Univesidade de São Paulo Biblioteca Digital da Podução Intelectual - BDPI Depatamento de Sistemas Eletônicos - EP/PSI Comunicações em Eventos - EP/PSI 014-03-6 Paametic study of etangula patch antenna using denim textile mateial http://www.poducao.usp.b/handle/bdpi/4485 Downloaded fom: Biblioteca Digital da Podução Intelectual - BDPI, Univesidade de São Paulo
Paametic Study of Retangula Patch Antenna Using Denim Textile Mateial Macus Gilo and Fatima Salete Coea Depatment of Electonic Systems Engineeing Polytechnic School, Univesity of São Paulo São Paulo- Bazil macus.gilo@usp.b fcoea@lme.usp.b Abstact This pape pesents a textile patch antenna designed fo WBAN applications at.45 GHz ISM band. The antenna uses denim as substate and conductive fabic fo the gound plane and adiato layes. The main pupose of this pape is to analyze the influence of typical deviation of denim popeties and patch adiato dimensions on the pefomance of the antenna. The paametes consideed in the analysis ae the elative pemittivity and thickness of denim and the width and length of the ectangula patch adiato. The dependence of the cental opeation fequency of the antenna on those paametes was studied using the antenna eflection coefficient obtained fom EM simulations. Rules of thumb fo one-shot design wee deived and applied to design a ectangula patch antenna. An antenna pototype was fabicated and measued, demonstating a 10 db impedance band of 4.8 % centeed at.45 GHz, in good ageement with simulated esults. Keywods Micostip Patch Antenna, textile antenna, denim, WBAN. I. INTRODUCTION In ecent yeas, flexible and textile antennas have becoming eve moe attactive due to thei application in weaable systems [1,]. The new wieless potocols fo body aea netwoks and pesonal aea netwoks (BANs and PANs) opeating in the.45 GHz band ae highly attactive as futue solutions fo adio fequency identification (RFID), smat home and militay applications [3-5]. Weaable antennas need to be hidden and of low pofile. This equies a possible integation of the antenna elements within eveyday clothing. The micostip patch antenna is a epesentative candidate fo any weaable application, as it can be made confomal fo integation into clothing [6-8]. Conductive textiles ae commecially available and have been successfully used in textile antennas [9-11]. Fabics have been used as dielectic substates. Effects of antenna bending on its impedance and adiation chaacteistics ae pesented in [1-14]. These esults shows shift in esonance fequency and changes input impedance of the antenna. Howeve, little infomation is found on the electomagnetic popeties of egula commecialized textiles. Paametes as elative pemittivity and thickness of the dielectic mateial ae cucial in the design of antennas, detemining the dimensions of the antenna adiato. The choice of the thickness of the dielectic mateial is a compomise between efficiency and bandwidth of the antenna [14]. Moeove, the thickness of the substate also influences the geometic sizing of the antenna. Howeve, fabics ae flexible and compessible mateials, which thickness and density might change with low pessues. Futhemoe, the fact that the antenna be sewn can esult in vaiations on its thickness. All these featues ae somehow difficult to contol in eal applications of textiles. Theefoe it is impotant to know how they may influence the behavio of the antenna, in ode to minimize any unwanted effects. This pape pesents a study of the impact of vaiations of these paametes on the chaacteistics of patch antenna manufactued on denim substate, with metal layes of Pue Coppe Polyeste Taffeta Fabic (PCPTF) [15]. This pape is oganized as follows. Section II descibes the textile mateials used fo an antenna manufactuing. Section III pesents the design of a textile ectangula patch antenna fo.45 ISM band and an analysis of the influence of deviations on the textiles fabics chaacteistics and adiato dimensions in the antenna pefomance. One-shot design ecommendations ae suggest along this section. Section IV pesents the antenna manufactuing pocess and section V compaes measued and simulated antenna fequency esponse. Conclusions ae dawn in section VI. II. A. Micostip patch antenna DESIGN METHODOLOGY A micostip patch antenna consists of a metallic adiating patch on top of a dielectic substate which has a gound plane on the othe side as shown in Fig. 1. Fig.1. A micostip patch antenna.
Micostip patch antennas ae appopiate to be poduced out of textile mateial mainly because of thei compact geomety and plana pofile. The adiating elements may be squae, ectangula, cicula, tiangula, and elliptical o some othe common shape [14]. B. Design specification The antenna was designed to opeate at the unlicensed.45 GHz ISM band. C. Textile Mateial The textile mateials employed in the designed antenna ae denim as the dielectic substate and PCPTF fo the conductive adiato and gound plane. The selection of an appopiate conductive textile is an impotant step in the design of a textile antenna. Popeties of PCPTF povided by supplie ae given in table I. TABLE I. Popeties Suface esistance Conductivity Thickness (t) PROPERTIES OF THE CONDUCTIVE FABRIC USED - PCPTF. Conductive Textile Mateial 05 Ohm/sq.5 x 10 5 S/m 08 mm Weight 80 g/m PCPTF is made using 35% of coppe and has the advantages of being flexible, light weight, easy to be cut, sewn and meant to be used in electomagnetic shielding applications. The denim fabic is one of the best candidates fo use as substate on textile antennas, since it is duable, inelastic, low thickness, comfotable fo use and low cost mateial. The pemittivity and the loss tangent of denim substate wee extacted using the techniques explained in [16] with the aid of electomagnetic simulation, esulting in ε = 1.6 and tanδ = 01 at.45 GHz. The substate thickness was measued along a 10 mm x 10 mm denim sample using a calipe ule and h = 7 +/- 1 mm was obtained. D. Antenna Design A.45 GHz ectangula patch antenna was designed using the expeimental data obtained fo the denim and the chaacteistics of the conductive fabic PCPTF fom the manufactue datasheet. The sensitivity of the designed antenna was analyzed assuming vaiations of the elative pemittivity and thickness of denim. The impact of deviations on the dimensions of the adiato, which usually esults fom the pocess of cutting and fastening the conductive layes on the denim substate, was also simulated. Using the design pocedue pesents in [14], a simple ectangula micostip antenna was designed, using the following equations: whee, L c f 1 W (1) 1 1 1 h ef 1 1 () L ( 41.h. ( f ef ef c ef 3). W h 58) L W h W 64 8 The antenna dimensions wee optimized using HFSS [17] and ae pesented in Table II. TABLE II. Paamete DIMENSIONS OF TEXTILE PATCH ANTENNA. Dimension (mm) W 51 L 46 y o 3.6 d 5 The designed antenna pesented a simulated 10 db impedance bandwidth of 3.% centeed at.45 GHz. Fig. pesents the simulated fa-field adiating pattens of the designed antenna at.45 GHz. The simulation indicated a maximum antenna gain of.1 db. This elatively low gain is due mainly to the conductive losses on the adiato and gound plane. Fig.. Simulated nomalized adiation patten. The influence of the substate chaacteistics and of the adiato dimensions on the fequency esponse of the designed antenna was studied by means of compute simulations, consideing typical deviation on these antenna paametes. (3) (4)
III. PARAMETRIC STUDY OF PATCH ANTENNA A. Influence of deviations on elative pemittivity ε and loss tangent tg The elative pemittivity of the antenna substate has a significant ole in the antenna design and deviations on its value will esult in an antenna opeating bellow o above the specified cental fequency. Since micostip antennas ae naowband, a elatively small eo in the value of the substate elative pemittivity can esult in an antenna opeating out of the desied band. The impact of denim ε on the antenna cental fequency was evaluated fom compute simulations of the antenna caied out fo ε anging fom 1.5 to 1.8 with 1 steps, keeping all othe paametes constant. As shown in Fig. 3, vaiations of +/- 1 in ε have a significant impact in the antenna opeating band. The elative pemittivity of denim epoted in technical liteatue vaies fom 1.6 to 1.8 [1-4]. Thus, the accuate measuement of dielectic chaacteistics of textiles should be pefomed befoe the antenna design fo one-shot design pocedue. Fig.4. Simulated etun loss of the antenna x fequency as a function of denim thickness h. C. Influence of deviations on adiato dimensions L and W The influence of deviations on the ectangula patch adiato dimensions on the cente fequency of the antenna was analyzed by means of compute simulations, vaying independently the width W and length L of the adiato aound thei design values. It was assumed that the pocess used to cut the adiato conductive fabic have an accuacy of +/- 5 mm at each edge cut, esulting vaiation up to +/- 1 mm on W and L. Fig. 5 shows the simulated antenna etun loss fo a width fixed (W = 51 mm) and length vaiable (L = 46 +/- 1 mm). As expected, the length of the adiato L has a stong influence on the antenna cental fequency. In the pesent case it was found that a vaiation of +/- 1 mm in L caused a displacement of +/- 4.4% on the antenna cente fequency, indicating the need to contol stongly the length of the adiato duing the antenna fabication pocess. Fig.3. Simulated etun loss of the antenna vesus fequency as a function of denim elative pemittivity ε. Values of loss tangent in the ange of 01 to 05 wee also consideed duing the antenna simulations. The deviation on the loss tangent had a second ode effect on the antenna cente fequency, affecting mainly the etun loss magnitude. B. Influence of deviations on denim thickness h The thickness of textile substates pesents some vaiation along thei extension. Fo example, the measued thickness of denim substate studied in this pape is h = 7 +/- 1 mm. Futhemoe, the pocedue of sewn the conductive layes of the antenna on the substate poduces a slight eduction of its thickness, and affects the antenna behavio. Fig. 4 pesents the simulated esults fo the antenna etun loss when fo substate thickness vaies fom 5 to 8 mm. As shown in Fig. 4, the thickness of the substate will affect the opeating fequency band of the antenna and cause degadation on its etun loss. Fig.5. Simulated etun loss of the antenna x fequency as a function of the adiato length L. Fig. 6 shows the simulated antenna etun loss, when vaying the adiato width (W = 51+/-1 mm) and keeping its length constant (L = 46 mm). The adiato width of the ectangula patch antenna is elated to its adiation efficiency and has a second ode effect on its esonance fequency. Simulation esults indicated a shift of +/- 4% at the opeating cente fequency of the antenna fo a vaiation of +/-1 mm in the adiato width W, confiming that its impact on antenna cental fequency is not significant.
Fig.6. Simulated etun loss of the antenna x fequency as a function of the adiato width W. The design of textile antennas employing metallic fabics should focus on antennas using adiato shapes that ae easy to be cut using conventional fabic cut tools. The use of adiatos with ectangula geomety facilitates the cutting of the patch, hence inceasing the accuacy of its physical dimensions. D. Influence of the adiato and gound plane layes conductivity on the feed point y 0 The feed point position is elated to the input impedance of the antenna and is stongly affected by the conductivity of the metallic layes used to fabicate the antenna adiato and gound plane. This effect is not significant in patch antennas using conventional micowave substates plated with coppe layes, due to its high conductive. Howeve, conductive fabics used on textile antennas have conductivity about one o two odes of magnitude lowe than coppe, affecting stongly the ideal antenna feed point. Fig.7 compaes the pefomance of the designed patch antenna, consideing the metallic layes using PCPTF (ζ =.5x10 5 S/m) and coppe (ζ = 5.7x10 7 S/m). The ideal feed point fo PCPTF layes is y o = 3.6mm, esulting in -40 db of simulated antenna etun loss at.45 GHz. Consideing metallic layes of coppe, the same feed point gives a simulated antenna etun loss of -8.9 db, causing impedance mismatch. In this case, the feed point has to be moved to y o = 11.5 mm in ode to ecove a good impedance matching condition. The lowe conductivity of the conductive fabic esults in antennas with lage bandwidth than the ones using coppe, as shown in Fig. 7, at the expense of degadation of antenna gain. These esults show that it is essential to know the conductivity of the mateial used to fabicate the adiato and gound plane in ode to do a one-shot design. IV. ANTENA FABRICATION PROCESS The antenna pesented in section II.D was designed accoding to the citeia of one-shot design deived in this pape. In the antenna design, the measued values of the physical chaacteistics of the substate, ε and tanδ, wee used. The adiato shape is ectangula, and is made of conductive fabic, cut to pecise dimensions. This antenna was fabicated and chaacteized, and its expeimental pefomance was compaed to simulation esults in ode to validate the design ules pesented hee. Fig 8 pesents a photogaph of the antenna pototype. The gound plane and the adiato wee fist attached to the denim substate using a glue stick, and afte they wee additionally sewn in ode to obtain a unifom thickness. The stuctue is fed by a 50 Ω SMA connecto positioned at the bottom of the antenna stuctue. The cente conducto of the connecto eaches the adiato though a hole in the substate. The conductive fabic accepted conventional solde pocess fo the connecto without damages. (a) Font (b) Back Fig.8. Antenna photogaph. V. RESULTS AND ANALYSIS A. Retun Loss The simulated and measued etun loss of the poposed antenna ae shown in Fig. 9. The measuements wee pefomed using an HP87D netwok analyze. Fig.7. Simulated antenna etun loss as a function of the metallic layes mateial and feed point y 0. The use of metallic layes with lowe conductivity will incease the antenna losses, educing its adiation efficiency. Fig.9. Simulated and expeimental etun loss of antenna as a function of the fequency. Both simulated and measued esults indicated that the antenna opeates at cente fequency of.45 GHz, with excellent impedance matching. The bandwidth of measued esults is slightly highe than the simulated one. The 10 db etun loss bandwidth of the simulated antenna is 3.% (.41.49 GHz) and the measued one is 4.8%
(.39.51 GHz). Good ageement between the expeimental and simulated values is obseved and, in both cases, the antenna showed to have sufficient bandwidth to cove the ISM systems equiement. Despite all the uncetainties associated with the manufactuing of textile antennas, the expeimental esults met the design specifications, so that the goal of oneshot design was achieved. VI. CONCLUSIONS A textile antenna fo.45 GHz ISM band application is pesented, analyzed and fabicated. The antenna substate was denim and conductive fabic was used as the ectangula adiating element and gound plane. The pefomance of the designed antenna was simulated consideing typical deviations of the denim elative pemittivity and thickness. Deviations on the adiating element dimensions due to the pocess used to cut the conductive fabic wee also simulated. The analysis of the simulation esults indicated that an eo of +/- 1 on the elative pemittivity of denim used on the antenna design could esult in an antenna pototype opeating out of the.45 GHz ISM band. The same poblem could be poduced by an eo of +/- 5 mm on the cut of each edge of the adiato element length duing the antenna fabication. Typical vaiations on denim thickness and on the adiato width had second ode effects on the simulated fequency esponse of the antenna. The effect of the conductivity of the fabic used on the adiato and gound plane was also evaluated and affects in a significant way the ideal antenna feed point and the antenna bandwidth. Fo one-shot design it is impotant to chaacteize the textile substate detemining its elative pemittivity (and loss tangent) at the antenna opeating fequency. It is also impotant to contol adiato cut pocess so that its length met the design value accuately. The conductivity of the mateial used on the adiato and gound plane layes has to be account on the antenna design, fo a pecise design of the ideal antenna feed point and bandwidth. These ecommendations fo one-shot design wee applied to the design of the ectangula micostip patch antenna. The measued fequency esponse of the antenna pototype was in vey good ageement with the simulated esults and met the antenna specifications, without the need of tuning o new design steps. Refeences [1] M. Mantash, A.-C. Taot, S. Colladey, and K. Mahdjoubi, Investigation of flexible textile antennas and AMC eflectos, Intenational Jounal of Antennas and Popagation, vol. 01, aticle ID 36505, 10 pages, doi:11155/01/36505. [] M. E. Jalil, M. K. A. Rahim, and M. A. Abdullah, Compact CPW-fed Ulta-Wideband (UWB) Antenna Using denim Textile Mateial Poceedings of ISAP01, Japan, 01, pp.30-33. [3] R. J. Langley, K. L. Fod, and H. J. Lee, Switchable on/off-body communication at.45 GHz using textile micostip patch antenna on stipline Antennas and Popagation (EUCAP), 6th Euopean Confeence, 01, pp. 78-731. [4] H. A Rahim, M. F. Abd Malek, I. Adam, K. M. Juni, and M. I. M. Saleh, Basic chaacteistics of a textile monopole antenna fo Body-Centic wieless communications IEEE Symposium on Wieless Technology and Applications (ISWTA),01, pp.7 75. [5] T. Kaija, J. Lilja, and P. Salonen, Exposing textile antennas fo hash envionment, in Poceedings of 010 Militay Communications Confeence, San Jose, CA, USA, 31 Octobe 3 Novembe010; pp. 737 74. [6] Z. Wang, L. Zhang, D. Psychoudakis, and J. L. Volakis, Flexible textile antennas fo body-woncommunication, in Poceedings of IEEE Intenational Wokshop on Antenna Technology,Tucson, ZA, USA, 5 7 Mach 01; pp. 05 08. [7] B. Gupta, S. Sankaalingam, ands. Dha, Development of weaable and implantable antennas inthe last decade, A Review. In Poceedings of Mediteanean Micowave Symposium (MMS),Guzelyut, Tukey, 5 7 August 010, pp. 51 67. [8] M. A. R. Osman, M. K. A. Rahim, N. A. Samsui, H. A. M. Salim, and M. F. Ali, Emboideed fully textile weaable antenna fo medical monito in gap plications, Pogess In Electomagnetics Reseach, 011,vol. 117, pp.31-337. [9] M. I. Jais, M. F. Jamlos, M. Jusoh,T. Sabapathy, M. R. Kamaudin, R. B. Ahmad, A. A. H. Azemi, E. I. Azmi, P. J. Soh, 4, G. A. E. Vandenbosch and N. L. K. Ishak, A novel.45ghz switchablebeamtextile antenna (SBTA) fo out doo wieless body aea netwok (WBAN) applications, Pogess In Electomagnetics Reseach, 013, vol. 138, pp.613-67. [10] I. Loche, M. Klemm, T. Kistein, andg. Töste, Design and chaacteization of puely textile patch antennas. IEEE Tansactions on Advanced Packaging, vol.9, pp.777 788, 006. [11] H. A Rahim, M. F. A. Malek, I. Adam, K. M. Juni, and M. I. M. Saleh, Basic chaacteistics of a textile monopole antenna fo body-centic wieless communications, IEEE Symposium on Wieless Technology and Applications (ISWTA), 01, pp.7-75 [1] S. Dey, N. Saha and S. Biswas, Design and pefomance analysis of UWB cicula disc monopole textile antenna and bending consequences, Poceedings of the 5th Euopean Confeence on Antennas and Popagation (EUCAP), 011, pp.119 1133. [13] S. Sankaalingam and B. Gupta, Development of textile antennas fobody weaable applications and investigations on thei pefomance unde bent conditions, Pogess In Electomagnetics Reseach B,010,vol., pp.53-71. [14] C. A. Balanis, Antenna Theoy: Analysis and Design, nd ed. New Yok: Wiley, 1996. [15] http://www.lessemf.com/11.pdf [16] Y. H. Chou, M. J. Jeng, Y. H. Lee, and Y. G. Jan, The measuements of RF dielectic constant, dielectic loss coefficient, and conducto loss coefficient in PCB, Pogess In Electomagnetics Reseach Symposium,010,pp.64-67. [17] High Fequency Stuctue Simulato (HFSS), Ansoft Copoation Pittsbugh, PA.