Improving damage tolerance of aerospace composites using 3D printing technologies

Carbon fibre reinforced epoxy composites generally have high specific mechanical properties such as strength, stiffness and fatigue resistance. Additionally, their low thermal expansion coefficient and excellent resistance to corrosion makes them a suitable structural material for aerospace applications. However, carbon-epoxy composites are susceptible to barely visible impact damage (BVID) and have characteristically poor impact damage tolerance, which limits their efficiency as a structural material. Techniques used to improve the damage tolerance of aerospace composite materials include adding tough thermoplastic particles to the epoxy matrix phase. Thermoplastic veils and scrims are common selective toughening techniques for composites and they typically introduce 1-8 wt% of thermoplastic filaments into the interlaminar regions. However, veils generally provide modest improvements in impact damage tolerance to composites. 3D polymer printing technologies provide a unique opportunity to create tailored thermoplastic veils to increase the interlaminar toughness and impact damage resistance of composites. The wide variety of 3D printing techniques provides multiple avenues for creating interlayer architectures. The high-level aims of this PhD project are three-fold. The first aim is to characterise the thermoplastic materials used as feedstocks for the 3D printer and understand how environmental conditions (humidity) influence the processability and morphology of the printed thermoplastics. The second aim is to develop a mathematical model to describe a novel 3D printing process that has been tailored for the purpose of creating 2D flat fibrous veils. The third aim is to understand how cure profiles, thermal preforming conditions and variations in the interlaminar veil architecture influence the impact damage tolerance of carbon-epoxy composites. A comprehensive literature review is presented in this PhD thesis outlining the current capability of commercially available polymer 3D printing techniques and the current state-of-the-art in composite interlayer toughening. A critical review of 3D printing technologies concludes with the selection of a suitable family of technologies (based on Fused Deposition Modelling), but not specifically a suitable 3D printing technique for the purpose of creating fibrous interlayer veils. Hence, a novel 3D printing technique called ¿Drop, Draw and Extrude (DDE) 3D printing (3DP)¿ is presented as an alternate method of creating thermoplastic veils. A mathematical model that incorporates both fundamental physics relationships and experimentally derived relationships provides the foundation for understanding the DDE 3DP process. The model predicts the drawn/extruded filament diameter produced using the DDE 3DP process across the operating range of the 3D printer. The model gives a reasonable prediction of the experimentally determined average filament diameters with an average discrepancy of less than 5% for the two different thermoplastic feedstock materials that were studied (PA645 and PA12). Three key printing process inputs were evaluated including head translation speed, extruder speed and nozzle temperature. However, a range of secondary variables were also found to influence the filament diameter, including nozzle offset distance from the carbon fabric substrate, the anchor spacing and the fibre draw length. Additionally, experimental evidence of the variation of filament diameter between the beginning and end of a print move is presented. Rheological and thermal properties of the thermoplastic feedstock materials were experimentally determined to assist modelling and provide a basis for selecting the operating range of key 3D printing process variables. Mathematical relationships between the printer extruder speed, processing temperature and moisture content for each thermoplastic feedstock material were determined, and then used as inputs for the DDE 3DP process model. Other studies that assisted in setting process limits included melting temperature characterisation, drying profiles and moisture absorption profiles in constant temperature and humidity environments. A study was conducted on composite materials containing 3D printed veils to determine which preforming temperature provides the greatest improvements to the impact damage tolerance. Cure profiles were selected so that two conditions could be compared: (I) the thermoplastic architecture would not melt during cure, and (II) the thermoplastic architecture would ¿just melt¿ during cure. Preform temperatures of 25°C (room temp), 170°C, 180°C and 190°C were compared on the basis of damage area and compression-after-impact (CAI) strength of the composites for each cure profile. The result for the higher cure temperature was a linear decreasing trend, whereby the lowest preforming temperature produced the greatest improvements and the highest preforming temperature the poorest improvements. The damage resistance of laminates manufactured with lower cure temperature displayed minimal changes with respect to preforming conditions. The greatest overall improvements were found for the higher cure temperature and lowest preforming temperature; this combination was then used as a constant in the 3D printed design experiment. The mechanical performance of DDE 3DP veil toughened composites was also investigated and compared to a non-toughened control and a toughened commercial grade material. The distribution of fibre diameters in veils manufactured using the DDE 3DP process and the commercial grade material all displayed normal distributions, centred around the nominal average filament diameter. The efficacy of each veil¿s toughening ability was determined by comparing damage suppression in the toughened composite to the non-toughened control material. The effects of 3D printed veil architectural variables were systematically investigated through use of a designed experiment. The three primary variables of the 3D printed veils were (I) fibre diameter (30, 41 and 54 ¿m), (II) fibre spacing (printed at 1.5 mm, 3 mm and 6 mm) and (III) fibre angle (printed at ±5°, ±25° and ±45°). This test matrix resulted in veil areal weights ranging from 0.42 gsm to 6.13 gsm. Minitab 18 statistical software was used to create regression models and to determine the relative influence of each veil variable on the impact damage tolerance of the composite materials (i.e. impact dent depth, impact damage area, CAI strength). Interaction effects were also shown and discussed, if present. A direct comparison between a DDE 3DP veil and a commercial grade veil is also presented. Each had a 3.61 gsm veil at each interlaminar region in the composite, with an average fibre diameter of 41 ¿m. This comparison indicated that irrespective of the type of architecture (random vs. ordered), the damage area and CAI strength is almost identical for both types of veils. However, structured veils with areal density as low as 1.35 gsm (30 ¿m fibre) displayed over 50% reduction in crack surface area compared to the non-toughened control, which is almost identical to the performance of composites with the commercial grade veil but with a much higher areal weight of 3.61 gsm (41 ¿m fibre). Additionally, the highest toughening efficiency per unit thickness of laminate results from veils with 30 ¿m fibre, ±25° fibre angles and 1.5 mm spacing. Therefore, there may be an opportunity to improve the current commercial grade veil by using smaller fibre diameters produced via the DDE 3DP process. Optical micrographs of the fracture surfaces indicate that impact damaged composites with veils induce crack arrest features, crack tip blunting, crack bridging and crack diversion mechanisms, thereby forcing a longer crack path and causing the crack to interact with more energy absorbent materials as it propagates. This PhD project demonstrates, for the first time, the ability to manufacture tailored veil architectures of various fibre diameters, designs and areal weights, using engineering polymers using the DDE 3DP process. The 3D printed thermoplastic veils improve the impact damage tolerance and CAI properties of carbon-epoxy composites in a controlled manner. This PhD project is an industry-focussed activity aligned with the future research and business requirements of the Boeing company. The research aims and scope were defined by the Boeing company. The intellectual property arising from this project is fully owned by Boeing and for commercial security reasons no publications (e.g journal papers, conference proceedings or book chapters) have been made.