Skip to main content

Industrialization of Manufacturing Technologies for Composite Profiles for Aerospace Applications

Final Report Summary - IMAC-PRO (Industrialization of manufacturing technologies for composite profiles for aerospace applications)

Executive Summary:

Due to the fact, that the demand on the efficiency of air transportation is increasing more and more, the use of carbon fibre reinforced plastics (CFRP) is continuously rising in the field of aircraft structures. Almost the complete load carrying structure of civil aircrafts is realised by profiles and so a length of more than 2 000 meters of CFRP profiles in total could be obstructed for an all composite aircraft. For recently developed long range aircrafts, CFRP structures were mainly realised with state of the art technologies with high production costs. The technological objective of IMAC-PRO was the development of a complete integrated process chain for the cost effective serial production of optimised CFRP stiffener profiles for all kinds of aircrafts, based on textile technologies in combination with advanced injection and curing technologies. The project is separated in two main profile routes; in the realisation of stringer on the one hand and frames and beams profiles on the other hand. The details of the profiles are:

1. Stingers for fuselage applications with a length of up to 20m mainly with a constant cross section and straight alignment. In some cases stringers with variable cross section and a partial curvature will also be investigated.
2. Heavy profiles like frames and beams which may have variations in wall thickness, profile cross section and fibre orientation along its length.

For the stringer preforming route, three different technologies were developed. One stringer preforming route is the forming of non-crimped fabric material with the help of a discontinuous hot pressing process (HP). A second approach is the production of an integrated preform tape out of single patches cut from unidirectional glass (UD)-tapes to be formed to a T-stringer. A third approach was to produce stringers out of tubular braided sleeves, predestined for curved stringers. For the curing of stringers, pultrusion and infusion technologies were investigated.

The preferred preforming method for heavy profiles like frames and beams is the UD-braiding technology. Two types of profiles were selected to be developed within IMAC-PRO: a JF-profile and a C-profile. For these kinds of preforming technologies different solutions for braiding mandrels have to be realised. For each profile type, a new curing tool was realised. To achieve an efficient production, a gap infusion technology was developed where a small gap between preform and tool surface allows a very short infusion time before closing the mould completely for wetting and curing.

Several quality assurance technologies for preforms and curing were developed, to check the excellence of the realised technologies.

At the end of the project all these profiles and manufacturing technologies were evaluated in terms of high volume demands stated at the beginning of the project. As far as the stringer preforming is concerned one outcome of the project was that the HP technology has the advantage of a very high level of maturity, but the fibre patch preforming (FPP) technology has a higher potential for high volume production related to a higher production speed. Also the thinner layers of the FPP material have a better material performance and therefore a weight saving potential which leads to lower fuel consumption at the end.

The UD-braiding showed that it is a very suitable manufacturing process to realise complex curved preforms for primary structures in a very cost effective way with a low scarp rate.

Project Context and Objectives:

The technological objective of IMAC-PRO is the development of a complete integrated process chain for the cost effective serial production of optimised CFRP stiffener profiles (e.g. frames, stringer, struts, floor beams) for all kinds of aircrafts (passenger planes, helicopters, fighters) based on textile preforming combined with an advanced injection and curing approach.

Profiles are one of the most important structural components in aerospace as well as in many applications in ground transport, marine and mechanical engineering. Together with the skin, they form lightweight structures with high stiffness and strength and additional functionality (for example aerodynamic shape). Various other profiles are used to build the inner structure of an aircraft fuselage. PAX floor cross beams and longitudinal beams, partly with integrated seat rails and the z-struts, used to support the floor grid, can be mentioned as representatives of profiles which are necessary in a very high quantity.

Therefore, the optimised design and cost effective manufacturing of profiles is of very high interest, which will significantly grow in the future due to the rapidly growing aircraft market and the demand for lightweight designs to improve the ecological compatibility of planes and helicopters with respect to a reduction of aircraft production and development costs.

Depending on the geometrical requirements and the loads, profiles can be straight with constant cross section (for example fuselage stringers), straight with varying cross section (for example wing spars) or complex curved (for example fuselage frames).

State of the art are integrally milled aluminium structures, but the highest potential for lightweight design is offered by composite materials, because they allow an optimised design regarding geometry, local thicknesses and local fibre direction.

This has been proven in many research and development projects, mainly by using prepreg (pre-impregnated, unidirectional reinforced layers). Thin single prepreg layers are placed in the required fibre direction (0, 90 and 45 degree) in a mould and cured in an autoclave. This manufacturing method is very time consuming and expensive, because the potential for automation is limited to more or less flat geometries and the semi-finished prepreg material is very costly. In addition, the high performance of the prepreg material cannot be utilised in complex profiles, because the number of single individual patches would be very high and the waste may reach 60% of the prepreg or even more. Also, the area of the profile on which the optimum fibre direction can be realised is very small, especially on curved profiles. For this reason, textile preforming technologies in combination with resin injection and curing are very promising for the cost effective manufacturing of high performance profiles.

One of the most efficient textile technologies which should be used for the benefit of this project is braiding. EADS, USTUTT and SGL Kuempers are well experienced with the over-braiding of complex shaped mandrels guided through the centre of a circular braider by an industrial robot.

The applicability of braiding for high performance CFRP parts should be supported by the UD-braiding technology, developed in recent times by EADS. The simple basic idea of UD-braiding is the replacement of the carbon roving on the bobbins of one moving direction with very thin, so called 'support yarns'. The only purpose of these support yarns is, to keep the reinforcement fibres on their desired positions on the mandrel. Initially, they don’t have any function in the cured component, but they may also be used for the homogenous distribution of auxiliary materials, i.e. resin modifiers.

The remove of the fibre undulation lead to a significant increase of the mechanical properties of the material.

A new preforming technology, FPP, allows the fully automated production of complex fibre preforms. The technique has been developed for shell geometries with varying fibre angle and wall thickness. A special fibre tape of only 80g/m2 stabilised with a binder material is used. The tape is cut in patches with defined length and a lay-up device picks them up. Each patch can be placed individually with defined orientation. The lay-up device works with a silicon rubber stamp, which presses the fibre homogeneous even on a curved surface area.

With this technique, the production of very complex shaped components is possible. Parts that would cause wrinkles or gaps in normal textiles can be built up with patches. The preforms are net shaped and fibre waste is reduced to a minimum. The design and realisation of a FPP machine to produce endless FPP preforms to be folded to straight profiles was planned in this project.

The second production step is the infiltration and curing of the preforms. The processes have to be quick enough to keep pace with the preform production. Fast curing realised with pre-heated tools or application of microwaves are possible solutions. Additionally, continuous curing with pultrusion technique should be also in the focus.

Infiltration tools should be investigated to improve their mass production applicability with respect to a fast and secure preform loading and optimised resin flow. This work should be supported by injection and cure simulation. A cure monitoring system should be used to investigate the influence of variations in the process parameters and finally to secure part quality.

Pilot manufacturing machines should be established and several realistic components should be manufactured and tested.

The components of a so called cargo floor unit will represent most of the major challenges for profile production which should be addressed in IMAC-PRO:

1. curved profiles
2. variable profile cross sections
3. local reinforcements
4. high variability to realise different profile types.

For all process steps quality assurance concepts will be developed and the philosophy of a qualification process will be taken into consideration.

All process steps will also be consequently evaluated concerning the balance between structural performance and affordability. This on-going evaluation, in which all partners will be involved, will also allow reviewing the overall results of the project compared to the goals defined at the beginning.

The overall benchmark is, to realise profiles with at least the same structural performance as prepreg based parts. Some applications, for example curved frame profiles may even reach a higher performance due to the optimisation of the fibre orientation. 20% weight reduction compared to aluminium structures at 25% lower overall costs are the challenges at a glance.

The project consortium is in good balance and involves university institutes, research establishments, small and medium sized enterprises (SMEs) and application partners represented by the respective technology leaders from all over Europe. This is not only a guarantee for the quick transfer of the scientific and technological results to applications in the aerospace industry but also for a broad utilisation in the whole composite world (machine development, software tool development, design guidelines, education etc.).

Therefore, the project will have a high impact on competitiveness of Europe in this important technology field. The competition for the best black fuselage design for the next generation of passenger aircrafts is only one example.

Project Results:

In the following section a detailed comprehensive summary of all activities realised within IMAC-PRO is given. It is divided into single paragraphs, sequentially describing the technical topics. The related work packages (WPs) are listed in brackets.

T-stringer preforming (WP 3.1 6.1)

In IMAC-PRO three different T-stringer preforming routes and two different stringer designs (concept A and B) have been realised. The concept A stringer has a classical cross section geometry and the concept B stringer is a new approach to reduce the stress level in the radius area.

One approach for manufacturing is the forming of non-crimped fabric (NCF) material by applying a discontinuous hot pressing process. Conventional NCF material is cut into ribbons equipped with binder and stacked together in the dedicated fibre orientations. Afterwards the complete stack is feed into a hot forming die to drape the T-stringer.

This technology delivers T-stringer preforms in excellent and reproducible quality. These kinds of profiles were realised in concept A geometry.

The final production speed could be increased up to 25 m per hour after performing a number of optimisation loops.

A second approach is the production of a preform tape composed of single patches cut from a UD-tape. This method allows the production of a highly flexible ribbon, which is suitable for curved stringers. This flexibility comes from the fact, that the single patches are bonded together only on selected areas. To produce this new type of flexible tape, a complete new FPP tape production machine was designed and manufactured.

The process starts with a 0° layer, transported on a continuous conveyor defining the overall speed of the process. Several feeders, oriented in different angles along the conveyor are positioning patches in fixed angles on top of the previous layer. The FPP tape production machine can realise a four-layer tape (0°, 90°, +45°, - 45°) and works very reliable with a speed of 60m/h. Several tapes were then combined by preforming to a multi-layer tape and formed to a T-stringer preform. The principle of an automated forming device could be shown within the project but has to be refined for a possible high volume approach. Within IMAC-PRO only a four layer tape could be realised with the FPP tape production machine but for serial production the 60m/h could also be realised with an increased number of feeders. The stated production speed is absolutely suitable for even high volume demands for single aisle aircrafts. An additional positive effect is that the FPP material with its only 80g /m2 leads to parts with low undulation and high flexibility in stacking. This technique has been used to produce the stringers of concept B.

The third second approach of concept A stringer preforming was the use of tubular braided sleeves. A low cost, pre-manufactured braided sleeve material is stabilised with binder and cut in longitudinal direction on the edges. The resulting sub-preforms are then formed in a 90° angle to a textile T-stringer.

Due to the high flexibility of the preform this technology is very suitable to realise partly curved stringers which are typical for an application in the rear fuselage.

T-stringer curing (WP 4.1 6.2)

For the curing of straight T-stringers two processes, pultrusion and an infusion where selected. The pultrusion technology is normally used for straight parts with a very high percentage of 0° fibres because high tension forces are necessary to pull the preform through the die. In this project it was planned to process preforms with a balanced stacking, which means that most of layers are not aligned in process direction. For conventional pultrusion, the single fibres are wetted in a resin bath before going into the die. In our approach the preform is already very compressed. Together with the fact, that we use a qualified single component infusion resin instead of a dedicated pultrusion resin, the approach was very challenging.

Therefore a complete new die design had been developed and tested. An infusion area in the first zone of the tool should bring the resin into the right area. First tests showed that the resin does not reach the inner layers of the preform. So a redesign of the tool was realised which allowed higher pressure. With this modification fully impregnated short sections could be realised but now the necessary pulling forces on the preform got too high and the preform was torn off after several centimetres.

Finally the process worked not stable over several meters and the void content was still too high inside the parts. The result of the pultrusion development was that with two L-preforms instead of a complete T-stringer the process will work reliable with sufficient quality.

Stringers with concept B geometry were made exclusively with FPP technology. The manufacturing device to produce stringers out of FPP tape material was a completely new development for IMAC-PRO able to produce in principle endless stringer preforms. The possibility to produce endless stringer preforms was successfully demonstrated with the FPP device, but due to various limitations (space in lab, size of forming device) only stringer preforms with a length of 800mm were produced and used for infiltration. No continuous curing of concept B stringers could be made in IMAC-PRO. An infusion tool for 700mm long stringers was designed and manufactured which was sufficient to manufacture the test samples.

The infusion of the FPP stringers was made with resin transfer moulding six (RTM6) resin in an autoclave without additional pressure. The curing could also be realised in a simple oven. With this curing technology in total 13 parts were successfully cured. The parts were mainly used for mechanical tests. NDI test made with computer tomography showed a good material quality. Only in the filler area some voids could be found. This can be tracked back to an insufficient fit of the filler. The problem can be prevented easily by using fillers with a well-adapted shape.

To realise stringers with a partly curved geometry, which is representative for fuselage stringers, two manufacturing approaches were investigated. The first was a self-heating carbon infusion tool which was developed by Alenia. With this tooling, stringers with a total length of three meters and a curved section of one meter were realised. Preforms with tubular braided sleeves were used for this kind of infusion. Due to the braided structure of the preform the infusion worked pretty well and the part quality was good.

A second approach was to produce curved stringers continuously. Therefore the pultrusion line was combined with a guiding device placed directly after the pultrusion die. The principle behind was to cure the resin just that far, that the cohesion of the fibres is guaranteed, but the resin is still flexible. While the resin was in this pre-cured state the stringers could be bended successfully with the guiding device in the required radius. The stringer was then finally cured with infrared heating elements.

For these production trials a +/- 90° fabric material together with a dedicated pultrusion resin was used to guarantee a stable production process. Alternative preforms with a more open structure for a better wetting together with a pultrusion resin was used.

Frames and beams preforming (WP 3.2 6.3)

The preferred preforming method for large profiles like frames and beams is the UD-braiding technology. UD-braiding offers a very good alignment of the fibres with low undulation. Two types of profiles were selected to be developed within IMAC-PRO: a JF-profile and a C-profile.

The approach for the C-beam preforming was an over-braided, square aluminium mandrel. Different braiding layers combined with winding layers and UD-tape were laid up with binder in between and preformed with heat and pressure afterwards. After this, the preform was cut in half in longitudinal direction and two C-profiles were gained.

This process worked very reliable and the braiding and winding process ran fully automated which is mandatory for a high volume production.

For the JF-preform two mandrels were over braided separately. One mandrel was also aluminium. The second one was a newly developed inflatable so called FX core. The speciality of this concept is that this mandrel is stiff enough to be used for the braiding process, but has additional integrated functionalities supporting the infusion and curing process.

After the separate over-braiding of both mandrels, the two were combined and over-braided again together.

Another profile which was produced by braiding was the Falcon spar from Dassault Aviation. Currently the part in the aircraft is made out of NCF material. Within IMAC-PRO, it should be demonstrated, that this part can be made also out of braiding technology. The braiding of this part was very challenging, because several variations in cross section and thickness had to be realised.

After braiding of several profiles, it can be stated, that this preforming technique is predestined for high volume production, due to its high degree of automation and high output rate.

Frames and beams curing (WP 4.2 6.4)

For most profile types, new curing techniques were realised within this project. For the C-beam, an innovative RTM tool for very short cycle times was developed. This so called compression RTM technique consists of an RTM tool with two closing positions. In the first position the tool is already resin and vacuum tight but realises a small gap of about 2mm between cavity and preform. Within this gap, the resin can be transported over the complete length of the profile very quickly. After the resin distribution, the tool is closed completely and the resin has to penetrate only through the thickness. With this process the infusion time can be cut down below one minute.

The C-beam tool was working quite well. With the gap injection process it was possible to manufacture parts of good quality. After a redesign of the inlet and outlet gate and further tests, compression-RTM was possible with the tool. Due to the fact, that a higher output rate can be achieved by the gap injection process, this process is suitable for high volume curing.

For the JF-frame, also a RTM tool was developed. Also in this case, a short infusion time was in the focus. After over-braiding of both mandrels together, the complete set-up was placed into the tool. Then the preform was cut on the top and the aluminium mandrel was removed. Now the both flanges could be bended for 180°. After this, a so called 'intensifier' was placed on to the preform to realise the sufficient consolidation of the part. After infusion of the resin, also a gap infusion like for the C-beam, the cavity on top and the FX-core were pressurised with air during curing.

After several optimisation loops, the part quality could be increased, but it was finally not possible to manufacture parts completely free from dry spots. The main problem was the tool tightness (vacuum and pressure tightness). Main leakages were assumed between intensifier and mould, as well as between FX-core and mould. The complete sealing concept of the tool needs to be modified. Nevertheless, the basic functionality of the tool could be proven and with some modification this concept is assumed to be reliable for high volume curing of this profile type.

The curing of the Falcon spar was made in one of the RTM production tools existing at Dassault aviation.

The preform could be infused successfully and the quality was sufficient to use sections for mechanical testing.

Microwave curing (WP 4.1 6.2)

Microwave curing is a completely different approach for curing. This technology is based on the heat generation inside or on the surface of a dielectric material (e.g. carbon fibre reinforced plastic). The energy which is required for the heat generation is transported by high-frequent electromagnetic waves. The presence of a strong electromagnetic field is a requirement for the mould and the influence of this new and (until now) unusual requirement has to be evaluated.

Two different mould concepts for microwave curing of C-struts were developed. Taking into account the main requirements - accessibility of composite materials to the applied electromagnetic field and injection of liquid preheated resin - the focus was laid on a one-sided mould.

Such moulds consist of a solid metallic tool adapted to be used inside microwave field and auxiliary materials (peel-ply, vacuum bag, etc.) which are transparent to microwave fields and fix the fibre material on the tool.

The manufacturing concept connected with this mould concept is described below:

1. Step one: A mandrel with dimensions of the C-strut is made of a metallic material. Injection vents are drilled into the mandrel;
2. Step two: C-strut is braided on the mandrel;
3. Step three: peel-ply and vacuum tube are placed over the mandrel with braided C-strut on it,
4. Step four: material is heated up to the injection temperature;
5. Step five: liquid resin is injected through the injection vent on the one side of the mandrel and flows through the material to the opposite vent;
6. Step six: the microwave field power and therefore the material temperature is increased until compete curing of the composite;
7. Step seven: the temperature is monitored by optic thermocouples fixed on the vacuum bag.

With this technology, C-struts used for the cargo frame unit could be cured successfully. Visual inspection of the cured parts approves complete impregnation of the fibres with resin and a sufficiently high degree of curing.

Additionally, the continuous curing of profiles with microwave technology was investigated.

The first step was the definition of the requirements applicable to the equipment and the process. All major requirements can be subdivided in two groups: safety-related and quality-related. The designed equipment consists of three main components:

1. Microwave chamber and feeding waveguide
2. Damping waveguide ('choke')
3. Mould.

The technical concept behind the equipment design is as follows: The composite material is transported through the microwave chamber by a set of rollers. There is an electromagnetic field inside the chamber that interacts with the material and causes the temperature raise. The field is generated by an external source (magnetron) that is connected to the chamber via a feeding waveguide. In order to assure compatibility of the designed system with the existing equipment all the waveguides and chamber have a cross-sectional rectangular area of 86mm x 43mm that corresponds to the normalised waveguide size WR340. Attached damping waveguides ('chokes') prevent microwave field from escaping the chamber. Three major challenges were defined for this kind of technology: Microwave radiation leakage, contamination of equipment and quality of cured parts. Point one and two could be partly fulfilled with the developed test setup.

A reliable heating process was not established for all mould material combinations. The main problem was uncontrolled heating and arcing of the multiaxial composite material, leading to occasional destruction of the mould and material to be processed. Arcing occurred whether a metallic bottom mould (poor contact between mould and carbon fibres) or a ceramic or a polymer one (no grounding of conductive fibres available) was used. In addition to that, conducting fibres became stuck between the bottom and top mould reducing the process quality even more. A lot more development work has to be done to realise this challenging approach of continuous curing by µ-wave technology.

Preform quality assessment (QA) (WP 3.1 3.2)

To check the quality of the produced preforms, two different quality optical technologies were developed within this project. One system was used to inspect the surface of the preforms after production and one to measure the fibre angle during braiding.

The preform QA-system is able to detect gaps between rovings, fibre angle variances, undulations, fibre fracture, splices, fuzz balls and foreign objects on the surface. This system is placed on a robot. After teaching the geometry of a specific profile, the robot is able to do the inspection completely automated.

The existing two-dimensional (2D) preform measurement system has been extended to test new image processing functions developed by fibre. Additionally, further improvements to the control software have been implemented to increase both speed and usability.

To test complex three-dimensional (3D) preforms, a new sensor system has been developed. The image processing library provided by Fibre and the sensor head provided by iSAM allow the detection of all relevant material defects with high accuracy.

The quality assurance sensor has a very wide field of view with respect the small, lightweight housing. Thus, even the inner sides of small stringer profiles can be inspected.

Due to the modular architecture of the sensor head, the system can be adapted for other stringer geometries as well. Only the illumination device and sometimes the camera lens have to be changed. All knowledge to develop adapted devices has been gained within this project.

The whole system is used to rapidly examine complex stringer profiles from all accessible sides. After only one teaching process for each profile geometry all parts can be scanned automatically without human intervention.

The testing results can be sorted by the relevance of defects to detect defect parts immediately.

iSAM provided a software-interface that can easily be used to control the material inspection process and to evaluate the results.

The project has shown that it is possible to produce stringer profiles with a high production rate and an increased quality without manual process control. That way, production costs can be lowered and the product reproducibility is guaranteed.

The second developed system was a fibre angle measurement system to measure and control this parameter directly on the braiding machine.

The fibre angle of the braided material is a very critical parameter, because loads can only be taken in the direction of the fibres. Even small angle deviations result directly in significant loss of properties of the product.

Due to variations in the braiding process itself a closed loop control system is necessary in order to guarantee an exact angle on the out coming material.

Thus, a QA system is required to determine the angle of the fibre rovings and, at the best, to control the angle during the braiding process itself. A complete real-time fibre angle detection system has been developed. Two approaches have been tested. Results have shown that the grey image detection algorithm should be used.

It fits the need of fast and reliable online angle detection. The maximum possible frame rate is dependent on the computer hardware and the field of view of the camera system. In the current system, a frame rate of 1.4 images per second has been measured. The field of view is on its large dimension 33mm long. Thus, a distance of 46 millimetres can be scanned within one second. The feed motion of the braiding machine is approximately 16 millimetres per second. As can be seen, the detection speed is sufficient and no angle deviations within the field of view will be ignored by the system.

The system is currently being used within the production line for frames and beams at EADS-IW/Stade. Tests have shown a high accuracy of the measured angles. A deviation of the detected angles from the angles, which have been manually measured, could not be detected. Therefore the angle output of the system has been put into the braiding machine control. It is successfully used to adjust braiding parameters online.

Curing QA (WP 4.1 4.2)

The response of polymeric materials to external electric fields has been used for many years as a technique for studying a range of physicochemical processes. The ultimate aim of any successful cure monitoring system is to provide the means for controlling the curing process through material state signals that will be acquired online during the cure. The first step for achieving control is to link the cure monitoring signal to a material state parameter.

In IMAC-PRO, the complex impedance signal has been correlated to the degree of cure of the thermoset system RTM6.

Within the project, a dielectric sensor including a simple graphical user interface was developed to be used in several tools that were realised.

A first simple flat tool was designed and manufactured by Alpex, for implement the dielectric sensor and monitoring system. During work package progress, the QA system developed by Inasco had to be implemented into more complex tools for JF-frame and C-beam production. It was decided the dielectric sensor to be positioned into the C-beam mould.

An RTM6 cure kinetics simulator was developed by Inasco, where standard and alternates cure cycles can be simulated and final properties of the fully cured composite can be obtained. Cure optimisation can be performed, resulting in time and cost saving of the manufacturing process.

Dielectric measurements were performed at flat laminates for quality assessment of the samples and were conducted at DLR facilities. Curing was conducted at temperatures 80°C, 120°C and 150°C, while further data analysis that was performed with the use of cure kinetic simulator tool indicated that curing at 80°C and 120°C leads to low degree of cure in the final product. Temperatures above 150°C are most indicative when RTM6 resin system is used in order to reach high conversion values.

An additional approach was to integrate the dielectric sensor in the µ-wave cure process. First dielectric measurements under microwave heating were performed and further analysed. Final conclusions cannot be made as dielectric signal had a lot of noise and curing procedure could not be traced or evident through the plot. Phase takes positive values at random time indicating that microwave heating affects significantly sensor signal. Another possible QA concept on curing machine of µ-wave curing would include the dielectric measurements to take place after microwave curing as an indicator of the final cured composite part or in the mid-time, when microwaves are not transmitted.

Evaluation of test results (WP 5, 7.1 7.2)

D-level tests (specimen)

Suitable test methods were selected and defined to evaluate the material and sample quality of the technologies developed in this project. Material evaluation is normally based on coupons made out of flat plates. They do not represent the real properties of net shaped textile preforms. Property deviations due to geometrical and process related influences are not covered. Alternative specimen dimensions or ways of test material extraction out of components have been considered. Therefore special test methods, applied to samples directly taken from the profiles were figured out and correlated to test standards.

The test program has been proposed to allow determination of material, profiles and subcomponents (spar and panel) properties. Strength analyses have been undertaken on the extracted test specimens to evaluate tensile, compression, impact, interlaminar shear, bearing and pull-through performance. The specimen origin within a structure and effect of environmental conditions are considered. The following results have been acquired:

1. For the JF-beam, there was significant difference between the web and the flange performance confirmed by tensile and pull-through tests. The web part tensile strength was higher but the pull pull-through strength was lower when compared with the flanges. This behaviour shows lowered performance of the web under loading in the transverse direction which has to be noticed especially when using bolts in this part of the JF-profiles.
2. The C-beam profile has significantly better strength of the cap region in comparison to the web measured in tensile, interlaminar shear strength (ILSS) and compression tests.
3. The matrix performance mostly determines the strength when test types such as ILSS, bearing and compression are used. For these tests, the measured strength generally decreases with increasing temperature as the matrix strength and toughness decreases. The lowest strength values for these three types of tests measured for hot conditions (70°C) after wet ageing.
4. For test types where is the coupon strength determined by fibre performance, such as for the tensile test, the measured strength decreases with decreasing temperature. The lowest strength values have been measured for the cold conditions (-55°C). This behaviour is connected with increasing fibre and matrix brittleness for low temperatures.

The braided samples extracted from the Dassault spar could be directly compared with serial production values from the real Falcon spar. The tension and compression values for strength were comparable with the values of the serial parts and the stiffness was even better.

C-level tests (profiles)

To evaluate the different T-stringer designs and manufacturing concepts, 4-point-bending tests were performed with and without impact damage on concept A HP and Alenia braided stringers and concept B FPP stringers. Stringers were cured by EADS, Secar, HAI and Alenia. The following results could be stated:

1. Based on impact screening investigations can be stated that the concept B stringer (FPP) is most sensitive to the impacts already from low impact energy. The delamination and indentation depth is significantly higher than for Alenia and Secar concept A technology. Alenia stringers are most resistive against impact damage.
2. Maximum values of deflection at failure are significantly higher (60-68 mm) for Alenia profiles than for other manufactured stringers (both for concept A and B) (5 - 13 mm). By contrast the maximum achieved forces at failure are significantly lower (more than twice) for Alenia profiles than for other stringers. This appearance can be caused due to the missing of 0° layers in the lay-up.
3. Decreasing of bending strength of Alenia stringers due to impact damage (3J) is significantly lower (about 5%) than for all other stringers.
4. Decreasing of bending strength due to 3J impact energy application is 30% for Secar HP stringers, 37% for HAI stringers and 55% for EADS FPP stringers.
5. The minimum weight and volume has EADS FPP stringer concept B. FPP stringer has a higher density in comparison with other stringers manufactured by different process. Secar HP stringer has a higher weight and volume in comparison with the other concept A stringers, but its density is minimal.

The overall result is that the FPP stringers have in relation to their weight the best bending performance. Due to the FPP manufacturing technique, the straight aligned fibres are very vulnerable to impacts on the front surface.

To compare the HP and FPP technology additional compression tests on run out and continuous stiffeners were performed, also under impact damage. The relevant test results were the load level and the failure mode of such structures.

The continuous stiffener test series had permitted to establish a classification between the different impact configurations that can be envisaged on continuous stiffeners: the most critical impact configuration regarding the residual strength is the impact at the top of the stiffener one, then the drainage hole at the bottom of the stiffener one (for which the impact on the aerodynamic surface is not influent), then the less critical configuration is the impact on aerodynamic surface one. The stiffener run-out test series permitted to learn that the previous hypothesis that consists in using the failure CAI strain at the ¼ of the thickness on the delaminated width of the stiffener run-out was not appropriate. Then this test series permitted to understand the stiffener run-out behaviour regarding the failure (the neutral fibre offset generates a pressure load that stabilises the delaminated plies) and to suggest a new sizing hypothesis which consists in applying the failure CAI strain on the external side of the skin on the delaminated width of the stiffener run-out area.

Additionally, a three-point bending test on a falcon spar section was performed to learn more about the behaviour of a structure using a preform manufactured by UD-braiding process. This test permits to evaluate the failure mode of such a structure when loaded with a bending mode and the associated failure mode.

The analysis of the test results revealed that this kind of technology can generate steered lay-ups that need to be taken into account for the sizing. Nevertheless, with using a manufacturing process adapted to the UD-braiding solution (RTM injection before cutting operation then automatised cutting operation) this technology could be much more automatised and so much more cost saving.

B-level tests (panels)

To see also how the two different stringer concepts react in a realistic environment, stiffened panels with three stringers were manufactured and tested. First set of T-stringers was manufactured using HP stringer (concept A), the second set of T-stringers was manufactured using FFP technology (concept B). Concept A stringers were co-cured with the skin and concept B stringers were co-bonded with the skin. The panel dimensions are 700 mm (length) x 450 mm (width). Overall panel dimensions including end potting plates are 700 mm (length) x 500 mm (width) x 100 mm (depth). The each panel consists of a 2.91 mm thick skin, three T profiles/stringers.

The following conclusions can be stated on the basis of data listed above:

1. The ultimate strength of panel concept B is about 6% higher than panel concept A. The actual maximum failure load of panels is about 25% lower than the predicted failure load by IAI. The predicted stiffness of this panel is also higher than actual value. The actual and predicted maximum displacement at failure is nearly the same.
2. The high frames per second (FPS) video recording shows that at first, the waviness initiates in the middle stringer and then delamination of a skin/stringer interface originated. In the end the stringers failed and in consequence panel lost its stability and compression capacity.
3. Buckling modes initiated for both panels approximately in the same forces. The first non-linarites occurred under 300 kN and buckling modes start at 330 kN.
4. The comparison of the buckling mode between numerical model and optical measurements shows some differences in the buckling mode near the panel edges.

A-level tests (cargo unit)

To evaluate the different profiles realised within IMAC-PRO under realistic conditions, a cargo floor unit was designed as close as possible to the real application. Cargo floor unit consists of a curved frame profile, a straight cross beam and struts.

The testing equipment and associated facilities for the data acquisition from different sensors were developed. Mechanical performance of the cargo floor unit has been evaluated through the three stiffness static tests and the one static test under limit load. The complete design of the test, test fixture manufacturing, test frame and hydraulic systems, facilities for data record and other planed work associated with the test conducting and data evaluation were manufactured and prepared. One cargo floor unit could be manufactured but the quality of the first part was not good enough to be tested. Due to fact, that the curing tool was finished very late, no additional part could be manufactured and so unfortunately, the cargo frame unit could not be tested as planned.

Potential Impact:

The status in manufacturing of aerospace structural parts at project start was aluminium processing for single aisle aircrafts and mostly expensive prepreg manufacturing in selected areas of long range passenger airplanes. The aim of IMAC-PRO was to realise parts with 20% weight reduction compared to aluminium structures at 25% lower overall costs compared to prepreg manufacturing. Within the project, extensive development work has been done to adopt textile technologies to the needs of manufacturing of fibre reinforced plastics. This work, together with new curing approaches, aimed on the reduction of the overall manufacturing costs of CFRP structural aircraft parts and to proof the ability to use these technologies for high volume production.

As far as the cost situation is concerned, it is still nearly impossible to gain representative costs for serial aerospace prepreg parts. On the one hand comparable parts are not in service so far, on the other hand prices are not published by manufacturers.

Due to the fact, that in most of the high volume sectors where CFRP specimen are used like automotive or sports goods the textile technologies are on the rise, so the potential of cost saving is obvious.

Due to a very comprehensive test programs including D-, C- and B- level test, the suitability of our developed textile technologies for the stated requirements was proven and therefore also the weight saving compared to aluminium.

As far as the high volume demand is concerned the T-stringer manufacturing was evaluated with regard to two scenarios of possible end-users as a baseline; the demand of T-stringers of Dassault and Airbus. The need of Dassault is estimated to about 4 000 meters of profiles per month and the need of Airbus about 72 000 meters per month for a typical single aisle aircraft. The two most promising T-stringer production technologies were both investigated.

With the HP technology, in total 57 meters of preforms have been realised. The basic textile material used for the preforms was a non-crimped fabric material with IMS carbon fibres. The material and the profile design are typical for stringers used for fuselage applications.

During the project the manufacturing speed of the HP process could be increased up to 25m/h after several optimisation loops. The machine used for the preforming at Xperion was a prototype and no production machine. A final production speed of 30m/h seems reasonable after several optimisation steps and an optimisation of the process parameters. With the assumed production speed and the defined profile amount for both scenarios, the following machine demand can be determined.

The demand of Dassault of 4 000 meters of preforms could be realised with one production machine, working only in one shift. The Airbus demand is more challenging. In total six HP production machines have to work in three shifts to cope with the demand. Due to the fact, that the investment for six of these machines is not too high, the demand of manual work is low and the use of online QA-systems leads to a high quality and stable process, the production scenario is realistic.

The second production approach to realise T-stringer preforms is the FPP process. The FPP production machine processes endless UD tape material and produces a multilayer tape. Within IMAC-PRO we realised an automated tape production machine to investigate this complex approach. The multilayer tape has to be folded into an L-shaped preform afterwards. For this folding process, the principle of an automated machine was developed and a prototype was manufactured but not used for the folding of the preforms for the project. The investigation showed, that the folding process is not the bottleneck of the FPP production approach and therefore it is was not necessary to do a more intensive investigation to determine the output of this sub-process. To evaluate the manufacturing output of the FPP route, we only focus on the tape production machine.

With the FPP technology in total 250 meters of tape which were processed to 13 preforms could be produced. The basic textile material used for the preforms was a UD tape material with 80g/m2 out of IMS fibre. The manufacturing speed of the process could be increased up to 60m/h. This production speed is valid for the tape production of a three layer tape. To realise a tape with 19 single layers per side, nine revolver feeder and 10 0°-tape un-winding devices are necessary. The number of feeders has no influence on the production speed. With the assumption of an increase of productivity of 20% for real production, a preform production speed of 72m/h seems to be realistic. The 19 feeders/un-winding devices to produce one L-preform could also be used to produce the symmetric counterpart by changing the position of the 45°-feeders but with a reduction of the monthly output of 50%. In this case it is a question of investment to have a production speed of 72m/h or only half of it. The demand of Dassault of 4 000 meters of preforms could be realised with only one FPP production machine in one shift working only 8 days per month. In total three FPP tape production machines have to work in three shifts but only for 16 days per month, to cope with the demand of Airbus single aisle aircraft. Also for this manufacturing route, the possible high volume production scenario is realistic. The investment is tolerable and the share of manual work is also very low. An automated QA system could also be used and every single layer could be checked after positioning.

Summarised it can be said, that the HP technology has a high level of maturity but the FPP technology has a higher output, a lower scarp rate which is very important concerning environmental aspects and has weight specific a significant better performance.

The following evidence for exploitation of results can be stated:

A Falcon business jet spar from the HTP has been manufactured by Dassault-Aviation, with an UD-braided preform and RTM6 resin in a RTM mould:

1. visual aspect and material health were verified by NDI,
2. mechanical characteristics were investigated, showing a strength comparable to the currently used material.

This study has demonstrated the potential of the UD-braiding process up to a TRL6/7 in terms of manufacturing and mechanical properties. However further studies should be performed:

1. a partial spar test and associated stress analysis in order to verify that loads can be carried out with the same material mass,
2. additional test coupons to investigate the repeatability in terms of mechanical properties and consolidate the mechanical characteristics with a significant test campaign,
3. an improvement of the braiding process robustness and repeatability (lays and cutting, 90° lay deposition, fibre angles, preform expansion etc.).

Finally, a detailed cost effectiveness study should be performed to compare the current and UD-braiding range of manufacturing. Several modifications should be envisioned to use a UD-braided preform with the current range, such as the use of the braiding mandrel as a female injection tool, or the cutting to shape after injection and curing.

The promising results achieved in the field of stringer preforming will lead to further investigations in this topic. The FPP tape production machine which worked well for the project will be adopted to meet more the demands of Airbus. The stringer design changed compared to the year 2008 when the project started from T-to ?. So the machine will be modified to process wider tapes which are necessary to produce ?-stringers. Also the investigations made in the folding of the tapes will be helpful to design and realise an automated folding device for ?-stringers. An additional benefit of the adopted stringer design is the fact that the ? design is much more robust against impacts caused by tool drop which was the only drawback of the FPP material.

The work that has been done in IMAC-PRO in the field of frames braiding brought the following perceptions. The braiding of complex integrated profiles like JF-frames is possible and due to the good alignment of the fibres based on the UD-braiding process also with the required performance. The most challenging production step is the folding of the flange and the preform handling in general. But the investigations showed promising solution approaches also for high volume production. Due to the complex geometry of the JF Frame or an integrated profile in general it is still not possible to realise this kind of profile in a serial prepreg production process. But the weight saving aspect of this kind of profile compared to a state of the art profile like a C-frame is significant due to the possible release of hundreds of clips. So the motivation of the use of integrated frames in a composite aircraft is high and the UD-braiding technology offers an opportunity. Due to the work done in IMAC-PRO, the possible end users in the consortium like Sabca, Alenia, Dassault and EADS have a great interest to pursue the activities to reach a weight reduction in the fuselage and therefore reduce the carbon dioxide emission of future aircrafts.

The use of textile technologies like braiding or FPP makes it mandatory to use optical QA systems to check the fibre angles and other parameters to guarantee a reproducible quality in the parts. Two different systems could be developed within this project and their functionality could be proven. The beneficiaries who worked on this topic could increase their level of experience significantly and have therefore a good marked position with their future products to serve the rising marked in QA for textile carbon fibre (CF) preforms.

To cure these entire textile preforms infusion technologies are necessary. To react on the high volume demand we have in IMAC-PRO, we invented technologies like the compression RTM technology to increase the speed for resin infusion significantly. The development of the compression resin transfer process has progressed greatly due to this project via the development of the flat plate, C-beam, FX Core and JF frame tools. Whilst the process is not ideal for continuous frames and stringers, the fast impregnation that can be achieved due to the reduction in through thickness permeability of an un-compacted preform, is extremely well suited to highly automated production of parts with varying cross section (e.g. automotive industry).

The work conducted in this study has formed a basis to gain deep understanding of tool and process design. Now these developed approaches are being further developed for various aerospace and automotive structures, using a new generation of resin systems - termed 'snap cure' systems which may be designed to cure in three minutes. The cure reaction, hence viscosity may be activated to occur at very fast speeds in these resins - triggered by a critical temperature. Therefore, fast impregnation as well as the ability to control the temperature precisely, whilst also understanding the rheo-kinetics of a thermoset are vital - each of these aspects were developed during IMAC-PRO by collaboration between FHNW, Alpex and RUAG.

The development of a lightweight preforming and infusion tool for the complete cargo frame was very extensive and time consuming. The tool turned out to be much more complex than expected. As a result, the original plan to assemble the individual tools for the cargo floor cross beam and the fuselage frames (used in the development phase) to a single, integrated tool for the complete cargo unit had failed. An additional specialised tool to cure the cargo unit, which was not originally planned, had to be designed and manufactured. The final design was out of a combination of foam cores and pre-cured carbon templates wrapped with braided material and cured with epoxy resin. This is a complete new approach which worked in the end very well. Originally it was not planned to develop and realise a new lightweight tool concept but which is now from great interest for the tool manufacturer and end users in IMAC-PRO to be progressed.

Various dissemination activities have been performed during the duration of IMAC-PRO. In total six European and international conferences have been visited and more than 16 presentations were held with technical topics of IMAC-PRO. A complete session at the 33nd International Technical Conference and Forum, SEICO 12 in Paris, the Europe-wide audience of experienced CFRP manufacturers and users received a comprehensive overview over the project with four different presentation held by project beneficiaries.

Finally the SETEC conference in Lucerne (nearly at the end of the project) offered the opportunity to present the project results with six technical presentations and a panel session combined with a video presentation. Additionally two articles in technical magazines were published referring to the European project IMAC-PRO.

List of Websites:

To present the project to a wide audience, the project homepage (see http://www.imac-pro.eu) gives an overview over the project objectives and the project partners.