A novel processing solution for the production of spatial three-dimensional stitch-bonded fabrics

Abstract

Spatial three-dimensional (3D) stitch-bonded preforms offer a huge potential and an optimal utilization of the fiber mechanical properties for their usage in various technical applications. These preforms form an efficient alternative to the existing two-dimensional textile preforms due to their load-bearing fiber orientation offered by their geometrical configuration. The fundament of this research activity, based on the stitch-bonded technology, is the development of a new processing solution for producing 3D spatial open stitch-bonded textile preforms with minimal processing stages. Based on a newly developed concept with variable warp yarn delivery and differential doffing, this technology is principally enhanced so that various helical, circular and curved grid structures, along with 3D spatial structures, could be produced on a single machine with minimal setup, which hitherto has not been realized with any other existing processing technique. These novel form-based textile preforms are found to be suitable for application as reinforced components in complex spatial composites with a mineral or a plastic-based matrix, wooden composites and elastomers.

Keywords

stitch-bonded textiles three-dimensional spatial preforms novel production method

Composite structures with high-performance fiber reinforcement are well established in various technical applications. However, a comprehensive industrial usage has not yet been achieved. The primary reason is the existing “Cost-Effectiveness Gap” in their production. In general, high-performance fibers embedded in a matrix are the principal load-carrying members in the composite parts. In order for the composite structure to effectively withstand impact loads, intricate orientation of the fiber layup in accordance to the loading direction is required. Complex spatial three-dimensional (3D) composite parts contain textile fabrics as reinforcing components. However, they require further processing in order to obtain the required geometry. Most of the textile preforms are available in the form of planar fabric sheets and need to undergo multi-stage processing to attain a spatial 3D form. Complex and labor-intensive, slow intermediate processes together with the existing economically limited automation capabilities are the basic hindrance in their large-scale production. One step for reducing the production cost is to decrease the existing intermediate processing stages that form the basis of the complex spatial 3D preform fabrication.¹ This would not only offer an economically viable solution, but also lead to improved mechanical properties in the composite parts.²

Planar and spatial textile preforms with pre-defined geometries and load-conform fiber orientations can currently be produced due to various advancements in braiding, weaving, knitting and tailored fiber placement technologies. The existing braiding technique allows the production of profile sections with varying thickness or variable shape geometries^3,4 with modularly constituted bobbin drive elements. It also offers the possibility of a load-conforming yarn insertion into the preform.⁵ The structural variability that could be achieved by this technique is currently held against its low productivity. In addition, the size of the braid and the corresponding preform size are severely restricted by the number of braiding coils. The modified weaving processes, such as the shape-weaving process,^6,7 make it possible to produce shell-shaped preforms in addition to the existing two-dimensional (2D) contour-based woven. The desired 3D contour is obtained by the local differential changes in the weft density produced through the differential doffing of warp yarns with graded working speed.⁸ Owing to the nature of the process, inhomogeneity occurs in the structure, which leads to undesirable anisotropies⁹ in the composites. The flat knitting process¹⁰ can successfully integrate more reinforcing fibers with different orientations in 2D and 3D shaped knitted fabrics. The single needle control in this technology opens unlimited shaping possibilities with respect to the theoretically feasible preform geometries. Nevertheless, there exists limitation as in the case of the shape weaving, due to the fact that the differences in warp yarn lengths are realized by the doffing force behind the tie-up unit. Furthermore, it is only suitable for lower production rates and also has limited capabilities for producing open preforms. The knitting technology (warp knitting, stitch bonding) with bi- and multi-axial types is being increasingly used for preform production, as in the case of both plastic and mineral matrix reinforcements.^11,12 Although the currently available warp-knitting technology is well developed, there is still a lack of technological basis for producing structures of varied 2D and 3D geometrical shapes. The multi-axial fabric production offers many advantages, such as undulation-free production, a better usage of the fiber mechanical properties and load-oriented fiber layout. However, the production of fabrics with varied contour is still highly constrained and the production of preforms with spatial 3D geometry has not been realized until now. A gradually graded doffing of the warp yarns can produce planar curved surfaces. The radius of the curvature, however, in this case is solely dependent on the conicity of the doffing rolls. Any change in the curvature radius requires new doffing rolls with corresponding conicity. Furthermore, solutions exist for the production of the warp-knitted fabrics with foldable areas, such as Double-T profiles.^13,14 However, the extent of the cost-effectiveness in the process solutions on this basis has not yet been extensively studied. As an alternative to these techniques, the tailored fiber placement offers a possibility to fix the load and shape-conforming reinforcement fibers on the base material.¹⁵ A freely selectable in-plane arrangement of the single filament allows the application of this method in various 2D geometrical shapes. However, in the case of preforms with 3D geometrical shape, additional steps are required. This is especially true for the material studies conducted by Meyer and Gessler¹⁶ in their fiber-patch preforming.

The above state of the art indicates that the existing textile technological possibilities for having a unified process in the production of 2D and spatial 3D textile preforms are either highly limited or the property needs of the preforms with respect to their production efficiency are not fully sufficient. Moreover, no known textile manufacturing process exists for the realization of open, net-shaped spatial 3D preforms. Hence, the focus of the research activity is the development of a single-step fabrication process for producing spatial 3D preforms that are close to the final part geometry. Stitch-bonding technology, a special warp-knitting process, is a well-developed technology regarding the quality and quantity of the produced fabric sheets. The initial work on producing spatial textile preforms at the Institute of Textile Machinery and High Performance Technology, Technische Universität Dresden, was successfully demonstrated on the basis of the crochet galloon technique, a warp-knitting process analogous with stitch-bonding arrangements. This paper presents and discusses a newly developed solution and its concept, where this stitch-bonding technology is fundamentally enhanced to produce 2D and 3D net-shaped structures with negligible machine setup requirements. A new technology of variable warp yarn delivery makes it possible to realize doubly curved spatial stitch-bonded fabrics, in addition to the existing planar fabric manufacturing, thereby enabling the production of complex textile preforms (Figure 1) in very few steps.

Figure 1.

Sample net-shaped spatial three-dimensional preforms.

Theoretical concept for technical realization

The basic concept for the production of 3D stitch-bonded fabrics is based on a variable measurement and delivery of the individual warp yarns. Figure 2 illustrates the variable warp yarn delivery concept, where differential delivery of the yarns can be ensured by varying the feed-in speeds. Through a pre-defined difference in the length of a given warp yarn section, the stitch length and the distance between the weft yarns of the fabric could be specifically varied in each knitting cycle. This principle is in general applicable for other processes as well. A technical enhancement of the stitch-bonding technology is carried out for the implementation of this concept, wherein the lengths of the delivered reinforcement yarns (warp yarns) are varied and are independent from one another. Together with a synchronous differential doffing system, this principle can offer a better alternative in reducing the inhomogeneities and anisotropies in the preforms when compared to the existing differential doffing where the doffing force is the prime factor for differences in warp yarn lengths.

Figure 2.

Theoretical concept of variable warp yarn delivery.

Another salient feature of this concept is the geometrical representation of the 3D preforms in a mathematical model and the corresponding analysis for their technical realization. In order to make computational predictions for their production feasibility, the determination of the individual warp lengths with respect to various parameters, such as yarn properties and material, machine/process parameters are required. It would help in advance to understand and optimize product designs, their subsequent manufacturing operations and product reliability, thereby ensuring judicious usage of material and energy efficiency.

Realization of a three-dimensional stitch-bonded fabric

The conformance of the concept for the 3D stitch-bonded fabric realization is demonstrated here with an experimental production of sample net-shaped textile reinforcement for a section of Torus. Figure 3 shows the preform production specifications for the sample Torus cross-section. A technology enhancement can only be ensured by a seamless integration of the new concepts and theory in the existing process. Hence, the developed solution, based on the above concept, first needs to cater for the existing and future production requirements. These requirements based on two criteria – (1) Mandatory (M) and (2) Optional (O) as proposed by Pahl and Beitz¹⁷ – are illustrated in Table 1.

Figure 3.

Production specifications for the sample Torus cross-section.

Table 1.

Requirement criteria

Key feature	Requirement	Criterion
Geometry	• Computational predictions and production feasibility analysis of the 3D preforms	M
	• Process integration on the existing stitch-bonding machine	M
	• Usage of full machine width	O
Kinematics	• Continuous active feed of the warp yarns	M
	• Feed-in speed of the individual warp yarn can be varied	M
	• A complementary synchronous differential doffing unit	O
	• Dynamic changing of the feed-in speed	M
Material	• Processing of different yarn material and counts	M
Material	• Efficient processing of the yarn material	M
Ergonomics	• Processing of 2D and 3D preforms with no mechanical changes in the machine	M
Efficiency	• Setup and integration complexity	O
Efficiency	• Optimal usage of material and energy efficient production	M

M – Mandatory, O – Optional.

The textile reinforcements for the sample section are split into four spatial parts, each enclosing the top and bottom and the left- and right-hand side. While the top and bottom parts are symmetrical in nature, the sides differ in their direction of the curvature (Figure 4). These preform variants consisting of top, right- and left-hand cross-sections are first built using a mathematical model and then their production feasibility is analyzed. The individual warp yarn lengths are then calculated for each section, taking the process and machine parameters into account.

Figure 4.

Variants for the sample section.

A mechanical graded differential warp yarn delivery unit, along with a basic synchronous doffing system, is designed and constructed on the basis of the computed lengths for the production on a bi-axial stitch-bonding machine for the purpose of validation. The produced fabric preform is finally analyzed and evaluated for the geometrical accuracy with physical model part geometry. Based on the attained results, a modified universal solution with a new NC-Drive warp yarn controlled delivery unit and new doffing system is being currently developed.

Geometrical model and computation

An efficient and sustainable production can only be achieved when a continuous process chain from computer-aided design (CAD) to manufacturing exists where the producibility of the preform geometries can be effectively analyzed offline, production parameters determined and finally the fabrication of varied geometries can take place on the machine. Although various CAD software packages for modeling and studying the yarn behavior with respect to the geometrical shape exist, they still do not have provisions to study the behavior with respect to the process/machine parameters.¹⁸ A study in relation to the production parameters can only determine whether the geometry is actually producible. This is essential for an optimal and cost-efficient production.

Most of the existing methods (with the exception of braiding and winding) for producing spatial preforms are based on the manipulation and calculation of the reinforcement yarn lengths in order to attain the 3D geometrical shape. Various computational methods used for determining the filament yarn length were examined for their suitability with respect to the universality and transferability. In the case of shape weaving, the commonly used calculation method is cutting the surface in the warp and weft yarn directions perpendicular to the ground plane and distributing the resulting cross-section curve in equal distance across the geometrical surface. In contrast to their linear course in the planar fabrics, the weft and warp yarns evenly cover the surface here. The lengths of the warp and weft yarn are calculated from the distance between the overlapping points on the surface. The disadvantage of this approach is that the producibility of the structure with respect to the yarn course (the angle between warp and weft threads and the length of the thread sections may be not too small) must be subsequently verified. Moreover, this calculation method is not universally valid, and the proposed sample Torus cross-section variants cannot be calculated with this method. Alternatively, each spatial surface can be flattened as a planar surface and, subsequently, its dimensions measured. This method has been used for computation in an integrated knitting process.¹⁹ This procedure is indeed transferable to other surfaces, but still does not offer a universal solution, as each surface has to be analyzed individually. The creation of a suitable process is very complex²⁰ and more complicated for free-form, doubly curved surfaces. Hence, a new offline computational software package is written and developed for the geometrical model representation and computation analysis. The mathematical calculations are performed directly in 3D space and are universally applicable for various structures.

The 3D geometrical surface here is mathematically represented in subsets of $R^{3}$ as point clouds and is described by the following function:

S = (\begin{matrix} x y f (x, y) \end{matrix})

(1)

where S is the surface.

The third dimension of the surface is here represented as a function of two variables, x and y. These variables are directly linked to the machine parameters for an eventual optimization of surface geometry based on the producibility analysis.

In this case a quadratic form representation is assumed:

f (x, y) = (x, y)^{T} \cdot A \cdot (x, y)

(2)

The curvature of the surface here can be controlled by the Eigenvalues of the matrix A. Here A is defined as a diagonal matrix for easy computation, as the corresponding Eigenvalues are then simply the elements on the diagonal. These Eigenvalues are used as the control parameter for the radius of curvature in the software to determine the permissible ranges of curvature for a given geometrical shape and material type that can be processed on the machine. Besides that, they are also used to create different preform shapes from a given initial plane of length L and width W. Figure 5 depicts the different curvatures that can be generated by varying the Eigenvalues for a given geometrical shape. The surface of the geometrical model is represented in the form of point clouds²¹ generated from the above functions, and is stored in $n \times m$ rows and columns representing the length and width of the geometry. The resolution of the point cloud determines the accuracy of the computed warp yarn length.

Figure 5.

Curvatures generated from a given initial plane and respective Eigenvalues 0.

For the computation of the warp yarn lengths, the geometrical model surface is taken as an initial reference. A central axis is defined in plane as a reference. Figure 6 shows the generated geometrical model with a central axis for the sample torus cross-section. An initial point P on this central axis plane is then projected on the surface P₁. The surface normal N at point P₁ is then determined (Figure 7).

Figure 6.

Geometrical model with central axis for the Torus cross-section.

Figure 7.

Geometrical model with central axis for the Torus cross-section.

The warp yarn distribution on the surface is then determined on the basis of the machine gauge K. The number of required warp yarns can be determined by the following equation:

n_{KF} = \frac{W_{PF}}{K}

(3)

where W_PF is the absolute surface width [mm] and K is the machine gauge [mm].

Surface normal N, N₁, N₂ for all the n_KF warp yarns are then determined in intervals defined by the machine gauge K across the weft direction. The warp yarns are equally distributed on both side of the central axis. The surface normals determined for each of the n_KF warp yarns are then moved across the surface in the warp yarn direction until the minimum weft yarn distribution distance S is achieved for all warp yarns. The total length of all the intermediate points for each weft yarn is then calculated. This length represents the warp yarn length that has to be delivered in a single working cycle. This procedure is iteratively performed for n_SF number of weft yarns, as defined by the user or alternatively for the whole geometry.

The maximum and minimum weft yarn distances are calculated using the following equations:

n_{SF} = \frac{l_{KF, min}}{S} S_{max} = \frac{l_{KF, max}}{n_{SF}}

(4)

where n_SF is the number of weft yarns,

l_{KF, min}, l_{KF, max}

are the minimum and maximum warp yarn lengths [mm] and

S, S_{max}

are the minimum and maximum weft yarn distances [mm].

The characterization of the geometries based on the above principle was analyzed in accordance with the machine parameters, as illustrated in Table 2. Figure 8 shows the computed model for two fabrics with absolute widths of 66 mm and 186 mm. Tables 3(a) and 3(b) provide the corresponding calculated warp yarn lengths. These values are stored in an ASCII file format. Based on the calculated warp yarn lengths, the required dimension of the variable warp yarn delivery unit is determined. Figure 9 shows the intermediate stages involved in the design and integration of the graded delivery unit on a stitch-bonding machine.

Figure 8.

Computed model for two fabric samples of varying width.

Figure 9.

Integrated warp yarn graded delivery unit (Source: Lotzmann²²).

Table 2.

Machine parameter for Malimo 14022

Machine parameter		Data for Malimo 14022
Width of the fabric	W_PF	66 mm or 186 mm
Minimum warp yarn length that can be delivered		> S
Weft yarn distribution	S	6.4 mm (Gauge E4)
Machine gauge	K	7.2 mm (Gauge F3.5)

Table 3.

(a) Calculated warp yarn lengths for a 66 mm fabric width. (b) Calculated warp yarn lengths for a 186 mm fabric width

n_KF\n_SF	1	2	3	4	5	6	7	8	9	10	11	12	13
(a)
1	5.618	5.618	5.618	5.619	5.619	5.619	5.619	5.619	5.62	5.620	5.620	5.620	5.620
2	5.696	5.696	5.696	5.696	5.697	5.697	5.697	5.697	5.697	5.697	5.698	5.698	5.698
3	5.756	5.757	5.757	5.757	5.757	5.757	5.758	5.758	5.758	5.758	5.758	5.759	5.759
4	5.840	5.840	5.840	5.840	5.841	5.841	5.841	5.841	5.841	5.841	5.842	5.842	5.842
5	5.903	5.903	5.904	5.904	5.904	5.904	5.905	5.905	5.905	5.905	5.905	5.906	5.906
6	5.989	5.989	5.989	5.989	5.990	5.990	5.990	5.990	5.990	5.991	5.991	5.991	5.991
7	6.074	6.074	6.075	6.075	6.075	6.075	6.076	6.076	6.076	6.076	6.076	6.077	6.077
8	6.138	6.138	6.138	6.139	6.139	6.139	6.139	6.139	6.14	6.140	6.140	6.140	6.140
9	6.221	6.221	6.222	6.222	6.222	6.222	6.222	6.223	6.223	6.223	6.223	6.223	6.224
10	6.282	6.282	6.282	6.283	6.283	6.283	6.283	6.283	6.284	6.284	6.284	6.284	6.284
11	6.359	6.360	6.360	6.360	6.360	6.361	6.361	6.361	6.361	6.362	6.362	6.362	6.362
12	6.415	6.415	6.415	6.415	6.416	6.416	6.416	6.416	6.41	6.417	6.417	6.417	6.417

(b)
1	5.240	5.240	5.240	5.240	5.241	5.241	5.241	5.241	5.241	5.242	5.242	5.242	5.242
2	5.248	5.248	5.248	5.249	5.249	5.249	5.249	5.249	5.249	5.250	5.250	5.250	5.250
3	5.262	5.263	5.263	5.263	5.263	5.263	5.264	5.264	5.264	5.264	5.264	5.264	5.265
4	5.283	5.283	5.283	5.284	5.284	5.284	5.284	5.284	5.284	5.285	5.285	5.285	5.285
5	5.309	5.309	5.309	5.31	5.310	5.310	5.310	5.310	5.310	5.311	5.311	5.311	5.311
6	5.352	5.352	5.352	5.352	5.352	5.353	5.353	5.353	5.353	5.353	5.353	5.354	5.354
7	5.389	5.389	5.390	5.390	5.390	5.390	5.390	5.391	5.391	5.391	5.391	5.391	5.391
8	5.446	5.447	5.447	5.447	5.447	5.447	5.448	5.448	5.448	5.448	5.448	5.448	5.449
9	5.494	5.494	5.494	5.494	5.495	5.495	5.495	5.495	5.495	5.496	5.496	5.496	5.496
10	5.563	5.563	5.563	5.563	5.564	5.564	5.564	5.564	5.564	5.564	5.565	5.565	5.565
11	5.618	5.618	5.618	5.619	5.619	5.619	5.619	5.619	5.62	5.620	5.620	5.620	5.620
12	5.696	5.696	5.696	5.696	5.697	5.697	5.697	5.697	5.697	5.697	5.698	5.698	5.698
13	5.756	5.757	5.757	5.757	5.757	5.757	5.758	5.758	5.758	5.758	5.758	5.759	5.759
14	5.840	5.840	5.840	5.840	5.841	5.841	5.841	5.841	5.841	5.841	5.842	5.842	5.842
15	5.903	5.903	5.904	5.904	5.904	5.904	5.905	5.905	5.905	5.905	5.905	5.906	5.906
16	5.989	5.989	5.989	5.989	5.990	5.990	5.990	5.990	5.990	5.991	5.991	5.991	5.991
17	6.074	6.074	6.075	6.075	6.075	6.075	6.076	6.076	6.076	6.076	6.076	6.077	6.077
18	6.138	6.138	6.138	6.139	6.139	6.139	6.139	6.139	6.14	6.140	6.140	6.140	6.140
19	6.221	6.221	6.222	6.222	6.222	6.222	6.222	6.223	6.223	6.223	6.223	6.223	6.224
20	6.282	6.282	6.282	6.283	6.283	6.283	6.283	6.283	6.284	6.284	6.284	6.284	6.284
21	6.359	6.360	6.360	6.360	6.360	6.361	6.361	6.361	6.361	6.362	6.362	6.362	6.362
22	6.415	6.415	6.415	6.415	6.416	6.416	6.416	6.416	6.417	6.417	6.417	6.417	6.417
23	6.484	6.484	6.484	6.484	6.485	6.485	6.485	6.485	6.485	6.486	6.486	6.486	6.486
24	6.531	6.531	6.532	6.532	6.532	6.532	6.533	6.533	6.533	6.533	6.533	6.534	6.534
25	6.588	6.588	6.589	6.589	6.589	6.589	6.590	6.590	6.590	6.590	6.591	6.591	6.591
26	6.626	6.626	6.626	6.627	6.627	6.627	6.627	6.628	6.628	6.628	6.628	6.628	6.629
27	6.669	6.669	6.669	6.669	6.67	6.670	6.670	6.670	6.670	6.671	6.671	6.671	6.671
28	6.695	6.695	6.695	6.695	6.696	6.696	6.696	6.696	6.696	6.697	6.697	6.697	6.697
29	6.715	6.715	6.715	6.716	6.716	6.716	6.716	6.717	6.717	6.717	6.717	6.718	6.718
30	6.73	6.730	6.730	6.730	6.731	6.731	6.731	6.731	6.731	6.732	6.732	6.732	6.732
31	6.738	6.738	6.738	6.738	6.739	6.739	6.739	6.739	6.739	6.740	6.740	6.740	6.740

The computed warp yarn lengths form the basis for the construction in the case of a mechanical warp yarn delivery system and for determining the speed of drive units in the case of a drive-based yarn delivery system.

Stitch-bonding machine Malimo 14022

The textile technological requirements for the realization are fulfilled by the high-performance stitch-bonding machine Malimo 14022 from Karl Mayer Textilmaschinenfabrik GmbH, Germany (Figure 10). The technical specification of the machine is illustrated in Table 4. An initial production of the fabric sample with the width of 66 mm for the given Torus cross-section was carried out on this machine. The weft yarns are hung on both ends of the transport chain and delivered to the stitch-bonding unit. The yarns are then immediately cut off from the rolls after the stitch bonding. The yarn feed system of the machine has been modified to accommodate carbon fiber yarns up to 10,000 Tex.

Figure 10.

Malimo 14022 stitch-bonding machine.

Table 4.

Machine specifications of Malimo 14022 stitch-bonding machine

Malimo 14022 stitch-bonding machine
Specifications		Range
1	Working width	Up to 1500 mm
2	Machine gauge	7F/10F/18F
3	Stitch length	0.5–5 mm
4	Working speed	up to 1400 min^–1

Variable warp yarn delivery unit

A fixed-graded variable warp yarn delivery unit was designed and integrated for realizing the variable warp yarn delivery. The term “fixed-graded delivery unit” is defined here as a delivery unit in which the feed-in speed of each warp yarn is graded and remains fixed during the production. The delivery unit layout, as depicted in Figure 11, consists of graded disc rolls with varying diameters mounted on an existing shaft of the stitch-bonding machine. These disc rolls have grooves across their circumference where the warp yarns are carried through. The pressure discs unit ensures that warp yarns are held together in their respective grooves. The shaft is driven by a servo motor. The different diameters of the discs result in a varying delivery speed of the warp yarns to the stitch-bonding unit. The advantage of this delivery unit is its simple design and integration possibility into the machine. Any fluctuations in the drive speed do not influence the surface building, as the speeds are defined by the disc diameters.

Figure 11.

Illustration of a graded delivery unit.

The pressure disc unit consists of individual discs with pressure springs. The pressure for each disc could be preset separately, thereby providing a better control over the slippage of the warp yarns on the disc rolls. The pressure on the individual discs is set manually with an adjustable screw. In addition, these pressure discs can be used in combination with other disc roll sizes as well, as the distance between them can be adjusted. Thereby the need for redesigning the pressure disc in accordance to the disc roll sizes is made redundant. However, this hardware setup can only be used for a specific geometry and needs to be newly designed for different shapes.

The required grading and diameter of the disc rolls are calculated from the computational model and used for construction. A rubber coating is applied onto the grooves of the disc rolls for a better static friction between the yarn and the surface. This reduces slippage.

Synchronous doffing unit

In order to have an efficient stitch bonding, a synchronous doffing in accordance with yarn delivery feed needs to be ensured. The existing doffing unit in the machine is designed only to handle planar fabrics. As the 3D fabrics build a spatial surface and cannot be flattened into a planar surface, a new supplementary doffing system is designed and integrated. The doffing system consists of two parallel mounted rolls in such a way that the one roll is actively driven. The other roll is passive and driven through friction. The produced fabric is passed between the two rolls and doffing occurs by driving the rolls. The doffing of the fabric here must take place immediately after the stitch-bonding unit. A winding of the 3D fabric into rolls here is not possible. The 3D sample was produced with glass fiber roving yarn with varying stitch density, as can be seen in Figures 12(a) and (b). Table 5 presents the production parameters for the sample. The produced sample is limp and no coating is applied.

Figure 12.

Produced three-dimensional fabric sample with varying stitch density. (a) Closed grid, (b) Open grid.

Table 5.

Production parameters for the three-dimensional fabric sample

Warp yarn	Glass fiber rovings 2400 Tex
Weft yarn	Glass fiber rovings 640 Tex
Sewing thread	Polypropylene textured 20 Tex
Machine speed	100 min⁻¹
Stitching length	3 mm
Speed of warp yarn shaft	1.7 mm/machine revolution
Weft feed	1 full/2 empty
Bonding	Tricot

Evaluation of the produced fabric and analysis

A commonly available pipe bend structure, which represents the Torus section form and has the same dimensions as calculated by the geometrical model, is used for validating the geometrical accuracy of the produced sample. As a first step, the produced 3D sample is laid on the pipe surface for geometrical shape verification. As seen in Figure 12(b), the fabric does not have any surface distortions or folds and has the same geometry as the pipe bend structure. Hence, a draping of the fabric is not required. The fabric can be directly removed from the production, laid on the surface and lightly pressed.

The surface curvature formation, that is, the shape form, can be easily recognized immediately after doffing. The computed curvature and geometrical accuracy pertaining to the structural arrangement of the warp yarn is validated by first marking the pipe surface with a grid indicating the position of the weft and warp yarns as generated by the mathematical model. The fabric is then laid on the grid in such a way that the center warp yarn of the fabric and the center of the grid coincide, that is, no deviations occur at the center. The deviation between the laid fabric and the grid determines the production accuracy. Deviations of warp yarn were found here at the edges of the fabric from the outer grid. The outermost warp yarn at the inner pipe radius and the three warp yarns starting from the edge of the outer pipe radius exhibit deviations (Figures 13(a) and (b)). The maximum deviation from the actual grid is found to be 3 mm. The 3D fabric in principle demonstrates good handling capabilities. However, in the absence of coating it exhibits limpness.

Figure 13.

Integrated warp yarn graded delivery unit (a) Outer pipe radius, (b) Inner pipe radius.

An analysis of the experimental production clearly indicates that the concept of variable warp yarn delivery can be used for the realization of 3D spatial preforms. In addition, a synchronous differential doffing system is essential. The maximum preform size that can be produced on the machine is also be predicted by the computational model and is found to be 950 mm. The current results serve as a testimony for demonstrating the fundamental suitability of the concept in producing 3D net-shaped, stitch-bonded fabrics. However, the existing setup is geometry-specific and a flexible solution is still required for producing planar and 3D geometry on the same machine with no hardware modifications. This is essential for future technology standardization.

Modified solution for a flexible production of planar and three-dimensional fabrics

Based on the above results and analysis, a modified solution for the production of various geometries has been developed recently. The most significant modification here is the replacement of the graded variable warp yarn delivery unit with an NC-Drive-controlled warp yarn delivery unit. The warp yarns here are delivered in small groups through the NC-Drives. Although the machine gauge K remains fixed in both methods, the latter offers the possibility to produce both planar and 3D fabrics without any machine modifications. Thus, the full functionality of the machine is retained. Besides that, 3D fabrics with various curvatures can also be produced without any hardware changes, unlike the graded delivery unit. A modified synchronous doffing system based on differential doffing and adjustable slippage is also developed. Figure 14 shows the schematics of the new modified solution.

Figure 14.

Modified stitch-bonding machine concept.

The modified warp yarn delivery unit (Figure 15) consists of several small NC-Drive units, each carrying a set of warp yarns that can be mounted on a creel. The unit is driven by a synchronous servo motor. Each set of warp yarns is doffed by doffing roll units immediately after the stitch-bonding unit. These doffing roll units (Figure 16) are based on the “two-roll principle” similar to the previous doffing unit. The slip in the rolls can be set accordingly so that the doffing corresponds to the delivered warp yarn lengths. This ensures that the warp yarns are held in tight during the stitch-bonding process. A cutting unit guarantees that fabric is cut off immediately after the stitch bonding for an efficient doffing.

Figure 15.

Warp yarn delivery with NC-Drive unit.

Figure 16.

Differential doffing unit.

The mechanical integration and production of this new method is proposed on a larger and specially modified multi-axial stitch-bonded machine, Malimo 14024, from the company Karl Meyer Textilmaschinenfabrik GmbH.

Summary and outlook

The focal point of this research work is to achieve and demonstrate a single-step production of spatial 3D preforms to be used as textile reinforcements in fiber-reinforced composites. In order to have a sustainable and economically efficient production, a continuous process chain from CAD to manufacturing needs to be established. The currently existing mathematical model and the computations with respect to the various process variables allow an offline generation of the process data for the corresponding product and serve as an input in the production line. The viability of this model has already been demonstrated with the production of the sample Torus section. Based on this model, a fixed-graded warp yarn delivery unit and a synchronous doffing unit were developed and were found to be suitable for the production of the spatial 3D fabrics. In order to have geometry-independent production, a non-graded warp yarn delivery system (NC-Drive) and a differential doffing is required.

Through the proposed modified solution of an NC-Drive warp yarn delivery unit and a differential doffing unit, both planar and 3D fabrics can be produced with no hardware changes. The integration and realization of this solution will be carried out in the near future. The computational prediction accuracy of the existing mathematical model will also be enhanced by integrating the yarn property and the process parameters (machine speed, machine gauge). The realization of this forms the basis for an automated and cost-effective production of spatial 3D stitch-bonded fabrics. Thereby, the existing markets can profit from cheaper fiber-reinforced composites and new application areas can be acquired.

Footnotes

Acknowledgements

The authors are grateful to their colleagues Mr Martin Waldmann and Mr Karsten Trips for their valuable guidance and support.

Funding

This work was funded by the German Research Foundation at the Technische Universität Dresden (grant number DFG CH 174/20-1).

References

Gutowski

. Advanced composite manufacturing, Cambridge: Wiley Interscience, 1997.

. 3-D fibrous assemblies: properties, applications and modeling of three-dimensional textile structures, Manchester: Textile Institute, 2008.

Miravete

. 3-D textile reinforcements in composite materials, Cambridge: Woodhead, 1999.

Mungalov

Bogdanovich

. Complex shape 3-D braided composite preforms: structural shapes for marine and aerospace. SAMPE J 2004; 40: 7–21.

Büsgen WA. Neue verfahren zur herstellung von dreidimensionalen textilien für den einsatz in faserverbundwerkstoffen. PhD Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 1993.

Büsgen

. Shape weaving - new invention for producing 3D-shaped shells. Melliand Int 1995; 3: 183–185.

Chiu

Cheng

. Weaving method of 3D woven preforms for advanced composite materials. Textil Res J 2003; 73: 37–41.

Chiu

Cheng

. Weaving mechanisms for spiral woven fabrics. SAMPE J 2001; 37: 44–50.

Sengupta

Sarkar

. Anisotropy in some mechanical properties of woven fabrics. Textil Res J 1972; 42: 268–271.

10.

Cebulla

Diestel

Offermann

. Fully fashioned biaxial weft knitted fabrics. Autex Res J 2002; 2: 8–13.

11.

Hörsting K. Rationalising the production of long fibre reinforced composite materials using multiaxial multiply fabrics. Aachen: Shaker, 1998.

12.

Hausding

Engler

Franzke

. Concrete reinforced with stitch-bonded multi-plies – a review. In: Aldea

C-M

(eds). SP-244: Thin fiber and textile reinforced cementitious systems, Farmington Hills, MI: American Concrete Institute, 2007.

13.

Arnold R, Hufnagl E and Helbig R. Textile semi-finished fabric product, for the fiber reinforcement of plastics profiles. Intellectual Property Right DE 101 61 240 C1, Germany, 2003.

14.

Arnold R, Hufnagl E and Helbig R, et al. Textile blank for large-area fiber-reinforced plastic with integral reinforcements. Intellectual Property Right DE 101 61 241 C1, Germany 2003.

15.

Feltin D. Entwicklung von textilen Halbzeugen für Faserverbundwerkstoffe unter Verwendung von Stickautomaten. PhD Dissertation, Technische Universität Dresden, 1997.

16.

Meyer O and Gessler A. Fiber-patch preforming technology. In: SETEC 02/-07, proceedings of the SAMPE Europe technical conference and “table-top” exhibition, Universidad Politecnica de Madrid, 6–7 September 2007, pp.231–245.

17.

Pahl

Beitz

. Engineering design: a systematic approach, 7th edn. London: Springer, 2005.

18.

Teng

. Computation fabric mechanics – present status and future trends. Finite Elem Anal Des 1996; 21: 225–237.

19.

Cebulla H. Formgerechte zwei- und dreidimensionale Mehrlagengestricke mit biaxialer Verstärkung - Entwicklung von Maschine, Technologie und Produkten. PhD Thesis, Technische Universität Dresden, 2004.

20.

Cherif CH, et al. Development of a process chain for the realization of multilayer weft knitted fabrics showing complex 2D/3D geometries for composite applications. Textil Res J. Epub ahead of print 16 January 2012. DOI: 10.1177/0040517511429602.

21.

Chiu

Cheng

. Weaving method of 3D woven preforms for advanced composite materials. Textil Res J 2003; 73: 37–41.

22.

Lotzmann, H. Entwicklung zweidimensionaler konturgerechter und dreidimensionaler Fadenlagennähwirkstoffe. Diploma Thesis, Technische Universität Dredsen, 2007.