step changes each time a ply is dropped to reduce thickness. This not only causes a stress concentration due to the step change in thickness and reinforcement level but also causes a resin rich area immediately after the ply drop which itself causes structural problems. The use of small discrete rods of the instant invention helps reduce ply drop issues in two ways. First it is possible to spread the drop over a significant length by dropping one rod at a time rather than having to eliminate a complete layer and secondly the end of each rod can easily be tapered to further reduce the stress concentration. Fig. 10 shows a curved and tapered (in both directions at the tip 23) spar cap 20 incorporating the above-mentioned construction wherein some of the rods 22 are shortened or tapered before the resin 21 is flowed around the rods 22 and then cured to finish the spar 20.

Most current blades and spars are produced by some variation of vacuum infusion.

The use of this technique in the preparation of the samples of the instant invention discussed above gives a high level of confidence that this manufacturing method would be a suitable means to manufacture full scale blades according to the instant invention. A good general overview of existing manufacturing processes can be found in Gurit's blade design manual.

Today glass fiber is the predominant reinforcement in wind blades accounting for approximately 98% of the reinforcement usage. However other fibers, particularly carbon are receiving increasing attention because of their higher specific properties (higher strength & stiffness and lower density). The main limiting factor is generally considered to be cost; carbon fibers, depending on grade typically cost 8-10 times more than E-glass. However as blade sizes continue to increase, the weight and performance improvements make higher performance fibers increasingly viable. However carbon fibers also have other limitations. In particular although they have roughly twice the tensile strength of glass, they have barely the same compressive strength. Also their compressive fatigue performance is

correspondingly poor. Given the significant improvement in compression fatigue that the material of the instant invention demonstrates that the instant invention comprises pultruded rods comprising carbon fibers. There are further advantages related to carbon fibers. In general carbon fibers, especially uni-axial, have lower diffusivity than glass and are significantly harder to infuse. As a result, when carbon is used it is primarily in the form of pre-preg which further increases cost and necessitates the use of heated molds. Using pre-pultruded carbon fiber rods according to the instant invention simplifies manufacturing, allowing conventional unhealed molds to be used Manufacturing is critical for any new material. Regardless of the achievable properties new materials are of little interest unless they can be used to produce

components at least as effectively as prior art material options. The two dominant prior art blade manufacturing techniques are vacuum infusion and pre-preg layup. Vacuum infusion is certainly the most used though pre-preg is gaining ground because of improved properties and consistency as well as the difficulty of using infusion with carbon.

Vacuum infusion typically involves laying up multiple layers of dry glass fabrics into a mold with peel ply, breather and flow media laid on top followed by a vacuum bag sealed all the way around the edges of the mold. Multiple feed ports are punctured through the vacuum bag at successive stations along its length. Infusion begins at one end and continues until the resin has been 'sucked' through the fabrics as far as it will go at which point the next vacuum port is opened and a further length infused, repeating until the complete part is filled. Vacuum infusion has many drawbacks and sources of

inconsistency. The speed and ability of the resin to flow through the reinforcements is proportional to the permeability of the reinforcement stack, the applied pressure gradient and the inverse of the resin viscosity. These variables are often difficult to control accurately, for example applied pressure depends not only on the vacuum system used but also on atmospheric pressure on a given day, and resin viscosity depends strongly on resin, ambient and mold temperatures. Controlling viscosity using temperature is a double edged sword since while higher temperatures reduce viscosity; it also speeds up cure time, reducing the time available to fill the mold before the resin begins to gel. Prior art manufacturing techniques walk a fine line in balancing these and many other variables to produce a blade of acceptable quality at minimum cost.

Pre-preg layup is often preferred for spars because the improved consistency and higher axial fiber loadings. The pre-impregnated sheets of mainly uni-axial fibers are laid into the mold, already containing the optimum (or slight excess) of resin. The laminate stack has a peel ply, perforated release film and breather applied over it and is vacuum bagged. Heat and pressure/vacuum is then applied to cure and consolidate the structure. Although many of the inconsistencies of vacuum infusion are removed in this process there remain significant challenges - removing all entrained air from between the plies, adequately controlling temperatures to reach full cure in the minimum time without overheating in thicker sections and dealing with the limited usable life of the pre-preg are all problematic for large scale manufacturing.

As a very broad generalization, vacuum infusion is more commonly used in the prior art to produce the more geometrically complex, but less structurally critical blade shells while pre-preg is gaining ground for the manufacture of spar caps.

A number of manufacturing methods are viable in the instant invention. Although manual handling and placement of rods is possible (and in many ways easier than manual placement of fabrics) automatically or semi-automatically placing the rods is preferred. When the rods are positioned then they need to have the secondary matrix introduced, consolidated and cured. There are several possible options to do this. All three of the following techniques are believed to be feasible: (a) vacuum infusion; (b) B-stage coating of rods; and (c) rod laying machine with fast cure adhesive applied between layers.

The sample preparation work described above shows that vacuum infusion is an adequate method of manufacture. In principle this is very similar to prior art vacuum infusion techniques. The reinforcement (in this case pre-manufactured rods instead of dry fabrics) is assembled into a mold. The lay-up is covered with a peel ply and the complete assembly is vacuum bagged and infused as normal. However there are also significant differences, some of which may be of considerable benefit in speeding up manufacture and improving consistency.

In prior art infusion, the resin must flow in the very small channels available between the reinforcement fibers. Since the fibers are of the order of 20 micrometers diameter and are forced together under vacuum this means that the individual flow paths for the resin are of a similar scale. A surface flow media is required to carry resin along the surface of the reinforcement for easier flow, and is then sucked down vertically/diagonally into the reinforcement. Even with this technique it is still necessary to have multiple vacuum feeds along the length of a spar, typically every 3 to 6 meters. The flow media and the multiple vacuum feeds both contribute to waste and possibility for inconsistency in the final part. The flow media and multiple vacuum lines and ports along the component are consumables, extra cost that is scrap once the part has been produced. The resin remaining in the flow media and lines is an additional expense. The additional complexity of these items adds to the probability of defects. These additional items are necessary simply because the prior art fabrics are so tightly packed that it is not possible to infuse directly through the fabric. With C-Squared material of the instant invention the flow path between the rods is much larger, and is very consistent, in effect it is a series of channels, of similar size to the rod diameter which the resin is able to racetrack down. This improved flowpath eliminates the need for a flow media and will greatly reduce or even eliminate extra vacuum feeds. This can represent a significant saving in time, consumables and wasted resin using the instant invention..

When dry fabrics are laid up, multiple inconsistencies are introduced. The fabrics may be slightly misaligned, minor winkles disrupt the reinforcement direction, crimps caused by weaving or stitching result in consistent misalignment, tows tend to bunch where they cross each other and spread in other places leading to localized inconsistencies in fiber volumes which can be further exacerbated by the resin flow itself moving fibers. All of these inconsistencies contribute to local defects which have the potential to be weak points or stress concentrations leading a spar to fail locally at loads far below its predicted capability. The use of pre-manufactured rods according to the instant invention essentially eliminates all of these defects.

Typically the resin system is designed around the time required for infusion. Ability to infuse faster can allow a faster reacting system to be utilized saving time on both the infusion step and the curing phase. Additionally the presence of the pre-cured pultruded rods has the potential to act as a temperature stabilization mechanism. When epoxy is processed in thick sections (such as blade roots) where the volume to surface area ratio is high, there is always the danger of a runaway exotherm which can cause overheating and degradation of the matrix. The presence of the pre-pultruded rods can help to prevent this as the rods minimize the concentrated volume of epoxy and act is internal heat sinks in the system.

Although the use of conventional vacuum infusion is relatively simple in the method of the instant invention it is also believed that the use of B-stage resin systems can be used in the instant invention. As with existing pre-pregs, while the application of the B-stage resin is an additional manufacturing step, the resulting final assembly can be further simplified and increased in consistency. A first option would be to continue to use round rods and to coat the rods with just sufficient B stage adhesive to fill the gaps between them. The amount of resin can be accurately calculated to ensure that as the resin liquefies during the cure process a small excess of resin is bled out leaving a perfectly consolidated component. Simple geometry shows that a circular coating applied over the rod should be 1.05 times the diameter of the rod in order to supply enough resin to fill the space between circular rods. In practice a slightly thicker coating would be used to allow a small excess of resin to bleed out, ensuring complete filling of all voids.

A second option would be to use rectangular pultruded rods with a minimum coating of B-stage resin. This would allow closer packing of the rods and even higher properties. The option of using rectangular rods is probably not so readily applicable if vacuum infusion is used as there is too great a risk of ending up with dry spots in the infusion where the rods push together.

The B-stage coating can easily be applied to the rods on site. This would eliminate the current shelf life / refrigeration issues that exist with current systems. The application of the B stage resin to rods can be considerably simpler than preparation of conventional pre-preg. Once the desired shape and thickness of resin has been determined it can be applied in a process analogous to wire coating or over-extrusion coating. A line to apply a pre-coat of epoxy onto the rods can be easily established with a wire coating die and tube mandrel to ensure even coating thickness around the rod.

Vacuum infusion and B-stage coating are the first choices discussed above primarily because they are essentially modifications to prior art techniques already well known within the wind-blade industry and therefore can probably be implemented most readily. While vacuum infusion and B-stage coating methods according to the instant invention

significantly improve on manufacturing deficiencies associated with the prior art and offer improved manufacturing cycles there are potential processes which could improve manufacturing cycles according to the instant invention even further by, for example, using a rod laying technology to position the continuous individual pultrusions, co-apply a UV- curable matrix, followed by a directed UV light source and pressure roller in a combined head, essentially positioning and curing the layer in a continuous operation. A head configured to position a 300mm wide by 1mm thick layer at the current pultrusion speed can, it is believed, deliver over 3000 kg of material an hour. Variations of the application technology allow for alternative matrix and cure energy sources. Pultrusions are critical in the instant invention. It is not proposed to elaborate significantly on the process of pultrusion as it is well known and well documented in the literature. However a few comments are helpful related specifically to the requirements in the context of this project. The typical spars will utilize considerable lengths of the proposed small diameter rods. A typical 1.5MW blade would use over half a million feet of rod in a full spar set. A typical production time for a blade is around one day. The following two options are suggested to achieve the required material: (a) conventional pultrusion with multiple exit stacked dies; and (b) to utilize u/v curing pultrusion which can operate significantly faster with line speeds up to 150 meters per minute or more.

In order to make full use of the improved spar properties according to the instant invention it is necessary to have a good connection to the hub. The root joint is a critical component of the blade structure as it transfers all of the loads from the blade to the turbine. Root joints may be made as a separate component and subsequently joined to the spar and blade or may be produced integrally. Separate manufacturing is simpler and reduces the risk of defects but adds a critical bonding step. A typical approach is to bond studs or 'T' bolts into holes drilled into the thick section of the root allowing the blade to be bolted onto the hub. The root section is typically very thick, 10cm or more, and uses multiple bi-axial layers of reinforcement to match the complex stress states that are present in the hub. Since the applied loads are so high, it is important to maximize fiber continuity and minimize section changes which would cause stress concentrations.

It would be possible to produce a separate hub as is done with some existing blades and avoid the complexity of utilizing the rods in a hub. However the ability to continue the very high strength reinforcing rods directly to the hub is potentially very valuable. A joint in which the multiple rods which are combined to form the spar are split and spread out to surround the stud anchor points allowing a very direct load path is believed to be beneficial in the instant invention. Obviously this adds a degree of complexity in terms of rod placement and manufacturing. However it is felt that the potential benefits of such an arrangement where the studs are connected more directly to the high strength pultrusions probably justifies this approach. The objective is to produce a joint in which the shear area and strength of the bond is sufficient to cause tensile failure of the stud rather than a pull out of the stud. Figs 11-14 depict the above-mentioned construction. Fig. 11 is a side view showing the stud 24 and the rods 25. Fig. 12 is a cross sectional view of the stud 24. Fig. 13 is a cross sectional view of the construction at a location where the rods 25 surround the stud 24. Fig. 14 is a cross sectional view of the construction at a location before the rods 25 surround the stud 24.

A more conventional approach of simply drilling into the thick walled composite can, of course, also be used in the instant invention. This approach is less elegant and certainly less structurally efficient however it avoids the added complexity of attempting to place the continuous rods around the studs.

Almost all wind blades use a conventional T section, 'C section or box section spar with flanges that are as flat within the limitations of the airfoil profile. This maximizes the spacing between the spacing between the flanges thus maximizing the second moment of area for a given amount of material. It also happens to be very convenient for laying down multiple layers of stitched or woven mat to form the spar caps. For maximum performance these spar sections should be designed such that bending failure results from pure tension or compression in the flange. In practice close to failure, some buckling of the spar flanges is typically observed. This buckling may be prevented or reduced by slightly modifying the shape of the flange to create slightly thicker spots at the outer edges to minimize buckling. This is relatively easy to achieve using the instant invention. Fig. 15 shows a simplified cross sectional view of a wind blade 26 having an "I" section spar comprised of an upper spar cap 29, a lower spar cap 28, a spar web 30 and an airfoil skin 27.

There has been much discussion and experimentation with hybrids of glass and carbon fibers. Such hybridization using the instant invention is very simple by substituting individual rods in different materials as the lay-up stack is produced. It is also much easier to tune a section by placing higher property rods at critical positions. As both this technology and understanding of overall blade structural performance improves, this hybridization can be extended, allowing gradually varying properties across the spar, e.g. transitioning materials (carbon to S glass to E glass) or transitioning fiber loadings, to maximize the advantage of high performance expensive materials or to allow a softer transition between spar properties and fairing properties.

Most blades are limited by flap-wise bending with chord-wise bending being less of an issue due to the large chord dimension, however as blades become larger and higher in aspect ratio chord-wise bending can become an issue. The instant invention can be a relatively simple means to incorporate 'mini-spars' in the leading and trailing edges if necessary to improve chord-wise properties.

In many instances spars and blade shells are produced separately and bonded together. This simplifies the handling of different reinforcements, allows better control of flow for reinforcements with very different diffusivities and allows more careful consideration for thick sections in spar roots which may otherwise create exotherm issues. It is contemplated that the combination of rods of the instant invention and broadgoods of the prior art will allow a more efficient approach to combining spars and blade shells in a single component.

Addition of components with functionality other than structural is facilitated by the use of the instant invention. Such components could include rods incorporating strain sensors, hollow tubes to allow for in situ sensors such as accelerometers within the blade or conductive elements for lightening protection. These elements can be incorporated either by substituting them for an individual rod or by incorporating them directly into the pultrusion.

In its broadest scope it should be understood that the instant invention is any spar or beam comprising a plurality of separate composite rods bonded together in a mold to form the spar or beam including but not limited to aircraft wings, aircraft rotors, marine spars, marine masts, and civil engineering applications such as bridge beams.

CONCLUSION

While the instant invention has been described above according to its preferred embodiments, it can be modified within the spirit and scope of this disclosure. This application is therefore intended to cover any variations, uses, or adaptations of the instant invention using the general principles disclosed herein. Further, the instant application is intended to cover such departures from the present disclosure as come within the known or customary practice in the art to which this invention pertains.