Automated manufacturing processes for secondary structure aerospace composites

Key, Ross A. (2016) Automated manufacturing processes for secondary structure aerospace composites. PhD thesis, University of Nottingham.

[img] PDF (Thesis - as examined) - Repository staff only - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Download (47MB)

Abstract

As projected manufacturing rates for commercial aircraft increase to levels of multiple ship sets per day from individual manufacturing facilities, GE Aviation have expressed the need for a shift in composites secondary structure manufacturing philosophy. Traditional manufacturing processes tend to be touch labour intensive and hence costly. The manual placement of large numbers of individual ply profiles, lengthy debulking operations and complex cure cycles, result in excessive component lead times and manufacturing costs. As a result, direct labour cost is a major factor in the total economies of production processes.

The implementation of industrial robotics has proved highly successful in automotive manufacturing, and various methods for automating individual aspects of the composites manufacturing process have been suggested. Technical cost modelling has been used to anticipate the production costs of a prototype secondary structure component, as supplied by GE Aviation, through direct simulation of the existing manufacturing process. This work has clearly highlighted the potential for cost and cycle time reductions if process automation can be successfully introduced.

Observation of the existing manufacturing process has allowed three alternative manufacturing scenarios to be considered with respect to cost-effectiveness and feasibility, whilst highlighting long term cost benefits. Investigations have been undertaken to identify and evaluate alternative material and processing methodologies ranging from resin infused woven dry fabrics to UD prepreg tape and tow. In addition, candidate processing routes have been systematically evaluated using design of experiments techniques, which focussed on assessing the feasibility and technology readiness of robotic deposition and consolidation methodologies, including pick and place and debulking. Process automation in these areas has the potential for total component cost and cycle time reductions in the order of 2.8 to 21.6 and 0.6 to 63.4 per cent respectively.

The quasi-static mechanical testing of a range of face sheet materials has provided a performance assessment based on tensile, compressive and shear properties and laminate Vf. Findings suggest that materials offering increased suitability for automation typically have reduced mechanical performance when compared to candidate prepregs; tensile modulus and strength reductions of 5 and 34 per cent were reported when comparing a 6k woven 2X2 twill fabric and equivalent prepreg respectively. Furthermore, 26 and 4 per cent reductions in tensile modulus and 38 and 40 per cent reductions in tensile strength were observed for 179 and 318gsm UD NCF, when compared with a candidate UD prepreg.

Data has also been presented on the effect of varying the traditional consolidation frequency and methodology. While earlier findings suggest that debulking has little effect on the laminate tensile modulus; ply compaction level varies considerably. Furthermore, it has been shown that on-the-fly consolidation, using a robotically mounted, roller-based end effector has the advantages of mechanical performance retention, cycle time reduction and repeatable laminate post cure thickness. In addition, when compared with candidate woven and UD prepreg laminates manufactured using the traditional vacuum bagging approach; equivalent tensile modulus, strength and fibre volume fraction have been observed and with less variability.

Handling characteristics inherent to vacuum and needle grippers, including pickup performance, defined as the pickup or holding force required to overcome fabric weight, shear force performance; the maximum force that can be exerted on the fabric before the onset of slip, and the accuracy with which non-rigid-materials (NRMs) can be handled, have also been considered. The achievable positional accuracy of robotically pick and placed prepreg plies greatly exceeds that of dry fabrics in all cases and with less variability, irrespective of the gripping mechanism used. Vacuum grippers exhibit more uniform positional error and increased positional accuracy when handling dry fabrics, whilst needle grippers outperformed the vacuum alternative when handling prepregs, irrespective of form. Robotic pick and place solutions offer low variability in ply positional error with a guaranteed placement accuracy of ±0.8mm and ±2.3mm for prepregs and dry fabrics respectively.

Characterisation of the gap type defect and butt and overlapping joining methodologies has provided a performance trend based on ply positional error. Quasi-static mechanical testing has revealed that laminates with equivalent tensile modulus to an un-spliced control could be achieved. However, significant reductions in the tensile strength and an increase in overall laminate thickness and thickness variation highlighted the negative effect of ply splicing on laminate performance. However, it has been shown that a robotic placement accuracy of ±0.8mm gives rise to acceptable tensile strength reductions in candidate prepreg laminates.

The up-scaling of laminate level robotic manipulators has been discussed and addressed in conjunction with the commissioning of a flexible robotic manufacturing cell, facilitating the manufacture of full-scale secondary structure aerospace components. Comparisons have been made between a benchmark prepreg panel, manufactured using traditional manual methods and alternative dry fabric and prepreg panels manufactured using increased levels of process automation. In each case, manufacturing feasibility, mechanical performance and component geometric accuracy have been assessed.

It has been shown that there are significant advantages to be gained from the implementation of robotic automation within the traditional manufacturing process. Component cost and cycle time reductions, coupled with the processing and performance advantages and increased suitability to automation of woven dry fibre materials are clear. Findings which support a key driver of this project, which seeks to justify alternative dry fabrics as a viable alternative to traditional prepreg broadgoods for the manufacture of secondary structure aerospace components.

Item Type: Thesis (University of Nottingham only) (PhD)
Supervisors: Turner, T.
Harper, L.T.
Keywords: Cost-effective manufacture, Automated component manufacture, Industrial robots, Composite materials
Subjects: T Technology > TL Motor vehicles. Aeronautics. Astronautics
Faculties/Schools: UK Campuses > Faculty of Engineering
Item ID: 33572
Depositing User: Key, Ross
Date Deposited: 22 Oct 2018 13:42
Last Modified: 09 Sep 2019 11:57
URI: https://eprints.nottingham.ac.uk/id/eprint/33572

Actions (Archive Staff Only)

Edit View Edit View