Final Report Summary - ACHSO (Automated Chemical Stitching and Preforming)
Executive Summary:
The Chemical Stitching und Continuous Preforming technologies had been developed independently and needed to be combined into an integrated process. This necessitated extensive design work and the manufacture and assembly of a new preform production line.
The forming, stitching and curing units though had to be designed specifically for the final product(s) to be produced on the device. The first step in the production is the cross section forming. To learn more about the possibilities of the cross sectional several concepts for draping mechanisms were considered and tested in the form of lab scale prototype devices (D2-2) to help choose the best technology for the final design. The two possibilities that exist are preforming with a solid forming tool and preforming using a set of rollers. For the preform geometry with L sections and flat doublers, the forming method of the roller based technique was chosen for best quality.
The rollers gradually change the flat textile into the L shape. After the forming and before the stabilisation of the stitching the shape is kept by a snug fitting solid tool, that is able to guide the preform but is too narrow for the flat doubler. The doubler preform is transported up and over the roller section of the preforming unit and passes the section thus without any clearance issued.
The chemical stitching was also advanced. First, concepts were developed how multiple stitching points can be set simultaneously. Also it was considered, what kind of injectors could be used in the final design. The simple injector used before was replaced with one that can deliver the controlled resin amounts necessary for the industrial application of this technology, this new injector was tested in a lab scale prototype of the stitching unit to determine the best process parameters concerning adhesive amount, curing time, curing temperature etc. In the detailed design of the production line the stitching unit was designed to reflect the production of the two different profile geometries.
The manufacturing and assembly of the new preforming line was completed on time. The integration of the resin injectors was the more challenging task. The injectors must move at the same speed than the textile layers through the preforming device. At the same time, they must also move up and down. These individual movements must be synchronised for the process to function. Other parameters like the resin viscosity, which depends on the amount of curing agent, also influence the setting of these movements. The viscosity influences the injection time, which is limited by the maximum pressure the injectors can achieve. The amount of curing agent curing agent must be set according to the speed of the textile, because this determines the amount of time the resin points will remain in the hot curing zone.
After lengthy optimisation of these parameters and the chemical stitching device, a functioning, stable and high quality process was achieved. Due to the automated and continuous manner of the preform production line, even a 24 hour usage of the device is viable with minimal downtime. Under these conditions, a production rate of 1000 m of stacked, shaped, stabilised, cut and infusion ready preforms is not unrealistic.
Project Context and Objectives:
Lightweight requirements have always accelerated innovation in the aviation industry and have found their current highpoint in the extensive application of composite materials in the Airbus A350 and Boeing 787. Lightweight structures help save fuel and CO2 making transport by aircraft at the same time more economical and more ecological. Increasingly though, in order to further European competitiveness, cost requirements also play a large role in the airplane design. This leads to two developments: firstly the use of dry fibre textiles and liquid composite moulding (LCM) technology instead of prepreg processes; and secondly the need for highly automated processes. These goals are somewhat in contrast to each other, since today’s prepreg processes are to a large degree automated while dry fibre textiles are more difficult to handle in automated processes. From this arises the need to develop fully automated dry fibre preform processes.
Fully automated preform processes are especially needed for the production of small parts which occur in high numbers. To mind come profiles such as stringers, frames and spars for which several kilometres of profile preforms need to be produced for each aircraft, resulting in a huge number of necessary preforms is one considers the future rates of aircraft production of more than 40 aircraft per months for single aisle aircraft. Such profile preforms are well fitted for continuous automated processes. Today’s state of the art continuous preforming technologies such as pultrusion however, do not offer the necessary quality and process control for aerospace applications. New preform processes must be developed to accommodate the demand for a fully automated preforming process for the production of 3D-shaped dry fibre profiles for the aerospace industry.
Today’s preforming technologies are largely manual, thus increasing the costs of Liquid Composite Moulding (LCM) technologies such as Resin Transfer Moulding (RTM). Systems that can drape 3D profiles automatically and continuously, such as the one developed by FIBRE, have just left the development stage but are in need of further development to in-crease productivity and quality. The project’s objective is to develop a fully automated dry fibre preform process for 3D-shaped composite profiles that uses the energy efficient chemical stitching technology.
The chemical stitching (CS) technology offers a way to reduce the lead time by replacing the time consuming binder application with localised adhesives application. For the application, needles are be used which inject either a thermoplastic (hot melt) or thermoset (microwave curing) adhesive between the layers. This will also improve permeability and thus the quality of the finished part. Furthermore is the chemical stitching technology more energy efficient because melting or curing of the adhesive is restricted to minimal areas and volumes.
In previous investigations within CleanSky EcoDesign project, the feasibility of chemical stitching approach was demonstrated by Fraunhofer Gesellschaft with ultraviolet (UV) and microwave (MW) curing adhesive.
The process starts with positioning of the adhesive application head on the textile structure which should be fixed. At the same time the textile structure is fixed mechanically. After-wards thin injection needles are driven into the textile structure to implicate minimal amount of the chosen binder adhesive between the layers. Following step is to cure the implemented adhesive.
The local adhesive implementation via Chemical Stitching approach shows a lot of advantages compared to areal bindering. These advantages are:
• Two times faster VARI infiltration of the preform (nearly the same as unbindered) due to a better permeability
• Same peeling forces by using 3 times lower amount of binder
• Same preform compressibility as a nonbindered textile structure.
• Higher peeling strength by using same amount of binder,
• Furthermore is the chemical stitching technology more energy efficient because melting or curing of the adhesive is restricted to minimal areas and volumes.
The CS technology is today only available as a lab scale test bench based on a gantry device. In this project the CS technology shall be advanced to level where it can function in a relevant industrial environment alongside other proven technologies. For this the process must move away from a single injector gantry device to a process that can stitch in parallel with very high speeds to match the line speed of the preforming process. In the lab device, the fabric is fixated with a vertical stamp. In the continuous process this will have to be removed to make the continuous movement of the fabric possible. This removal is easily implementable, because in the continuous preforming device, the fabrics are fixated in any case by the rollers that form and transport the fabric.
With the technology previously developed by FIBRE it is possible to produce full scale aerospace preform profiles of different geometries including complex LCF-frames, sidestepping any problems that will arise from up scaling. Up scaling is extremely difficult for fabric related processes because the weave cannot be scaled up or down without significantly changing the drapeability of the fabric. The prototype uses sets of rollers to form the flat fabric into the desired shape. The line speed can reach up to 10 m/min. Dependent on the profile geometry will the fabric layers fixed to each other before or after the forming sequence, using powdered binder across the entire area of the preform. The prototype includes devices for producing curved preforms, automatic cutting, and automatic removal of the preforms at the end.
This 3-D preforming technology can be combined with the chemical stitching technology to create an automated solution for a continuous, energy efficient process for the manufacturing of 3-D shaped composite dry fibre profiles. This requires developing the chemical stitching technology further from the current lab-scale batch process to an industrial scale that can function continuously. The stitching speed also has to be synchronised with the speed of the preforming process. In order to apply the adhesive in parallel to the preforming process an array of needles will inject the adhesive at different places simultaneously. Because the cycle of filling the injector, positioning it, injecting the adhesive, and retracting the injector is too time consuming, several injectors have to be arranged in sequence. To match the chemical stitching speed to the preforming speed several sets of needles need to be used. Another solution would be to move the needles horizontally to match the velocity of the fabric. The CS device is arranged past the preforming device to fix the fabric in its new shape.
In a second phase after the current project, it can be matched with a continuous curing technology. The basic technology of the continuous curing is has been developed also at FIBRE, but to bring it to industrial maturity further development is needed. This applies specifically to profile geometries that vary along their length e.g. due to ramps, steps or joggles.
This project's objective was the development of a preforming process that can produce reinforcement profiles preforms that are stabilised by chemical stitching. Chemical stitching is a technology to fixate the textile layers using very small amounts of locally applied liquid adhesive. The industry standard way is to use powdered binder adhesive across the entire textile surface. Replacing this with the minimum amount of adhesives used in chemical stitching improves infusion characteristics and laminate quality.
The process is fully automated and was validated by producing spar preforms, of high enough quality to integrate them into the CS EDA WP2 Demonstrator A4 “Aileron of Do 228”.
Project Results:
Concepts for draping mechanisms
To study the process of continuously preforming of a flat textile tape into a three-dimensional profile geometry a continuous preforming prototype was developed and build. The prototype works like the later to be build final preforming device but allows more freedom to change the preforming technique used. A proposed setup uses a solid metal forming tool, to form the flat textile tape on the right into a three-dimensional profile. This technique is investigated as an easier, less complicated way to shape the textile than the previously used technique with rollers. The continuous preforming prototype can, nonetheless, also be used with rollers in which the roller positions and attitudes can be changed in a greater degree than a final version that will be optimised towards an ideal shape.
Concepts for Stitching design
For the stitching injector a two-component injector head will be used. It can be filled with the two components of an adhesive which will be mixed at the moment of injection. This way no pre-mixed adhesives have to be used. Pre-mixed adhesives always have a limited use time, because the adhesive is slowly curing inside the adhesive reservoir. This can be avoided with the two-component head. It is also capable of injecting the highly viscous adhesive in a very short time. High viscosity is needed to achieve a localised disposition of the adhesive. If the viscosity is chosen too low it will be soaked up by the surrounding fibres instead of staying at the injection point. High viscosity on the other hand is detrimental to a short injection time. The chosen injector type is able to achieve a short injection even at the high viscosity that is needed. Before the final choice is made, the injector head will be tested with the actual adhesive to be used in the project.
To further increase the production speed, a multi-needle setup will be used. The injector head is going to be equipped with multiple needles. This way, several stitching points can be set at the same time and no movement of the injector head in the transversal y-axis is required.
Selection of technology for draping (rollers) and stitching (injector specified)
Even though a new draping mechanism using a solid tool was studied with a lab scale device, the known technique of gradually forming the flat textile into the desired shape. The rollers give better control of the preform especially considering the switch to L-profiles and even flat preforms (if they can be called as such). A C-preform is stabilised when it runs through the preforming device because its two flanges grip around the positive toll and so it cannot move left or right. An L-profile is not stabilised as such. Therefore it was decided revert back to the roller based setup planned in the beginning.
For the stitching unit an injector, that fulfilled the criteria for industrial use was selected. Up to this point, the chemical stitching used an injector that could only be used for very slow stitching of the textile, and would not be acceptable in an industrial application. The chosen injector head is the Viscotec Viscoduo V- 4/4. It is capable of placing very small amounts of resin at a speed that is usable in the integrated process. This is important because exact control over the deposited amounts of resin is necessary to produce preforms that are both stable but also have only a minimal amount of stitching resin in them. A too low amount of deposited resin in a stitching point would lower the preform stability necessary for subsequent handling and injection steps. A too high amount of resin would block the bath of injection resin in this area and would therefore be detrimental to preform permeability and also laminate mechanical properties.
Final design of integrated preforming production line achieved (MS2)
The most significant achievement is the final design for the integrated preform production line. It combines two very innovative technologies. The design also incorporates all the knowledge previously gained in this project. The tasks that preceded the final design advanced the preforming and stitching technology so that it could be integrated into an industrially usable machine.
An overview of the complete preform production line is given in Fig 6, in which the subunits that are combined to form and fixate the preform are shown. The subunits demonstrate the modularized, flexible design of the preforming production line. To the right is the material store unit where the different textile tapes for the different geometries are stored as coils. To the far left is the drive unit that pulls these tapes through the production line. In between the entire forming and fixation process takes place, completely automated. The flat tapes are formed into their 3-dimensional geometry (in case of the L-sections) in the preforming unit and fixated in their shape by chemical stitching in the subsequent stitching and curing units. The report will now take a closer look at each individual subunit.
Storage Unit
As previously stated, the production of preforms begins with the material storage unit. In this unit four tape reels can be stored to be used in the preform production. The number of individual coils is not just governed by the amount of plies necessary to reach the desired laminate thickness, but also by the laminate design itself. This means that in components with different mechanical loads, taped of different fibre orientations will be used in order to tailor the mechanical properties of the laminate. For the spar preforms 0° and ±45° non-crimp fabric (NCF) carbon fibre tapes will be used. The 0° tapes will be used predominantly in the spar caps (the L-sections), where tensile and compressive loads dominate. The straight 0° fibres of the NCF are best suited to carry these loads. In the spar web, the doublers, shear loads dominate. Thus, ±45° tapes are used for the doublers since the main load direction of a shear force is in the ±45° direction. However, because the tapes are pulled through the preforming line, a few 0° glass fibres are introduced into the ±45° tapes to stabilize them when being pulled. With the four coils in the material storage unit, all necessary material can be stored in the machine at all times. In the future, the material storage could be expanded to include more coils, if a wider variety of material is to be combines in one preform.
Preforming unit
The preforming unit as is one of the central aspects of the automated preform production line. The preforming line will be able to produce two different geometries, the L-Caps and the flat double webs. This will be possible without any changes to the preforming unit because two paths through the unit are possible. The lower, straight path forms the L sections, while for the flat doublers, for which no forming is necessary, the tapes are routed above the forming tool and back down at the end of the unit before the stitching unit. Compared to the final design of the substructure the path of the flat doubler preform was changed to integrate it better with the other subunits. Because the doublers and caps use different tapes as described in the previous paragraph, the production can be changed from one geometry to the next quickly. The L-section is formed by a set of rollers and a solid metal forming tool. First the rollers change the flat textile gradually from the flat tape into the final shape by increasing the angle of the L from 0° to the final 100° (outside angle). The solid forming tool then stabilises the preform until it can be fixated in the stitching and curing units.
Stitching unit
That two component geometries are being fabricated on the same machine is also reflected in the design of the stitching unit, as seen in a top-down view in Fig 9. During the stitching process two stitching heads inject small amounts of a 2-component resin between the textile layers. Because of the constant movement of the textile substrate, the injection heads must move in a synchronized movement along the axis of production. The injection heads are therefore attached two linear drives that are synched up with the machine through a computer. In order to eliminate any chance of collision between the stitching heads, the linear drive are mounted behind each other even though the rows of stitching points made by each are parallel. The horizontal movement f the stitching heads in the machine x-direction (the direction of production) is overlaid with a motion orthogonal to the substrate. This is necessary to push the needle in between the textile layers, where the resin is to be deposited and pull it out again after all resin is at the correct place. This motion is either in the z-axis or the y-axis depending on the geometry of the preform. For the flat doublers, both injector heads perform virtually identical motions in the x and z-directions. For the L-sections though, because the two flanges are at an angle close to 90°, one injector heads has to work in the y-direction (horizontally). For this, one injector head can be swung down and arrested in a second position to set the row of stitching points on the vertical frame.
Curing unit
The curing unit is necessary to cure the resin injected as a form of preform stabilisation in the previous step. It consists of compaction section and the heating section. In the first section a series of rollers compact the textile layers again after the stitching process, and before the curing. After the curing, the injected resin points are hardened and the textile layers cannot be moved again. It must be made sure therefore, that before the resin is cured the preform is compacted to its desired thickness. The stitching points are cured by a set of infrared heaters. The length of the heating zone has been chosen that with the given power output of the heaters, the resin is fully cured while moving continuously along through the heating zone at the desired production speed.
Drive unit
At the very end of the preform production line, sits the drive unit. The drive wheels are pressed down on the preforms and pull all textile layers through the entire production line. This way, the textile layers always under tension, which works against the formation of undulations due to compressive forces in the textile. It requires a small amount of 0° fibres in ±45° layers though, because usually pulling a ±45° textile in the 0° direction means strong deformation because ±45° fibres cannot transmit the tensile loads of a 0° pulling action.
Material and Lay-Up
As lay-up a symmetrical stacking of (±45/0)s was chosen for the spars. The doublers will contain the same material with a lay-up of (±45/0/±45). By choosing these stackings and textiles it was realised to use the same supply-coils for spars and doublers. The resulting fibre-volume content will be 54% for the spars and 59% for the doublers.
Preform production line
An overview of the complete preform production line is given in Fig 6, in which the subunits that are combined to form and fixate the preform are shown. The subunits demonstrate the modularized, flexible design of the preforming production line. To the right is the material store unit where the different textile tapes for the different geometries are stored as coils. To the far left is the drive unit that pulls these tapes through the production line. In between the entire forming and fixation process takes place, completely automated. The flat tapes are formed into their profile geometry (in case of the L-sections) in the preforming unit and fixated in their shape by chemical stitching in the subsequent stitching and curing units.
Assembly of Prototype
In order to validate the function of the basic preforming operation of the preform production line, the shape giving features of the device were tested. The Spar Caps (L-geometry) consist of four layers of non-crimp fabric (NCF). These layers are being reeled off the material store reels and combined into a flat stack before being fed into the preforming unit. The single layers have to be aligned precisely to form an even stack. Any misalignment in this part of the production line will result in problems in the later units. The stack is being fed through solid shaping tools in the later steps. Also, misalignment of the single layers cannot be corrected later in the production line so that any geometry deviation will manifest in out-of-tolerance parts in the end. The reels carrying the material are braked to ensure a constant tension on the layers.
After the layers are combined the preforming unit forms the flat stack into an L-shaped profile. A set of rollers is used for a gradual forming of the flat tapes. The rollers only use little down force, to avoid damaging the fabric. Their function is also not consolidation of the preform, which happens later in the production line but just shaping. The roller’s shapes change from the right to the left, gradually increasing the angle of the “L” shape. The down force of the rollers can be adjusted manually and is calibrated at the start of the production cycle. The down force is set so that the preform can be pulled through the entire production line without damaging the textile while at the same time forcing the flat tape into the desired shape.
The preform is being passed through a solid shaping tool after the section with the rollers that has very little tolerance on any geometry deviations. If the preform can be fed through the tool it has exactly the necessary shape and is ready to be used in further processes.
The maximum line speed that can be reached in the device is 1 m/min. The limiting factor is the synchronisation between the drives that move the injector head and the drive that pulls the preform through the line. The chemical stitching unit and subsequently also the curing unit were not used at this stage, because the special injectors for the chemical stitching could not be delivered in time. These steps will be done in the beginning of 2014.
Preform Optimisation
After the successful assembly and initial operation of the prototype preform production line, the process parameters could be adjusted to achieve a fast and stable process that produces high quality parts. The optimisation parameters are mostly interdependent, which mostly stems from the fact that two technologies, chemical stitching and continuous preforming, were combined in this project. The process parameters that were varied in the course of the process optimisation are the following:
- Line speed
- Continuous / semi continuous process
- Needle geometry
- Resin viscosity
- Stack consolidation /spread
- Injection pressure
- Injection time
- Other injection parameters
In an integrated process such as the combination of automated preforming and chemical stitching in one device that was the centre of the project AChSo, the individual parameters are rarely independent of each other. Changing one parameter often results in better performance in one part of the production line but worse performance in another. The interdependencies can often be estimated qualitatively but the exact influences of one on another are not known for new processes such as this. A lot of experience is necessary to achieve control over the process. This experience cannot be derived through other means as trial and error experiments.
Line speed is obviously important for the economic production of preforms. Continuous production processes are deployed, when a high number of parts are needed. The line speed was chosen as the overall goal parameter in this process. As in many processes, production speed is inversely coupled with production quality. At low speeds an exact control of all parts of the integrated process line is possible, but production is not economic. Increasing the speeds lets one usually produce quality parts until a certain cut off point is reached and the produced parts do not comply with the requirements anymore. Following the philosophy of “as exact as necessary and not as exact as possible” this cut off point must be found. Depending on the stability of the process and the tolerances of the requirements the set of parameters that bring one closest to this red line with a stable process must be found.
The programming of the drives for pulling the fibre preform and for moving the injector heads let the user choose between two process states. Firstly a continuous process that required parallel synchronised motion of the fibre preform substrate and the resin injector heads, this is the one that should be used for an economic process. Secondly a semi-continuous process in which the movement of the fibre preforms was halted during the chemical stitching. No x-Movement of the injector is necessary in this case. This mode should only be used in the set-up stage as it is considerably slower than the continuous one.
However, it lets the user decouple the preforming and the chemical stitching processes. On the one hand this makes it easier to find the right parameters for each singular process. On the other hand, once going back to the continuous process the interdependencies between the two processes must be found, as described above. Because the injectors that are responsible for the chemical stitching are a very complex system in itself, the semi-continuous process was used to understand their behaviour and produce stitch points that were considered acceptable. The next step was to produce these stitch points in the continuous process.
In the stitching process alone several parameters can be adjusted. The resin itself can be varied in its viscosity. The reason to increase the viscosity is the behaviour of it in the textile preform. With the usual viscosity, the resin can be injected very easily between the preform layers, but is dispersed there in a wide area due to the capillary forces between the filaments. It is, in effect, sucked up by the fabric substrate like water by a cloth and cannot form a localised stitch point. The viscosity of the resin can be adjusted by adding calcium carbonate to the resin. The median size of the CaCO particles is 2.7 µm so it can be dispersed in the liquid resin easily. Increasing the resin viscosity, will make it harder and harder to transport it through the injectors and the needles into the preform. This will increase the time necessary to stitch one point, slowing down the process. It might also result in injection pressures that are too high for any of the parts involved, the weakest usually being the static mix tube. The concentration of CaCO was increased until the viscosity was high enough to sufficiently stay at the stitch point. The chosen mixture was resin weight to CaCO 1:1.4.
To lower the injection pressure, a larger diameter needle might be used. The limit for the needle geometry is the ability to penetrate the preform. If the needle is too thick it might either not perforate the upper layers of the textile at all or damage the punctured fibres in a way that is detrimental to preform quality. The needle can penetrate the upper layers more easily it has a chamfered tip instead of a flat tip. This is turn smears the resin over a larger area in the z-direction (thickness), which can be positive or negative depending on the degree of consolidation that is chosen for the preform at this point. If the preform is strongly compacted the chamfering might result in spilling most of the resin on top of the preform because part of the needle opening sticks out of the textile stack. If, on the other hand, the layers are spread out the chamfering can aid in spreading the resin regularly between all the layers. A good compromise was found with needles of 0.84 mm diameter and a 20° chamfered tip.
The consolidation of the stack can be adjusted by adding rollers or spreaders in the area directly in front of the stitching area. In the forming part of the preforming line, a set of rollers is used to bring the flat material into the desired L-shape. This also leads to a pre-compaction of the textile stack. The consolidation of the textile stack during the stitching influences the final preform in several ways. With a high consolidation, the resistance for the resin to flow through the layers and wet each layer is high. The chance of failure to stitch all layers rises. On the other hand does a high degree of consolidation lead to a good, localised application of the resin and a small stitch point. If the resin in injected into a less consolidated stack later consolidation tend to smear the resin across a wider area. The lower consolidation raises the permeability of the stack and makes it thus easier for the resin to flow through the preform. This lowers injection pressures and injection time and ultimately leads to higher line speeds. To spread the layers after the pre-consolidation in the forming section, a comb can be placed directly in front of the stitching area. This separates the layers and creates a space between each of them into which the resin can flow without much resistance. In the end the comb was chosen over the extra roller to achieve greater separation and a better overall process.
Production of preforms
In the report of the preceding deliverable D6-1, the process optimisation was described in detail. A number of interdependent parameters were adjusted to improve the preforming quality. A final improvement was done subsequently to further improve process stability and reproducibility of the preforms. Before, the resin was injected into the middles of the stack and would then flow through the textile to reach all layers. While this produced preforms with all layers bonded by the resin, the amount and placement of the resin depended on the flow processes of thee resin.
To improve upon this, the z-axis that moves the injector up and down was redesigned to be adjustable in speed. An electric drive and a controllable pneumatic drive for the z-axis were considered. Finally, a mechanical solution was chosen that lets the user control the vertical speed by changing the inclination of an iglidur® glide ramp. This simple solution achieved the most consistent results and reduced the complexity of the overall device and control, compared to the pneumatic or electric variants.
With the controllable z-axis speed, the deposition of the resin between the layers could be controlled more exact. The amount of resin between each layer can be adjusted by changing the angle of the ramp, with a lower ramp angle letting the injector stay longer between two layers and depositing more resin. The position of the resin is better reproducible, because it is no longer dependent on the complex flow process of the resin through the textile.
The position of the stitch points was chosen to stabilize the preform geometry and the laminate stack. A row of stitch points on each edge keeps the textile layers together at the area where they are most vulnerable to being damaged. The second needle of each injector is positioned close to the middle bend of the L shaped profile and stabilizes the geometry of the preform. By injecting resin in this area of the preform, the preform is kept from relaxing into the flat shape of the textile raw materials.
The continuous preforming process ran stable and reproducible. Profile preforms were produced in a continuous manner and cut into the desired length for the resin infusion process without stopping the production line. Once set up, the process can run endlessly. If the textile is running out, a new strip cab spliced onto the end of the previous one without stopping the process, although the part of the preform with the seam must probably be discarded. The resin can be refilled during the running process by using two resupply containers for each component that can be switched with a valve. Due to the automated and continuous manner of the preform production line, even a 24 hour usage of the device is viable with minimal downtime. Under these conditions, a production rate of 1000 m of stacked, shaped, stabilised, cut and infusion ready preforms is not unrealistic.
Potential Impact:
Final result and Impact
The final result of the project is the development of an innovative preforming process and the demonstration of this technology in a prototype preform production line. A functioning, stable and high quality process was achieved. The process is suited to produce profile preforms with a large variation of geometries and material (both laminate and roving material). Such profile preforms are used in the aerospace industry for reinforcement structures such as stringers and spars, but potentially also be used in other fields that use profile structures. Automotive and marine applications are possible but also civic uses for buildings or bridges. The process can be enhanced beyond the state established in this project to produce curved preforms and preforms that change their profile geometry lengthwise. Due to the automated and continuous manner of the preform production line, even a 24 hour usage of the device is viable with minimal downtime. Under these conditions, a production rate of 1000 m of stacked, shaped, stabilised, cut and infusion ready preforms is not unrealistic.
The goals of ecological improvements set by EU partner countries can arguably only be achieved when they do not impair, or at best, even improve upon, European competiveness. Composite materials are well suited to lower the carbon dioxide emissions through lighter structures and thus lower fuel consumption. Industrial composite processes that produce the high quality components necessary for aerospace applications are not able to produce the large amount of party needed for a large scale use of composites.
The continuous preforming technology demonstrated in this project can deliver such a process. It combines the high performance material and high quality preforms necessary for aerospace applications with low operational expenditures. It achieves this by moving from a manual process to a highly automated one. This increases European competitiveness two-fold.
Firstly, the developed technology is superior to the ones available from non-European companies. European automation companies that are able to develop such a process will have an advantage over their competitors. Secondly, through the networking that is perpetuated by EU programs such as CleanSky, European airframe structure producers will sooner be informed about such innovative products. This and the close connections formed through the EU projects also reduce the time to market for emerging technologies. Thirdly, the higher degree of automation and lower operational expenditures make it more likely for industrial process to remain within the EU. More manual processes are more likely to be outsourced to countries with lower personnel cost. More complex, automated processes need more highly trained personnel, who are available more likely in EU than in competing countries.
List of Websites:
Dissemination
Dissemination activities of the project results took several shapes. For scientific dissemination, a paper was published and presented. To further the industrial application of the developed processes, the results will prepared to be shown at trade fairs. On the project side, the technologies that were part of this project will be the focus of future projects that will build upon the results of AChSo.
A paper called “Implementation of Automated Chemical Stitching into the Continuous Preforming Process” was presented at the Greener Aviation Conference that took place from 12th - 14th March 2014 in Brussels, Belgium. The paper has been published in the conference proceedings. The Greener Aviation conference is specifically set up to show the work done in the CleanSky program. This is a good opportunity to meet other members of the European industrial research community. The few opportunities for networking between the CfP-Partners and the core partners of CleanSky can be equalized somewhat by events like this. The presentation and discussion of the results in Brussels were a good way to show the developed process and gather input from fellow researchers for future development of this technology.
To show the project results to the industry, so that the developments might be implemented in industrial serial production, presentations at trade shows is an effective way of exposing the project result to an interested public. The most important international trade show for composite products and processes is the JEC in Paris, France. The continuous preforming technology was presented at the 2014 JEC and the integration of the Chemical Stitching into it will be presented at the JEC 2015. A movie is being produced at the moment to show the working process and the results. This movie will also be shown at other opportunities and be released through FIBRE’s internet presence (www.faserinstitut.de ).
The project results have been published (http://www.textilforschung.de/en/project?id=1) in the information portal of the German Textile Research Board (FKT Forschungskuratorium Textil). The FKT explicitly connects textile researchers in academia with textile users in the industry. It is active by organizing dissemination events in which textile research institutes can present their innovations to leaders and decision-makers and funds cooperational research by institutes and industry players. The research board can thus function as a great multiplier to spread the results of this project.
The technologies used in this project will be further pursued in future projects. For example, the process demonstrator will be used to develop a process that builds on top of the chemical stitching technology but for thermoplastic matrix systems. The current chemical stitching uses thermoset stitch points for use in thermoset LCM processes. For cases where thermoplastic matrix materials are favoured, a different set-up is needed, but it will retain the basic concept with localized fix points, instead of the fixation agent across the entire area of the textile.
Participating in the pan-European CleanSky program was a very good way to meet new partners on a transnational level. The combination of two innovative technologies in one project was a very interesting approach. All partners were able to gather new know-how from each other and also utilize each other’s network. Because more interested parties will be exposed to the technologies developed within a joint program such as CleanSky, it is more likely that research will find its way into the serial production of aerospace components. FIBRE will try to enter in the CleanSky2 program as a partner.
The Chemical Stitching und Continuous Preforming technologies had been developed independently and needed to be combined into an integrated process. This necessitated extensive design work and the manufacture and assembly of a new preform production line.
The forming, stitching and curing units though had to be designed specifically for the final product(s) to be produced on the device. The first step in the production is the cross section forming. To learn more about the possibilities of the cross sectional several concepts for draping mechanisms were considered and tested in the form of lab scale prototype devices (D2-2) to help choose the best technology for the final design. The two possibilities that exist are preforming with a solid forming tool and preforming using a set of rollers. For the preform geometry with L sections and flat doublers, the forming method of the roller based technique was chosen for best quality.
The rollers gradually change the flat textile into the L shape. After the forming and before the stabilisation of the stitching the shape is kept by a snug fitting solid tool, that is able to guide the preform but is too narrow for the flat doubler. The doubler preform is transported up and over the roller section of the preforming unit and passes the section thus without any clearance issued.
The chemical stitching was also advanced. First, concepts were developed how multiple stitching points can be set simultaneously. Also it was considered, what kind of injectors could be used in the final design. The simple injector used before was replaced with one that can deliver the controlled resin amounts necessary for the industrial application of this technology, this new injector was tested in a lab scale prototype of the stitching unit to determine the best process parameters concerning adhesive amount, curing time, curing temperature etc. In the detailed design of the production line the stitching unit was designed to reflect the production of the two different profile geometries.
The manufacturing and assembly of the new preforming line was completed on time. The integration of the resin injectors was the more challenging task. The injectors must move at the same speed than the textile layers through the preforming device. At the same time, they must also move up and down. These individual movements must be synchronised for the process to function. Other parameters like the resin viscosity, which depends on the amount of curing agent, also influence the setting of these movements. The viscosity influences the injection time, which is limited by the maximum pressure the injectors can achieve. The amount of curing agent curing agent must be set according to the speed of the textile, because this determines the amount of time the resin points will remain in the hot curing zone.
After lengthy optimisation of these parameters and the chemical stitching device, a functioning, stable and high quality process was achieved. Due to the automated and continuous manner of the preform production line, even a 24 hour usage of the device is viable with minimal downtime. Under these conditions, a production rate of 1000 m of stacked, shaped, stabilised, cut and infusion ready preforms is not unrealistic.
Project Context and Objectives:
Lightweight requirements have always accelerated innovation in the aviation industry and have found their current highpoint in the extensive application of composite materials in the Airbus A350 and Boeing 787. Lightweight structures help save fuel and CO2 making transport by aircraft at the same time more economical and more ecological. Increasingly though, in order to further European competitiveness, cost requirements also play a large role in the airplane design. This leads to two developments: firstly the use of dry fibre textiles and liquid composite moulding (LCM) technology instead of prepreg processes; and secondly the need for highly automated processes. These goals are somewhat in contrast to each other, since today’s prepreg processes are to a large degree automated while dry fibre textiles are more difficult to handle in automated processes. From this arises the need to develop fully automated dry fibre preform processes.
Fully automated preform processes are especially needed for the production of small parts which occur in high numbers. To mind come profiles such as stringers, frames and spars for which several kilometres of profile preforms need to be produced for each aircraft, resulting in a huge number of necessary preforms is one considers the future rates of aircraft production of more than 40 aircraft per months for single aisle aircraft. Such profile preforms are well fitted for continuous automated processes. Today’s state of the art continuous preforming technologies such as pultrusion however, do not offer the necessary quality and process control for aerospace applications. New preform processes must be developed to accommodate the demand for a fully automated preforming process for the production of 3D-shaped dry fibre profiles for the aerospace industry.
Today’s preforming technologies are largely manual, thus increasing the costs of Liquid Composite Moulding (LCM) technologies such as Resin Transfer Moulding (RTM). Systems that can drape 3D profiles automatically and continuously, such as the one developed by FIBRE, have just left the development stage but are in need of further development to in-crease productivity and quality. The project’s objective is to develop a fully automated dry fibre preform process for 3D-shaped composite profiles that uses the energy efficient chemical stitching technology.
The chemical stitching (CS) technology offers a way to reduce the lead time by replacing the time consuming binder application with localised adhesives application. For the application, needles are be used which inject either a thermoplastic (hot melt) or thermoset (microwave curing) adhesive between the layers. This will also improve permeability and thus the quality of the finished part. Furthermore is the chemical stitching technology more energy efficient because melting or curing of the adhesive is restricted to minimal areas and volumes.
In previous investigations within CleanSky EcoDesign project, the feasibility of chemical stitching approach was demonstrated by Fraunhofer Gesellschaft with ultraviolet (UV) and microwave (MW) curing adhesive.
The process starts with positioning of the adhesive application head on the textile structure which should be fixed. At the same time the textile structure is fixed mechanically. After-wards thin injection needles are driven into the textile structure to implicate minimal amount of the chosen binder adhesive between the layers. Following step is to cure the implemented adhesive.
The local adhesive implementation via Chemical Stitching approach shows a lot of advantages compared to areal bindering. These advantages are:
• Two times faster VARI infiltration of the preform (nearly the same as unbindered) due to a better permeability
• Same peeling forces by using 3 times lower amount of binder
• Same preform compressibility as a nonbindered textile structure.
• Higher peeling strength by using same amount of binder,
• Furthermore is the chemical stitching technology more energy efficient because melting or curing of the adhesive is restricted to minimal areas and volumes.
The CS technology is today only available as a lab scale test bench based on a gantry device. In this project the CS technology shall be advanced to level where it can function in a relevant industrial environment alongside other proven technologies. For this the process must move away from a single injector gantry device to a process that can stitch in parallel with very high speeds to match the line speed of the preforming process. In the lab device, the fabric is fixated with a vertical stamp. In the continuous process this will have to be removed to make the continuous movement of the fabric possible. This removal is easily implementable, because in the continuous preforming device, the fabrics are fixated in any case by the rollers that form and transport the fabric.
With the technology previously developed by FIBRE it is possible to produce full scale aerospace preform profiles of different geometries including complex LCF-frames, sidestepping any problems that will arise from up scaling. Up scaling is extremely difficult for fabric related processes because the weave cannot be scaled up or down without significantly changing the drapeability of the fabric. The prototype uses sets of rollers to form the flat fabric into the desired shape. The line speed can reach up to 10 m/min. Dependent on the profile geometry will the fabric layers fixed to each other before or after the forming sequence, using powdered binder across the entire area of the preform. The prototype includes devices for producing curved preforms, automatic cutting, and automatic removal of the preforms at the end.
This 3-D preforming technology can be combined with the chemical stitching technology to create an automated solution for a continuous, energy efficient process for the manufacturing of 3-D shaped composite dry fibre profiles. This requires developing the chemical stitching technology further from the current lab-scale batch process to an industrial scale that can function continuously. The stitching speed also has to be synchronised with the speed of the preforming process. In order to apply the adhesive in parallel to the preforming process an array of needles will inject the adhesive at different places simultaneously. Because the cycle of filling the injector, positioning it, injecting the adhesive, and retracting the injector is too time consuming, several injectors have to be arranged in sequence. To match the chemical stitching speed to the preforming speed several sets of needles need to be used. Another solution would be to move the needles horizontally to match the velocity of the fabric. The CS device is arranged past the preforming device to fix the fabric in its new shape.
In a second phase after the current project, it can be matched with a continuous curing technology. The basic technology of the continuous curing is has been developed also at FIBRE, but to bring it to industrial maturity further development is needed. This applies specifically to profile geometries that vary along their length e.g. due to ramps, steps or joggles.
This project's objective was the development of a preforming process that can produce reinforcement profiles preforms that are stabilised by chemical stitching. Chemical stitching is a technology to fixate the textile layers using very small amounts of locally applied liquid adhesive. The industry standard way is to use powdered binder adhesive across the entire textile surface. Replacing this with the minimum amount of adhesives used in chemical stitching improves infusion characteristics and laminate quality.
The process is fully automated and was validated by producing spar preforms, of high enough quality to integrate them into the CS EDA WP2 Demonstrator A4 “Aileron of Do 228”.
Project Results:
Concepts for draping mechanisms
To study the process of continuously preforming of a flat textile tape into a three-dimensional profile geometry a continuous preforming prototype was developed and build. The prototype works like the later to be build final preforming device but allows more freedom to change the preforming technique used. A proposed setup uses a solid metal forming tool, to form the flat textile tape on the right into a three-dimensional profile. This technique is investigated as an easier, less complicated way to shape the textile than the previously used technique with rollers. The continuous preforming prototype can, nonetheless, also be used with rollers in which the roller positions and attitudes can be changed in a greater degree than a final version that will be optimised towards an ideal shape.
Concepts for Stitching design
For the stitching injector a two-component injector head will be used. It can be filled with the two components of an adhesive which will be mixed at the moment of injection. This way no pre-mixed adhesives have to be used. Pre-mixed adhesives always have a limited use time, because the adhesive is slowly curing inside the adhesive reservoir. This can be avoided with the two-component head. It is also capable of injecting the highly viscous adhesive in a very short time. High viscosity is needed to achieve a localised disposition of the adhesive. If the viscosity is chosen too low it will be soaked up by the surrounding fibres instead of staying at the injection point. High viscosity on the other hand is detrimental to a short injection time. The chosen injector type is able to achieve a short injection even at the high viscosity that is needed. Before the final choice is made, the injector head will be tested with the actual adhesive to be used in the project.
To further increase the production speed, a multi-needle setup will be used. The injector head is going to be equipped with multiple needles. This way, several stitching points can be set at the same time and no movement of the injector head in the transversal y-axis is required.
Selection of technology for draping (rollers) and stitching (injector specified)
Even though a new draping mechanism using a solid tool was studied with a lab scale device, the known technique of gradually forming the flat textile into the desired shape. The rollers give better control of the preform especially considering the switch to L-profiles and even flat preforms (if they can be called as such). A C-preform is stabilised when it runs through the preforming device because its two flanges grip around the positive toll and so it cannot move left or right. An L-profile is not stabilised as such. Therefore it was decided revert back to the roller based setup planned in the beginning.
For the stitching unit an injector, that fulfilled the criteria for industrial use was selected. Up to this point, the chemical stitching used an injector that could only be used for very slow stitching of the textile, and would not be acceptable in an industrial application. The chosen injector head is the Viscotec Viscoduo V- 4/4. It is capable of placing very small amounts of resin at a speed that is usable in the integrated process. This is important because exact control over the deposited amounts of resin is necessary to produce preforms that are both stable but also have only a minimal amount of stitching resin in them. A too low amount of deposited resin in a stitching point would lower the preform stability necessary for subsequent handling and injection steps. A too high amount of resin would block the bath of injection resin in this area and would therefore be detrimental to preform permeability and also laminate mechanical properties.
Final design of integrated preforming production line achieved (MS2)
The most significant achievement is the final design for the integrated preform production line. It combines two very innovative technologies. The design also incorporates all the knowledge previously gained in this project. The tasks that preceded the final design advanced the preforming and stitching technology so that it could be integrated into an industrially usable machine.
An overview of the complete preform production line is given in Fig 6, in which the subunits that are combined to form and fixate the preform are shown. The subunits demonstrate the modularized, flexible design of the preforming production line. To the right is the material store unit where the different textile tapes for the different geometries are stored as coils. To the far left is the drive unit that pulls these tapes through the production line. In between the entire forming and fixation process takes place, completely automated. The flat tapes are formed into their 3-dimensional geometry (in case of the L-sections) in the preforming unit and fixated in their shape by chemical stitching in the subsequent stitching and curing units. The report will now take a closer look at each individual subunit.
Storage Unit
As previously stated, the production of preforms begins with the material storage unit. In this unit four tape reels can be stored to be used in the preform production. The number of individual coils is not just governed by the amount of plies necessary to reach the desired laminate thickness, but also by the laminate design itself. This means that in components with different mechanical loads, taped of different fibre orientations will be used in order to tailor the mechanical properties of the laminate. For the spar preforms 0° and ±45° non-crimp fabric (NCF) carbon fibre tapes will be used. The 0° tapes will be used predominantly in the spar caps (the L-sections), where tensile and compressive loads dominate. The straight 0° fibres of the NCF are best suited to carry these loads. In the spar web, the doublers, shear loads dominate. Thus, ±45° tapes are used for the doublers since the main load direction of a shear force is in the ±45° direction. However, because the tapes are pulled through the preforming line, a few 0° glass fibres are introduced into the ±45° tapes to stabilize them when being pulled. With the four coils in the material storage unit, all necessary material can be stored in the machine at all times. In the future, the material storage could be expanded to include more coils, if a wider variety of material is to be combines in one preform.
Preforming unit
The preforming unit as is one of the central aspects of the automated preform production line. The preforming line will be able to produce two different geometries, the L-Caps and the flat double webs. This will be possible without any changes to the preforming unit because two paths through the unit are possible. The lower, straight path forms the L sections, while for the flat doublers, for which no forming is necessary, the tapes are routed above the forming tool and back down at the end of the unit before the stitching unit. Compared to the final design of the substructure the path of the flat doubler preform was changed to integrate it better with the other subunits. Because the doublers and caps use different tapes as described in the previous paragraph, the production can be changed from one geometry to the next quickly. The L-section is formed by a set of rollers and a solid metal forming tool. First the rollers change the flat textile gradually from the flat tape into the final shape by increasing the angle of the L from 0° to the final 100° (outside angle). The solid forming tool then stabilises the preform until it can be fixated in the stitching and curing units.
Stitching unit
That two component geometries are being fabricated on the same machine is also reflected in the design of the stitching unit, as seen in a top-down view in Fig 9. During the stitching process two stitching heads inject small amounts of a 2-component resin between the textile layers. Because of the constant movement of the textile substrate, the injection heads must move in a synchronized movement along the axis of production. The injection heads are therefore attached two linear drives that are synched up with the machine through a computer. In order to eliminate any chance of collision between the stitching heads, the linear drive are mounted behind each other even though the rows of stitching points made by each are parallel. The horizontal movement f the stitching heads in the machine x-direction (the direction of production) is overlaid with a motion orthogonal to the substrate. This is necessary to push the needle in between the textile layers, where the resin is to be deposited and pull it out again after all resin is at the correct place. This motion is either in the z-axis or the y-axis depending on the geometry of the preform. For the flat doublers, both injector heads perform virtually identical motions in the x and z-directions. For the L-sections though, because the two flanges are at an angle close to 90°, one injector heads has to work in the y-direction (horizontally). For this, one injector head can be swung down and arrested in a second position to set the row of stitching points on the vertical frame.
Curing unit
The curing unit is necessary to cure the resin injected as a form of preform stabilisation in the previous step. It consists of compaction section and the heating section. In the first section a series of rollers compact the textile layers again after the stitching process, and before the curing. After the curing, the injected resin points are hardened and the textile layers cannot be moved again. It must be made sure therefore, that before the resin is cured the preform is compacted to its desired thickness. The stitching points are cured by a set of infrared heaters. The length of the heating zone has been chosen that with the given power output of the heaters, the resin is fully cured while moving continuously along through the heating zone at the desired production speed.
Drive unit
At the very end of the preform production line, sits the drive unit. The drive wheels are pressed down on the preforms and pull all textile layers through the entire production line. This way, the textile layers always under tension, which works against the formation of undulations due to compressive forces in the textile. It requires a small amount of 0° fibres in ±45° layers though, because usually pulling a ±45° textile in the 0° direction means strong deformation because ±45° fibres cannot transmit the tensile loads of a 0° pulling action.
Material and Lay-Up
As lay-up a symmetrical stacking of (±45/0)s was chosen for the spars. The doublers will contain the same material with a lay-up of (±45/0/±45). By choosing these stackings and textiles it was realised to use the same supply-coils for spars and doublers. The resulting fibre-volume content will be 54% for the spars and 59% for the doublers.
Preform production line
An overview of the complete preform production line is given in Fig 6, in which the subunits that are combined to form and fixate the preform are shown. The subunits demonstrate the modularized, flexible design of the preforming production line. To the right is the material store unit where the different textile tapes for the different geometries are stored as coils. To the far left is the drive unit that pulls these tapes through the production line. In between the entire forming and fixation process takes place, completely automated. The flat tapes are formed into their profile geometry (in case of the L-sections) in the preforming unit and fixated in their shape by chemical stitching in the subsequent stitching and curing units.
Assembly of Prototype
In order to validate the function of the basic preforming operation of the preform production line, the shape giving features of the device were tested. The Spar Caps (L-geometry) consist of four layers of non-crimp fabric (NCF). These layers are being reeled off the material store reels and combined into a flat stack before being fed into the preforming unit. The single layers have to be aligned precisely to form an even stack. Any misalignment in this part of the production line will result in problems in the later units. The stack is being fed through solid shaping tools in the later steps. Also, misalignment of the single layers cannot be corrected later in the production line so that any geometry deviation will manifest in out-of-tolerance parts in the end. The reels carrying the material are braked to ensure a constant tension on the layers.
After the layers are combined the preforming unit forms the flat stack into an L-shaped profile. A set of rollers is used for a gradual forming of the flat tapes. The rollers only use little down force, to avoid damaging the fabric. Their function is also not consolidation of the preform, which happens later in the production line but just shaping. The roller’s shapes change from the right to the left, gradually increasing the angle of the “L” shape. The down force of the rollers can be adjusted manually and is calibrated at the start of the production cycle. The down force is set so that the preform can be pulled through the entire production line without damaging the textile while at the same time forcing the flat tape into the desired shape.
The preform is being passed through a solid shaping tool after the section with the rollers that has very little tolerance on any geometry deviations. If the preform can be fed through the tool it has exactly the necessary shape and is ready to be used in further processes.
The maximum line speed that can be reached in the device is 1 m/min. The limiting factor is the synchronisation between the drives that move the injector head and the drive that pulls the preform through the line. The chemical stitching unit and subsequently also the curing unit were not used at this stage, because the special injectors for the chemical stitching could not be delivered in time. These steps will be done in the beginning of 2014.
Preform Optimisation
After the successful assembly and initial operation of the prototype preform production line, the process parameters could be adjusted to achieve a fast and stable process that produces high quality parts. The optimisation parameters are mostly interdependent, which mostly stems from the fact that two technologies, chemical stitching and continuous preforming, were combined in this project. The process parameters that were varied in the course of the process optimisation are the following:
- Line speed
- Continuous / semi continuous process
- Needle geometry
- Resin viscosity
- Stack consolidation /spread
- Injection pressure
- Injection time
- Other injection parameters
In an integrated process such as the combination of automated preforming and chemical stitching in one device that was the centre of the project AChSo, the individual parameters are rarely independent of each other. Changing one parameter often results in better performance in one part of the production line but worse performance in another. The interdependencies can often be estimated qualitatively but the exact influences of one on another are not known for new processes such as this. A lot of experience is necessary to achieve control over the process. This experience cannot be derived through other means as trial and error experiments.
Line speed is obviously important for the economic production of preforms. Continuous production processes are deployed, when a high number of parts are needed. The line speed was chosen as the overall goal parameter in this process. As in many processes, production speed is inversely coupled with production quality. At low speeds an exact control of all parts of the integrated process line is possible, but production is not economic. Increasing the speeds lets one usually produce quality parts until a certain cut off point is reached and the produced parts do not comply with the requirements anymore. Following the philosophy of “as exact as necessary and not as exact as possible” this cut off point must be found. Depending on the stability of the process and the tolerances of the requirements the set of parameters that bring one closest to this red line with a stable process must be found.
The programming of the drives for pulling the fibre preform and for moving the injector heads let the user choose between two process states. Firstly a continuous process that required parallel synchronised motion of the fibre preform substrate and the resin injector heads, this is the one that should be used for an economic process. Secondly a semi-continuous process in which the movement of the fibre preforms was halted during the chemical stitching. No x-Movement of the injector is necessary in this case. This mode should only be used in the set-up stage as it is considerably slower than the continuous one.
However, it lets the user decouple the preforming and the chemical stitching processes. On the one hand this makes it easier to find the right parameters for each singular process. On the other hand, once going back to the continuous process the interdependencies between the two processes must be found, as described above. Because the injectors that are responsible for the chemical stitching are a very complex system in itself, the semi-continuous process was used to understand their behaviour and produce stitch points that were considered acceptable. The next step was to produce these stitch points in the continuous process.
In the stitching process alone several parameters can be adjusted. The resin itself can be varied in its viscosity. The reason to increase the viscosity is the behaviour of it in the textile preform. With the usual viscosity, the resin can be injected very easily between the preform layers, but is dispersed there in a wide area due to the capillary forces between the filaments. It is, in effect, sucked up by the fabric substrate like water by a cloth and cannot form a localised stitch point. The viscosity of the resin can be adjusted by adding calcium carbonate to the resin. The median size of the CaCO particles is 2.7 µm so it can be dispersed in the liquid resin easily. Increasing the resin viscosity, will make it harder and harder to transport it through the injectors and the needles into the preform. This will increase the time necessary to stitch one point, slowing down the process. It might also result in injection pressures that are too high for any of the parts involved, the weakest usually being the static mix tube. The concentration of CaCO was increased until the viscosity was high enough to sufficiently stay at the stitch point. The chosen mixture was resin weight to CaCO 1:1.4.
To lower the injection pressure, a larger diameter needle might be used. The limit for the needle geometry is the ability to penetrate the preform. If the needle is too thick it might either not perforate the upper layers of the textile at all or damage the punctured fibres in a way that is detrimental to preform quality. The needle can penetrate the upper layers more easily it has a chamfered tip instead of a flat tip. This is turn smears the resin over a larger area in the z-direction (thickness), which can be positive or negative depending on the degree of consolidation that is chosen for the preform at this point. If the preform is strongly compacted the chamfering might result in spilling most of the resin on top of the preform because part of the needle opening sticks out of the textile stack. If, on the other hand, the layers are spread out the chamfering can aid in spreading the resin regularly between all the layers. A good compromise was found with needles of 0.84 mm diameter and a 20° chamfered tip.
The consolidation of the stack can be adjusted by adding rollers or spreaders in the area directly in front of the stitching area. In the forming part of the preforming line, a set of rollers is used to bring the flat material into the desired L-shape. This also leads to a pre-compaction of the textile stack. The consolidation of the textile stack during the stitching influences the final preform in several ways. With a high consolidation, the resistance for the resin to flow through the layers and wet each layer is high. The chance of failure to stitch all layers rises. On the other hand does a high degree of consolidation lead to a good, localised application of the resin and a small stitch point. If the resin in injected into a less consolidated stack later consolidation tend to smear the resin across a wider area. The lower consolidation raises the permeability of the stack and makes it thus easier for the resin to flow through the preform. This lowers injection pressures and injection time and ultimately leads to higher line speeds. To spread the layers after the pre-consolidation in the forming section, a comb can be placed directly in front of the stitching area. This separates the layers and creates a space between each of them into which the resin can flow without much resistance. In the end the comb was chosen over the extra roller to achieve greater separation and a better overall process.
Production of preforms
In the report of the preceding deliverable D6-1, the process optimisation was described in detail. A number of interdependent parameters were adjusted to improve the preforming quality. A final improvement was done subsequently to further improve process stability and reproducibility of the preforms. Before, the resin was injected into the middles of the stack and would then flow through the textile to reach all layers. While this produced preforms with all layers bonded by the resin, the amount and placement of the resin depended on the flow processes of thee resin.
To improve upon this, the z-axis that moves the injector up and down was redesigned to be adjustable in speed. An electric drive and a controllable pneumatic drive for the z-axis were considered. Finally, a mechanical solution was chosen that lets the user control the vertical speed by changing the inclination of an iglidur® glide ramp. This simple solution achieved the most consistent results and reduced the complexity of the overall device and control, compared to the pneumatic or electric variants.
With the controllable z-axis speed, the deposition of the resin between the layers could be controlled more exact. The amount of resin between each layer can be adjusted by changing the angle of the ramp, with a lower ramp angle letting the injector stay longer between two layers and depositing more resin. The position of the resin is better reproducible, because it is no longer dependent on the complex flow process of the resin through the textile.
The position of the stitch points was chosen to stabilize the preform geometry and the laminate stack. A row of stitch points on each edge keeps the textile layers together at the area where they are most vulnerable to being damaged. The second needle of each injector is positioned close to the middle bend of the L shaped profile and stabilizes the geometry of the preform. By injecting resin in this area of the preform, the preform is kept from relaxing into the flat shape of the textile raw materials.
The continuous preforming process ran stable and reproducible. Profile preforms were produced in a continuous manner and cut into the desired length for the resin infusion process without stopping the production line. Once set up, the process can run endlessly. If the textile is running out, a new strip cab spliced onto the end of the previous one without stopping the process, although the part of the preform with the seam must probably be discarded. The resin can be refilled during the running process by using two resupply containers for each component that can be switched with a valve. Due to the automated and continuous manner of the preform production line, even a 24 hour usage of the device is viable with minimal downtime. Under these conditions, a production rate of 1000 m of stacked, shaped, stabilised, cut and infusion ready preforms is not unrealistic.
Potential Impact:
Final result and Impact
The final result of the project is the development of an innovative preforming process and the demonstration of this technology in a prototype preform production line. A functioning, stable and high quality process was achieved. The process is suited to produce profile preforms with a large variation of geometries and material (both laminate and roving material). Such profile preforms are used in the aerospace industry for reinforcement structures such as stringers and spars, but potentially also be used in other fields that use profile structures. Automotive and marine applications are possible but also civic uses for buildings or bridges. The process can be enhanced beyond the state established in this project to produce curved preforms and preforms that change their profile geometry lengthwise. Due to the automated and continuous manner of the preform production line, even a 24 hour usage of the device is viable with minimal downtime. Under these conditions, a production rate of 1000 m of stacked, shaped, stabilised, cut and infusion ready preforms is not unrealistic.
The goals of ecological improvements set by EU partner countries can arguably only be achieved when they do not impair, or at best, even improve upon, European competiveness. Composite materials are well suited to lower the carbon dioxide emissions through lighter structures and thus lower fuel consumption. Industrial composite processes that produce the high quality components necessary for aerospace applications are not able to produce the large amount of party needed for a large scale use of composites.
The continuous preforming technology demonstrated in this project can deliver such a process. It combines the high performance material and high quality preforms necessary for aerospace applications with low operational expenditures. It achieves this by moving from a manual process to a highly automated one. This increases European competitiveness two-fold.
Firstly, the developed technology is superior to the ones available from non-European companies. European automation companies that are able to develop such a process will have an advantage over their competitors. Secondly, through the networking that is perpetuated by EU programs such as CleanSky, European airframe structure producers will sooner be informed about such innovative products. This and the close connections formed through the EU projects also reduce the time to market for emerging technologies. Thirdly, the higher degree of automation and lower operational expenditures make it more likely for industrial process to remain within the EU. More manual processes are more likely to be outsourced to countries with lower personnel cost. More complex, automated processes need more highly trained personnel, who are available more likely in EU than in competing countries.
List of Websites:
Dissemination
Dissemination activities of the project results took several shapes. For scientific dissemination, a paper was published and presented. To further the industrial application of the developed processes, the results will prepared to be shown at trade fairs. On the project side, the technologies that were part of this project will be the focus of future projects that will build upon the results of AChSo.
A paper called “Implementation of Automated Chemical Stitching into the Continuous Preforming Process” was presented at the Greener Aviation Conference that took place from 12th - 14th March 2014 in Brussels, Belgium. The paper has been published in the conference proceedings. The Greener Aviation conference is specifically set up to show the work done in the CleanSky program. This is a good opportunity to meet other members of the European industrial research community. The few opportunities for networking between the CfP-Partners and the core partners of CleanSky can be equalized somewhat by events like this. The presentation and discussion of the results in Brussels were a good way to show the developed process and gather input from fellow researchers for future development of this technology.
To show the project results to the industry, so that the developments might be implemented in industrial serial production, presentations at trade shows is an effective way of exposing the project result to an interested public. The most important international trade show for composite products and processes is the JEC in Paris, France. The continuous preforming technology was presented at the 2014 JEC and the integration of the Chemical Stitching into it will be presented at the JEC 2015. A movie is being produced at the moment to show the working process and the results. This movie will also be shown at other opportunities and be released through FIBRE’s internet presence (www.faserinstitut.de ).
The project results have been published (http://www.textilforschung.de/en/project?id=1) in the information portal of the German Textile Research Board (FKT Forschungskuratorium Textil). The FKT explicitly connects textile researchers in academia with textile users in the industry. It is active by organizing dissemination events in which textile research institutes can present their innovations to leaders and decision-makers and funds cooperational research by institutes and industry players. The research board can thus function as a great multiplier to spread the results of this project.
The technologies used in this project will be further pursued in future projects. For example, the process demonstrator will be used to develop a process that builds on top of the chemical stitching technology but for thermoplastic matrix systems. The current chemical stitching uses thermoset stitch points for use in thermoset LCM processes. For cases where thermoplastic matrix materials are favoured, a different set-up is needed, but it will retain the basic concept with localized fix points, instead of the fixation agent across the entire area of the textile.
Participating in the pan-European CleanSky program was a very good way to meet new partners on a transnational level. The combination of two innovative technologies in one project was a very interesting approach. All partners were able to gather new know-how from each other and also utilize each other’s network. Because more interested parties will be exposed to the technologies developed within a joint program such as CleanSky, it is more likely that research will find its way into the serial production of aerospace components. FIBRE will try to enter in the CleanSky2 program as a partner.
