Development and testing of material extrusion additive manufactured polymer–textile composites

The adoption of Additive Manufacturing (AM) has gradually transformed the fashion industry through innovation and technology over the last decade. Novel AM systems and techniques are continuously being developed, leading to the application of AM polymers with textiles and fabrics in the fashion industry. This work investigates the development and testing of polymer–textile composites using polylactic acid (PLA) filaments on synthetic mesh fabrics using direct material extrusion (ME). An aspect of this paper highlights the appropriate combination of printing material, textile substrate, and printer settings to achieve excellent polymer–textile adhesion. Details of the printing process to create polymer–textile composites are described, as are the interfacial strength results of the T-peel tests, and the observed failure modes post-testing. The peel strengths for different ME bonded polymer–textile composites are examined and used to identify the compatibility of materials. This work visualised the potential of direct ME of polymers onto textile fabrics as a material-joining approach for new textile functionalisation, multi-material composite explorations and innovative aesthetic print techniques. This work also adds to the limited knowledge of AM polymer–textile composites, which can provide helpful information for designers and researchers to develop new applications and facilitate future research development in smart embedded and programmable textiles.

the fight against COVID-19, through the development of personal protective equipment (PPE) and other medical equipment (e.g. test swabs, ventilators) (Singh et al. 2020).
More precisely, the rise in the adoption of AM has led to a significant transformation of the fashion and textile industry through innovation and technology. One of the pioneers of this adoption is the influential designer, Iris Van Herpen. In 2010, she showcased her first 3D printed dress, which led to greater awareness and exploitation of the technology being employed in the fashion industry (Van Herpen 2010). The role of AM has continually evolved with increasing awareness and interest in the technology from researchers and designers. The number of research publications on "3D Printing Textiles" has continuously increased over the past few years with over 4000 publications on Google Scholar in 2019. This figure shows that AM will potentially open up new opportunities in fashion and textile innovation, promoting localised production of on-demand and personalised garments, allowing smaller batches or home production to compete in the market (Table 1).
The combination of digital manufacturing techniques gives the possibility for a textile to be three-dimensionally manufactured without tedious labour work, complex patterncutting, stitching, or the use of a specific mould. This approach also promotes a more environmental conscious and sustainable future for materials used in the fashion industry (Flynt 2019;Kim et al. 2019;Mageean 2018; Van der Velden et al. 2015;Zapfl 2019). However, the production of AM textiles is machine-intensive which require extensive understanding of the materials, the design and modelling programs, and the printing production process.
This paper focuses on the development and testing of material extrusion (ME) AM polymer-textile composites, which involves direct printing of thermoplastics onto conventionally manufactured textile fabric substrates. This novel material-joining technique highlights the synergy between conventional manufacturing processes and AM process to encourage a new vision of polymer-textile functionalisation and multi-material exploration in the textile industry. This study contributes to new knowledge and understanding of ME polymer-textile composites to facilitate future research development and integration of novel AM technology in textile design. For instance, Functionally Graded Additive Manufacturing (FGAM) with the integration of digital materials using PolyJet technology can offer sophisticated localised graded colours and different properties on a single piece of textile by varying the material organisation at a precisely defined area (Bader et al. 2016;Loh et al. 2018;Oxman 2011). New materials such as shape-memory materials can be used to create programmable or stimulus-responsive textiles that can transform or morph from one form to another when subjected to an external stimulus, known as 4D Printing (Leist et al. 2017;Pei and Loh 2018).
This paper discusses three key interconnected factors (i.e. printing material, textile substrate and printer settings) affecting the production and overall quality of the polymer-textile composites. This paper also gives details of the manufacturing process, as well as the experimental setup, procedures, and analysis techniques used to quantify the adhesion properties for different orientations of bonded ME printed polymer-textile composites. This study investigates the effect of varying textile substrate parameters (i.e. different fibre types, structure, and weights) on the polymer-textile adhesion force. The printing material used, and ME printing parameters were kept constant. Different combination of ME printed polymer-textile composites using PLA (printing material) and Nylon and Polyester (textile substrates) were manufactured to evaluate their manufacturing feasibility and assess their mechanical properties. The manufacturing demonstration and experimental results adds to the current limited knowledge of developing and testing of ME printed polymer-textile, which provide useful information for designers and researchers to facilitate further research and increased uptake towards industrywide applications.
Factors affecting ME printed polymer-textile composites ME as a category of AM process described in ISO-ASTM 52900 (ISO/ASTM 2017), often known as "Fused Deposition Modelling (FDM)" and "Fused Filament Fabrication (FFF)" is the predominant method of manufacturing polymer-textile composites (Chatterjee and Ghosh 2020). The AM process involves the material from a spool of filament loaded into the printer, melted above its glass transition temperature (T g ) for amorphous polymers and above its melt temperature (T m ) for semi-crystalline polymers. The polymer is then selectively dispensed through the heated extrusion nozzle and deposited onto the build platform at a predetermined location (Loh et al. 2020;Redwood et al. 2017). This technology of additively building up material by selectively dispensing through a nozzle or orifice allows AM parts to be built directly on the surface of the textile substrate. Sanatgar et al. (2017) described it as a thermal welding method for joining of the printing material (adhesive) and the textile substrate (adherent) during the ME process (see Fig. 1).
There are three major interconnected factors that affect the fabrication, polymertextile adhesion and the overall quality of ME polymer-textile composites (Loh and Pei 2019;Melnikova et al. 2014;Pei et al. 2015).
a. Printing material, b. Textile substrate, and c. Printer settings.

Printing material
ME makes use of thermoplastics in the form of filaments, which are typically 1.75 mm or 3 mm in diameter (Redwood et al. 2017). ME processes allow a wide variety of materials with diverse characteristics and properties to be used, ranging from commodities, engineering, to high-performance thermoplastics, composites, and functional materials (Loh et al. 2020). Table 2 gives details of some of the common thermoplastics used in ME, their key material characteristics, cost as well as their printing parameters including the nozzle and build platform temperature. The printing temperature, performance and cost of the materials increase through each material category from PLA to Polyetherimide (PEI) (Redwood et al. 2017;Rigid.Ink 2019). In general, the better the engineering properties of thermoplastics, the higher the temperature required to heat the material to a deformable state, and therefore, the more difficult the material is to print (Redwood et al. 2017). The use of materials with Fig. 1 Desktop ME printer (Original Prusa i3 MK3S 3D Printer) setup for ME onto the textile substrate lower printing temperature is recommended to avoid damaging or burning the textile substrate during direct deposition of the polymer.    Table 3 describes the variables of the textile substrates that affect the polymer-textile adhesion of ME printed polymer-textile composites. These properties include, but not limited to, the types of fibres, fabric weight, weave pattern, weft density and surface properties. These variables determine the type of print structure layout appropriate for the chosen textile substrate proposed in Table 4. Print layout A involves embedding the textile substrate between two print layers to form a laminated composite, while print  layout B involves a one-sided print, deposited directly on the textile substrate. Print layout A is suitable for printing on "open" mesh or perforated textile substrates, whereas print layout B is suitable for "closed" tightly woven textile substrates (Meyer et al. 2019).

Printer settings
Desktop Cartesian ME printer uses a system of X-Y-Z coordinates to determine the location of the extrusion nozzle, which allows direct ME onto the textile substrate. The setup is shown in Fig. 1 using an Original Prusa i3 MK3S ME machine with a single extruder. ME machines with multiple extruders can be used to create multi-material AM components, achieved by swapping the filament materials at a predetermined location or between layer changes. Table 5 identifies some of the ME processing parameters taken into consideration and the settings used during the fabrication of polymer-textile composites. These results are based on preliminary tests and literature review. The processing parameters include the Z-distances, printing temperature (Table 2), layer height, printing speed, fill settings, extrusion width, flow rate as well as build platform surface. The printer settings have a great impact on the visual and haptic finishing of the printed structure (Pei et al. 2015). The Z-distance has a significant effect on the adhesion of polymers to the textile substrate and quality of the print. An increment in Z-distance (build platform to extrusion nozzle adding fabric thickness) must be applied while printing on mid-weight to heavyweight or textured textile substrate to compensate the fabric thickness. An optimum Z-distance adjustment should prevent the extruder nozzle from getting caught on the fabric but close enough to press the extruded polymer into the textile substrate with no gaps between deposited parameters. In line with an optimum Z-distance between the nozzle and the build platform used, 0.1 mm and 0.2 mm layer height can usually provide good linear surface finishing with no scars on the top surface, messy first layer or gaps between infill and outline (Loh et al. 2020). A layer height greater than 0.2 mm exhibited a negative effect on dimensional accuracy and adhesion force (Spahiu et al. 2017). The printing temperature and printing speed have the largest effect on the adhesion force (Sanatgar et al. 2017). High nozzle temperature can reduce the material viscosity, allowing deeper and stronger material penetration into the textile substrate (Spahiu et al. 2017). For printing taller or larger components, the nozzle temperature can be adjusted back to the suggested temperature after five print layers on top of embedded textile to prevent overheating. Although Sanatgar et al. (2017) claimed that the build platform temperature does not affect the adhesion force, an optimum build platform temperature can provide better first layer adhesion to build platform and prevent warping. The extrusion width should be set at 100% or 150% greater than the default nozzle diameter (> 0.4 mm) in order to generate enough material to penetrate into the textile fabric (Spahiu et al. 2017). The study by Spahiu et al. (2017) also revealed that increasing the printing speed and polymer flow rate showed no substantial effect on the polymer-textile adhesion. Loh et al. Fash Text (2021)

Materials and AM process
This work explains the procedure of direct ME off-the-shelf PLA on selected mesh fabrics using print layout A. Three different combinations of polymer-textile-polymer composites were produced as shown below: a. PLA-Nylon (net structure)-PLA, b. PLA-Polyester (voile structure)-PLA, and c. PLA-Nylon (voile structure)-PLA.
The printing material was Prusa PLA filament with a diameter of 1.75 mm. PLA is cost-effective and relatively easier to print at a lower nozzle and build platform temperature, without burning the textile substrates. This is because PLA has a melting point of 260 to 270 °C (Callister and Rethwisch 2018). It has relatively low warping and stringing properties, leading to high detail finishing and better overall aesthetical quality (Tyson 2020c). The three different types of textiles substrates used were namely Nylon (net structure), Polyester (voile structure) and Nylon (voile structure). Table 6 gives the properties of the three lightweight mesh fabrics. The woven Polyester and Nylon voile shared relatively similar properties, comprising fabric thicknesses, non-stretch properties, fine pore sizes with smooth and sheer surface texture. On the other hand, the knitted Nylon net fabric had larger thicknesses (almost double), stretchable horizontally, relatively larger pore sizes of approximately 2 × 1.5 mm with rough surface texture. Adhesion affects the durability and quality of the final product (i.e. polymer-textile composite). Therefore, it was deemed appropriate to investigate the mechanical properties (via T-peel tests) of the bonded ME polymer-textile composites, to determine the optimum printing material and textile substrate combination and orientation. The ME polymer-textile composites designed for the T-Peel test were in line with British Standards (BS) EN ISO 11339 (2010) (Fig. 2). To manufacture the polymertextile composites using ME, the CAD design of the printed structure (L × W × H of 200 mm × 150 mm × 0.5 mm) was created using SolidWorks, exported as an STL.file, imported into Slic3r for slicing and exported as a G-code for printing. The polymertextile composites were manufactured using an Original Prusa i3 MK3S 3D Printer with a 250 mm by 230 mm build platform and a nozzle diameter of 0.4 mm, using the printer settings specified in Table 5. The nozzle temperature to print Prusa PLA was 220 °C, while the build platform temperature was set at 60 °C.
As far as the printing procedure for ME polymer-textile composites (print layout A) was concerned, the first or base PLA layer (0.2 mm layer height) was printed on the build platform using the calibrated Z-distance between build platform to extrusion nozzle. While the first layer print was almost complete, the nozzle height was increased to a new Z-distance to offset the fabric thickness. The printer was paused immediately once the first layer was completed. The Z-axis will automatically be lifted from the build platform by the system (Table 7A). To create the T-Peel test panel, a section of blue tape was applied on the surface of PLA layer to create a 50 mm unbonded area, separating with the subsequent print layers during the printing process (Table 7B). Thereafter, the textile substrate, which was cut prior, to match the size of the build platform was placed above the blue tape separator, secured, and tensioned using binder clips to remove any wrinkles or crease. It is extremely important to position the clips carefully to prevent any obstruction in the path (top-bottom and both sides of the built platform). Afterwards, the printer was resumed to complete the print (Table 7C).
The T-Peel test panel was printed as a whole sheet, then cut into six individual strips in 200 mm (L) × 25 mm (W) × 0.5 mm (H). Six T-Peel test panels for each polymertextile combination were created, producing a total of 18 specimens to be tested. The unbonded area was pull separated by hand to form a "T" angle for the T-Peel specimen to be fixed to the top and bottom clamps on the testing machine for the T-peel test (Fig. 2). The adhesion value of the separated section will not be considered in the result as the unbonded area of the T-Peel test panel was designed to be clamped on the universal testing machine. Therefore, the net structure belonging to the upper or lower part of the unbonded area would not affect the adhesion result of this experiment.

Test procedures
T-peel tests were carried out on the bonded ME printed polymer-textile composites to determine the peel force and peel strength required to separate the bonded polymers. Figure 3 shows a schematic diagram and dimensions of the bonded ME printed polymer-textile composites in line with BS EN ISO 11339 (2010).
All the specimens had the same nominal dimensions. The peel force was divided by the width (25 mm), to compute the peel strength. The tests were carried out using a universal testing machine (Instron 5969), which had a maximum load capacity of 500 N. The grips of the machine were used to secure the ends of the specimens so that they were subjected to uniform tension. The crosshead displacement was applied at a rate of 10 mm/min, based on the guideline given in BS EN ISO 11339 (2010). A digital camera was used to monitor the failure modes of the specimens. Table 7 The printing procedure involved in manufacturing ME polymer-textile composite (print layout A) to create T-Peel test panel 1. First layer calibration for optimum Z-distance between build platform to extrusion nozzle 2. Print first or base PLA layer using 0.2 mm layer height 3. Add nozzle height increment while the first layer print was almost complete 4. Pause printer immediately once the first layer was completed 5. Apply blue tape on the printed PLA layer 6. Place and secure the textile substrate above the layers using binder clips 7. Resume the printer to complete the subsequent print layers 8. Cut the T-Peel panel into six individual strips

Results and discussion
PLA in general printed well on Polyester and Nylon textile substrate with good linear and haptic finishing. Due to a limited volume of turquoise coloured Prusa PLA filament, a white coloured Prusa PLA filament was used to create the remaining test specimens. The material properties remained the same whist the material colour would not have an impact on the T-Peel test results. Figures 4, 5 and 6 show the test setup on and failure modes for the PLA-Nylon (net structure)-PLA, PLA-Polyester (voile structure)-PLA and PLA-Nylon (voile structure)-PLA orientations, respectively. The failure mode classification was based on BS EN ISO 10365 (1995). For the two orientations with Nylon net structure (Fig. 4) and voile structure (Fig. 6), both were the failure of an adherend, caused by the fracture of printed PLA layer (cohesive substrate failure). On the other hand, the orientation with Polyester voile structure showed an adhesion failure mode, delamination of printed PLA layer (substrate) from the textile, shown in Fig. 5c.   Fig. 10 shows a representative force versus extension responses for the three different combinations of polymer-textile composites. Similar to their failure modes, the force versus extension responses for PLA on Nylon net structure and voile structure were similar, far better results as compared to Polyester. For both PLA-Nylon composites, the initial force exceeded 40 N and included a few force peaks, up to a maximum extension of about 20 mm, which subsequently dropped leading to failure of the PLA polymer, reflecting a relatively stronger bond compared to PLA on Polyester textile. In comparison, the force versus extension responses for the PLA-Polyester (voile structure) composite showed an initial linear response until about 7 N. After that, there were several force peaks (i.e. undulating curve), reflecting the gradual separation of the Polyester textile from the PLA polymer, until the end of the test. For all the PLA-Polyester (voile structure)-PLA T-Peel specimens, there was no damage to both printed polymer layer and textile substrate.
Following the recommendation given in BS EN ISO 6133 (2015), the average peel force was determined based on the midpoint of the minimum and maximum peak force values, whilst ignoring the initial rise at the start of the test. Table 8 gives the average peel forces, peel strengths and coefficients of variation for the three bonded ME printed polymer-textile composites. The peel strengths for the PLA-Nylon (net structure)-PLA, PLA-Polyester (voile structure)-PLA and PLA-Nylon (voile structure)-PLA polymer-textile composites were 1.3 N/mm, 0.4 N/mm and 1.4 N/mm, respectively.
The results show that the average peel forces and strengths for both PLA-Nylon composites (net and voile structure) were about three times greater than PLA-Polyester composite (voile structure), which explained the breaking of the deposited layer at the beginning of extension in all samples during the T-Peel test. The statistical analyses show that the PLA-Nylon (voile structure)-PLA orientation had the lowest coefficient of variation of 9% and the PLA-Polyester (voile structure)-PLA orientation had a coefficient of variation of 18%, representing a relatively larger variation across the tested specimens.
PLA printed on Polyester textile did not show high peel strength result, which revealed that both materials were less compatible. According to the diffusion theory, the fine pore properties of voile structure decrease the amount of infiltration between the two polymer layers for polymer-polymer adhesion. As a result, the deposited polymer cannot protrude through the textile layer to create a form-locking connection (Eutionnat-Diffo et al. 2019;Sabantina et al. 2015;Unger et al. 2018). However, this theory was challenged when comparing the results obtained from both voile structures and the two PLA-Nylon composites. Despite both nylon textiles have different mesh structure (net and voile), weave type, thickness and pore sizes, there were no substantial differences on their peel force and strength. It can be concluded that the fibre type has a predominant effect on the interfacial bonding strength between the printing material and textile substrate due to the chemical nature of both and interpolymer polar interactions (Van der Waals dipole-dipole interactions) across phase boundaries as explained by Sanatgar et al. (2017). The compatibility between the printing material and the textile substrate fibre type has a significant effect on the polymer-textile adhesion. With respect to the textile stretchability, there was no direct and substantial effect of the textile stretch on the peel resistance, nevertheless, working with low level or non-stretch textile substrate improves the ease of printing process. It can be equally stretched in both vertical and horizontal directions to be secured on the build platform and no pre-strain to cause irregular pore circularity and pore area which will cause an inconsistent amount of infiltration of the printed polymer at the time of printing, which will correspondingly affect the peel strength (Narula et al. 2018).
Concerning the limitations of this study, the ME printing production process is demonstrated and tailored for Cartesian XZ hot end, Y bed ME desktop printer (i.e. Prusa i3 MK3S). Although the manufacturing concept is similar, there will be slight differences in the calibration and printing steps to accommodate the function of each ME printers, such as the Cartesian XY hot end, Z bed ME desktop printer (I.e. Ultimaker) (3D Printing Beta 2020). This study examined the mechanical and adhesion properties of ME polymer-textile composites using basic structure design to meet the standards' requirement BS EN ISO 11339 (2010). Future work could explore printing different geometrical structures on more variety of textile substrates of different properties (i.e. weight and texture). PLA with different performance, mechanical properties and flexural characteristics mentioned in Table 2 can be explored. For instance, PLA Plus+ can be used for greater mechanical performance and resistance than regular PLA and has lower printing temperature compared to ABS and PETG (Gregurić 2020). Flexible PLA can be used to create soft and flexible prints that can drape according to the fluidity of the textile fabric (Griffin 2019).

Conclusion
In this paper, the three key interconnected factors (the printing material, textile substrate, and printer settings) which affect the production, printed quality and adhesion strength of the polymer-textile composites were discussed. The experimental setup, procedures, and analysis techniques to quantify the adhesion properties of polymer-textile composites have been described, and the results were compared and discussed.
This study investigated the influence of varying textile substrate parameters (i.e. different fibre types, structure, and weights) on polymer-textile adhesion force. The printing material used, and ME printing parameters were kept constant. Different ME printed polymer-textile composites were manufactured using PLA (printing material) and Nylon and Polyester (textile substrates), to evaluate their manufacturing feasibility and assess their mechanical properties. The ME printed polymer-textile composites included (a) PLA-Nylon (net structure)-PLA; (b) PLA-Polyester (voile structure)-PLA; and (c) PLA-Nylon (voile structure)-PLA. Based on the results from the T-peel tests, it can be concluded that the compatibility between the printing material and the textile substrate fibre type has a dominant effect on the peel resistance of ME polymertextile composite. The average peel forces and strengths for both printed PLA on Nylon textiles composites were nearly three times stronger than Polyester textile despite the differences in in their mesh structures, pore properties and weave type.
Finally, the work reported in this paper has not only added to the current limited knowledge of developing and testing of ME printed polymer-textile but also demonstrated and visualised the potential of direct ME of polymers onto textile fabrics as a material-joining technique for the development of new textile functionalisation and multi-material composite explorations. This new AM technique can exploit and promote new visions of personalised, localised and scalable garments with added functionalities, and boost the uptake of innovative and sustainable models for the textile industry. The principles and procedures uncovered from this study can be also applied for new applications or be extended to necessitate future research textile development.