Effect of bending and padding on the electromagnetic performance of a laser-cut fabric patch antenna

Abstract

A fabrication and characterization procedure is detailed for a flexible planar antenna integrated into textiles by interfacing thin metal-coated fabric sheets on a polyester fabric substrate. From the full-wave electromagnetic simulations and measurements, it is observed that the low dielectric dissipation in the porous woven polyester enables the fabric antenna to achieve a high gain of 8.4 dBi. It is comparable to other antennas fabricated with engineered substrates of low-loss polymer composites. Using this antenna, the impact of cylindrical concave bending deformation is observed in terms of the impedance matching and radiation performance. The simulated and measured results agree reasonably well. A 1.2% frequency shift is observed when the antenna is bent concavely along its length, while bending along its width showed only a marginal impact. On the other hand, the gain is reduced by as much as 1.0 and 0.5 dB when the antenna is bent along its length and width, respectively. The impact of padding layers was also investigated when placed above the radiating patch and below the ground plane. Because the textile padding layers have complex permittivity closer to air due to their highly porous structure, it is expected to observe only small influence on the radiation performance. However, the simulations and measurements show that padding the radiating patch lowers both the operating frequency and the realized gain by up to 1.6% and by up to 0.9 dB, respectively, due to dielectric loading and dissipation.

Keywords

textile antenna conformal patch antenna protective covering electromagnetic simulation antenna measurement

With growing interest in wireless body area network (WBAN) technology for military, aerospace, healthcare, and entertainment applications,^1–3 planar antennas are being widely investigated for wearable electronics since they can be easily implanted into clothing.^4–6 In addition to the ease of integration, certain textile substrates, such as polyester, Lycra®, Quartzel®, etc., have dielectric properties⁷ desirable for the fabrication of antennas. The porous nature of these substrates results in a lower dielectric dissipation and a dielectric constant close to that of air, which enables construction of antennas with high gain and high efficiency.

The fundamental patch antenna consists of a rectangular radiating patch mounted on a grounded substrate. In order to realize this geometry with textile materials, previous studies incorporated foils,⁸ conductive threads,^4,5,9–13 and plated fabrics^1,14 as the conducting constituents of antennas. Although foils have the best electrical conductivity among these options, they gradually develop microcracks after the repetition of conforming and stretching and thus may not be suited for wearable applications. On the other hand, textile conductors constructed by weaving,^5,9–11 knitting,¹³ or embroidering¹² conductive threads could offer a better fabric hand and seamless integration into the finished products than the foil. However, because of the inferior mechanical flexibility of the conductive threads, a high packing density is challenging to attain, and thus the resulting conductors exhibit insufficient electrical conductivity for a high-performance antenna. Conventional woven fabrics coated with metal using an electro- or electroless plating method are simple in fabrication and offer sufficiently low sheet resistance^15–17 and great mechanical flexibility.¹⁶ Also, thermoplastic polymer-based conductive fabrics are generally compatible with laser cutting, and highly accurate and clean edges can be obtained from laser cutting without an additional finishing step.¹ Thus, metal-coated thermoplastic fabrics have advantages for high-efficiency antenna applications.

To compose a dielectric substrate, knits,¹³ wovens,^4,5 and nonwovens^1,14,18,19 have been studied for their dielectric properties and used to design patch antennas in the literature. Knit fabrics are advantageous in offering excellent stretchability and flexibility resulting from their loop structures, but woven and nonwoven fabrics, being interlaced or bonded, provide a higher stability and geometrical accuracy as a frame of the antenna, and thus may be preferred for patch antennas.⁷ In selecting fibrous polymers for the antenna dielectric, hydrophobic polymers such as poly(ethylene terephthalate) (PET) and poly(ethylene) (PE) are often chosen, because hygroscopic polymers such as polyamides (PAs) and many natural polymers, including cellulose and polypeptides, absorb moisture from the atmosphere, causing instability in the electrical permittivity during antenna operation.²⁰

When embedded into clothing, textile antennas are subject to a variety of physical conditions, including bending, stretching, and compressive strain caused by movement of the wearer. In wearable antenna design, the directions of these loads along with their magnitudes must be taken into consideration due to sensitivity of the antenna parameters to flexing. It has been reported that both impedance matching and radiation performance of textile patch antennas are highly sensitive to the magnitude and direction of flexing. For instance, the cylindrical bending in certain directions alters the effective length of the radiating patch – specifically, a patch convexly bent along its length could demonstrate a shortened effective length, while a patch concavely bent along its length would show an extended effective length.²¹ Also, another research⁵ reported that as the radius of the cylindrical curvature decreased, a larger frequency shift and a more critical degradation in the radiation pattern was observed due to a more severe change in the effective length of the convex patch, indicating a great sensitivity of the textile antennas to conformal deformation.

Another important consideration in wearable antenna development is the placement of additional insulating layers on both sides of the antenna for packaging purposes. This type of treatment is especially important to protect the wearable antenna components from abrasive wear and tear as well as skin secretions. In the literature,^22–25 various kinds of conformal coatings and laminates have been studied for the electromagnetic (EM) influence and it has been reported that both impedance matching and radiation performance of the patch antennas could be impaired by coatings and laminates.

The current research features a textile patch antenna fabricated from a polyester woven fabric. Polyester has the largest share in the global fiber market²⁶ and is one of the most common textiles worldwide. Polyester was chosen as the dielectric substrate due to its dielectric stability²⁰ under the presence of moisture. Also, the thermosetting property of polyester enables antenna fabrication with clean edges by laser cutting. Another Ni/Cu-plated polyester woven fabric was chosen as an antenna conductor. The reasons for this selection were not only accurate patch fabrication capability with laser cutting,¹ but also a good mechanical flexibility and an adequate electrical conductivity.⁷ In order to mount the conductive polyester layers onto the dielectric substrate, a PA fusible tape was used. This bonding method provides a simple but secure adhesion of the fabric layers without causing electrical shorting.⁷

While the impact of bending on film-based antennas has been intensively studied in the literature, few reports have been published on the EM performance of a concavely bent antenna built with textile materials. In particular, the study of the radiation performance is a new topic for concave fabric patch antennas. Thus, aiming at the development of model systems, this work investigated the effect of the concave bending on the impedance matching and radiation performance of the fabric patch antenna.

Another unique component of this research is the incorporation of additional fabric layers for antenna packaging and protection purposes. The fabric padding could provide not only a better wearability and comfort than the conformal coatings and laminates, but also low-profile integration into clothing with a natural look. Moreover, with a low electrical permittivity originated from their porous structures, fabric padding could offer a better microwave transparency. For these reasons, fabric padding could be advantageous for smart clothing applications. Nevertheless, the EM influence of the conventional fabrics for padding has not yet been reported in the literature. Although textile fabrics generally have very small electrical permittivity, dielectric loading^22–25 could still have some influence on the impedance matching and radiation pattern. Thus, this research proposes the possibility of using fabric layers as a packaging component by investigating the influence of those layers on the antenna performance.

Therefore, the two goals of this research are to provide comprehensive analysis on the effect of the cylindrical concave bending and the padding on the impedance matching and radiation performance of the textile patch antenna. Firstly, cylindrical concave bending was investigated by bending the fabric patch antenna along two bending axes – along the length and width with three radii. Then, the impact of padding of the antenna in its near-field region was investigated by placing additional layers of woven polyester on the radiating patch and the ground plane. In both experiments, the changes in the operating frequency and the realized gain were examined through EM simulations and measurements. The three-dimensional (3D) full-wave EM simulations were performed using a commercial finite element method (FEM) solver, and the measurements were carried out using a vector network analyzer placed in an anechoic chamber.

Experimental details

Materials

A 0.08 mm-thick Ni/Cu-plated polyester ripstop fabric (Less EMF Inc.) with a sheet resistance of 0.03 Ω/sq. and a fabric weight of 90 g/m² (density of 1,125 kg/m³) was selected as a conductor for radiating and ground planes. For the dielectric substrate and the padding layers, a 0.32 mm-thick polyester plain-weave fabric with a fabric weight of 192 g/m² (density of 600 kg/m³ and 56% porosity) was used. This fabric was composed of 64-tex warp (45 ends per inch) and 57-tex weft (33 picks per inch) yarns. A PA fusible web tape (Bostic Inc.) was used to mount the conductive sheets on the dielectric substrate. A panel-mount SubMiniature version A (SMA) connector (Pasternack Enterprises Inc.) was used to excite the antenna using the direct feeding method.

The dielectric properties of the polyester fabric were characterized by the resonant cavity method.^27–29 The resonant cavity method employs a copper cylindrical cavity to estimate the dielectric properties of a material. By filling the cavity with the polyester fabric, the resonant frequency for the TM₀₁₀ mode was observed with an Agilent E5071C ENA Series Network Analyzer. The dielectric constant (real part of the relative permittivity; ɛ_r’) was calculated²⁷ by observing a shift in the resonant frequency between the air-filled and substrate-filled cavity (shown in Figure 1), while the loss tangent (dissipation factor; tanδ) was calculated by observing the reduction in the quality factor (Q) of the cavity after introducing the substrate. The measured dielectric constant and the loss tangent were 1.55 and 0.0087, respectively, at 2.26 GHz. Although these values are significantly smaller than those of bulk PET,³⁰ the differences came from the presence of air in the fabric structure.^7,31–33 These characterized dielectric properties are assumed to be constant across a small (less than 10%) variation in the frequency.

Figure 1.

The woven polyester fabric filled into the cylindrical resonant cavity for permittivity characterization.

Antenna design and simulation

The dimensions of the designed antenna are shown in Figure 2. The antenna consists of a rectangular radiating patch, a ground plane, 10 layers of polyester fabric as the dielectric substrate, and a coaxial feeding pin. This geometry was simulated and designed using an EM simulator (HFSS, ANSYS Inc.) under the given condition of the conductor and dielectric properties. To match the antenna impedance to the 50 Ω source, a feed inset was located 12 mm from the edge of the patch.

Figure 2.

Schematic diagram (not to scale) of the designed multi-layer fabric antenna in the (a) perspective, (b) top, and (c) side views. PET: poly(ethylene terephthalate).

The optimized antenna had a considerable (3.2 mm) thickness. Generally, the dielectric thickness is one of the most critical factors for high-performance planner patch antennas, and a thicker substrate could help to decrease conductor loss and broaden the bandwidth.²⁹ Therefore, the stacked structure of fabric layers was designed to achieve a reasonable separation between the patch and the ground plane.

Bending model simulation

To investigate the effect of bending on the operating frequency and the radiation pattern of the antenna, simulation models with a variety of curvatures were generated. Figures 3(a) and (b) show these models designed to conform antennas to the cylindrical concave surfaces. The different radii (r) of curvature used for this conformation were 105, 85 and 70 mm. The selections of these bending conditions did not intend mimicking specific situations during wear accurately, but rather to serve as model systems for analysis purposes. As depicted in Figures 3(a) and (b), the designed antenna was bent about two different bending axes – the x-axis (along the patch width) and the y-axis (along the patch length).

Figure 3.

The simulation setup for (a) bending along the patch width (about the x-axis), (b) bending along the patch length (about the y-axis), (c) padding above the patch, and (d) padding above the patch and below the ground simultaneously. PET: poly(ethylene terephthalate).

Padding model simulation

To create simulation models of the padded antenna, the same polyester fabric that was used as the antenna substrate was placed on the antenna. Padding of one, four, and six polyester fabric layers was added as a protective component on the top of the rectangular patch (P1, P4, and P6), and both on the patch and the ground plane (P1G1, P4G4, and P6G6). These simulation models with padding are depicted in Figures 3(c) and (d).

Antenna fabrication

The Ni/Cu-plated conductive fabric and the polyester substrate were prepared in the dimension acquired during the antenna simulation (Figure 2). In order to achieve the required thickness of 3.2 mm, 10 layers of polyester fabric were joined together by a lockstitch sewing machine (ISO 301) along the grain direction. The same procedure was applied to prepare padding having four and six layers of the polyester fabric. The distance between each sewing line was approximately 5 mm, and the stitch count was 10 stitches per inch (SPI). The thread tension was adjusted during stitching to have the correct amount of stitch tension in the sewn fabrics.

Then, the stitched polyester fabrics and the Ni/Cu-plated conductive fabrics were cut with a laser cutter (Universal Laser Systems VLS 6.60) into the given dimensions. This cutting technique allowed the preparation of the fabric pieces with accurate and fray-free edges. Both the polyester substrate and the conductive fabrics were immersed in water before cutting to avoid burn marks on the edges.¹ The power and speed settings were adjusted to load the minimum amount of energy required for cutting, as given in Table 1.

Table 1.

Optimized power and speed settings for the laser cutter

Materials	% power	% speed
10-layered polyester fabric (wet)	100	4
6-layered polyester fabric (wet)	60	8
4-layered polyester fabric (wet)	40	10
Single-layered polyester fabric (wet)	20	30
Ni/Cu conductive fabric (wet)	50	50

The laser-cut conductive fabrics were mounted on the dielectric substrate. A PA fusible web tape was applied to attach the conductive sheets to the dielectric polyester. This interfacing method provided a simple but secure and uniform mounting of the conducting planes onto the dielectric layer. Finally, a panel-mount SMA connector was soldered to the patch and ground planes. Figure 4 shows a sample of the fabric patch antenna.

Figure 4.

Fabricated antenna sample with stacked polyester fabrics (substrate), conducting patch, and ground planes.

Measurements

The antenna performance was characterized in terms of the reflection coefficient and the radiation pattern. The reflection coefficient was measured through the scattering (S)-parameters using a calibrated vector network analyzer. The realized gain was calculated from the radiation pattern measured in the E-plane (xz-plane) and the H-plane (yz-plane) inside an anechoic chamber at every 10° increment at 2.45 GHz. For the measurements under the different bending conditions, curvature templates were made by 3D printing of poly(lactic acid) (PLA) using a Fusion3 F400 3D Printer (Fusion3 Design, LLC), as shown in Figure 5(a), to control the radius of curvature into three different levels – 105, 85, and 70 mm. Those templates were placed next to the ground plane to minimize their impact on the measurements. For the measurements with padding layers, the fabricated antenna was covered by one, four, and six layers of polyester fabrics. Six different testing conditions were implemented: covering the radiating patch with one, four, or six layers (P1, P4, P6) and covering both the patch and ground plane with one, four, or six layers (P1G1, P4G4, P6G6). Finally, the Pearson correlation coefficient³⁴ was calculated to evaluate the correlation between the simulated and measured data sets.

Figure 5.

(a) Three-dimensional printed concave curvature template with radii of 105, 85, and 70 mm, and antenna samples bent about (b) the x-axis and (c) the y-axis.

Results and discussion

Performance of the antenna

Figure 6 shows the simulated and measured reflection coefficients of the designed antenna. The impedance of the fabricated antenna was well matched to the source impedance of 50 Ω at the operating frequency of 2.45 GHz with a S₁₁ of –24.9 dB compared to the simulated value of –34.1 dB. The measured fractional bandwidth of the antenna (|S₁₁| ≤ –10 dB) was 3.57%, while the simulated value was 4.08%.

Figure 6.

Simulated and measured reflection coefficient of the fabric antenna.

Successful impedance matching at the target frequency of 2.45 GHz (Figure 6) verifies that the design of the patch antenna is valid and well-fabricated, as expected under the given conditions. For planner patch antennas, the operating frequency is predominantly determined by the dielectric properties.²⁹ Thus, from a good agreement between the simulated and measured operating frequencies, it was confirmed that the packing densities of polyester in the resonant cavity and in the antenna sample were close to each other, and the permittivity of the fabric substrate was well-characterized by the resonant cavity method.

The co-polarized E-plane (xz-plane) and H-plane (yz-plane) radiation patterns measured at the operating frequency of 2.45 GHz are shown in Figure 7. The realized gain of the antenna was 7.69 dBi in simulation, whereas the realized gain of the fabricated sample was measured 17% higher at 8.39 dBi. The calculated efficiency of this antenna was high (82%), and one attribute for this high efficiency would be the low dielectric constant and low-loss tangent of the dielectric polyester. The measured 3 dBi beamwidth of the antenna was 80° for the E-plane and 70° for the H-plane.

Figure 7.

The simulated and the measured co-polar radiation patterns of the fabric antenna in (a) the E-plane and (b) the H-plane at 2.45 GHz.

Comparing the realized gain values in the simulation and measurement, we observed that measurement acquired a 0.70 dB higher gain than simulation. In general, the rectangular patch antenna has several different types of losses, and major losses include the dielectric loss, conductor loss, and surface wave loss, and all these losses could influence the antenna gain.^29,35 Since the permittivity of the polyester fabric was accurately characterized, as confirmed by the reflection coefficient measurement (Figure 6), it is probable that the conductor loss and/or surface wave loss were significantly higher in the simulation.

One possible explanation for the higher gain in the measurement is lower resistive loss in the antenna sample. The plated woven conductor used in the antenna sample could have a significantly higher surface area compared to the solid conductor designed in the simulator. Generally, at high frequencies, the volume where the electric current can flow is limited to very near the surface of the conductor (the skin effect).²⁹ Thus, a conductor of a larger surface area, such as the textile conductor used in this work, could allow more flow of current and thus generate less conductor loss compared to the materials of smaller surface area having the same direct current (DC) resistivity.

Effect of bending

Figure 8 shows the simulated and measured reflection coefficient of the same patch antenna when it was bent with the templates of varying curvature. The Pearson correlation coefficient of the simulated and measured operating frequencies (sample size n = 8) was 0.63, indicating a strong correlation. It was also calculated that the x-axis bending group (n = 4) alone returns a highly negative correlation (R = −1.0), while the y-axis bending group (n = 4) returns a highly positive correlation (R = 0.87). This could be interpreted that the negative correlation of the x-axis bending was masked by the positive correlation of the y-axis bending because the effect of the x-axis bending was much smaller than that of the y-axis bending. In fact, it was observed from the simulation and measurements (Table 2) that the operating frequency changed minimally, less than 1%, when bent along the patch width (about the x-axis). However, more obvious detuning was observed when bent along its length (about the y-axis). As shown in Table 2, the operating frequency was lowered gradually up to ∼4% as the bending radius decreased. Since the bandwidth of the antenna sample is less than 4%, this result indicates that bending along the patch length measurably impacts the impedance matching.

Figure 8.

(a) Simulated and (b) measured reflection coefficient plots with a variety of curvatures.

Table 2.

Shift in the operating frequency and the realized gain at 2.45 GHz on bending about the x-axis and the y-axis. Percentage (%) shift in operating frequency and gain in dB are calculated from comparison to the flat case

Radius (mm)	Bending axis	Simulated				Measured
Radius (mm)	Bending axis	Operating frequency (GHz)	Frequency shift (%)	Gain (dBi)	Gain change (dB)	Operating frequency (GHz)	Frequency shift (%)	Gain (dBi)	Gain change (dB)
∞ (flat)	–	2.45	–	7.69	–	2.45	–	8.39	–
105	x-axis	2.46	+0.41	7.23	−0.46	2.44	−0.41	8.30	−0.09
85	x-axis	2.46	+0.41	7.15	−0.54	2.44	−0.41	8.02	−0.28
70	x-axis	2.47	+0.41	6.85	−0.84	2.43	−0.82	7.93	−0.46
105	y-axis	2.44	−0.41	7.22	−0.47	2.43	−0.82	7.85	−0.54
85	y-axis	2.42	−2.45	6.98	−0.71	2.43	−0.82	7.60	−0.79
70	y-axis	2.40	−4.08	6.74	−0.95	2.42	−1.22	7.42	−0.97

These bending effects could be analyzed with the bending model and the transmission line theory in the literature. Galehdar and Thiel²¹ presented that convex bending could reduce the antenna length while concave bending expends the antenna length. As the resonant frequency ( f ) of the intact patch antenna (and hence its operating frequency) in the TM₀₁₀ mode has an inversely proportional relationship with the length of the patch based on the transmission line theory,^29,35 convex bending would increase the operating frequency, while concave bending would lower the operating frequency. Thus, our observation of the decreased operating frequency by bending along the patch length agrees with the previously proposed model.²¹ Another observation of the minor frequency changes with concave bending along the patch width is also supported by the transmission line theory that the width of patch has little impact on the resonant frequency.^5,29,35

Figure 9 illustrates the simulated and measured E-plane and H-plane cuts of the radiation pattern for the concave patch antenna at the frequency of 2.45 GHz. The Pearson correlation coefficient was 0.82, indicating a very strong correlation between the simulated and measured gains. As expected, the realized gain decreased as the bending radius increased. The simulation yielded the realized gain reduced by up to 0.84 and 0.95 dB by bending along the patch width and length, respectively (Table 2). This was verified with the measurement data shown in Figures 9(c) and (d), indicating a decrease of up to 0.46 and 0.97 dB in the realized gain by bending along the patch width and length, respectively.

Figure 9.

Simulated (a) E-plane and (b) H-plane radiation patterns, and measured (c) E-plane and (d) H-plane radiation patterns, monitored at 2.45 GHz with a variety of curvatures.

One of the major causes of the gain degradation could be the antenna detuning brought about by the frequency shift. However, because the reflection coefficients at the feeding port stayed within the acceptable range (S₁₁ ≤ –10 dB) even after the extreme bending cases, antenna detuning cannot be the only attribute for the 0.46 dB (r = 105 mm about the x-axis) and 0.96 dB (r = 105 mm about the y-axis) lowering of the realized gain. From the radiation patterns (Figure 9), it can be seen that bending the patch antenna also caused broadening of the main beam. Therefore, because the antenna radiation spread to the broadside directions, the maximum realized gain became lower.

From the simulation, it was observed that the bending direction also influenced the realized gain. When the patch antenna was bent about the x-axis, the gain change was smaller compared to that of the y-axis (Table 2). This tendency was also seen in the measurement. It could be understood that the lesser gain change for x-axis bending was attained due to the smaller shifts in the operating frequency during x-axis bending (Figure 8).

Effect of padding

The simulated and measured reflection coefficient with padding layers is depicted in Figure 10. The Pearson correlation coefficient (R = 0.94) indicates a very strong correlation between the simulated and measured operating frequencies. Dielectric loading of the polyester padding layers decreased the operating frequency, and severe detuning up to 2% was observed with more padding layers (Table 3). Considering the very narrow bandwidth (3.57%) of the bare antenna, this frequency shift caused by six layers of padding fabrics might not be negligible. This indicates that dielectric loading to the patch needs to be precisely planned in advance during antenna design and is not subject to change for the extreme padding case. On the other hand, the additional padding layers placed over the ground plane did not pose any impact on the operating frequency because of the presence of the ground plane.

Figure 10.

(a) Simulated and (b) measured reflection coefficient plots with padding.

Table 3.

Shifts in the operating frequency and maximum gain at 2.45 GHz by padding. Percentage (%) shift in operating frequency and the gain change in dB are calculated from comparison to the bare case

Number of padding layers	Simulated				Measured
Number of padding layers	Operating frequency (GHz)	Frequency shift (%)	Gain (dBi)	Gain change (dB)	Operating frequency (GHz)	Frequency shift (%)	Gain (dBi)	Gain change (dB)
0 (bare)	2.45	–	7.69	–	2.45	–	8.39	–
P1	2.43	−0.82	7.56	−0.13	2.44	−0.41	8.25	−0.14
P1G1	2.43	−0.82	7.61	−0.08	2.43	−0.82	8.29	−0.10
P4	2.40	−2.04	7.24	−0.45	2.42	−1.22	8.06	−0.33
P4G4	2.40	−2.04	7.40	−0.49	2.42	−1.22	8.06	−0.33
P6	2.40	−2.04	7.13	−0.56	2.41	−1.63	7.56	−0.83
P6G6	2.40	−2.04	7.07	−0.62	2.41	−1.63	7.53	−0.86

Figure 11 depicts the radiation patterns of the fabric patch antenna sample padded with one, four, and six layers of polyester fabrics, simulated and measured at 2.45 GHz. A reduction in the realized gain was up to 0.62 dB (simulated) and 0.86 dB (measured), as shown in Table 3, and there was a very strong correlation (R = 0.95) between the simulated and measured gain values. Two possible reasons for the gradual decrease in the realized gain are antenna detuning by the dielectric loading and the increased dielectric dissipation by the additional polyester layers. Because the polyester fabric had a low dielectric constant (1.55) and a low-loss tangent (0.0087), the impact was not significant when a single padding layer was applied (P1 and P1G1). However, greater impact was observed as the number of padding layers increased. Despite the low dielectric constant and the low-loss tangent of the padding layers, this study suggests that a multi-layer antenna structure could, in fact, measurably impact the antenna matching and the radiation performance.

Figure 11.

Simulated (a) E-plane and (b) H-plane radiation patterns, and measured (c) E-plane and (d) H-plane radiation patterns, monitored at 2.45 GHz with varying padding layers.

Conclusions

With a detailed description of the fabrication procedures, a high gain fabric antenna was fabricated using 100% textile materials. Woven polyester fabrics were used for the patch and the substrate of the antenna as well as the ground plane. The impedance matching and radiation performance were measured and the measured results reasonably agreed with the numerical calculations, indicating that the antenna characterization and fabrication procedures were valid.

Concave bending was found to lower the operating frequency by 1.2% when the patch antenna was bent along its length, while bending along the antenna width resulted in a marginal effect on the operating frequency (less than 1%). On the other hand, a clear degradation in the realized gain was observed when the patch was bent along either direction (–0.46 dB along the patch width and –0.97 dB along the patch length). Although the fabricated antenna exhibited a high gain of 8.39 dBi, the experimental results suggest that textile antennas could experience measurable impact on the realized gain due to bending.

Textile padding layers also influenced the antenna performance. Padding the ground plane caused no measurable impact on the operating frequency and the measured gain. However, padding the patch led to dielectric loading and brought significant impact on the operating frequency (up to 1.63%) of the patch antenna. This was associated with the lowering of the realized gain (up to 0.86 dB). These findings suggest that antenna packaging, such as the placement of the additional clothing layers on top of the radiating patch, could deteriorate the antenna performance, and therefore this condition must be considered during the antenna design stage.

Footnotes

Acknowledgements

The authors would like to thank Mr Ming Wang for providing the polyester fabric samples. Our sincere appreciation also goes to Ms Yi Cao for her support during the laser cutter operation. The authors are thankful for having assistance from Mr Daniel M Hawkins during the 3D printing. Our appreciation also goes to Dr Hasan Shahariar and Mr Clifford Muchler for their support during the radiation pattern measurements. This work was performed in part at the NCSU Nanofabrication Facility (NNF), a member of the North Carolina Research Triangle.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Haagenson

Noghanian

Leon

P de

et al.

Textile antennas for spacesuit applications: Design, simulation, manufacturing, and testing of textile patch antennas for spacesuit applications. IEEE Antennas Propag Mag 2015; 57: 64–73.

Gil I and Fernández-García R. Wearable GPS patch antenna on jeans fabric. In: 2016 progress in electromagnetic research symposium (PIERS), (ed. JA Kong), Shanghai, China: IEEE, 8–11 August 2016, pp.2019–2022.

Soh PJ, Vandenbosch GAE and Schreurs DMMP. On-body characterization of textile antennas for biomedical health monitoring systems. In: 2013 IEEE topical conference on biomedical wireless technologies, networks, and sensing systems (eds S Pinarello, S Leuschner and WT Khan), Austin, TX, USA, 20–23 January 2013, pp.19–21.

Wang

Zhu

et al.

Simulation and experimental study of double-element antennas based on a three-dimensional woven structure with various curvature radii. Text Res J 2017; 87: 216–223.

Zhu

et al.

Electromagnetic performance of a three-dimensional woven fabric antenna conformal with cylindrical surfaces. Text Res J 2017; 87: 147–154.

Tak J and Choi J. Design of an all-fabric higher order mode circular patch antenna for wireless body area network application. In: 2014 international workshop on antenna technology: small antennas, novel EM structures and materials, and applications (iWAT) (eds K Esselle and G Gay), Sydney, Australia: IEEE, 4–6 March 2014, pp.29–32.

Salvado

Loss

Gonçalves

et al.

Textile materials for the design of wearable antennas: A survey. Sensors 2012; 12: 15841–15857.

Declercq

Rogier

Hertleer

. Permittivity and loss tangent characterization for garment antennas based on a new matrix-pencil two-line method. IEEE Trans Antennas Propag 2008; 56: 2548–2554.

Yao

Qiu

. Design and fabrication of microstrip antennas integrated in three dimensional orthogonal woven composites. Compos Sci Technol 2009; 69: 1004–1008.

10.

Yao

Wang

et al.

Effect of wire space and weaving pattern on performance of microstrip antennas integrated in the three dimensional orthogonal woven composites. Appl Compos Mater 2012; 19: 21–30.

11.

Yao

Qiu

. Design and electromagnetic properties of conformal single-patch microstrip antennas integrated into three-dimensional orthogonal woven fabrics. Text Res J 2015; 85: 561–567.

12.

Zhang S, Chauraya A, Whittow W, et al. Embroidered wearable antennas using conductive threads with different stitch spacings. In: 2012 Loughborough antennas propagation conference (LAPC), Loughborough UK: IEEE, 12–13 November 2012, pp.1–4.

13.

Zhang S, Chauraya A, Seager R, et al. Fully fabric knitted antennas for wearable electronics. In: 2013 USNC-URSI radio science meeting (joint with AP-S symposium), Lake Buena Vista, FL, USA: IEEE, 7–13 July 2013, p.215.

14.

Hertleer

Tronquo

Rogier

et al.

The use of textile materials to design wearable microstrip patch antennas. Text Res J 2008; 78: 651–658.

15.

Mallory

Hajdu

. Electroless plating: Fundamentals and applications, Orlando, FL: American Electroplaters and Surface Finishers Society, 1990.

16.

Guo L, Bashir T, Bresky E, et al. 28 - Electroconductive textiles and textile-based electromechanical sensors—integration in as an approach for smart textiles. In: Koncar V (ed.) Smart textiles and their applications. Oxford: Woodhead Publishing, 2016, pp. 657–693.

17.

Sen

. Coated textiles: Principles and applications, 2nd ed. Boca Raton, FL: CRC Press, 2008.

18.

Shahariar H, Soewardiman H and Jur JS. Fabrication and packaging of flexible and breathable patch antennas on textiles. In: Southeastcon 2017 (ed. L Stogner), Charlotte, NC, USA: IEEE, 30 March–2 April 2017, pp.1–5.

19.

Shahariar

Soewardiman

Muchler

et al.

Porous textile antenna designs for improved wearability. Smart Mater Struct. 5 March 2018.

20.

Morton WE and Hearle JWS. Physical properties of textile fibres. Boca Raton, FL: Woodhead Publishing, 2008.

21.

Galehdar A and Thiel DV. Flexible, light-weight antenna at 2.4 GHz for athlete clothing. In: 2007 IEEE antennas and propagation society international symposium (eds JW Wallace and J Hirokawa), Honolulu, USA: IEEE, 9–15 June 2007, pp.4160–4163.

22.

Biswas

Sen

. Design and development of rectangular patch antenna with superstrates for the application in portable wireless equipments and aircraft radome. Microw Opt Technol Lett 2014; 56: 883–893.

23.

Lowes

Day

Korolkiewicz

et al.

Performance of microstrip patch antenna with electrically thick laminated glass superstrate. Electron Lett 1994; 30: 1903–1905.

24.

Ammann MJ. A comparison of some low cost laminates for antennas operating in the 2.45 GHz ISM band. In: IEE Colloquium on Low Cost Antenna Technology, London, UK: IEEE, 24 February 1998, pp.3–3.

25.

Heikkinen JJ, Laine-Ma TT and Kivikoski MA. Flexible fabric-base patch antenna with protective coating. In: 2007 IEEE antennas and propagation society international symposium (eds JW Wallace and J Hirokawa), Honolulu, HI, USA: IEEE, 9–15 June 2007, pp.4168–4171.

26.

Borneman J. Man-made fibers continue to grow. Textile World, http://www.textileworld.com/textile-world/fiber-world/2015/02/man-made-fibers-continue-to-grow/ (accessed 28 January 2018).

27.

Bharambe V. Analysis of liquid metal vacuum filling approach to develop 3D printed antennas. North Carolina State University, Raleigh, NC, 2016.

28.

Yaghjian

Best

. Impedance, bandwidth, and Q of antennas. IEEE Trans Antennas Propag 2005; 53: 1298–1324.

29.

Balanis

. Advanced engineering electromagnetics, 2nd ed. Hoboken, NJ: Wiley, 2012.

30.

Wei

Delgado

Hawley

et al.

Dielectric analysis of semi-crystalline poly(ethylene terephthalate). MRS Online Proc Libr Arch 1994; 347: 735–735.

31.

Bal

Kothari

. Permittivity of woven fabrics: A comparison of dielectric formulas for air-fiber mixture. IEEE Trans Dielectr Electr Insul 2010; 17: 881–889.

32.

Bal

Kothari

. Study of dielectric behaviour of woven fabric based on two phase models. J Electrost 2009; 67: 751–758.

33.

Sihvola AH. Electromagnetic mixing formulas and applications. London: Institution of Electrical Engineers, 1999.

34.

Kallner

. Laboratory statistics: Methods in chemistry and health sciences, Amsterdam: Elsevier, 2018.

35.

Balanis

. Antenna theory: Analysis and design, Hoboken, NJ: Wiley, 2016.