Formability Characterization of Stitched Preform Toward Applying for Composite Structures

doi:10.21203/rs.3.rs-2195623/v1

Download PDF

Research Article

Formability Characterization of Stitched Preform Toward Applying for Composite Structures

https://doi.org/10.21203/rs.3.rs-2195623/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Characterization of the stitched textile preform is required for the prediction of textile preforms forming in the fabrication of composite structures for aerospace applications. The behavior of the preform in compression during resin injection into the preform is one of the critical issues in achieving a high-quality composite for space structures. This paper concentrates on the experimental and finite element analysis of compaction for textile preform. The investigation of mechanical properties of preforms during compression tests revealed that formability parameters of a preform can be altered by the stitching process. The load-deformation response, which is depicted in detail, had the greatest influence on the preform deformation. Compression of the stitched preform resulted in less tow undulation in the plane direction and more stitching thread undulation in the thickness direction, while the stitching thread stimulated the preform to endure more pressure. The results provided more detailed insights into the effects of compaction on other properties such as permeability and composite mechanical behavior, which is based on the internal geometry of 3D textile preforms.

Textile Preforms

Composite Structure

Formability

Performance

High-quality composite materials have been used for decades in space applications, and mostly can be found in human spacecraft, satellite structures and space launch vehicles [12, 7]. They are used for a wide range of applications in launch vehicles such as solid rocket engines and fuel and gas pressure vassels. Many composite materials used as a thermal protection system for vehicles to re-enter the atmosphere. Carbon fiber composites are commonly used on satellites structures and their payload systems [24]. The bus structure of the satellite is made of aluminum honeycomb cores with composite skins. Other structures that require dimensional stability are constructed from reinforced composite materials. Figure 1 depicts example application of composite materials for advanced space structures and how to determine their performance when those are subjected to hypervelocity debris. These composites materials aid in the preservation of extreme dimensional stability at extreme temperatures in space [13]. The demand for larger composite structures has prompted the development of high-quality composite structures that can fabricate these components with fewer joints, thereby increasing the benefits of using composite structures [1].

Three-dimensional (3D) textile preforms and composites have been used in a wide range of aerospace applications over the last decade, including pressure vessels, rotor blades, composite fuselages, external fuel tanks for commercial aircraft, and the development of next-generation aircraft [8]. 3D textile processing techniques and resin infiltration are used to create textile composites based on textile preforms. Braiding, Weaving, Stitching, and Knitting are the four major groups of 3D textile preforms based on their manufacturing techniques. Stitched composites are made by inserting a fiber in the direction of thickness. These composites play critical roles not only in the development of faster and lighter aircraft but also in the reduction of production costs significantly.

According to the literature, aerospace companies like NASA, Boeing, and Airbus are looking for ways to build large transport aircraft that are more efficient and effective to fly [10, 21]. They are attempting to replace current commercial aircraft designs with tubular fuselages and attached wings with more efficient aircraft known as Blended Wing Body (BWB), which are fixed-wing aircraft with no line between the wings and the main body of the aircraft. Based on the design, manufacture, and testing of stitched composite structures, a fuselage entirely integrated and blended with the aircraft wings significantly reduces the use of fasteners and rivets in aircraft [17]. To address the shortcomings of two-dimensional (2D) laminate composites, 3D textile composite structures for aerospace applications have been developed [22].

3D textile composites comprising 3D textile preforms have the following properties: improved strength and stiffness in the thickness direction, and the ability to design and manufacture near-net-shapes. 3D textile composites have been shown to improve composite damage resistance by increasing bond strength and damage content, such as delamination and eliminating interlaminar weakness due to the integrated structure [15]. To improve out-of-plane properties, a high-strength yarn is threaded through the thickness of textile fibers using the stitching method. Through stitch parameters such as stitch yarn diameter, stitching pattern, and tension, the insertion of the yarn through the thickness can tailor the mechanical properties. Experiments have revealed that it is unclear whether mechanical properties decrease, increase, or remain unchanged.

On the other hand, it is recognized that stitching sites serve as stress concentration zones that initiate damage. Some studies have suggested that stitching several layers together and curing using liquid molding technology can reduce manufacturing costs, but stitching will occasionally result in lower plane properties [11, 2]. The rupture and displacement of fibers caused by needle penetration within textile fabrics, as well as the presence of gaps and voids around the thickness fibers of the troughs, can result in this decrease. Nonetheless, most experimental research suggests that fiber oriented in the direction of thickness effectively improves the resistance to damage of thick laminates rather than thinner composites.

The stitching process increases interlaminar fracture strength while decreasing mechanical properties due to damage caused by fiber distortion and waviness. Furthermore, the literature suggests that stitching geometry has a significant influence on delamination mechanisms for stitched composites subjected to low and high velocity impact. The use of textile fabrics in the aerospace industry allows for the production of complex shapes components using liquid composite molding. The structure of the textile preforms influence properties such as deformability, permeability, and porosity, as well as the final composite mechanical properties [8, 6]. Characterization of local fiber variation during forming, fiber volume fraction, and fiber thickness are important parameters for describing textile fabric performance.

Different deformation modes occur during the textile preform forming including: straightening, stretching, intera-ply slippage, inter ply slipping, shearing, and wrinkling. The behavior of preforms in compression is an important issue during the resin injection process [4]. When vacuum for resin injection process is applied, a flexible film constrains the impregnated textile preform. The film's pressure is sometimes supplemented by additional weights. When a stitched preform is laid on a resin injection tool, compression influence the textile architecture and the fiber volume fraction, resulting in a variation of stitched preform properties. Textile fabric reconfiguration can result in a significant increase in preform thickness.

Therefore, due to the high sensitivity of textile fabrics to external loads, and multiple potential deformation modes at the fabrics are the main challenges in textile preform modeling. Predicting the results of even a simple compacted textile preform appears to be problematic, although many finite element models have been used to predict compaction behavior, there has not been much research on modeling of stitched preforms due to their complexity [16, 17]. For modeling of textile fabrics deformations under compression, the fibers act as bending units accompanied by fiber twisting and frictional slippage between the fibers.

The applied pressure P to the fabric volume V can be related using Eq. 1, where the constant is affected by fiber youngs modulus E, fiber mass m, fiber density df, and the variable K [18, 14]. The value of K is affected by crimp value and fiber orientation, but is not dependent by fiber diameter d_f. Eq. 2 shows a relationship between the applied compressive pressure P and the fiber volume fraction V_f. The constant k is influenced by the fiber orientation and bending rigidity. The stacking sequence of textile fabric layers and their fibre orientation, the fiber volume content in the textile have a significant impact on compressibility. This is important in defining the resin flow pattern during textile composite fabrication processes that use mold pressure.

$$P=\frac{a}{{V}^{3}} where a=KE{\left(\frac{m}{\delta }\right)}^{3}$$

$$P=k {\left(Vf\right)}^{n}$$

Using the 3-D finite-element process, the complex 3D stitched fabric geometry and the spatially different yarn properties can be modelled [23, 5]. The development of suitable material models and a geometry description can, however, be tremendous challenges and no published studies on this subject are currently available. A lot of research has been done on the analysis of textile deformations, however, no detailed general modeling study was conducted on the 3D stitched preform compression.

The current paper discusses preform compaction mechanisms and the importance of geometrical factors of stitched and unstitched preforms in the response to applied compression load is investigated. A finite element model is used to predict thickness as a function of compression pressure accompanied with an experimental test based on pressure displacement. The highest emphasis is on the importance of stitched preform geometrical and mechanical factors influencing compression behavior. An application of this approach to 3D stitched preform demonstrates the approach's capability, particularly in terms of predicting the evolution of the reconfiguration of the textile fabric structure.

The textile preforms comprised 11 layers of woven textile fabrics were laid down in quasi-isotropic configuration. The weight of the textile fabric is 305 g/m2. These textiles were stitched with a twisted 240 Tex polyester thread using a KUKA 6-axis robot equipped with a commercial tufting head. Figure 2 shows a thread-stitched preform structure accompanied with simple woven textile fabric. The thread was almost perpendicular to the fabric surfaces in these stitched preforms in order to create a stable stitched preform.

An MTS testing machine with a load cell capacity of 100 KN was used to perform compression tests on preform specimens. The compression test setup consisted of moving the upper circular punch down and compressing the stitched preform at a constant rate. When the compressive force reached the pre-described limit, the punch automatically reversed its movement. During the testing, the force and displacement were measured. The punch movement speed was 1 mm/min. The upper load limit was set at 20 KN. For analysis purposes, the thickness at which the compressive load reached 20 KN was used as the final thickness, and a typical load/deflection response curve for compression testing was recorded. In addition to the experiment testing, we attempted to simulate the compression test using the finite element technique. The use of the three-dimensional finite element method at multiscale, as shown in Fig. 3, allows for more accurate modeling of the complex fabric geometry. An appropriate unit cell simulation was performed to investigate the compacted textile geometry to evaluate the effects of fabric compaction on preform permeability and composite mechanical behavior. The unit cells of the stitched and unstitched woven fabrics were discretized using appropriate elements, and the rigid elements were used for the compression platens.

The unit cell model was placed between the platens as shown in Fig. 4. The lower plate was restricted fully, and the upper plate was compressed at a constant rate of displacement. Multiple contacts between the yarns and unit cell surfaces with the plates were defined by the surface-surface approach. All necessary steps to perform the simulation including geometry, material property, mesh generation, and boundary conditions were performed.

Textile preforms are modeled as a transversely isotropic component with a mesoscale elastic material having linear properties. The geometry of the basic weave structure is driven by the actual sample measurement and Table 1 lists the property of a woven textile fabric.

Table 1

Material properties used in simulation [20].
Textile fabric	${{G}}_{12}$ =3.5 (Gpa) ${{E}}_{22}$=10.4 (Gpa) ${{X}}_{{T}}$=1140 (Mpa) ${{X}}_{{C}}$=620 (Mpa)	${{G}}_{12}$ =4.3 (Gpa) ${{\nu }}_{12}$= 0.28 ${{Y}}_{{T}}=39$ (Mpa) ${{S}}_{12}=89$ (Mpa)	${{E}}_{11}$=41 (Gpa) ${{\nu }}_{23}$ =0.5 ${{Y}}_{{C}}=128$(Mpa) ${{S}}_{23}=89$(Mpa)

Several tests were conducted to achieve adequate repeatability and establish a typical material response. The pressure curves of stitched and sun-stitched preforms are depicted in Fig. 5. The stitched preforms had the same features as those which were not stitched. Stitching preforms had the highest loading values at the same displacement when compared with preforms without stitching. The lower value of compression energy for unstitched preforms indicates that the preform forms easily and deforms better than the stitched preforms according to the compression energy defined by the area under the load-deflecting curve. The deformation curves of the two preforms are found to be nearly identical and each curve consisting of three parts: two linear parts and an exponential part. Linear part can be related to preform compactions by decreasing porosity and gaps between filament and fiber bundles, and cross-sectional deformation of fibers and fiber bending is relevant to the exponential part.

Figures 7 and 8 show the simulated deformed unit cells for the unstitched and stitched preforms. As When an unstitched preform is compressed, the stress is concentrated in a small area in contact with the upper platen as shown in Fig. 7. First, a flat region is formed on a yarn cross-section in two yarns, causing the cross-section to deform and the yarns to be compacted. The volume fraction of the fibers will vary, while the volume fraction increasing in the contact area but not at the yarn edges. Therefore, these simulations confirm that during compaction, all sections of yarn do not carry the same load at the same time and that the external load is carried only at fibers crossovers.

The above-mentioned modeling techniques are easily applicable to high-performance stitched textiles. Figure 8 depicts a deformed FE model of textile compaction caused by the downward movement of the top platen. While the preform geometry at the end of the compression loading, the applied compressive load is mainly concentrated in the stitching area, and the through-thickness yarn crimp changes significantly through the thickness direction. The compaction of the stitched fabric is a function of the compression pressure and depicts the nonlinear compaction behavior of the textile. Therefore, it can be said that during the initial compression loading, a change in the thickness of the preform can be caused by the compression of the stitching thread and a change in the crimping of the thread.

Figure 9 shows the compression behavior force against displacement curves for stitched and unstitched preforms. Our findings clearly show that the modeling method can reasonably predict the mechanical response during compression and is very suitable for predicting all preform changes in the thickness direction. There may be a small difference in the stitched preform subjected to compressive load during modeling, such as a sudden increase in force due to misalignment and bending of the thread through the thickness. The difference between the predicted from modeling and measured compressive behavior can be explained by the fact that the yarns are modeled as solids, which cannot provide the slippage of the filaments inside the yarns during the compression loading.

To achieve a high level of accuracy in the predicted mechanical behavior, accurate geometric models and measured mechanical data of the fiber bundles are required as input to the constitutive material model as well as the correct implementation of the finite element model. In addition, it is observed that the compressive force intensity in the stitched preform has a higher-order which causing pressure inside the yarns and during the compression, the trough thickness yarns tend to buckle out. The complete compression deformation process for a preform can be explained in two stages, according to the pressure curve: initially, the yarns are compressed against one another and deformed.

In the second stage, as soon as the compression deformation begins, the yarns begin to slip over each other until compaction begins at a certain force, and the stiffness rises as the adjacent yarns start to compress each other. This process will be repeated until the yarns are no longer easily compressible. If the deformation is allowed to continue after the yarns have been locked, the fabric will buckle out of the plane, significantly increasing stiffness. The modeling results show good validity with experimental results for stitched and unstitched fabrics that were found to be reasonable, and the current model's results are expected to be fairly good.

Figure 10 depicts the unit cell's energy curve, which is related to fabric deformation during the compression process. According to the work-energy theorem, as the workload increases as shown in the graph, so does the energy. A sudden increase in energy indicates fiber crimping and non-slippage between fibers due to the presence of stitching thread in the stitched preform. Because higher crimping in fibers is accompanied by higher deformation constraints, the increasing energy indicates that higher crimping in fibers is exposed to greater compression loading.

Figure 11 depicts the application of localized compressive pressure to a small region of yarn crossovers in the stitching area, which defines the overall compressive load-bearing capacity of the preform. When compared to other parts of the fabric, the fiber volume fraction is greatest and the cross-sectional area is smallest at these points. The uniform compression pressure applied to the top platen is distributed unevenly across the fabric yarns. The stitching area sections that come into direct contact with the plate are subjected to the most pressure.

The deformed and reconfigured unit cells were previously discussed. It was found that in the case of stitched and unstitched compression, the stress limited in a small area is in contact with the top platen and that not all parts of a fiber bundles support the same load as the external load. Only at yarn intersections, the compressive force changes the cross-sectional dimensions of the yarn as well as its crimp height, thus reducing the thickness of the fabric. In addition, the compressive force is distributed over larger areas of yarn contact, therefore the stitched preform has a larger area to withstand external loads than the unstitched preform.

The relationships between through-thickness yarns and pressure in stitched preform lead the most difficult to compact, while the unstitched unit cell was easiest to compact. The stitching yarns leads to a bending deformation of the yarn and more bending rigidity contributes much to resisting meso unit cell compaction. As a result, increasing the stitching thread height can devrease fabric compressibility. Furthermore, through-thickness yarn height has a direct influence on yarn crimp interchange, which is significant because it affects in-plane stiffness and geometric nonlinearity. Clearly, due to meso-bending deformation, the contribution of the stitching yarn to unit cell compaction differs significantly from that of the unstitched unit cell.

The compression test of stitched preforms was subjected to experimental and modeling characterization. For stitched and unstitched textile preforms, a finite element unit cell compression model has been developed. The Finite element model captures the primary fabric compression response, which includes geometric and preform nonlinearities as well as yarn interactions. The study of mechanical properties of preforms during compression tests revealed that stitching can change the formability parameters of a stacked textile fabric. The most significant effect on the deformed textile preform was the load-deformation response, which was depicted in detail. According to our observations, compression of the stitched preform resulted in less fiber bundle undulation in the plane direction and more stitching thread undulation in the thickness direction, while the stitching thread stimulated the preform to initiate pressure. The application of compression load to preforms resulted in an increase in fiber volume fractions because of fiber bundles flattening and a decrease in fiber crimp. During the compression test, the energy absorption in the stitched preform was found to be high. Future research will focus on the effects of fabric compaction on preform permeability at the meso and micro scales, as well as the investigation of stitched preforms and composite in the whipple shield configurations against space debris for satellite structures.

STATEMENTS AND DECLARATIONS

No funding has been used for the current research.

The author declares that he has no conflict of interest.

All data has been included in the paper and no other data is available.

All work has been done by the author.

ACKNOWLEDGMENT

This research was supported by the Space Technology Research Laboratory at the department of Mechanical and Aerospace Engineering, Nazarbayev University, Kazakhstan, and BHS composite in Canada. We thank our colleagues from Nazarbayev University, Istanbul Technical University and BHS composite, and CTT group in Canada for providing facilities that greatly assisted our research.

Anz-meador P., Opiela N., Shoots D., History of on-orbit satellite fragmentations: NASA/TM-2018-22003, Johnson Space Center, NASA, 2018.
Barile C., Casavola C., Pappalettere H., The influence of stitching and unconventional fibres orientation on the tensile properties of CFRP laminates, Compos. Part B Eng., Vol. 110, pp. 248–254, 2017.
Deconinck p, Abdulhamid h, Héreil P., Experimental and numerical study of submillimeter-sized hypervelocity impacts on honeycomb sandwich structures, J. Procedia engineering, 204:452–459 2017.
Dixit A., Misra R., Mali H., Compression Modeling of Plain Weave Textile Fabric Using Finite Elements, Mat.-wiss. u. Werkstofftech, 45(7), 2014,.
Guzman-Maldonado E., Wang P., Hamila N., Boisse P., Experimental and numerical analysis of wrinkling during forming of multi-layered textile composites. Compos Struct, 2019.
Huang J., Nahiène Hamila and Philippe Boisse, « Draping modelization of stitched composite reinforcements », ESAFORM, 26 August 2021.
Ivanov D. S., S. Lomov, Modelling the structure and behaviour of 2D and 3D woven composites used in aerospace applications, Polymer Composites in the Aerospace Industry, Woodhead Publishing, 2014.
Iwata A., Inoue T., Naouar N., Boisse P., Lomov V., Coupled meso-macro simulation of woven fabric local deformation during draping. Composites Part A: Applied Science and Manufacturing, 118:267–280, 2019.
Mahadik, Y., Hallett S., Finite element modelling of tow geometry in 3D woven fabrics. Composites Part A: Applied Science and Manufacturing, 41(9), 2010.
NASA Contractor Report, final report for contract NNL13AA11C, No: 218149, 2104.
Pierreux G., Van-Hemelrijck D., Massart T., RVE computations for evaluation of microdamage initiation in structural stitched non-crimp fabric composites. Journal of composite materials, Vol.54, No.30, pp.4751–4771, 2020.
Reinhart J., Overview of composite materials, in Handbook of Composites, Springer, 1998.
Riedel W., Nahme, H., White D.M., Clegg R., Hypervelocity impact damage prediction in composites: Part II—experimental investigations and simulations. Int J Impact Eng 33, (1–12):670–680, 2006.
Robitaille F., Gauvin R., Compaction of textile reinforcements for composites manufacturing: review of experimental results Polym. Compos. Vol.19, 198–216, 1998.
Rolfe E., Quinn R., Sancho A., Kaboglu C., Johnson A., Liu H., Hooper P., Blast resilience of composite sandwich panels with hybrid glass-fibre and carbon-fibre skins, Multiscale and Multidisciplinary Modeling, Experiments and Design, Vol.3: 197–210, 2018.
Schirmaier F., Dörr D, Henning F, Kärger L. A macroscopic approach to simulate the forming behaviour of stitched unidirectional non-crimp fabrics (UD-NCF). Compos Part Appl Sci Manuf 2017.
Shen H, Wang P, Legrand X, Liu L, Soulat D. Influence of the tufting pattern on the formability of tufted multi-layered preforms. Compos Struct, 228:111356, 2019.
Sherburn M., Geometric and Mechanical Modelling of Textiles. Ph.D. Thesis, Nottingham University, 2007.
Strong B., History of Composite Materials-opportunities and necessities. Brigham Young University, 2002.
Thrash P., Manufacturing of a Stitched Resin Infused Fuselage Test Article, CAMX SAMPE, 2014.
Tomich A., Clarke S., Integration of Complex Geometry, 3D Woven Preforms via Innovative Stitching Technique, Phase II, Space Technology Mission Directorate, NASA, 2019.
Unal, P.G., 3D Woven Fabrics. Woven Fabrics. ISBN, 2012.
Werner, H.O. Dörr, D. Henning, F. Kärger, L. Numerical modeling of a hybrid forming process for three-dimensionally curved fiber-metal laminates. AIP Conference Proceedings 2019.
Zheng s G., Yan j., A review on the space debris protection and protective materials, J. Space International,6: 49–53, 2014.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Formability Characterization of Stitched Preform Toward Applying for Composite Structures

Status:

Version 1

Abstract

Figures

Introduction

Experimental Material And Equipment

Results And Discussion

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1