Permeability Prediction Of Fabric Based On CT Image And Process Simulation Of Composite Exit Cone

In recent years,with the booming commercial aerospace industry and wide application of composite materials prepared by Resin Transfer Molding（RTM）in spacecraft parts,computer aided design of RTM using numerical simulation has become a focus of research.Permeability as an important parameter affecting the process has also attracted great attention of researchers.In this paper,the meso-structure and permeability prediction of 5-harness stain woven fabric were characterized based on Micro CT technology,and then numerical simulation is used to analyze the RTM forming process of the expansion section.Firstly,the ideal unit cell of the fabric geometry was converted into 2D image sequence at different resolutions.The effect of 2D slices with different resolutions on the characterization of fabric structure and permeability was investigated,so that the optimal resolution parameter of Micro CT scanning for the satin fabric was determined to be 15μm.Micro CT scanning experiments were conducted on 5-harness stain woven fabric using this resolution.Secondly,the meso-structure analysis and permeability prediction of satin fabric are carried out based on CT image.The slice orientation correction,slice reconstruction,image denoising,threshold segmentation and extraction of yarns are carried out by Fiji software and semi-automatic manual segmentation were carried out for the CT image of fabric.Then,the meso-structure of the fabric was analyzed by the processed CT images and pore network model.The results show that the fabric pore structure has the characteristics of both periodicity and discreteness.For example,the inter-tow porosity changes periodically along the three main axials of the fabric,and the unit cells at different positions have similar pore radius distribution,throat radius and length distributions,but different pore numbers,total pore volumes,pore connectivity and tortuosity.The out-of-plane fluctuation parameters and cross-sectional torsion angle of the warp yarns fluctuate periodically,and the variation of yarns cross-sectional shape parameters is mainly related to compression and nesting.The fiber volume fraction of the tow and fabric obtained from CT images are 69.8%and 58.2%,respectively.And the fabric fiber volume fraction obtained from macroscopic experiments is 57.8%.Due to the discrete structure of the fabric,the permeability in the thickness direction of the unit cell at different positions of the same fabric varies even under the same compression.The fabric unit cell with good pore connectivity,large porosity and small pore-tortuosity has better permeability.The average experimental permeability in the thickness direction is 1.25×10^-12 m²,and the predicted one based on CT images is1.18×10^-12 m²,which are in good agreement with only 5.6%error.Finally,the RTM process simulation of composite exit cone was carried out.For the RTM process,different schemes of one-stage injection,simultaneous injection at both ends,multi-gate sequential injection from smaller-diameter end,multi-gate sequential injection from lager-diameter end and sequential line injection from larger-diameter end were designed.Then process simulation was carried out by PAM-RTM to determine mold filling time and pore distribution for each injection scheme.The results show that the one-stage injection scheme has the lowest mold filling efficiency.Simultaneous injection at both ends tends to cause resin convection and intersection,making it difficult to discharge the pores.When multi-gate sequential injection is used,injection from the larger end of the exit cone gives higher product quality with lower porosity,but longer injection time is needed.High mold filling efficiency can be achieve by the sequential line injection from the larger-diameter end which leads to higher microscopic porosity undesirably.