Comparison of FDM and SLA printing on woven fabrics

ABSTRACT


Introduction
Additive manufacturing, also called 3D printing, belongs to the most interesting technologies of the last years.It enables producing objects in small numbers at reasonable prices as well as creating shapes that could not be produced by other techniques.On the other hand, most 3D printing techniques still have problems with relatively slow production and inferior mechanical properties to injection-molded parts [1].These mechanical problems are mostly based on air void and insufficient bonding between adjacent layers, but also on lower mechanical properties of the base materials for printing [2,3].The latter can be improved by developing new filament materials or adding nano-fillers [4,5].Alternatively, fibers or yarns can be included in the 3D printed objects [6,7].
Going one step further, several research groups investigated composites of polymers directly printed on textile fabrics in the last decade.Mostly, FDM printing is used for these approaches.The fiber-matrix adhesion depends on diverse parameters, such as the textile material and structure [8][9][10], nozzle and printing bed temperature [11,12], and strongly on the z-distance between nozzle and printing bed which defines the force with which the polymer is pressed into the textile fabric [13][14][15].The latter is necessary due to the high viscosity of the molten polymer in FDM printing which impedes penetrating into a textile fabric.
To overcome this challenge, a few attempts have been made in the last years to use low-viscous photopolymer-based resins to form textile-resin composites.Firstly, this was proven to be possible for stereolithography (SLA) [16], while more recently, PolyJet modeling (PJM) was also shown to enable direct printing on textile fabrics [17,18].For the SLA printing on textiles, however, no adhesion tests have been reported yet.
Here the adhesion reached by FDM and SLA printing on three different polyester (PES) woven fabrics is shown.The influence of printing parameters on the adhesion is discussed, and the advantages of SLA printing in case of densely woven, thin fabrics are underlined.

Materials and methods
The woven polyester fabrics used in this study are described in Table 1.They were chosen to compare fabrics of different thickness and surface roughness, while the textile material was kept constant.All fabrics were highly hydrophilic, impeding water contact angle and resin contact angle measurements due to very fast penetration of the drops into the woven fabrics.To perform adhesion tests, rectangles of 25 mm x 100 mm were printed on these textile fabrics.
For FDM printing, an Ender V2 (Shenzhen Creality 3D Technology Co., Ltd., Shenzhen, China) was used with a nozzle size of 0.4 mm.Printing of two layers was performed with the parameters given in Table 2, while the z-distance was varied to find the optimum value.The FDM filament was poly(lactic acid) (PLA) (Grauts GmbH, Löhne, Germany), which was found to have a higher adhesion than some other rigid FDM materials in previous studies [19,20].The textile fabrics were rigidly fixed on the printing bed using common adhesive tape, as shown in Fig. 1a.Printing starts with a skirt (the outer blue print in Fig. 1a), followed by the perimeters of the actual sample (inner blue part in Fig. 1a) which is subsequently filled (Fig. 1b), until the second layer is printed (Fig. 1c).SLA printing was performed with an Anycubic Photon S printer (Shenzhen, China) using grey "ABS-like Resin" (Anycubic) and the printing parameters given in Table 3.An overall printing thickness of 2 mm was chosen to avoid breaking of the samples during the adhesion tests, which were more brittle than the PLA specimens.The textile fabrics were fixed on a custom-made fabric holder to enable printing on them [21].All parts were washed in isopropanol for 4 min and cured under UV light for 14 min after the printing process (Wash & Cure station, Anycubic).Figure 2 shows a woven textile in the sample holder before/after printing.The adhesion of the 3D printed parts on the textile fabrics was tested by a Zwick/Roell tensile tester Z010 according to DIN 53530, as shown in Fig. 3, and evaluated according to ISO 6133.This standard describes a separation test on fabric plies bonded together.It is used to measure the adhesion force between two layers, in this paper between the textile fabric (fixed in the lower clamp in Fig. 3) and the polymer printed on it (fixed in the upper clamp), while the distance between the clamps is continuously increased.The relatively stiff SLA printed samples are held on one side of the clamps to avoid bending them too much at the beginning of the test, while the thin PLA sample shown in Fig. 3 is flexible enough to bend without any influence on the measured adhesion forces.Images were taken by an iPhone XR camera, a digital light microscope Camcolms2 (Velleman, Belgium) and a confocal laser scanning microscope (CLSM) VK-8710 (Keyence, Neu-Isenburg, Germany).

Results and discussion
The results of the adhesion tests for FDM printing on different PES woven fabrics as well as some printed samples are depicted in Fig. 4. Here, the adhesion forces are given in dependence on the zdistance between the nozzle and the printing bed (Fig. 4a).A z-distance of 0 mm means that the nozzle touches the printing bed, while it would typically have a z-distance of ~ 0.15 mm if the polymer should be placed directly on the printing bed.The values d1 and d2 show the thicknesses of the plain weave/twill fabrics and the Leno fabrics, respectively, i.e. at the distance d2 the nozzle just touches the Leno fabric, while it touches the other fabrics at a distance d1.Firstly, the adhesion forces of PLA printed twill 2/1 and plain weave fabrics (blue and green lines) shall be discussed.All of them are very low, maximally 1 N. Similarly low adhesion values have been found before for FDM printing on very thin fabrics [13].Here, however, the woven fabrics are not only relatively thin, but also quite densely woven, so that nearly no pores are available in which the PLA could be For a comparison of the apparent weave density, Fig. 5 depicts microscopic images of the textile fabrics under examination.While none of them shows large open pores, the weft threads the Leno fabric appear broader and thus easier movable to allow the molten polymer being pressed into the fabric.Indeed, the adhesion on the Leno fabric is much higher than on the other fabrics, which can also be attributed to its larger thickness, as shown in previous studies [9,13].For the adhesion tests on this Leno fabric, printing was performed in different z-distances between nozzle and printing bed.The fabric thickness of 0.48 mm, marked in Fig. 4a as "d2", indicates the starting point of the adhesion, i.e. printing with the nozzle not touching the fabric results in zero adhesion force.Printing with the nozzle being positioned lower and lower, across z = 0.0 where nozzle and printing bed would touch without the fabric between, leads to increasing adhesion forces, until a value of approx.-0.3 mm is reached.Here, the counterforce from the fabric starts clogging the nozzle, and for even lower nozzle settings, the adhesion force is decreased again.While these findings have been reported similarly before [13][14][15], here a clear difference between both sides of the fabric is visible which indicates that not only thickness and porosity, but also the surface structure may influence the adhesion of an imprinted polymer.
Besides these adhesion measurements of FDM printed PLA on the fabrics under examination, the same tests were performed for SLA printed resin on these fabrics.In this technique, there is no optimization possible of the first layer position; a potential optimization by modifying the first layer thickness or curing time will be tested in the near future.The adhesion forces of the SLA printed samples as well as some exemplary SLA printed samples after the adhesion tests are given in Fig. 6, together with the highest values reached with FDM printing at the optimum z-distance.Even without any optimization approaches, the adhesion forces for SLA printed material on the textile fabrics are higher than the values found for FDM printing under optimized conditions.It is especially remarkable that both the plain weave and the twill fabric reached an adhesion force around 15 N, while less than 1 N was found for the adhesion between FDM-printed PLA and these fabrics.
These differences can mainly be attributed to the different viscosity of the molten PLA and the SLA resin, leading to a much better penetration of the latter into the textile fabrics.This is visible in Fig. 7, showing the back of the SLA printed fabrics after the adhesion tests.In all dark areas, there is a large amount of hardened resin at the back of the fabric which had completely flown through the textiles, in this way resulting in a very good form-locking connection.The differences between the adhesion values for the different textiles can probably be interpreted as differently large connection areas between the resin on both sides of the fabric, i.e. the adhesion forces may actually be breaking forces of the resin connections.This idea, however, cannot be verified with the recent samples, but needs samples with well-defined pore sizes, which will be tested in a forthcoming study.Besides adhesion measurements, CLSM images were taken of the surface structures of the printed fabrics, as depicted in Fig. 8.The highest waviness is visible for PLA printed on the Leno fabric with a zdistance of -0.35 mm (Fig. 8a,d), as can be expected due to the very thin first layer which partly penetrates into the fabric, so that the surface structure is similar to the first layer on a printing bed.For a relatively large z-distance of +0.20 mm (Fig. 8b,e), the neighboring strands are nearly interconnected, so that the waviness is nearly vanishing, while the roughness here is much higher.Finally, the SLA print (Fig. 8c,f) shows a smooth surface with very low waviness and roughness.This corresponds to the wellknown high resolution of SLA printing, as compared to FDM printing.Next, Fig. 9 shows cross-sectional images of these samples.For PLA printed at a low z-distance of -0.35 mm (Fig. 9a), the pink printed polymer is partly visible between the yarns of the Leno fabric, and the surface roughness is again clearly visible.
Printing PLA at a much larger z-distance of +0.20 mm (Fig. 9b) results in significantly thicker printed layers, and air voids between printed polymer and fabric are visible at some positions.Finally, the greyish ABS-like resin has fully impregnated the fabric (Fig. 9c), visible here as straight lower border in contrast to the fibrous fabric, as it could already been observed in Fig. 7.

Conclusions and outlook
In a recent study, 3D printing on different polyester woven fabrics was performed using FDM and SLA printing.By FDM printing of PLA filament, nearly no adhesion could be reached on both thinner and less porous textiles.For the thicker Leno fabric, the usual dependence of the adhesion on the z-distance between nozzle and printing bed was found.
Oppositely, SLA printing on these fabrics led to a significantly increased adhesion for all textiles.This could be attributed to the full penetration of the low-viscous SLA resin into all woven fabrics.
In the next studies, optimizing the adhesion of SLA-printed layers by varying the first layer thickness and curing time is planned as well as an investigation of the impact of the textile pore sizes on the adhesion force.

Fig. 1 FDM
Fig. 1 FDM printing on a woven fabric.(a) Starting with skirt and perimeters, (b) filling the first layer, (c) printing the second layer.

Fig. 2
Fig. 2 SLA printing on a woven fabric.(a) Fixing the fabric in the custom-made fabric holder, (b) fabric holder inserted in the SLA printer, (c) printed sample on fixed textile fabric before washing and curing.

Fig. 3
Fig. 3 Subsequent stages of an adhesion test according to DIN 53530.

Fig. 4 (
Fig. 4 (a) Adhesion forces gained by FDM printing on different textile fabrics.The distances d1 and d2 correspond to the thicknesses of plain weave/twill weave (d1) and Leno fabric (d2), respectively; (b) photographic image of samples after printing.

Fig. 6
Fig. 6 Adhesion forces gained by SLA printing on different textile fabrics.The maximum values of FDM printing (cf.Fig. 4) are added for comparison; (b) photographic image of samples after testing.

Fig. 7
Fig. 7 Back of (a) plain weave, (b) twill 2/1, and the Leno fabric printed (c) from the back and (d) from the face, after adhesion tests.In the dark areas at the bottom, resin and fabric are not separated.

Table 1 .
Woven fabrics used in this study.