Skip to main content

Developement of CNT doped reinforced aircraft composite parts and a
ssociated tooling, using the Liquid Resin Infusion method

Final Report Summary - BULCANATU (Developement of CNT doped reinforced aircraft composite parts and a ssociated tooling, using the Liquid Resin Infusion method.)

Executive Summary:
This report performs a global overview on the research project named Bulcanatu, under Grant Agreement CS-GA-2013-325967.

This project has been performed between Technical & Racing Composites S.L. (project leader, TR from now on) and Universidad Rey Juan Carlos (project partner, URJC from now on).

The project objective was to investigate and develop a method to manufacture a composite bulkhead (D=2250mm) with the Liquid Resin Infusion system using a Carbon Nanotube doped Resin System. For this purpose a number of research activities were developed. Those can be summarized as follows:
• Research on Doped Resins (leaded by URJC).
• Research on Process Development (leaded by TR).

Current State of Art of aerospace composite structures show two main electrical conductivity requirements as a result of the isolation behaviour of composite materials (due to the resin):
• Lightning Strike Protection: the industry uses external layers (whose thickness depends on a zonation of the aircraft) to allow a proper dissemination across the structure of the energy of lightning strikes.
• Electrical Bonding: electrical and electronic equipment, as well as metallic systems such as high pressure hydraulic lines (normally manufactured with titanium) or other systems like actuators, etc. need an electrical bonding network (a global network on the inner side of all composite aerostructures) made of aluminium strips riveted or bonded to the structure and metallic cables that perform the static current discharge from the systems to the strips.

The objective of doping a resin with carbon nanotubes is to increase the electrical conductivity of the composite material, in order to explore applications that could comply with the mentioned electrical requirements.

However, one of the technical issues of a doped resin is achieving a proper distribution of the nanotubes and keeping a low viscosity to allow infusion of a dry preform. This, of course, limits the concentration of nanotubes and thus the resulting electrical conductivity.

Bulcanatu project investigated on a resin system doped with nanotubes that was suitable for production at large scale and able to be infused, and also to apply such resin on the development of a full-scale demonstrator of an aircraft structure. Rear Pressure Bulckhead of an A320 size aircraft with integrated omega stringers was the part selected.

Project Context and Objectives:
The main goal of the project is to manufacture a composite bulkhead based on Carbon Nanotube (CNT) doped Resin System. For this purpose, all investigations will be performed using industrial scalable techniques and aeronautical qualified materials. In addition, parts had to be manufactured and tested according to aeronautical quality standards and procedures.

CNT is an allotropic form of carbon, recently discovered by Ijimia in 1991. Its geometry is generated by rolling a graphene sheet, in which carbon atoms are placed in vertices of a honeycomb network bonded by covalent C-C bonds. According to the chiral angle, CNT presents three different chiralities: zig-zag, chiral and armchair. Mechanical properties of these materials depend on their number of walls (single-walled, double-walled or multi-walled CNT) and their lengths. In general, CNT materials own mechanical properties close to solids’ ones, also presenting the greatest Young’s modulus and tensile strength among all known materials. They have high stiffness and mechanical strength as well as high flexibility due to the high length and high aspect ratio (1000 or more). Other interesting properties of CNT are their excellent thermal and chemical resistance, high thermal conductivity and relative high thermal stability; their electrical conductivity strongly depends on the chirality, helicity and dimensions, being from semi-conducting to metallic.

Despite being excellent candidates as filler for composite materials, there is still not an ideal manufacturing method due to the nanoscale size of the filler, which presents several limitations:
1. High viscosity of CNT/polymer mixtures.
2. Heterogeneous distribution of CNT in the polymer.
3. Agglomeration of nanotubes.
4. Difficult alignment and orientation of CNT.


As mentioned in the introduction, a characteristic of CNT technology is to increase the conductivity of composite materials, potentially removing the need of a dedicated metallic electrical bonding network as first application.

As will be seen later on, conductivity levels that can be achieved of CNT doped composites would not fulfill lightning strike protection requirements.

Project Results:
1 Research on Doped Resin (URJC)

1.1 WP1: Manufacturing methodology of CNT doped resin dispersion
This first work package investigated about the following topics:
• Selection of materials through analysis of datasheet of potential materials.
• Dispersion characterization for definition of the CNT concentration, suitable for large scale production.
• Thermal, Mechanical and Electrical Behaviour: test campaign for CNT doped resin characterization.
• Dispersion Methodology and Procedure: development of the large-scale CNT doped resin production.

1.2 WP2: Manufacturing Methodology to produce an aircraft quality composite part
The second work package investigated about the following topics:
• Test standards and panels: set-up of the manufacturing method for insufing CNT doped resin into dry fiber carbon preforms.
• Microscopic Inspection: to determine the main morphological features of the manufactured multiscale composites.
• Mechanical Behaviour Report: test campaign for mechanical characterization of composite (carbon fiber and CNT doped resin). Main conclusions: 7-28% mechanical performance increase (depending on the test)

2. Research on Process Development (TR)
2.1 WP3: Definition of tooling configuration
2.1.1 Preliminary Design Review of Necessary Tooling
TR Composites developed a tooling concept fulfilling all project requirements, overcoming the Preliminary Design Review. Among different strategies analysed, Composite Master and Composite Mould was selected (manufacturing of a master model and lamination of the tool on it with special carbon fiber prepregs for tooling).

The alternative chosen included a female mould and co-infusion of skin and stringers needing the following tooling:
• Female tool for bulkhead.
• Tool for preforming of stringers.
• Devices to place stringers on the female tool allowing infusion and curing.

2.1.2 Report on infusion simulation
The main goal of the simulation was to define inlet areas and the overall infusion strategy to ensure an infusion time of less than 90 minutes (coinciding with the gel time of the resin), as well as to provide the type of tooling for co-infused carbon fibre stringers and the total volume of resin required in the process.

The software used for the simulations was RTM-Worx. This technology works with a Finite Element Method (FEM) combined with a Control Volume Method (CVM) to track the flow front.

The final outcome of the simulation was a fully impregnated preform in about 50 minutes, well below the maximum 90 minutes given by the resin properties.

2.1.3 Critical design review of necessary tooling and infusion system design
The final tooling design evolved considering information provided at PDR level and in parallel to the evolution of infusion simulations.

2.2 WP4: Manufacturing of Necessary Tooling
All tooling was manufactured according to TR Composites standards.

2.3 WP5: Quality Control on Necessary Tooling
This Work Package summarises the measurement results of the set of masters and moulds. Metrology tasks aim to provide geometrical deviations in terms of volume and drills’ location. Equipment used is: Leica Laser Tracker AT901 with T-Cam and T-Probe with the following specifications
Interferometer Distance Resolution 0.32 microns (0.000013”)
Interferometer Distance Accuracy ±0.5microns/m (±0.000006 ”/ft)
Dynamic Lock-On Accuracy ±10microns (±0.00039”)

Results obtained from the measurements of the Bulkhead Master are summarized as follows:
• The volume of Bulcanatu master is within 0.25mm. RMS calculated value is 0.064mm.
• All positioning holes of the Bulkhead master are within ±0.2mm tolerance, being 50% of the values below ±0.1mm.

Results from the measurements of the Stringers Master are:
• Global measurements indicate a tolerance within 0.3mm (+/-0.150mm). RMS calculated value is 0.061mm. Remember the master integrates 4 stringer moulds.
• Concerning the drills carried out on the master, regarding the X-Y vector deviation, all of them are located within 0.15mm around the theoretical position. Note these measurements considerably improve when performing them stringer by stringer.

As for the Bulkhead Tool, measurements provide the following information:
• BULCANATU mould volume is within ±1mm. RMS calculated value is 0.387mm.
• An overall symmetrical deformation is observed. TR’s assessment considers that the position of the substructure and the post-curing phase are the origin of this deviation. However, the value obtained is considered acceptable taking into account the purpose of this project.
• All positioning holes on the mould are within ±0.5mm.

2.4 WP6: Production of Bulkhead Demonstrators
2.4.1 Part Manufacturing.
After a first failed attempt, the second attempt was considered a success. Different 40 mm-side squares were trimmed on the bulkhead for further studies on microscopy and electric conductivity.

Finally, the third execution was considered a failure. Even though the material preparation and the overall process was improved respect to previous attempts, the bag blew out during the curing cycle affecting the final part obtained. The stringers were detached from the bulkhead and many areas were improperly cured. Furthermore, there was a lack of materials in this last attempt, such as T connectors, which caused several problems when assembling the infusion system.

2.4.2 Final strategy on the injection procedures
Considering all problems occurred in the previous executions, the final manufacturing process of BULCANATU demonstrators was redefined.

2.5 WP7: Non-Destructive Testing
2.5.1 Ultrasonic C-Scan Testing Report
Instead of performing task 7.1 and its associated report, the Topic Manager agreed with TRC to compensate this by sending all material available to produce another part at Topic Manager's facilities.

2.5.2 Microscopic Inspection Report
The quality of final part was determined, analysing the extracted coupons. The main goal of the inspection was the analysis of filtering effect, studying samples located at different distances from the LRI resin inlets. In addition, non-cured resin was collected from the resin inlet and outlet to confirm the presence of nanotubes in the whole multiscale material.

On the other hand, the density of the multiscale composite was also measured to analyse the quality of the material of the manufactured bulkhead. It allowed to study the presence of porosity of the final multiscale composite.

All the obtained results on this inspect report, together the previous analysis carried out on multiscale probes, allowed issuing the following conclusions:
• There is not a clear filtering effect on the multiscale composite of manufactured bulkhead. The presence of carbon nanotubes was confirmed on all areas of the whole bulkhead, determining the CNT presence in the inlet and outlet resin. Unfortunately, this analysis is only qualitative. It is not possible to determine variation on the CNT concentration.
• The z-electrical conductivity of the multiscale material of the bulkhead is close to 2 S/m. This value is higher than the electrical conductivity of neat composite with non-doped resin, but it is lower than the expected one. The areas with low electrical conductivity are associated to areas with low density.
• The general high density of the multiscale composite means a high quality of this material, with low level of porosity. In addition, the constancy on the density measurements in different areas indicates a high homogeneity of the material.

2.6 WP8: Quality Control of Demonstrators
The quality control steps according to internal procedures submitted to EN9100/ISO9001 quality system include:
• Inspection of carbon tools prior to LRI manufacture process.
• Airtightness check of tools prior to lamination.
• Fibre Placement process traceability. Cure and Post-cure Process traceability
• Inspection of manufactured LRI parts for imperfection and non-conform zones
• Metrology Inspection of parts for approval. It is going to be used the Leica Laser Tracker for this purpose, available at TRC facilities.

In summary, the infusion process has demonstrated the ability to produce a good part from an impregnation and curing points of view. However, without external access on the inner side, even if tolerances could not be compared to autoclave processes, the tooling concept provides a level of tolerances that depending on the application would not be acceptable.
3. Bulcanatu Lessons Learnt
After conclusion of the project, different lessons learnt can be extracted from the activities performed. This chapter depicts such know-how topic by topic:
3.1 CNT Doped Resin Development and Escalation
Regarding this topic, the following lessons were extracted:
• Properties obtained after characterization showed increase of performance compared to suppliers’ datasheets.
• The optimum CNT content which maximizes the mechanical properties of epoxy resin is in the range of 0.1 – 0.3 wt %.
• CNT does not modify thermal behaviour of the resin.
• Mechanical properties are improved by a fair amount thanks to adding CNT (15-20% in strength and up to 80% in elongation of resin only).
• The electrical conductivity increases with the CNT content.

3.2 Tooling Design and Process Simulation
Regarding this topic, the following lessons were extracted:
• Polyworx needs accurate values for drapability and permeability. Testing is needed.
• Polyworx spreadsheet for obtaining infusion parameters showed good accuracy.
• TR Composites applied internal procedures for composites tooling design with very good results: design of tooling, escalation factors, positioning pins, handling devices, etc.
• The caul plate concept showed a main issue: the vacuum bag must be constructed very carefully on the inner region of the bulkhead because the stringer tooling is floating on that area, being prone for suffering undesired lifting if a runner gets trapped under the caul plate.
• Regarding density of plates of the eggbox structure, the measurements performed on the main tooling recommend for future projects to have tighter pitch between plates. 250-300mm pitch for the plates would be a figure to bear in mind in the future.

3.3 Manufacturing of Tooling
Regarding this topic, the following lessons were extracted:
• This project was also useful for evaluating master material and tooling prepregs never used in the past by TR Composites.
Master material Ebablock M007 casted polyurethane from Ebalta: this medium density master material (0.82g/cm3) showed very good surface finish after fine machining and polishing. In addition, very few resin remained on the master surface after mould demoulding. Other materials used in the past showed a dry first ply on the mould. This did not happen in Bulcanatu.
Tooling Prepreg TR23 from Delta-Tech showed outstanding performance at a very affordable cost. This material is now baseline prepreg for tooling for TR Composites projects.
• Regarding the construction of the eggbox, handmade prepreg L-clips showed a very good result.
• High temperature silicone will keep on being used in the future for bonding eggboxes to carbon tooling.

3.4 Manufacturing of Aircraft Parts with LRI Process
Regarding this topic, the following lessons were extracted:
• Polyworx, showed very good correlation regarding the infusion process for a complex aircraft structure with doped resin: time difference within 10% between simulation and real part.
• After the failure of the first execution the investigation concluded that the supplier provided material not in accordance to the specifications. Thus, runners collapsed and the part was not infused. A lesson learnt to bear in mind is to perform a small trial with infusion material to verify their suitability for the full-scale part.
• In order to guarantee full impregnation, it is recommended to have at least 15-20% more resin of that calculated for the part + infusion runners.
• Semi-permeable membrane showed outstanding results for preventing air trapped in the part.
• Special care to be taken when positioning stringers to avoid trapping infusion tubes underneath.
• All ports to be fixed with flashbreaker. To avoid use of tacky tape for high temperature infusion because it becomes softer and is prone to block inlet ports (preventing the infusion).
• To fix the runner to the bag with tacky tape and not directly to the port to avoid the same as described above.
• To use runners with the same diameter than the inlet ports.
• Y-shape connectors make the infusion faster compared to T-shape connectors.
• Inner stringers’ bags shall have an extra length behind the tacky tape to avoid punctures.

4. Final Conclusions and Proposed Way Forward
4.1 Bulcanatu Conclusions
The Project allowed developing a resin doped with carbon nanotubes, enhancing resin properties, reaching electrical conductivity values suitable for electrical bonding applications, and with a resin production process able to supply resin quantities needed for the production of full-scale demonstrators.

The manufacturing tooling configuration has been developed taking into account the final shape of the component to infuse. Different combinations and strategies have been presented throughout the project, leading to an optimized one that gathers simplicity, reliability and final part quality. This way, the bulkhead tooling has been obtained from a casted master model plus a prepreg mould laminated on it. As for stringers, it has been decided to machine epoxy boards on which laminate the prepreg moulds. Furthermore, note the design has considered handling operations (including channels for the forklift) as well as assembly.

The RTM-Worx software has managed to design a tooling configuration that allows to infuse the bulkhead with the stringers within 90 minutes (before the resin starts to catalyse). Moreover, it has provided a timeframe to correct the infusion process in case something unexpected arises.

With all engineering work done, two parts have been manufactured with only one of them properly obtained. However, injection process of the bulkhead needs to be standardized to avoid making the same mistakes in further attempts. This way, the entire manufacturing process of BULCANATU has been defined step by step, from bag preparation to finishing operations.

Once both bulkheads have been manufactured, microscopic inspection has been carried out. Results show the presence of carbon nanotubes on all areas of the bulkhead at qualitative level. Moreover, the z-electrical conductivity of the multiscale material is close to 2 S/m. This value is higher than the electrical conductivity of neat composite with non-doped resin, but lower than the expected one (in the range of 2 to 5 S/m for electrical bonding applications). Finally, the high and homogeneous density of the multiscale composite together with a low level of porosity, confirms the good quality of the bulkhead.

On the other hand, part measurements performed at TR facilities shows high deviations (up to 6-7 mm) when comparing to CAD design. Although depending on the application this would not be acceptable, no tolerance levels were given as a requirement on this project. Thus, the observed results should be understood as a starting point for further attempts. Also to consider that the origin of such high deviations have been properly investigated and documented and were origin on manufacturing errors to be avoided in the future.

4.2 Proposed Way Forward
As seen above, the project demonstrated that doped resin can be produced at the volumes requested for medium aircraft structures, and also the final conductivity of the part produced is close to the requirements of electrical bonding.

Therefore, future projects on this topic need to focus on:
• Continuing the works on producing doped resin at larger volumes to support future serial production.
• To continue on the investigations about the infusion process to reach 5 S/m of conductivity in order to fulfil the most restrictive requirements for electrical bonding applications. Increase of concentration of nanotubes is needed, and in parallel an infusion process suitable for infusing a denser resin has to be developed.
• To investigate in mould concepts that prevents trapping of infusion elements under stringers and caul plates.
• To investigate in reusable bags for infusion in order to shorten production time and reduce risk of manufacturing mistakes.

Potential Impact:
The topic’s target is to evaluate important parameters, such as the Carbon Nanotubes (CNTs) content, mixing procedure, duration, type of epoxy and CNTs functionalization, for the Liquid Resin Infusion process (LRI) processes and develop techniques to allow manufacturing. After that, a demonstrator will be manufactured using the researched manufacturing techniques.
The incorporation of CNTs into a polymer matrix along with long fiber reinforcements for producing hybrid composites, have recently attracted significant attention. Due to its comparatively low manufacturing costs, the LRI is ideal for impregnating reinforcement fabrics with CNTs resin mixture, obtaining high performance composite parts that can lead to increase of fuel efficiency and lower costs of future CNT composite manufactured aircraft components.

In this context, Bulcanatu main resuts are:
1) a CNT resin was developed in a process suitable for scalation at aircraft primary structure quantity needs.
2) a full-scale demonstrator (under limited technical conditions) could be successfully infused with such resin.
3) under the limited conditions, this first attempt resulted in part conductivity levels in the range of conductivity requirements for electrical bonding.

In other words, Bulcanatu project is a successful first step towards eliminating the need of electrical bonding networks (metallic plates, ribbons, etc.) on aircraft structures. Achieving this final goal would imply:
a) Major cost savings at assembly (thousands of systems installation ours per aircraft and related grounding networks' materials)
b) Removing electrical bonding networks would imply major weight reduction opportunity for aircraft (the project partners do not have relevant information on this context). Moreover, this technology could become a major driver for further weight reduction in aircraft composite structures, thus strongly contibuting to suitanable aviation.
c) Less relevant, from a design standpoint, the activities for designing electrical bonding networks would not be needed.
List of Websites:
TR Composites: Daniel Claret (daniel.claret@enginyers.net)
URJC: Sílva González (silvia.gonzalez@urjc.es)