Low Cost Analysis of Industrially Representative Composite Defects for Propeller Design

  • Trevarthen, James (GE Aerospace (Dowty Propellers))
  • Baid, Harsh (AlphaSTAR)

Please login to view abstract download link

Composite propellers are in demand for the next generation of sustainable aircraft due to their high efficiency, low weight and long life. To deliver sustainable propulsion, composite propeller blades need to be light, efficient, and manufactured with minimal material and waste. This paper evaluates a method for high-speed, low-cost screening of composite materials during early design to deliver these aims. During initial design, composite blades are sized for compression, which is sensitive to defects - particularly through thickness wrinkles. Due to the complex manufacturing process and geometry of blades, defects can never be fully eliminated, despite extreme levels of process control. Hence, an acceptable defect level is defined during design, and blades are sized to accommodate a certain population of acceptable defects. Blades are inspected during production to identify and scrap parts with defects outside of these limits. The effect of defects has historically been challenging to predict by analysis, requiring extensive testing at a range of scales to determine defect limits for sizing and acceptance criteria. Improvements have been made with high fidelity analysis methods, but these are resource intensive. This poses a challenge during early design, when a large range of composite materials with different combinations of filaments, resins, fabric architectures and binders or interleaves would ideally be considered. Without existing test data or low-cost analysis methods to predict the impact of wrinkles, the result is either that designers are disincentivised to consider new materials or apply excessive conservatism. This has a negative impact on cost and sustainability. In this paper we evaluate a lower cost approach to predicting the performance of wrinkled composites, that is suitable to the initial screening phase of composite material design for new propeller products. Our approach uses a micro mechanics driven solution available in AlphaSTAR’s Material Characterization and Qualification (MCQ) commercial software to predict the effect of different wrinkles on a range of composite laminates’ compressive performance, using only standard lamina level coupon data. A comparison is made of both the results and the relative computational expense of the presented method to those of both empirical testing and higher fidelity FE models. Finally, the merits of these methods for propeller blade design is discussed.