Effect of rewinding twist on properties of braided composites
Publication Date
December 10, 2024
Creator
Abstract
State-of-the-art braiding simulation and processing capabilities use tow geometry approximations built on generalised assumptions. This results in discrepancies between the key process variables such as cover factor and braid thickness of the designed braid and manufactured preform. The challenge in predicting the properties of braided composites is one of the primary reasons for the limited application of braiding in composite manufacturing processes. Localised variations in braid architecture are detrimental to the final quality of the manufactured preform. This coupled with variance in key process metrics such as braid angle and cover factor between simulated and manufactured preforms leads to inaccuracies in the predicted mechanical properties. One of the ways tow widths can be influenced is by adding twist to fibre during the rewinding stage of the process.
It is accepted that the tow rewinding stage of the braiding process is critical to the resultant product quality. An additional measured level of twist can be added to fibres during the rewinding process, this is a technique used by some industry leaders to improve fibre processability. As the number of twists increases, a more compact tow structure is also produced resulting in a reduction in the degree of tow spread. This has a significant impact on the cover factor and braid thickness of a preform.
The main objective of this study to develop the understanding of the inclusion of a rewinding twist to braid architecture and the resulting impact on mechanical properties.
To address this objective a series of studies are presented, aiming to initially understand the influence of rewinding twist on the mechanical behaviour of braided composites. A higher variability in failure strength was observed with untwisted coupons in comparison to twisted coupons, which resulted in a high A-basis allowable for twisted coupons. Different stress-strain behaviour was observed. Untwisted coupons failed instantaneously with brittle failure, where twisted coupons saw an initial failure point but then continued to take on load before final failure occurred.
Braid architecture variability as a result of the inclusion of rewinding twist has been investigated through the braiding of cylindrical mandrels at different levels of rewinding twist between 0-10 twists per metre. The inclusion of 5 twists per metre was seen to reduce the tow width by 28% with the reduction in tow width resulting in an increase in tow thickness. This has been used to develop approximations for tow width and thickness as a result of rewinding twist, along with developing the understanding of the influence of rewinding twist on key process variables. Additionally, images of the braided preforms along with manufacturing observations indicate an increase in fibre damage as a result of braiding with untwisted fibres due to an increase in friction between yarns. This results in process interruptions and machine down time.
In addition to this, mandrel complexity has been investigated with the inclusion of converging and diverging frusto-conical sections to the mandrel. Observations demonstrate a significant impact in braid architecture variability and stability as a result of the introduction of a cross-sectional change. Braiding on diverging conical sections demonstrate the largest deviation in braid angle from predicted braid angle due to the effective reduction in stabilisation length. Whereas minimal angle deviation was noted when braiding on converging conical cross-sections. This angle deviation appeared to be independent of the level of rewinding twist on fibres. However, the addition of rewinding twist to fibres did reduce tow spread when braiding on a geometry of changing cross-section in either converging or diverging orientation.
These studies have enabled the investigation of the scale of the braiding machine using a 48-carrier and 192-carrier braider. Investigations into braid architecture including tow width and braid angle for both braiders on both circular and conical mandrels have showed minimal differences between the braiders. This is significant as it enables results from small scale trials and pre-production models to be used to approximate full-sized components. This reduces the requirements for expensive manufacturing trials and large-scale set-ups.
Item Type
ethesis
Thesis Type
EngD
Supervisors
Subjects (LC)
Associated Schools / Departments
Department of Mechanical, Materials and Manufacturing Engineering (UK)
eprints ID
78829
UoN Repository URI
Except where otherwise noted, this item's license is described as
File(s)![Thumbnail Image]()
Name
Grimes_Bethany_4342072_corrections.pdf
Type
Full-text
Description
Examined
Size
7.97 MB
Format
Adobe PDF
Checksum (MD5)
feb6803ee21f723591a36f2381f0d66c