BACKGROUND OF THE INVENTION When designing structures of composite materials, the optimum material with regard to strength, weight, E-modulus and cost etc. is often not the same for all parts of the structure. For example, in a spar for a wind turbine blade the preferred material for the base part may be a glass fibre reinforced composite due to low cost and limited me- chanical requirements whereas the preferred material for a load bearing outer part like a flange may be carbon fibre reinforced composite due to the higher stiffness and lower weight. The physical properties, for example stiffness and thermal expansion, are very different, however, and it is the general conception in the art that such parts cannot be effectively connected.

OBJECTS OF THE INVENTION It is the object of the invention to provide an effective permanent connection between two composite members for example being reinforced with different reinforcement fibres.

It is another object of the invention to provide a method for preparing such connections.

DISCLOSURE OF THE INVENTION The above and more objects are realised by the invention as described and explained in the figures, preferred embodiments and claims.

The connection according to the invention may provide a stable relative fixing of a first composite member and a second composite member having substantially the same

properties. It will be appreciated, however, that the connections may also be between composite members having rather different properties. For example composites being reinforced with carbon fibres, which typically have coefficients of thermal expansion (hereinafter CTE) near of even below zero, may be connected to composites being re- inforced with g) ass fibres, which have a substantial higher CTE.

The invention concerns a connection between a first composite member comprising first type fibres and a first resin, and a second composite member comprising second type fibres and a second resin. The connection comprises a transition zone having a layered structure. By transition zone is meant the volume of the connection that does not have the same composition and the same structure as the bulk of either of the connected composite members. The transition zone comprises the volume between the two com- posite members and may in some cases also include a part of one or more of the com- posite members.

By first composite member is referred to the part of the final connected structure having a structure substantially corresponding to an equivalent separately prepared first com- posite member and vice versa for a second composite member. Any one of the com- posite members may hence be prepared partially or fully in advance of the formation of the connection or the composite member may be prepared as an integrated part of the process of preparing the connection.

The volume between the composite members may optionally be fully or partially occu- pied by a transition member. This transition member may be prepared prior to the estab- lishing of the connection, it may be prepared directly in the transition zone as part of the establishing of the connection or any combination of these.

Each of the composite members may, independently, be uncured, pre-consolidated, partially or fully cured at the time of establishing of the connection. Likewise, a transition member, which is partially or fully prepared prior to the establishing of the connection, may be uncured, pre-consolidated, partially or fully cured at the time of the establishing of the connection. However, it is preferred that a transition is not fully cured at the time of the establishing of the connection.

The connected structure may be established (i. e. laid up) in only one operation. By one operation is meant, that both the composite members and the transition zone is pre- pared in connection to the other parts, typically but not necessarily followed by pre-con-

solidation or co-curing of the complete structure. This type of preparation is particularly useful when a transition member is not used.

Another way to establish the connection is to connect two members. Examples thereof may be : Positioning composite members near or in contact with each other and establishing a transition zone, optionally via a transition member, prepared'onsite'.

Connecting a composite member, which is prepared for connecting for example by hav- ing an incorporated partial transition zone, to another composite member which, option- ally, is also prepared for the connection ; Connecting a composite member, which is prepared with a partially or fully integrated transition member, to another composite member which, optionally, is also prepared for connection; Connecting a composite member to a transition member and preparing a further com- posite member directly onto the transition member; etc.

Alternatively, the connection may be established by combining three or more separate members (e. g. two composite members and one transition member, three composite members and no transition member, three composite members and one transition member, etc.).

Connections according to the invention may be used for connecting composite members in any direction relative to the main fibre orientation or orientations (e. g. parallel side by side, at an angle (orthogonal or any other) or parallel end to end). If fibres or layers comprising fibres are interlaced into a composite member, it is preferred, however, that the interlaced fibres are oriented at an angle to the interface between the composite members or at an angle to the interface between the composite member and the transi- tion member. This will tend to increase the mechanical strength of the contact relative to a parallel orientation.

The term composite member herein denotes any type of composite material comprising fibres, cured or uncured, irrespective of the structure being layered or not. Pre-forms and pre-consolidated pre-forms-cured or uncured-are important subgroups of com- posite members. In a preferred embodiment, a first composite member comprises first

type fibres and first type resin, and a second composite member comprises second type fibres and second type resin.

A transition zone as well as a transition member comprises a resin and fibres. The fibres may be provided in any suitable form including in prepregs, semi-pregs, woven or non- woven fabrics, mats, pre-forms, pre-consolidated pre-forms, individual or groups of fi- bres, tows, tow-pregs, etc. During lay-up (i. e. preparing up to the point before consolida- tion and/or curing of the resin) of a transition zone or a transition member, resin need not be comprised in the layers comprising fibres (e. g. a prepreg or semi-preg) or be- tween the layers comprising fibres. However, the resin should form a continuous matrix after the curing. Resin need not be comprised in or between two adjacent layers com- prising fibres. In a preferred embodiment an adhesive may in this case be provided be- tween at least some of such pairs of layers to at least temporarily and at least partially fix the adjacent layers comprising fibres.

By prepreg is meant a substantially or fully impregnated collection of fibres, fibre tows, woven or non-woven fabric etc. By semi-preg is meant a partially impregnated collection of fibres or fibre tows. The partial impregnation provides for enhanced removal of gas through or along the dry fibres during consolidation and/or curing. An example of a semi- preg is a layer of fibres (e. g. glass fibres or any other type of fibres mentioned herein) partially impregnated in the upper part and/or in the lower part. Woven and non-woven fabrics are collections of individual fibres or fibre tows which are substantially dry, i. e. not impregnated by a resin. Fibre tows are bundles of a large number of individual fi- bres, e. g. 1,000's, 10,000's or 100,000's of fibres. Tow-pregs are at least partially im- pregnated fibre tows.

It is within the scope of the invention to connect three or more composite members with the connection and the method according to the invention as this is considered a col- laboration of a number of connections according to the invention.

The transition zone or transition member may be prepared as a pre-form as described below. A pre-form is a composite material comprising fibres and-unless otherwise stated-an uncured resin. The fibres are preferably provided in layers of oriented fibres lilce for example individual or groups of fibres, fibre tows, fibre tow-pregs, prepregs, semi-pregs, woven or non-woven fabrics or mats. Individual fibres, fibre tows and fibre tow-pregs may in some cases be advantageous over prepregs, since the individual fi- bres are less bounded and hence may rearrange easier during subsequent processing.

Furthermore, individual fibres, fibre tows and tow-pregs may be advantageous over prepregs in that they may be provided in the pre-form with a greater freedom of mixing and orientation, the price is lower as well as the amount of waste may be lower. The pre-form preferably comprises at least three layers of oriented fibres. Pre-forms having a higher number of layers like e. g. 4, 5, 8, 10, 15, 20, 50, 100 or more layers may be used within the scope of the invention.

By fibres are hereinafter meant particles having an aspect ratio (length/equivalent di- ameter) of more than 10. By equivalent diameter is meant the diameter of a circle having the same area as the cross sectional area of the particle. However, in a preferred em- bodiment, the fibres are continuous fibres, i. e. fibres that substantially run from one edge of a pre-form or member to another. The properties of a fibre reinforced composite de- pend to a large extent on the properties of the fibres. However, the properties of different types of fibres vary considerably. For example, the coefficient of thermal expansion of carbon fibres is very low, and in some cases even negative. The first type fibres and the second type fibres may be any type of fibres having an influence on the properties of the composite member; it is preferred, however, that the fibres are selected from the group consisting of carbon fibres, glass fibres, aramid fibres, synthetic fibres (e. g. acrylic, polyester, PAN, PET, PE, PP or PBO-fibres, etc. ), bio fibres (e. g. hemp, jute, cellulose<BR> fibres, etc. ), mineral fibres (e. g. RockwoolTM, etc. ), metal fibres (e. g. steel, aluminium,<BR> brass, copper, etc. ), boron fibres and any combination of these.

In a preferred embodiment the first type fibres are carbon fibres and the second type fi- bres are glass fibres or the first type fibres are glass fibres and the second type fibres are carbon fibres. It is particularly interesting to connect composites with these fibre types as carbon fibres are very stiff and light and glass fibres are very affordable. In an- other closely related embodiment, the first type fibres are carbon fibres and the second type fibres are glass fibres or vice versa and both the first and the second type resin are epoxy-based.

By carbon fibres is hereinafter meant fibres where the main component is carbon.

Hence, by this definition carbon fibres comprise fibres with graphite, amorphous carbon or carbon nano-tubes. Thus, carbon fibres produced via for example a polyacrylonitril- route or a pitch-based route are comprised by this definition.

The fibres comprised in the transition zone and/or the transition member and/or the composite members being connected may be a mixture of more than one type of fibres.

For example, a combination of glass fibres and carbon fibres may be used, but any combination of two or more of the fibre types mentioned herein is feasible. The mixture may be homogeneous, with different concentrations in separate fibre layers or areas or with different concentrations of fibres within any fibre layer. Mixing of fibres may be ad- vantageous, since this opens for tailoring of material properties, for example from a combined stress/cost-perspective.

The resin for the composite members, the transition zone and the optional transition member may be provided as liquid, semisolid or solid resin. The resin may be a thermo- plastic or a thermosetting resin, it is preferred to use a thermosetting resin, however, for reasons of chemical and thermal stability as well as ease of processing. The resin may for example be based on unsaturated polyester, polyurethane, polyvinylester, epoxy, thermoplastics or similar chemical compounds, including combinations of these. In a preferred embodiment the first type resin and the second type resin have substantially the same composition. This is preferred as it reduces compatibility problems.

In a preferred embodiment of the invention, the resin is provided as a liquid and the resin is introduced by Resin Infusion, Resin Transfer Moulding, RTM, or Vacuum As- sisted Resin Transfer Moulding, VARTM, into an entity comprising several layers com- prising fibres (e. g. fibre tows or any other suitable collection comprising fibres mentioned herein). The entity comprising fibres may for example be a transition zone, a transition member, a composite member or any combination of one or more of these. Besides the layers comprising fibres, the entity may or may not further comprise a resin and/or an adhesive. In one embodiment, the entity comprises two adjacent layers comprising fi- bres having no adhesive or resin between the layers prior to the introduction of the liquid resin as described previously in this section. In another embodiment, adhesive and/or resin is provided between all layers comprising fibres in the entity prior to the introduc- tion of the liquid resin as described previously in this section.

The main function of the adhesive is to immobilise the fibres as they are placed on top of the adhesive. This can be achieved by having a tacky adhesive, whereby the fibres stick to the tacky adhesive. The adhesive may be any tacky material, or a solid with a tacky surface and the adhesive may for example comprise polyester, polyurethane, polyvinyl- ester, epoxy or similar compounds or a combination of these. It is within the scope of the invention to use any material or combination of materials having a tacky surface includ- ing solid materials with tacky surfaces. More than one type of adhesive may be used in

one member or transition zone. For example, it is within the scope of the invention to use the resin as an adhesive between layers of fibre tows, where a resin is provided, or to use a second type of resin below the first layer of fibre tows.

In another preferred embodiment, the resin is a solid. An entity comprising several lay- ers of oriented fibre tows, which may optionally have been immobilised previously during fibre laying by an adhesive, and a solid resin system is heated under vacuum in order to prepare a pre-consolidated or cured pre-form, which may be part of a transition zone, a transition member or a composite member.

In a further preferred embodiment, the resin is a semisolid and functions both as resin and as adhesive, i. e. during fibre laying, the resin will immobilise the fibres and during subsequent processing, it functions as a matrix material.

The resin may comprise more than one system, for example the resin may comprise two systems or even more systems. It may be advantageous to use more than one resin system to be able to optimise the properties of the resin for the subsequent steps of processing, for example with respect to viscosity and timing/controlling of the curing pro- cess. These systems may or may not be based on the same type of resin, however, it is preferred that such systems are based on the same type of resin such as two or more epoxy-based systems. In another preferred embodiment, the resin types differ but the resins are compatible. In a further preferred embodiment, the resin comprises two sub- stantially epoxy-based systems. The two epoxy-based systems may comprise a com- mon component. The common component may for example be a common catalyst, a common amine component or a common epoxy component, however, it is preferred that the common component is an epoxy component. A resin comprising two epoxy-based systems with a common epoxy component may comprise an amine-component of a first epoxy-based system that will react to the common epoxy component at a first relatively low temperature, such as below 50°C and preferably about room temperature. At this first temperature, a second epoxy-based system is preferably non-reactive or the reac- tion takes place at a very low rate. Since the reaction rate of the second epoxy-based system should be very low, the second epoxy-based system may advantageously be catalysed by a catalyst, which is non-active until activated. This activation may for ex- ample be by UV-light, by addition of a compound or by heating ; it is preferred, however, that the catalyst is activated by heating.

The resin may for example be distributed as discrete points, as random or organised

lines, continuous or non-continuous layers or areas or any combination of these. Fur- thermore, the resin or additional resin may be infused during processing.

Besides fibres and resin, composite members, the transition zone and-if present-the transition member may for example comprise one or more of fillers (e. g. a cheap inert material) and/or solvents and/or diluents and/or rheological agents and/or viscosity ad- justing agents.

Traditionally, gas enclosed in the pre-form prior to and during curing has been removed in the direction of the fibres, i. e. in the plane of a resin layer. Hence, the larger the structure, the longer the gas has to travel to be released from the structure. The risk that gas becomes trapped inside a cured structure is hence increased with the size of the structure. It appears that the problem with entrapped gas is particularly pronounced when the reinforcing with unidirectional fibres. It may be speculated that this is due to the very close packing of the fibres, which may arise in some areas of a composite re- inforced by unidirectional fibres. However, problems concerning entrapped gas may also be present in other types of fibre orientation e. g. biaxial or random orientations. In a preferred embodiment of the present invention the connections are therefore prepared to enhance removal of air from within the connections in a direction substantially or- thogonal to the surface. This may for example be realised by having a non-continuous layer of resin within the transition zone.

By gas is herein meant entrapped atmospheric air as well as gaseous products, by- products and starting materials related to the preparation process.

The method according to the invention may be adapted to automated processing. For example, in the production of a transition zone or a transition member, a robot may ad- vantageously distribute layers comprising fibres, resin and, optionally, adhesive. Auto- mation is facilitated by an at least partial immobilisation of fibres during laying, by an ad- hesive which will prevent or at least greatly reduce disturbance in the layers comprising fibres. Furthermore, when the adhesive is only applied to selected areas of the ground plan of the transition zone or the transition member, time is saved compared to distribu- tion of resin over the entire ground plan.

Resin systems may contain components, which may be irritant or harmful when in con- tact with naked skin, if ingested or inhaled. Avoidance of direct contact is therefore highly desirable. Since the processes according to the invention are particularly suited

for automation, the products and processes according to the invention represent a sig- nificant improvement to the working environment.

BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 shows the general conception of the angle between the overall plane of the inter- face between the transition zone and one of the composite members.

Fig. 2 shows a preferred embodiment of a connection according to the invention having macro grading.

Fig. 3 shows a preferred embodiment of a connection according to the invention having micro grading and details of a micro graded layer.

Fig. 4 shows a further preferred embodiment of a connection according to the invention having interlaced layers comprising fibres.

Fig. 5 shows yet another preferred embodiment of a connection according to the inven- tion.

Fig. 6 shows sketches of some other preferred embodiments of connections having in- terlaced layers comprising fibres.

Fig. 7 shows a sketch of the stress distribution near the end of a fibre.

DESCRIPTION OF THE DRAWINGS All the figures are highly schematic and not necessarily to scale, and they show only parts which are necessary in order to elucidate the invention, other parts being omitted or merely suggested.

In some of the figures separate layers comprising fibres are indicated. To increase clar- ity the number of layers shown is limited, and in real connections and composite mem- bers the number of layers may in some cases be considerably higher like for example 10's or 1 00's. The distance between the layers, the thickness of layers and the angles are examples of schematic parts.

The physical properties of fibre reinforced composites are to a large extent dictated by the reinforcement fibres. This includes properties like the coefficient of thermal expan-

sion, CTE, and Young's modulus. Hence, a thoroughly connection is for example rele- vant when a composite comprising carbon fibres are connected to a composite rein- forced by another type of fibres, since the CTE of carbon fibres is very low and may even be negative. However, the same type of connections may be used for strong con- nections between composites reinforced by other types of fibres. The fibres may be any of the fibre types mentioned previously in this description. A connection according to the invention may for example be used for connecting a carbon fibre reinforced composite to a glass fibre reinforced composite.

The connection according to the invention is best appreciated when a significant differ- ence in properties exists between the composite members to be connected, e. g. a first and a second composite member. For example the difference in CTE of the composites may be greater than 3x104°C-1 or even greater than 5x10-6°C-1 or the difference in Young's modulus may be greater than 25% of the lower value for the composites or even greater than 100% of the lower value. In many cases composites having even greater differences in properties may be connected by a connection according to the present invention. However, a connection according to the invention is also usable when the differences in the properties are smaller.

The transition zone should have a layered structure, i. e. comprise layers comprising fi- bres. The number of layers may vary considerably dependent on the design of the con- nection and the size and type of composite members. In some cases only a few, for ex- ample 2,3, 4,6 or 10 layers are used whereas in other cases a higher number, for ex- ample 20,30, 50,100 or more layers are needed to obtain the desired quality of the connection. The degree of laminated structure is often reduced during processing. For example pre-consolidation of a transition zone or a transition member tends to homoge- nise the structure. The structure of the transition zone is layered to enhance controlled orientation of the fibres, to facilitate optimisation of the reduction of stress between the composites as well as to facilitate ease of production like for example enhance suitability for automation of the production.

The fibres may be provided in any desirable orientation in the transition zone like for ex- ample unidirectional, biaxial or random. However, the fibres are preferably oriented to reduce the stress between the transition zone and the composite members and/or re- duce the stress between the composite members as well as to strengthen areas of the final structure which will be exposed to a higher stress during service.

The orientation of fibres may or may not be the same in all the layers comprising fibres within the transition zone ; however, in a preferred embodiment the fibres are oriented substantially the same way in all the layers of fibres. One or more layers of fibres may for example be oriented in another manner than other layers, if a stress analysis sug- gests a multi-axial fibre orientation. In another preferred embodiment, layers comprising first type fibres are mainly unidirectional and layers comprising second type fibres are mainly biaxial. Another way to strengthen an area of the final structure which will be ex- posed to higher stress during service, is to increase the amount of fibres in that area.

The resin may for example be distributed as discrete points, as random or organised lines (preferably at an angle to the fibre orientation), continuous or non-continuous lay- ers or areas or any combination of these. In a preferred embodiment, gas may escape orthogonal to the direction of the fibres through non-continuous layer of resin. Resin or additional resin may, furthermore, be infused during processing.

The resin may be provided between two layers comprising fibre for instance individual or groups of fibres, fibre tows, fibre tow-pregs, prepregs, semi-pregs, woven or non-woven fabrics or mats. This is the preferred positioning of resin and when this positioning is used, it is highly desirable that the resin is distributed in non-continuous layers. How- ever, the resin may also be provided in contact with only one layer comprising fibres, i. e. at the top or at the bottom of the transition zone or transition member. In this case, it is preferred to provide the resin at the bottom of the transition zone or transition member and the resin may be provided in a continuous layer as gas will not usually have to es- cape through the resin layer. In a preferred embodiment, resin is only provided at the top and/or the bottom of the transition zone or transition member, i. e. only adhesive is pro- vided between the layers comprising fibres. In another preferred embodiment, resin is only provided at the bottom of the transition zone or transition member, i. e. between the layers comprising fibres only adhesive is provided.

In a preferred embodiment, the resin is provided to form a non-continuous layer or lay- ers, as this facilitates removal of gas during a subsequent consolidation and/or curing of the transition zone, the transition member or the composite member. The resin is pref- erably semi-solid and may stick to and/or at least partially immobilise fibres of one or more layers. In another preferred embodiment, the resin is distributed to form a pattern of solid or semisolid particles, and these particles may for example be sprinkled over a layer of fibres. Alternatively, discrete resin or adhesive dots may for example be formed

from a resin provided as a liquid. A liquid resin may also be provided as a line or several lines, which may form an oriented pattern, a random pattern or a combined pattern. A different approach to a non-continuous layer of resin is a sheet of resin wherein a num- ber of through-going holes ! es are provided. As it is obvious from these exampfes of resin patterns, a person skilled in the art will be able to provide other patterns without depart- ing from the inventive idea of the invention.

Examples of preferred embodiments of continuous layers of resin are -layers prepared by distribution of a liquid resin, -layers prepared by infusion of resin into or between layers comprising fibres, - layers prepared from a sheet of solid resin, some but not all prepregs, etc.

The adhesive should at least partially immobilise the fibres in contact with the adhesive.

The adhesive may be any type of adhesive. The adhesive should be compatible to the resin, however, and is preferably a resin-type adhesive and more preferably the adhe- sive is related to the resin of the transition zone, the transition member or the composite member in the sense that it comprises compatible chemistry. One way to ensure com- patibility between the resin and the adhesive is to use substantially the same composi- tion. In a preferred embodiment, the composition of the adhesive is the same as the composition of the adhesive. It is within the scope of the invention to use more than one type of adhesive in a transition zone, a transition member or a composite member. For example some portions of adhesive may have the same composition as the resin, whereas other portions may have a different composition.

The adhesive may in principle be provided in the same patterns as the resin, however, it is preferred to provide a less dense pattern for the adhesive to save time. It is important to keep in mind that the purpose of the adhesive is to ensure that the layers comprising fibres are at least partially immobilised to facilitate fibre laying. Furthermore, the adhe- sive will often increase the mechanical strength and hence improve the handleability of an unconsolidated and uncured pre-form relative to the equivalent structure without ad- hesive by at least partially fixing adjacent layers of fibre to each other. One way to en- sure a facilitation of the fibre laying is to provide a strip of adhesive close to or exactly where the layers comprising fibres are initiated during fibre laying. Other embodiments are also feasible such as dots, broken or curved lines, etc. In principle continuous layers may be used; however, the advantage of using an adhesive over using a resin will be

reduced thereby. In some cases, automation may favour an adhesive pattern, where the adhesive is applied in a continuous line, for example as a zigzag or criss-cross pattern.

A person skilled in the art will appreciate the advantage of providing only a limited amount of adhesive compared to either a fu ! ! or nearly full layer of resin or cross-ply stitching, particularly with respect to the time saved during processing and ease of automation.

Three-dimensional forming may be applied for reducing the interfacial stress in a con- nection between composite members. A three-dimensional form may for example be re- alised by selective initiating and/or terminating of for example fibre tows or other types of layers comprising fibres during fibre laying. Traditionally, such three-dimensional forms will involve the creation of a large contact area orthogonal to the main stress direction in at least one axis. In Fig. 1 it is observed that the contact area may be increased by de- creasing the angle a between the overall plane of the interface 2 between the transition zone and a composite member 10 and a surface 4 of the composite member 12. A dis- tance much greater than the orthogonal distance between adjacent layers comprising fibres separates the terminations of the layers of fibres in the tapered section (2,4), which will tend to reduce the interfacial stress. Furthermore, if the angle a is sufficiently low, the end of a layer comprising fibres of the first composite member may be placed over an adjacent layer comprising fibres from the second composite member in the tran- sition zone. This may lead to a side-by-side coupling of the fibre layers, which is favour- able compared to an end to end coupling.

The fibres in a composite member near the transition zone may for example be oriented substantially parallel to the overall plane of the interface between the transition zone and that composite member, the fibres may be oriented substantially parallel to a surface of that composite member, the fibres may be oriented in a combination of these orienta- tions, the fibres may be provided with a random distribution, etc. It is preferred, however, that the fibres in the composite members near the transition zone are substantially par- allel to a surface of the composite member as this may provide an-even if limited- side-by-side coupling between the fibres. This is particularly the case when the angle a is low. Preferably the angle a should be less than about 10° but an even better connec- tion may by obtained when the angle is less than about 2°. If the utilised fibres are very stiff such as carbon fibres a may in some cases advantageous be as low as 0. 5° to 1 ° or even lower.

If thick composites are to be connected, low angles may be difficult to realise as this will require an unreasonable long transition zone. In such situations it may be advantageous to use a zigzag interface or to use a number of connections having the angle a on top of each other. This may for example be realised by separating the thick composite mem- bers into a number of thinner composite members or flanges and then connect the thin- ner pairs of composite member or flanges individually on top of each other or near each other.

The transition zone may be prepared in several ways. In a preferred embodiment, the transition zone comprises a transition member. The transition member is composite material comprising layers comprising fibres and a resin. In a preferred embodiment the transition member is prepared independently of the composite members to be con- nected. In this embodiment the transition member may for example be prepared as a pre-form and optionally be pre-consolidated prior to the connection to the composite members. A method for preparing a connection involving a transition member may com- prise the steps of: - providing at least one layer comprising first type fibres for the transition member - providing at least one layer comprising second type fibres for the transition mem- ber - providing a resin in contact with at least one of the layers comprising fibres for the transition member - connecting the transition member to the first composite member - connecting the transition member to the second composite member - curing the transition member, and -optionally co-curing the first and/or the second composite member with the curing of the transition member.

The layers comprising first type fibres or second type fibres, respectively, may be the same layer or layers if the fibre types are mixed. The resin may or may not have the same composition as one of the composite members, however, it is preferred that the resin is compatible with the resin of the composite members and more preferably the resin composition is substantially the same as the composition of one or both of the

composite members. The resin is preferably but not necessarily non-continuous in the sense that gas may be removed from the transition member orthogonal to a layer of resin. The transition member may for example also comprise an adhesive between two or more of the layers comprising fibres if this is desirable. One or more of the composite members may be co-cured together with the curing of the transition member. In a pre- ferred embodiment, the complete structure is co-cured in one operation.

In a preferred embodiment, the transition member is pre-consolidated, preferably prior to the connection of the composite members. Pre-consolidation of the transition member is particularly advantageous when at least one of the composite members to be connected is cured prior to forming of the connection, as pre-consolidation may reduce shrinkage during curing.

In a preferred embodiment an uncured or pre-consolidated transition member may be transported or stored for a prolonged period of time such as weeks or months without premature curing.

The layers of the layered structure of the transition zone are typically oriented substan- tially parallel to the overall plane of the interface between the transition zone and at least one of the composite members or the layers are oriented substantially parallel to a sur- face of at least one of the composite members. In some embodiments both options are possible and hence the design should be chosen depending on the desired properties of the final structure and the types of composite members involved.

One approach to the designing the transition zone is to provide a gradual change of the composition of the layers comprising fibres from the composition of the first composite member to the composition of the second composite member. In Fig. 2 this gradual change is realised by macro-grading. By macro-grading is meant that the transition zone comprises a number of layers each comprising fibres having either substantially the same fibre composition 14 as the first composite member 10 or substantially the same fibre composition 16 as the second composite member 12. In a preferred embodiment a first side of a stack of such layers is connected to the first composite member 10 and a second side of the stack is connected to the second composite member 12. Hence, the layers are typically oriented parallel to the overall plane of the interface between the transition zone and at least one of the composite members.

The transition zone shown in Fig. 2 are preferably prepared directly onto one or both of

the composite members, but other routes of preparation are feasible for example the route involving preparation of a separate transition member. The interface between the first composite member and the transition zone need not be parallel to the interface be- tween the second composite member and the transition zone, however in a preferred embodiment the interfaces are substantially parallel since this provides for an easier manufacturing.

The gradual change may for example be realised by changing the frequency of the two types of layers comprising fibres for example as it is shown in Fig. 2. Other examples are sequences like : Bulk A-B-A-Bulk B <BR> <BR> BulkA-B-A-A-B-A-B-B-A-BulkB<BR> BulkA-B-A-A-A-B-A-A-B-A-B-B-A-B-B-B-A-BulkB Here A indicates a layer comprising fibres having substantially the same fibre composi- tion as the first composite member, B indicates a layer comprising fibres having substantially the same fibre composition as the second composite member and Bulk indicates the composite member.

The sequence need not be symmetrical and the optimum sequences should be chosen with respect to properties, layer thickness, etc. in the particular situation. The examples of sequences is by no means exhausting for useful sequences and a person skilled in the art will be able to provide a variety of sequences without departing from the concep- tion of the inventive idea.

When the transition zone is prepared directly on one or more composite members the connection may be prepared by these steps: - providing a first composite member providing at least one layer comprising first type fibres on or in connection to the first composite member or on a previous layer comprising fibres in the transition zone - providing at least one layer comprising second type fibres on or in connection to the first composite member or on a previous layer comprising fibres in the transi- tion zone

providing a resin in contact with at least one of the layers comprising fibres in the transition zone connecting the structure to the second composite member curing the transition zone, and optionally co-curing the first and/or the second composite member together with the with the curing of the transition zone.

The transition zone may be prepared directly on both the first and the second composite member simultaneously if this is desirable.

A method for preparing connections comprising interlaced layers comprising fibre may for example comprise these steps: - providing layers comprising first type fibres - providing a means for at least partially immobilising the fibres, said mean for ex- ample comprising an adhesive or a resin - providing layers comprising second type fibres to be at least partially interlaced between layers comprising first type fibres, said layers comprising second type fi- bres extend beyond the first composite member - providing a resin in contact with at least one of the layers comprising fibres - connecting the structure to the second composite member -curing the transition zone and the first composite member, and - optionally co-curing the second composite member together with the with the cur- ing of the transition zone.

In a similar way the transition zone may be partially incorporated into the second com- posite member.

The transition zone may be pre-consolidated prior to the curing optionally the pre-con- solidation is conducted on the transition zone and at least one of the composite mem- bers simultaneously.

The fibres for the layers comprising fibres may for example be provided as prepregs, semi-pregs, woven or non-woven fabrics, mats, pre-forms, pre-consolidated pre-forms, individual or groups of fibres, tows, tow-pregs or a combination of these.

Another approach to gradually changing the composition of the layers comprising fibres involves micro-grading. By micro-grading is meant that the first type fibres and the sec- ond type fibres are mixed in at least one layer in the transition zone comprising fibres.

Alternatively, a layer comprising fibres or a mixture of fibre types having intermediate properties of first type fibres and second type fibres is present ; Particularly properties critical to the connection (e. g. CTE and Young's modulus) should be intermediate to those of the first type fibres and the second type fibres. An example of micro grading is shown in Fig. 3A where the connection has two micro-graded layers 20a and 20b be- tween the first composite member 10 and the second composite member 12. As it is in- dicated with the relative shading of 10,20a, 20b and12 in Fig. 3A it is preferred that the ratio of first type fibres to the second type fibres decreases gradually from the first com- posite member towards the second composite member. The micro-graded members 20, 20a and 20b may for example be prepared from a thick fibre mat comprising a homoge- neous mixture of first type fibres and second type fibres and a resin, but the person skilled in the art will know other ways to realise a micro-graded material. The micro- graded members may for example be prepared from a number of thinner layers com- prising fibres. These thinner layers being shown in Fig. 3B and C as dark lines 22.

These layers may generally be oriented parallel to a surface of the final structure as in- dicated in Fig. 3B or parallel to a larger surface of the micro-graded member as indi- cated in Fig. 3C. It is also within the scope of the invention for example to prepare the transition zone or a transition member directly from thinner micro graded layers com- prising for example one layer of fibres, one layer of fibre tows or a prepreg.

In a preferred embodiment of a micro graded connection, the fibres within a micro graded layer comprising a mixture of first type fibres and second type fibres are distrib- uted in a homogeneous or non-homogeneous manner within an individual layer. In Fig.

3D an example of a homogeneous distribution is shown. It is observed that the concen- tration of tows 15 of first type fibres and tows 17 of second type fibres is constant. The orientation of the first and the second type fibres need not be the same even it this is shown in Fig. 3D. In Fig. 3E an example of a non-homogeneous distribution of fibre tows is shown. It is observed that the concentration of tows 15 of first type fibres relative to tows 17 of second type fibres is much greater near the edge than near the centre of the

layer. Other sources for fibres may equally well be used instead of or together with fibre tows. In non-homogeneous layers, the concentration and/or the orientation of the first type fibres are preferably different from those of the second type fibres, but other exam- pies of inhomogeneous manners may be provided by a person skilled in the art in the light of these examples. This type of layers may for example be advantageous if more than two composite members are to be connected or if at least one of the composite members is inhomogeneous.

Fig. 4 shows a transition zone comprising two types of layers or sheets comprising fi- bres, one type 50 comprising fibres having substantially the same fibre composition as the first composite member 10 and another type 52 comprising fibres having substan- tially the same fibre composition as the second composite member 12. The two types of layers are partially interlaced in the sense that in a part of the transition zone at least one of the types of layers extend beyond the other type of layer. At least some of this part of the transition zone is used for the connection with the composite member having substantially the same fibre composition. Due to the interlaced layers the contact sur- face is very high and the connection is hence under the right conditions very strong. In a preferred embodiment both types of layers extend beyond the other in a part of the contact zone as shown in Fig. 4.

In a preferred embodiment the transition zone is integrated at least partially with one of the composite members or a pre-form. In Fig. 5 an example of such an embodiment is shown. The transition zone may for example comprise layers 62 comprising second type fibres initiated within the first composite member being reinforced by layers 60 compris- ing first type fibres and extending beyond the first composite member. The distance 64 from the ends of layers comprising first type fibres to the adjacent ends of layers com- prising second type fibres should preferably be great enough to prevent or reduce cou- pling of stress between the layers (see below). It is also preferred that the distance 66 between the ends of adjacent layers comprising first type fibres is great enough to pre- vent or reduce coupling of stress between the layers. The transition zone having inter- laced layers comprising second type fibres is preferably prepared as a part of the prepa- ration of the first composite. In a preferred embodiment, the layers comprising second type fibres are provided as prepregs. The prepregs may be unidirectional prepregs, however, experimental results suggest, surprisingly, that biaxial prepregs comprising the second type fibres provide a better basis for connecting of the pre-form to a structure reinforced by second type fibres. It may be theorised that this is due to better coupling

between layers where the relative main fibre orientation is not parallel, however, other effects may take part in this.

Experimental results have shown that the embodiments described in relation to Fig. 4 and Fig. 5 are capable of successfully connecting a glass fibre reinforced base of a spar for a wind turbine blade to a carbon fibre flange reinforced spar. The success in this high performance application indicates the span of the present invention.

In a preferred embodiment of a connection, the transition zone comprises layers, which comprise fibres, extending from the first composite member into the second composite member and/or layers, which comprise fibres, extending from the second composite member into the first composite member. This preferred embodiment may for example be established in one operation. More preferred embodiments having interlaced layers comprising fibres are shown in Fig. 6. Fig. 6A shows a connection where the layers of each composite member comprising fibres extend into the other composite members 10 and 12. This type of connection is typically prepared simultaneously with the fibre laying, i. e. preparation of both composite members and the transition zone in one operation. In Fig. 6A and B all layers comprising fibres extend into the other composite member, however, this need not be the case unless it is required to achieve the desired strength.

For example the transition zone may become thicker if all layers are continued into the other composite, which again may lead to bending of fibres entering into the transition zone. In Fig. 6B the transition zone 13 has been indicated by a broken line as an exam- ple of the extent of a transition zone.

Fig. 6B shows an integrated connection between two composite members having an interface which is substantially orthogonal to the surface of the composite members. To reduce the effect of the stress end-condition (see below), the ends of the fibre layers are provided in a zigzag pattern.

In a preferred embodiment of the connection shown in Fig. 6B, a zigzag pattern is also provided in the dimension of the interface orthogonal to the plane shown in the figure. In other words, the line defined by the end of a layer comprising fibres is not a straight line in this preferred embodiment.

Fig. 6C shows an example where the composite member 10 is prepared for connection with composite member 12 in that layers 40 having substantially the same fibre compo- sition as composite member 12 has been interlaced into composite member 10. In Fig.

6D the connection between the composite members is formed. It is observed that the layers 40 are being bent to form the layers 42. The bending is not desirable, but the bending angle may be greatly reduced by decreasing the angle a, which has been de- fined in relation to Fig. 1. In a preferred embodiment, a is typically kept below 100 as de- scribed elsewhere. The layers 42 may or may not comprise a resin (e. g. a prepreg or applied directly), but during curing of the transition zone a resin should be present to en- sure formation of a cured matrix of resin.

A method for preparing a connection as shown in Fig. 6A and B wherein the transition zone comprises layers comprising fibres extending from the first composite member into the second composite member and/or layers comprising fibres extending from the sec- ond composite member into the first composite member, may for example comprise the steps of : - providing layers comprising first type fibres, at least some of these layers extend into a part of the structure resembling the second composite member - providing layers comprising second type fibres, at least some of these layers ex- tend into a part of the structure resembling the first composite member -optionally providing a means for at least partially immobilising the fibres, said means for example comprising an adhesive or a resin - providing a resin in contact with at least one of the layers comprising fibres - optionally pre-consolidating the structure, and - curing the structure In a preferred embodiment, the structure is co-cured altogether, however, curing or pre- consolidating may be carried out on selected parts of the structure in individual opera- tions.

When a fibre reinforced material is stressed by a force F, the matrix material, i. e. the resin, is stressed equally along the layers comprising fibres. However, if a layer com- prising fibres is stressed differently for example as it is the case for interlaced, non-con- tinuous layers comprising fibres stress may build up in the resin. In Fig. 7B a schematic illustration of the shear stress in the resin between four layers comprising fibres are shown. The layers 30a, 30b and 30d comprising fibres are continuous layers such as

continuous layers comprising second type fibres in the second composite member whereas the layer 30c comprising fibres is a non-continuous layer such as an interlaced layer comprising first type fibres interlaced from a first composite member into a second composite member. The mainly vertical lines represent the local shear stress in the resin, i. e. a vertical line indicates that the resin experiences the average shear stress and an angled line indicates a difference in the experienced shear stress from the aver- age resin. It is observed that a large local difference in the shear stress is present near the end of the layers 30c.

In Fig. 7A the shear stress build up near the end of layer 30c is represented as a sche- matic plot of the variation in the shear stress in the resin along a line parallel to and near layer 30c relative to the average shear stress in the resin. It is observed that difference in shear stress is greatest at the end of the layer 30c. Furthermore, it is observed that at a certain distance away from the end of layer 30c the shear stress is about the average level of shear stress. When ends of layers comprising fibres are provided closer to each other than the extent 34 of the end-condition of the shear stress, the shear stress may couple between the layers and hence form a weak point, line or plane within the struc- ture. To reduce or prevent coupling of shear stress from the end of one layer to the end of a nearby layer, the distance between the ends of interlaced layers should be greater that the extent of the end-condition. Since the extent of the end-condition is difficult to establish, it is preferred to use a safety margin and hence separate the ends of two ad- jacent layers by at least two times the extent of the end-condition.

It is also reasonable to ensure that the distance between the nearest layer end of the same type of fibre should be separated by a distance corresponding to the extent of the end-condition, preferably with a safety margin and hence using a factor of two.

The actual extent of the shear stress concentration depends on a number of factors, such as thickness of layers comprising fibres and layers comprising primarily resin, the type of fibres, the type of resin, etc. The extent of the stress concentration may for ex- ample be established by modelling or by empirical methods.

When larger structures comprising pre-forms or composite members are to be prepared, this may follow a method wherein the not fully cured material is shaped at least partially plastically. The pre-form may be connected to further pre-forms before or after shaping to provide a larger structure. The pre-form may also be connected to other structures. It is preferred but not required that the connections involve a tapered part or layers com-

prising second type fibres. When the materials have substantially the same properties it may be sufficient to use a tapered section as shown in Fig. 1 to realise a reasonable connection, however, the greater the difference in properties the greater care should be taken.

Relevant curing processes and processing parameters are known to the person skitted in the art, however, it is preferred that a pressure is applied on the structure and/or a vacuum is applied to the structure as part of the curing procedure. The pressure may be applied via a vacuum in a vacuum enclosure surrounding all or part of the structure. It is preferred that the pressure and/or vacuum is applied on the structure before the curing reaction is commenced as this facilitate the removal of gas from the structure.

The connections and particularly separately prepared transition members as discussed previously may advantageously be pre-consolidated to provide a more homogeneous material with better performance. Transition members may often be considered as pre- forms.

In a preferred embodiment, the transition member or the transition zone (in the following sections referred to as the element) is treated by pre-consolidation to form a pre-con- solidated element as described in the following section. Pre-consolidation is particularly useful when the fibres are provided as individual or groups of fibres, fibre tows or fibre tow-pregs compared to fibres provided in prepregs as a lower viscosity may be realised during the pre-consolidation process. This will increase the redistribution of resin and/or fibres, which is highly desirable as it increases the homogeneity of the resulting product.

However, pre-consolidation of elements comprising fibres provided in prepregs, semi- pregs, woven or non-woven fabrics and/or mats may also be advantageous.

By pre-consolidation is herein meant a process, whereby gas inside an element is re- moved and a low porosity element is produced. Pre-consolidation involves redistribution of a resin and optionally a redistribution of fibres. Furthermore, pre-consolidation may involve a limited curing of the resin. Pre-consolidation is particularly useful as it pro- duces a dense element (hereinafter named a pre-consolidated element). Pre-consoli- dated elements and composites prepared from pre-consolidated elements will be appre- ciated amongst others due to good reproducibility, low porosity, high homogeneity, high strength, ability to plastically shaping, ability to be connected to other elements and/or other structures, suitability for automation and long shelf life without premature curing.

When the pre-consolidation involves a limited curing, this limited curing may involve a release of up to 50% of the energy that will be released by a complete curing of the resin. However, it is preferred that the extent of curing is limited to an extent that will al- low the pre-consolidatcd eIement to be deformed plastically. The degree of curing that<BR> will allow for plastical deformation of a pre-consolidated element depends amongst oth- ers on the resin as well as on the fibre type and fibre content. Generally, it is preferred that the limited curing involves less than about 20% of the energy that will be released by a complete curing of the resin and more preferably that the limited curing involves between 3 to 15% of the energy that will be released by a complete curing.

Generally speaking, the pre-consolidation process should reduce the porosity of the element and it is preferred that the resulting porosity of the pre-consolidated element is less than 5% by volume, preferably less than 2% by volume and more preferably less than 1% by volume. In some cases, a porosity of even 1% may reduce the properties of a composite considerably. In these cases, it will be appreciated that the method and the pre-consolidated element may be produced with porosities well below 1 %. For example, a reproduced porosity of about 0.2% by volume was realised for a composite with 60% carbon fibres in epoxy. The reduction of the porosity may for example be a result of ex- posing the element to a pressure and/or a vacuum in relation to the pre-consolidation process.

The porosity of the pre-consolidated element can not be established directly, as a den- sity is not known and may vary throughout the material. Hence, the porosity should be established by an optical method on a materialographic sample. Preparation of materi- alographic samples from an uncured pre-consolidated element is very demanding, since the material comprises both a very soft element (i. e. a resin) and a very hard element (i. e. the fibre). To establish a reproducible result, it is hence necessary to cure the ele- ment prior to materialographic preparation. This curing should be pressureless to ensure that the porosity is unaffected by the process.

To ensure handleability, the pre-consolidated element should be substantially unsticky, i. e. it should be easily releasable from any relevant surface and it should not leave ex- cessive amounts of resin on a surface when released.

To ensure a long shelf life and/or stability during transportation it is important that the rate of the curing reaction of the bulk of the resin is sufficiently low at room temperature and that a catalyst-if present-is not activated by accident. For example, if the catalyst

is activated by heating, it should be ensured that the activation temperature is consid- erably higher than the expected maximum temperature during storage.

In a preferred embodiment, the pre-consolidated elements are at least partially deform- able. This may for example be realised through a balanced and limited curing during the pre-consolidation process. In a preferred embodiment, at least a part of a pre-consoli- dated element is capable of being bent around an axis parallel to the main fibre orienta- tion with a diameter of more than 1 cm, however, in some cases a pre-consolidated element may be bent with a diameter of more than 5 cm by plastic deformation. The low bending diameters may be realised by rearranging of resin and/or fibres or by three-di- mensional forming of an element. By three-dimensional forming is herein meant that the thickness (e. g. the number of layers or amount of fibres and/or resin) and/or the form of the ground plan is adjusted for a part of the element relative to the bulk of the element.

Typically, only a part of the pre-consolidated element is prepared for very sharp bend- ing, whereas bending around an axis with larger diameters, e. g. 50 cm, may often be realised by all parts of the pre-consolidated element.

The stiffness of a element realised during a pre-consolidation process should ensure that the pre-consolidated element is sufficiently stiff to prevent relaxation of the pre-con- solidated element in the length direction of the fibres when placed on a non-flat surface and yet allow for plastic deformation around an axis parallel to the longitudinal direction of the fibres.

The pre-consolidation process often leads to an increase in viscosity of the resin in the element, for example by a partial curing. It is preferred that the viscosity at room tem- perature is increased by a factor of at least two and more preferably by a factor of at least five, as an increase in viscosity will enhance handleability, strength and un- stickyness. In some cases, the viscosity may be increased by a much higher factor of for example 10,100 or 1000. This is for example the case if part of the resin is injected into the element as a room temperature liquid. Another way to express the increase in vis- cosity is to look at viscosity directly. It is preferred that the viscosity of the resin in the unconsolidated element is between about 10 to 10, 000 cP at the temperature where the pre-consolidation process is conducted, preferably between about 500 to 5,000 cP.

The temperature where the pre-consolidation process is conducted may vary consid- erably depending particularly on the composition of the resin. Typically, the pre-consoli- dation temperatures for epoxy-based resin systems are 50 to 90°C and preferably 60 to

80°C, however, both higher and lower temperatures may be feasible in some systems.

The pre-consolidation process may lead to an increase in the glass transition tempera- ture, Tg, of the resin, for example by a partial curing. It is preferred that the Tg of the resin is increased during pre-consolidation by at least 2*C and preferably by at least 5°C, as an increase in Tg usually indicates an increase in the average molecular weight of the resin, which will enhance handleability, strength and unstickyness. In some cases, Tg may be increased more. This is particularly the case when Tg of the unconsolidated element is very low.

In a preferred embodiment a pre-consolidated element according to the invention with an epoxy-based resin system should typically have a Tg between-10 to +30°C and pref- erably a Tg between-5 to 10°C. In some cases, Tg of the resin of the pre-consolidated element is higher than about 0°C and preferably higher than about 3°C. For the unconsolidated element, Tg of the resin should be below about 5°C and preferably below about 2°C.

In some cases, curing of a pre-consolidated element without being exposed to a vacuum will result in a material with properties equivalent to a vacuum-cured element, since po- rosity has been eliminated or greatly reduced during the pre-consolidation process prior to the curing. This may for example be used when preparing a connection in a large structure where the establishing of a vacuum may be very time-consuming.

In a preferred method of pre-consolidating and/or curing an element, the element is placed on a reactor surface like for example a plate, a mould, etc. It is preferred that the reactor surface is flat and that it will withstand heating and/or vacuum. Then a pressure is applied to the element. The pressure may be applied by a press or-preferably-via a vacuum within a vacuum enclosure. If a vacuum is used, then a vacuum enclosure should be obtained prior to pressing. The vacuum enclosure may for example comprise a vacuum bag or it may comprise a reactor surface and a flexible cover connected in a vacuum-tight way to the reactor surface. Gas may for example be evacuated through the reactor surface or through an opening in the vacuum bag or flexible cover. Then pre- consolidation or curing is activated. The activation may take place before and/or during and/or after applying pressure. The activation comprises a reduction of the viscosity of the resin. This may for example be realised by physical means (e. g. heating, addition of solvent, pressure etc. ) and/or by a chemical reaction. During the pre-consolidation pro- cess, a limited curing may or may not take place. When the porosity has been reduced

to a desired level or another object of the pre-consolidation is obtained, the pre-consoli- dation process is terminated. The termination may for example be a result of exhaustion of a first resin system or cooling of the pre-consolidated element to a temperature, where the curing reaction is sufficiently slow and/or the viscosity is sufficiently high for the pre-consolidated element to achieve the stability needed for the desired shelf life.

In a preferred embodiment, the element to be pre-consolidated or cured is having at least one non-continuous layer of resin, through which gas may be removed during the pre-consolidation or curing process. Hence, the gas need not be removed from the ele- ment via one plane of a layer of resin or in a plane of one layer comprising fibres. Thus, the transportation distance and risk of having trapped gas inside the pre-consolidated element is greatly reduced. In a more preferred embodiment, all layers of resin-option- ally except from a layer on top of the top layer comprising fibres or below the bottom layer comprising fibres-are non-continuous.

An example of a method for ensuring that gas may continuously be removed from the element during pre-consolidation or curing involves a gradual activation of the pre-con- solidation process starting either from the centre of the element and moving towards the surfaces or from a side or edge and moving through the element. For example this may be realised by heating from the reaction surface only, thereby activating the pre-consoli- dation process gradually from the side of the element in contact with the reaction sur- face or by controlled microwave heating, thereby activating the pre-consolidation pro- cess gradually from the inside of the element and moving towards the surfaces.

The properties of a laminated structure having layers comprising oriented fibres to a large extent depend on the distribution of the main elements of the structure i. e. resin, fibres and porosity. The resin possesses a low strength compared to the fibres and may provide a route for crack propagation through the structure, if too thick layers of resin are present. Porosity may reduce the strength of the structure dramatically but the adversity depends on the size of pores, the shape and the distribution, e. g. the effect of small, isolated spherical pores is limited, whereas larger pores positioned in the interface be- tween resin and fibres may be fatal to the structure. Consequently it is vital to be able to control the distribution of the elements.

The extent of redistribution or homogenisation during pre-consolidation or curing de- pends on the viscosity of the resin during the compaction process, i. e. the lower the vis- cosity the easier the redistribution of the elements. By utilising a pre-consolidation pro-

cess the viscosity of the resin may be lowered more than what is feasible in the prior art, since the structure is not limited to supporting a particular shape during the process.

When the pre-consolidation involves a limited curing of the resin, the viscosity may be further reduced since the limited curing increases the handleability and reduces the sticking of the pre-consolidated element. Hence, pre-consolidation typically allows for redistribution of resin and/or fibres to a greater extent than what may be realised by di- rect curing. The resulting pre-consolidated elements may possess very low porosity as well as a more homogeneous structure. This may for example result in a composite structure having a less pronounced laminated structure, i. e. where the layers are less pronounced than a corresponding composite structure comprising only elements that were not pre-consolidated prior to curing.

Connections and methods according to the invention are useful for the connecting of two relatively incompatible composite members whereas composite members having com- patible properties may also be connected using a connection or a method according to the invention. The extent of the invention is best realised when trying to connect two in- compatible composite members. Examples of relatively incompatible composite mem- bers are composite members having a considerable difference in properties such as co- efficient of thermal expansion and/or Young's modulus, etc.

Connections and methods according to the invention are particularly useful for prepara- tion of wind turbine blades. For example in the preparation of a spar for a wind turbine blade, the connections may be used for connecting a mainly glass fibre reinforced part, e. g. the base, to a mainly carbon fibre reinforced part of the spar, e. g. the main part.

This is particularly useful for larger blades as the weight of a blade may be dramatically reduced, when part of the glass fibre reinforced composite is replaced by carbon fibre reinforced composite.

Table for identification 2 Overall plane of the interface between the transition zone and a composite member 4 Surface of a composite member a Angle 10 Composite member 12 Composite member 13 Transition zone 14 Layer comprising fibres having substantially the same composition as com- posite member 10 15 Tow of first type fibres 16 Layer comprising fibres having substantially the same composition as com- posite member 12 17 Tow of second type fibres 20 Layer having micro grading 20a Layer having micro grading 20b Layer having micro grading 22 Layer comprising fibres F Force 30 Layer comprising fibre 30a, b, d Continuous layer comprising fibres 30c Non-continuous layer comprising fibres 32 Lines indicating difference in the experienced stress from the average resin 34 The extent of the end-condition of the stress 40 Layers having substantially the same fibre composition as a composite member 42 Layers having substantially the same fibre composition as a composite member and being bend to follow the shape of the interface 50 Sheet or layer comprising first type fibres 52 Sheet or layer comprising second type fibres 60 Layers comprising first type fibres 62 Layers comprising second type fibres 64 Distance between layers comprising first type fibres and adjacent layers comprising second type fibres 66 Distance between adjacent layers comprising second type fibres