ANTENNAS IMPLEMENTED ON CERAMIC AND THERMO-SENSITIVE SUBSTRATES FOR RADIO FREQUENCY IDENTIFICATION, CELLULAR, AND CELLULAR VEHICLE TO EVERYTHING COMMUNICATIONS, USING LITHOGRAPHY AND ADDITIVE MANUFACTURING TECHNIQUES By Adamantia Chletsou A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Electrical Engineering – Doctor of Philosophy 2023 ABSTRACT This dissertation demonstrates the implementation methods and performance of antennas on dif- ferent substrates using the traditional lithography method and Additive Manufacturing (AM) tech- niques. The developed devices are used for biomedical applications and vehicular communica- tions. The effectiveness of using photonic sintering and reactive silver ink to develop 3D printed antennas on thermo-sensitive substrates is investigated. Intense Pulsed Light (IPL) is used to sinter silver nano-particle ink on the automotive Acrylonitrile Butadiene Styrene (ABS) and the vero- white polymer. Different sintering profiles of IPL are tested on the ABS and the vero-white to identify the optimal one. Development of antennas using lithography, Aerosol Jet Printer (AJP) combined with thermal sintering, AJP combined with photonic sintering, and AJP combined with reactive ink is investigated and their overall performance is compared. The first step of this dissertation is to explore the antenna design that is optimal for biomedical, Radio Frequency Identification (RFID) applications, operating inside human muscle and in free space. The next step is the development of a dual-band, planar antenna for automotive applications using lithography on a flexible, lightweight substrate and AM techniques on ABS. The antenna performance is tested on a real vehicle and the effects of the ground on the antenna radiation pattern are identified. Co-Planar Waveguide (CPW) lines are developed using the same procedure to identify the losses due to silver conductivity. Thereafter, an Electrically Small Antenna (ESA) is developed on a 3D printed hemisphere for vehicular communications. Prototypes of this antenna are tested on a real vehicle and a ground plane inside a near field system. The effect of the vehicle body on the antenna performance is evaluated. ACKNOWLEDGEMENTS I would like to express my gratitude, first and foremost to my advisor, Dr. Ioannis Papa- polymerou for all the guidance, support, and mentoring he provided me throughout my doctoral studies. His advice on my research and future career goals helped me advance as a person and a researcher. Moreover, I would like to thank Dr. Ahmet Cagri Ulusoy who served as my advisor during my first year of graduate studies and taught me how to conduct research. In addition, I would like to thank Dr. Tamara Bush, Dr. Premjeet Chahal, and Dr. Nelson Sepulveda for serving as members of my research committee and providing me with valuable suggestions and advice on how to proceed with my research. I am thankful to Dr. Jeffrey Nanzer, Dr. Edward J. Rothwell, and Dr. Balasubramaniam Shanker for their useful advice that helped me at critical points of my research. I want to express my gratitude to the MSU employee Brian Wright who taught me how to use most of the tools and assisted me with the experimental setups and measurements for my research. I would also like to thank the MSU employees, Michelle Stewart, Roxanne Peacock, Lisa Clark and Meagan Kroll who provided me with all the necessary resources. It has been my privilege and pleasure to work and take classes with many talented graduate students at MSU. First, I would like to express my deepest appreciation for Dr. Xenophon Kon- stantinou, Dr. Stavros Vakalis and Dr. Yuxiao He for their mentoring and advice during my time as a graduate student with regards to both research and transitioning from my graduate studies to employment. I would also like to express my gratitude to my former lab members who grad- uated before me, Dr. Si-Wook Yoo, Dr. Asad Ali Nawaz, Dr. Shih-Chang Hung, Dr. Jubaid Abdul Qayyum, and Ibrahim Kagan Aksoyak for their valuable advice. Also, I am grateful to have worked with Nick Sturim, Wesley Spain, and Ally Bannon while completing this dissertation. I also want to acknowledge Ford Motor Company and especially Mr. John Locke, Mr. Eric Newsom, and Dr. Varittha Sanphuang for funding my research and providing with useful insights on how to proceed. I want to thank my friends in US, Dr. Meltem Duva, Dr. Berk Can Duva, Sofia Stathoulia, iii Marianna Moustaka, Nikolas Fratzeskakis, Dr. Kathryn Lankford, and Dr. Serge Mghabghab who made the transition to living in US more fun and easy. I have been lucky enough to know Dr. Elias Strangas and Dr. Jane Turner, and I want to thank them for their kindness and for their assistance to adjust in the new environment. Also, I would like to thank my dear friends back home, Dimitra Koufou, Margarita Pantazi and Stacy Kokkini for persistently keeping in touch with me. Last but not least I want to express my gratitude to my son, Yannis, and my husband, Antonis, for their love, support ,and for always being here for me all these years. To my parents, Eleni and Kostas and my sister, Evi for believing in me, giving me the courage to proceed and making it possible for me to achieve my goals. iv TABLE OF CONTENTS LIST OF ABBREVIATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii CHAPTER 1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1 Additive Manufacturing of Antennas . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.2 Objective of this dissertation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.3 Dissertation Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 CHAPTER 2 RECTIFYING ANTENNA FOR RFID BIOMEDICAL APPLICATIONS . 8 2.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.2 Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.3 Fabrication Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.3.1 Antenna and muscle design . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2.3.2 Commercial chip EM4124, custom made energy harvesters and com- mercial chip EM4324 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.4.1 Simulation results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.4.2 Experiments results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 2.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 CHAPTER 3 AUTOMOTIVE ANTENNAS FOR CELLULAR AND C-V2X APPLICATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.2 Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 3.2.1 Antennas tested on automotive plastics . . . . . . . . . . . . . . . . . . . . 26 3.2.2 Antennas tested on real vehicle . . . . . . . . . . . . . . . . . . . . . . . . 28 3.3 Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 3.3.1 Antennas measured on automotive plastics without the presence of a vehicle 30 3.3.2 Antenna measured on real vehicle . . . . . . . . . . . . . . . . . . . . . . 46 3.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 3.4.1 Results of antennas measured on automotive plastics without the pres- ence of a vehicle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 3.4.2 Results of antennas measured on automotive plastics on a real vehicle . . . 58 3.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 CHAPTER 4 ADDITIVE MANUFACTURED AUTOMOTIVE ANTENNAS ON ABS . 65 4.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 4.2 Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 4.3 Fabrication Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 4.4 Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 4.5 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 4.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 CHAPTER 5 ADDITIVE MANUFACTURED CO-PLANAR WAVEGUIDE (CPW) LINES ON ABS WITH DIFFERENT SILVER THICKNESS . . . . . . . . . . . . 73 v 5.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 5.2 Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 5.3 Fabrication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 5.4 Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 5.5 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 5.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 CHAPTER 6 HELICAL ANTENNA FOR C-V2X APPLICATIONS . . . . . . . . . . . 81 6.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 6.2 Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 6.3 Fabrication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 6.4 Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 6.5 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 6.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 CHAPTER 7 FULLY 3D PRINTED HELICAL ANTENNA . . . . . . . . . . . . . . . . 94 7.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 7.2 Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 7.3 Fabrication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 7.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 7.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 CHAPTER 8 CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 APPENDIX A: FULLY 3D PRINTED ANTENNA FOR APPLICATIONS AT 28 GHZ . . 100 APPENDIX B: AEROSOL JET PRINTING MATERIALS . . . . . . . . . . . . . . . . . . 101 APPENDIX C: AEROSOL JET PROCEDURES . . . . . . . . . . . . . . . . . . . . . . . 103 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 vi LIST OF ABBREVIATIONS ABS Acrylonitrile Butadiene Styrene Ag Silver AJP Aerosol Jet Printing AM Additive Manufacturing AM radio Amplitude Modulation radio BalUn Balanced to Unbalanced C-V2X Cellular-Vehicle To Everything CPW Coplanar Waveguide GCPW Ground Coplanar Waveguide Cu Copper DAB Digital Audio Broadcasting DSRC Dedicated Short Range Communication DUT Device-Under-Test ETc Electronic Toll Collection FM Frequency Modulation FDM Fused Deposition Modeling GPS Global Positioning System IC Integrated Circuit IMDs Implantable Medical Devices IPL Intense Pulsed Light ISM Industrial Scientific Medical LCP Liquid Crystal Polymer LOM Laminated Object Manufacturing LRRM Line Reflect Reflect Match LTE Long Term Evolution MedRadio Medical Radio Communication Service vii mm-Wave Millimeter-Wave NP Nanoparticles PEC Perfect Electric Conductor PFN Pulse Forming Network PI Polyimide PILA Patch Inverted L Antenna RF Radio Frequency RFIC Radio Frequency Identification Circuit RFID Radio Frequency Identification RKE Remote Keyless Entry SAR Specific Absorption Rate SDARS Satellite Digital Audio Radio Service SGC Solid Ground Curing SLA Stereolithography Apparatus TPMS Tire Pressuring Monitor System V2I Vehicle To Infrastructure V2P Vehicle To Pedestrian V2V Vehicle To Vehicle WBAN Wireless Body Area Network viii CHAPTER 1 INTRODUCTION The dream of a connected world with wireless health monitoring, wearable devices, and au- tonomous driving has created an increase on the demand of antennas. New communication chan- nels have been developed to foster the growing market. The new antennas have to be compact, so that more than one antenna can be mounted on a single device, show adequate performance capabilities and serve more than one application at a time. In biomedical applications, implantable medical devices (IMDs) are envisioned to be used in a wide range of clinical applications to assist and optimize various medical procedures. Data from within the human body can be transferred wirelessly to the external world with the use of im- plantable antennas. Before using implantable antennas, the devices within the human body were powered and monitored from the external world using wires, through surgical procedures. RFID technology uses radio waves for data collection and transfer, without human intervention, for this reason it has been adapted by healthcare. An RFID system consists of hardware (electronic chip, antenna, reader, scanning antenna) and software components. The combination of the electronic chip and the antenna is known as a tag. Depending on the powering of the RFID tags, they are separated in two categories, the passive and the active. Passive tags are powered on by the elec- tromagnetic wave coming from the reader, as they do not have batteries; while active RFID tags have a battery power source to broadcast the response signal to the reader. An RFID reader is a device connected to the network, and uses radio waves to transmit signals that activate the tag. The tag sends a wave back to the reader antenna. The received wave is translated into data through the software. RFID combined with IMD can be used for tracking, identification and verification, [1], sensing [2], interventions, alerts and triggers, as stated in [3]. There are barriers to incorporate RFID technology to IMD. Antenna impedance has to match with the radio frequency identification chip (RFIC) to avoid losses. Human body consists of many biological materials that are charac- terized with different relative permittivity and conductivity values. For example, the skin, the fat, the blood, the bones, and the muscles have different dielectric properties. The different dielectric 1 properties of the human organs modify the antenna impedance, affecting the matching between the antenna and the RFIC and creating an increase in the overall losses of the system. Consequently, implanting an antenna inside the human body is a very challenging task. The design of the an- tenna is required to fulfill several conditions, such as miniaturization of the antenna dimension, biocompatibility, the satisfaction of the Specific Absorption Rate (SAR), and efficient radiation characteristics, as stated in [4]. There are many studies on the appropriate implantable antenna design for RFID applications. In this dissertation I present a loop and a dipole antenna connected with an energy harvester and operating inside a mix mimicking the human muscle and outside the human body. In the book [5], published in 2012, is stated that vehicles already have 24 antennas and this number will rise in the next years. Automotive antennas are used for entertainment systems like AM/FM radio, in-vehicle TV, digital audio broadcasting (DAB), satellite digital audio radio ser- vice (SDARS), for driver information, safety, and driving experience optimization like satellite navigation (GPS), tire pressure monitoring system (TPMS), remote keyless entry (RKE), cellular phone, Wi-Fi, bluetooth, electronic toll collection (ETC), vehicle to vehicle communications, vehi- cle to infrastructure communications, and collision avoidance radars. The antennas for automotive communications must fulfill particular radiation performance requirements, depending on the ap- plication. Common prerequisites for most automotive antennas mounting are the large distance from the ground, minimum coupling with other metallic structures or antennas, short distance to the receiver, large distance from spurious emissions, and specific polarization. The growing num- ber of the antennas in combination with the prerequisites for the antennas to operate successfully on a vehicle has led researchers investigating new mounting positions and materials. The most common place to mount automotive antennas is the vehicle rooftop that acts as a large ground for the antennas. Inside a shark fin can be found AM/FM, 4G LTE, GPS navigation and satellite radio antennas. Automotive radar sensors require directive antennas and are placed behind bumpers in the front and the back of the vehicle. Remote keyless entry antennas are most of the times placed in four different locations inside a vehicle. Most common areas are in the front, the back and the 2 two sides of the vehicles. It is understood that the amount of the antennas on a vehicle is not only affected by the numerous communication applications but also by the body of the vehicle. One application might require the mounting of several antennas in different locations on a vehicle to perform successfully. With the growing number of required antennas, automotive industry has started investigating new methods to mitigate this issue. 1.1 Additive Manufacturing of Antennas Development of antennas on new materials, using additive manufacturing techniques is a technol- ogy that can provide solutions to the automotive antenna mounting issue. Additive manufacturing (AM) has become one of the fastest growing fields of interest in the past decade with increased demands for rapid prototyping. Additive manufacturing was introduced in the 1980, [6], by Hideo Kodama, who invented two additive methods for fabricating 3D models. Hideo Kodama developed the first protoype machine based on the 1971 patend by Johannes F Gottwald on the Liquid Metal Recorder. The Liquid Metal Recorder was a continuous inkjet metal material device that could form a removable metal fabrication on a reusable surface, [7]. In 1984, Charles Hull developed the material Stereolithography Apparatus (SLA) and established the first 3D printing company in 1986, that produced the first 3D printing machine in the year 1987. In 1991, three new, different additive manufacturing technologies were commercialized. Scott Crump invented a technology called fused deposition modeling (FDM), and developed his company called Stratasys, as stated in [8]. Cubital company released the solid ground sintering (SGC) technology and Helisys intro- duced the laminated object manufacturing (LOM) AM technology. Numerous Additive Manufac- turing techniques have been employed in recent years for the implementation of antennas, most of which can be divided into two main categories: polymer/dielectric 3D printing and all-metal 3D printing. The most common fabrication processes are inkjet printing, aerosol jet printing (AJP), selective laser melting (SLM), fused deposition modeling (FDM), and stereolithography appara- tus (SLA), [9]. Additive Manufacturing allows the development of various antenna designs, such as patch antennas, [10], multi input multi output (MIMO) antennas, [11], fractal antennas, [12], 3 3D wireless local area network (WLAN) antennas, [13], Fabry–Perot resonator antennas, [14], vi- valdi antennas, [15], and foldable monopole antennas, [16]. The 3D printed antenna performance is affected by the quality of the printed parts. The resolution of the printer, any limitations on the structure, the finish of the printed surface are among many intangible factors that affect the antenna performance. Antennas with fine geometry details, 3D curved substrates and substrates with low melting temperature impose stringent requirements on the manufacturing methods. I investigated the use of Aerosol Jet Printer for the development of antennas on various substrates, experimenting with different silver inks. The development of antennas using silver (Ag) nanoparticles (NP), requires post-deposition sintering treatments to achieve the optimum electrical conductivity, because the growth mechanism involves coalescence of discrete nanoparticles. After Ag NPs based conductive ink is printed on the substrate, the sintering process is often needed to remove or decompose the protective agents from the surfaces of Ag NPs, enabling direct physical contacts between Ag NPs, and to establish a dense and conductive network throughout the printed feature, [17]. As the devices are usually printed on heat sensitive flexible substrates, it is crucial to keep the sintering in a moderate condition. Many efforts have been made to create a silver ink that can be sintered in low temperatures, [18], [19]. Research has also been developed around sintering methods that do not incorporate high temper- atures, like laser sintering, [20], plasma sintering, [21], microwave radiation, [17], and photonic sintering using Intense Pulsed Light [22]. I investigated the application of Intense Pulsed Light on white colored automotive plastic and vero-material to sinter three different silver inks. 1.2 Objective of this dissertation The objective of this dissertation is to expand on the state of the art antennas design, development and electromagnetic properties characterization of AM feedstocks for printing antennas and other RF devices, for biomedical and automotive applications. Through this research I tried to iden- tify the best antenna design for the application and select the appropriate substrate and additive manufacturing method to develop the designed device. 4 Antennas implanted inside human body incorporates losses that demean the antenna perfor- mance. RFID technology is utilized to detect antennas implanted within human body by an an- tenna external of human body. I investigated on the best antenna design that when connected to an energy harvester can be detected by a reader antenna in two extreme environments; a) when implanted inside human muscle and b) in free space. Automotive applications demand mounting of numerous antennas inside a single shark fin which eventually affects the antennas radiation characteristics. The limitation of space for antenna mounting has led researchers to investigate new materials and methods to develop them. Through the second study of this dissertation, I researched on a solution to accommodate the large amount of antennas required for automotive applications in combination with the limited space where an antenna can be placed on a vehicle due to its Perfect Electric Conductor (PEC)-like body. Multi- band antennas where a single antenna can cover more than one bands, as well as investigating placements of an antenna throughout the whole car body and not restrict the antenna placement only at the vehicle’s rooftop could solve this issue. Moreover, Additive Manufacturing (AM) can streamline the development of automotive antennas by implementing devices in less time and with less cost than the conventional methods. Printed nano-particle materials require a thermal sintering process to achieve the desired physical properties. The duration of thermal annealing can derail mass production and the required high temperatures forbid its use to automotive plastics like the Acrylonitrile Butadiene Styrine (ABS) that has a glass transition temperature of 102o C, as stated in [23]. In this work the Intense Pulsed Light (IPL) annealing method is tested on the development of antennas for automotive communications on ABS substrate. IPL facilitates a xenon lamp that emits radiation between the Ultra Violet (UV) and Infra-Red wavelengths. Moreover, reactive silver ink that needs minimum or no sintering to be conductive, is tested on vero-white substrate with a glass transition temperature of 50o C, [24]. 5 1.3 Dissertation Organization Chapter 2, expands on rectifying antennas (rectennas) for RFID biomedical applications. The best rectenna design to operate at two extreme environments; within human muscle and in free space is identified and implemented. The developed rectenna is connected with an energy harvester and its performance is measured inside a mix mimicking the human muscle and in free space. Part of this chapter has been published in [25] and [26]. In chapter 3 I introduce a dual-band antenna, covering cellular and Cellular- Vehicle -To- Everything communication bands for automotive communications. The antenna is designed on a planar flexible substrate and tested while mounted initially on various vehicular plastics and then on a real vehicle. The effect of the vehicular plastics on the antenna performance is measured inside a near-field measurement system. Then the antenna is transferred in an outdoor far field measure- ment system and tested on a rear vehicle. The effect of the vehicle body and the ground on the antenna radiation characteristics is measured. Part of this chapter has been published in [27], [28], and [29]. Chapter 4 proposes the use of photonic sintering on 3D-printed silver ink, to develop anten- nas on vehicular thermo-sensitive plastic. The antennas are printed using AJP on clear ABS and sintered using Intense Pulsed Light. The conductivity of the antennas and their performance is measured in a near field system under two setups; a) while the antennas are mounted on a stan- chion and b) while the one of the antennas is mounted on a mirror cover of a Bronco vehicle. The first part of this chapter has been published in [30]. Chapter 5 demonstrates the effectiveness of IPL when sintering different thickness of silver traces. Co-Planar Waveguide (CPW) lines with thickness varying from 4µm to 12µm are printed using AJP on clear ABS. The CPW lines are afterwards sintered using two different photonic profiles. The conductivity and performance are measured and compared. This chapter has been published in [31]. In chapter 6 I introduce a helical antenna for C-V2X communications. The antenna is tested in- side a Car High Mount Stop Light (CHMSL) of a real vehicle inside a near field system. The effect 6 of the vehicle body and the antenna mounting position on the antenna performance is identified. Chapter 7 explains the method to 3D print the helical antenna that is demonstrated in chapter 6. A fully 3D printed antenna is implemented using Additive Manufacturing techniques and its radiation properties are measured in a far field system. The measurement results are presented and compared to the simulations. Finally, chapter 8 gives a conclusion of the dissertation and a summary of possible future work. 7 CHAPTER 2 RECTIFYING ANTENNA FOR RFID BIOMEDICAL APPLICATIONS 2.1 Background As discussed in chapter 1, IMDs are used to transfer information from inside human body to an external receiver wirelessly, in real time and without human intervention. One dangerous medical implication that can be avoided using IMDs is gossypiboma. Gossypiboma is caused by medical sponges or laparotomy pads retained unintentionally inside the human body after a surgery. Doc- tors and nurses use numerous medical gauges to gather the blood of the patient during surgical procedures. When these gauges are forgotten inside the human body, the blood gathered at the sponge causes clots which are dangerous to human life. So far, nurses count the sponges before the surgery and re-count them after the surgery. If there is mismatch between the two numbers, the patient is going through X-Rays in order to identify the position of the forgotten gauge. To opti- mize the procedure of identifying forgotten sponges and avoid gossypiboma, commercial solutions have been available since 2007. These commercial devices use near field and RFID technologies. RFID chips are implanted in the gauges used during the surgery. After the surgery, the nurse, has to scan the patient using a hand held reader antenna and verify whether there are any gauges identified within the human body. There is not a commercially available solution using far field technology so far. The idea of having a set antenna reader, mounted above the patient bed, in a distance that will allow the doctors to perform the surgery and provide medical stuff with real time information regarding the sponges position inside the human body motivated us to perform this research. RFID systems consist of two main components: tag and interrogator. The interrogator is the RFID reader connected to a reader antenna. Tags are basically microchips connected with anten- nas. Tags are classified as passive and active depending on the method used to power on. Passive tags consist of an antenna and an energy harvester. They are powered on by a signal coming from the interrogator of the RFID system. Namely, the interrogator of a passive RFID system transmits an RF wave, which activates the tags. The energy of the wave transmitted by the interrogator is 8 received by the antenna of the tag and is rectified through a circuit on the electronic chip of the tag. If the power received at the terminals of the attached chip is higher than the sensitivity thresh- old of the chip, the backscattered signal is modulated by allowing the chip to alternate between two distinct states, having different values of input impedance. This pre-programmed switching between the two distinct states translates into a unique identification code of the tag, as described in [32]. Active tags have batteries and can emit radio signals to readers without the use of an energy harvester. RFID technology combined with far-field is especially useful in biomedical applications, as medical data can be collected by an implantable device and transmitted wirelessly to an external receiver, allowing continuous real time feedback, without the need of human intervention. How- ever, far-field detection is extremely challenging due to high tissue loss, reflections and changes of antenna behavior when surrounded by high-permittivity material. Antennas for IMDs have been developed for many applications. In [33], Gianmario Scotti et al. designed a loop antenna on a bio-compatible substrate, connected it with a commercial RFID chip and tested the detection distance under several environments, like pork, air and saline mix. The proposed antenna could be detected up to 2.5 cm away from the reader antenna when it was under pork, 0.8 cm away from the reader antenna when was submerged in the saline mix and 3.5 cm away from reader antenna when was left in air. Shubin Ma et al., in [34], proposed a spatially distributed antenna system to serve as the radio platform for RFID-inspired brain care applications. Their antenna system could enable remote powering and readout of a -18 dBm RFID microsys- tem at a distance of 1 m at 915 MHz with 10 mm implant depth. A broadband implantable loop antenna covering the medical device radio-communications service (MedRadio), and the Indus- trial, Scientific and Medical (ISM) bands is proposed in [35] by Rula S. Alrawashdeh et al. The gain of the rectenna, when placed in pork phantom, at 915 MHz was -21 dB. A four layer, triple band rectenna is suggested in [36], by Huang et al. The antenna was tested in minced pork of 65mm*92mm*50mm size and was detectable by an outside monopole antenna at a distance of 50cm. The antenna designed for this biomedical study is operating at 915MHz and has a higher 9 Ref. Antenna Dimensions Frequency (GHz) Gain (dB) in tissue Detection (mm2 ) Distance (cm) [35] 30 × 15 0.915 - 21 - [36] 10 × 10 2.45 -15 50 This work 38 × 12.7 0.915 -20 57 Table 2.1: Performance comparison of implantable rectennas. From [25] @ 2019 IEEE. gain than the one suggested in [35]. Compared to [36] it has a longer detection distance although it operates in lower frequency. The comparison of the antenna implemented for this research project is compared to the other two in Table 2.1. This is the first rectifying system that can be detected both submerged in human muscle and outside human body at a distance greater than 50cm, at 915 MHz, making it suitable for detecting medical gauges inside human body in the far field. 2.2 Design As part of this research project, three different RFID tags were designed and tested. The objective of this study was to design a rectifying antenna (rectenna), operating at 915 MHZ (Ultra High Frequency) that can exchange signals with the RFID interrogator both when it is implanted into muscle tissue of a human model and when it is left in free space. For this project, two antenna designs were implemented, one loop antenna and one dipole like antenna. The loop antenna de- sign was connected with the commercial EM4124 RFID chip and the dipole antenna design with a custom-made energy harvester, developed by Aksoyak and Dr. Ulusoy, and the commercial EM4324 RFID chip. Details on the custom made harvester can be found in [26]. The maximum distance between the implanted tag and the RFID interrogator at which the implanted tag could be detected was measured and compared with the detection distance that was analytically calculated using the Free Space Path Loss (FSPL) equation, 2.1. 2.3 Fabrication Process The design challenge is to maintain a reasonable antenna impedance for the two extreme condi- tions: a) in air and b) when submerged into a high-permittivity material, e.g. human muscle. For 10 this purpose, a single layer patch antenna was adopted as the radiating element. A commercially available chip was placed on the same substrate of the antenna and wire-bonded to it. The antenna design was optimized to deliver sufficient power to the IC in these two extreme operating condi- tions. The implemented loop rectenna is able to communicate with reader antenna at a distance of 6m in free space and of 0.5m-0.9m when placed inside a mix mimicking human muscle. The dipole-like antenna was designed and implemented on the same substrate but is smaller in size than the loop antenna. Two different experiments were performed on the dipole-like rectenna. For experiment (A) the dipole-like antenna was connected with a custom-made energy harvester and tested under these two extreme conditions, namely in free space and in the mix mimicking hu- man muscle. The output DC power of the energy harvester remained above 0.8V at a maximum distance of 125cm from the reader when the tag was in free space and at a maximum distance of 57cm when the tag was submerged in the human muscle model. For experiment (B) the dipole-like antenna was connected with with a custom-made energy harvester and the active EM4324 RFID chip. The tag could be detected at a maximum distance of 81cm from the reader in free space but could not be detected when it was submerged in the human muscle model. 2.3.1 Antenna and muscle design Antenna geometry The geometry of the first patch antenna that was designed is shown in Fig. 2.1. It consists of a loop in the middle of two patches. To achieve impedance matching with the chip, antenna impedance has to be inductive, which is the main motivation to choose a loop antenna design initially. Patches were added later to increase the antenna gain. The main design challenge is maintaining sufficient gain and desired impedance value (i.e., complex conjugate of the commercial IC) with minimum dimensions. For this purpose, multiple antennas were designed with varying size and geometry. The presented design, in contrast to similar designs with 10mm shorter length, achieves almost 3 dB higher gain when placed in human muscle. Furthermore, compared to shorter length designs, 11 Figure 2.1: The first proposed implantable rectenna with its dimensions as implemented with lithography on Rogers R3010 substrate. From [25] @ 2019 IEEE. the impedance of this antenna remains inductive in both extreme conditions, human tissue and air, delivering higher signal power to the chip. The antenna was implemented on a Rogers R3010 substrate using lithography. The R3010 substrate has a permittivity of ϵr = 10.2 and tanδ = 0.0022. The thickness of the substrate is 1.28 mm and of the antenna copper 6 µm. A superstrate of the same material and thickness was used to protect the antenna from corrosion when implanted in human tissue model. The chip was glued on the substrate using the Epo-Tek H20E silver epoxy and then it was wirebonded with the antenna. For the prototype implementation, the superstrate was attached on the substrate using tape, as illustrated in Fig. 2.2. The addition of the superstrate makes the antenna radiation more omnidirectional, improves its gain and decreases impedance changes when switching between free space and human tissue. The design parameters of the antenna are adjusted to ensure that the antenna impedance will match with the EM4124 chip impedance, 22- j261 Ohms, at 915 MHz. The second antenna was designed and connected with custom-made energy harvesters. The impedance of the dipole antenna was matched to the impedance of the energy harvesters with the use of commercial lumped inductors because each harvester has a different impedance. The new design looks like a dipole with extended arms, as shown in Fig. 2.3. The energy harvester was glued using thermally conductive epoxy on the same substrate of the antenna and then was wire-bonded with the antenna. This antenna was also protected using the same superstrate during experiments. 12 Figure 2.2: The tag with the superstrate. The superstrate was cut in an appropriate shape in order to accommodate the wire-bonds without breaking them. From [25] @ 2019 IEEE. (a) (b) Figure 2.3: The second proposed implantable rectenna: (a) with its dimensions as implemented with lithography on Rogers R3010 substrate; (b) The tag with the superstrate. The superstrate was cut in an appropriate shape in order to accommodate the wire-bonds without breaking them. Mix mimicking human muscle The tissue mimicking mix has electrical parameters similar to human muscle at 915 MHz. Accord- ing to [37] the human muscle has a permittivity of ϵr = 54.99 and conductivity σ = 0.95 at 916.5 MHz. The largest human muscle, as stated in [38], is the gluteus maximus with a cross section of 4842 mm2 . The human muscle mimicking mix is created in a cylindrical bowl with diameter of 100 mm and height of 60 mm. The mix consists of water, sugar and salt with quantities appropriate to match the human muscle properties. The dielectric properties of the mix were measured with the N1501A slim form dielectric probe kit of Keysight Technologies. 2.3.2 Commercial chip EM4124, custom made energy harvesters and commercial chip EM4324 The commercial EM4124 chip has an impedance of 22-j261 Ohms at 915 MHz. It operates at 860-960 MHZ and is powered by RF energy transmitted by the reader. This RF energy is rec- tified to generate a supply voltage for the IC. The read sensitivity of the chip is -19dBm under 13 Figure 2.4: The block diagram of the commercial EM4124 chip. Chip Number Measured Impedance (Ohms) Simulated Sensitivty (dBm) Number 1 10-j1120 -35.695 Number 2 10-j1150 -37.3761 Number 4 11-j922 -35.0432 Number 5 12.5-j496 -28.35 Number 7 11.8-j492 -20.5821 Number 8 45-j220 -20.5892 Number 9 46-j220 -27.2207 Number 10 12.5-j496 -26.4112 Table 2.2: Measured impedance and simulated sensitivity of custom made energy harvesters. complex-conjugate matching conditions. The block diagram of the EM4124 chip is illustrated in Fig. 2.4. The custom-made energy harvesters were designed in order to increase the detection dis- tance between the tag and the reader, by increasing the absolute value of the chip sensitivity. The custom made energy harvesters are fabricated in TSMC 65-nm CMOS technology and consist of a rectifier, a power management unit (PMU) and a load. The RF energy received by the antenna is converted into DC power and stored in a capacitor in order to be delivered to the load. The power management unit (PMU) decides the intervals for power delivery. On the same reticle there are 10 different chips designed as shown in Fig. 2.5. Each chip has a different impedance and sensitivity. The measured impedance and simulated sensitivity of each one of the chips are illustrated in the Table 2.2. The custom made energy harvesters cannot create and transmit an identification num- ber. For this reason an oscilloscope is used in order to measure their output DC power during the experiments. The commercial EM4324 chip was bought in order to be connected with the custom-made energy harvesters and provide the tag with the ability to send an identification number, since the custom-made energy harvesters do not have the circuit that creates an ID. 14 Figure 2.5: The custom made energy harvester. Ten different designs on the same reticle. The commercial EM4324 chip is both a passive and battery assisted UHF RFID chip. For our experiments it was used as passive chip and was powered on by the DC output of the custom-made energy harvesters. It has an impedance of 11-j155 Ohms at 915 MHz and is connected with the custom-made energy harvesters which provide the EM4324 chip with the required DC power. As mentioned before, the EM4324 is used in our design in order to provide the reader with an ID number, when the energy harvesters are used. 2.4 Results 2.4.1 Simulation results Antenna performance in free space In free space, the loop antenna, covered by the superstrate and substrate has a maximum gain of -3.8 dBi, as illustrated in figure 2.6a. The impedance of the antenna is 1.845+j22.55 Ohms and the reflection coefficient, S1 1, is -21.36 dB, at 915 MHz. The dipole-like antenna covered by the superstrate and substrate has a maximum gain of - 0.5 dBi in free space and an impedance of 0.275-j207.705 Ohms before adding the inductors as illustrated in Fig. 2.7. 15 (a) (b) Figure 2.6: Loop antenna’s maximum gain: (a) in free space, as simulated in HFSS; (b) in human tissue as simulated in HFSS. From [25] @ 2019 IEEE. (a) (b) Figure 2.7: Dipole antenna’s maximum gain: (a) in free space, as simulated in HFSS; (b) in human tissue as simulated in HFSS. Antenna performance in human muscle model For human muscle, a cylindrical box with radius of 50 mm and height of 60 mm is used with permittivity of ϵr = 54.99, conductivity σ = 0.95 and tanδ = 0.339. The antennas were placed in the middle of this cylindrical box, covered by superstrate and substrate. The maximum gain of the loop antenna is -20 dBi, as illustrated in Fig. 2.6b. The loop antenna impedance is 6.94+j228.955 Ohms. The input impedance of the commercial IC EM4124 is 22-j261 Ohms. 16 The dipole antenna’s gain is -24dBi and its impedance, before adding the commercial inductors is 8.42-j22.91 Ohms as shown in Fig. 2.7. 2.4.2 Experiments results For the experiment configuration, a Zebra Fx9500 fixed RFID reader and two Zebra fixed AN440 RFID reader antennas were used. The power output of the reader is 31–33 dBm. The reader’s antennas gain is 6 dBiL and they were placed one vertical and one horizontal to the mix so that polarization will not affect the transmission and reception of signals, as indicated in Fig. 2.8. When testing the tags with the commercial chips EM4124 and EM4324, the signal transmitted by the reader antennas is received by the designed antenna and the EM4124, EM4324 chips are activated. A signal containing the ID of the chip is transmitted through the designed antenna to the reader antenna. The ID of the chip is printed on the PC screen, using the reader software. To test the tag that consists of the designed antenna connected only to the custom-made en- ergy harvester, we connected the tag to the oscilloscope to measure the DC output of the energy harvester. The RF signal transmitted by the reader antenna is received by the designed antenna and transferred to the energy harvester. The energy harvester converts this RF signal to DC power which is measured by the oscilloscope as illustrated in Fig. 2.9. Antenna performance in free space The maximum detectable distance of the rectenna can be estimated through the free space path loss (FSPL) equation. 4π F SP L = 20log10 (d) + 20log10 (f ) + 20log10 ( ) − GT − GR (2.1) c The gain of the loop antenna is -3.8 dB according to simulations. The EM4124 chip sensitivity is -19 dBm and the loss due to impedance mismatch between the chip and the designed antenna’s tag is 11 dB, according to Matlab code. As a result, the maximum distance between the reader 17 Figure 2.8: Experimental configuration of reader’s antennas and tag inside the human mimicking muscle mix. The mix is centered in the middle of the two reader’s antennas. The rectenna is submerged in the bottom of the mix, in the middle of the cylindrical bowl. From [25] @ 2019 IEEE. antenna and the tag antenna is expected to be 4 m. During the experiments, the tag stopped being detected after 6 m away from the reader antenna. The dipole antenna for experiment (A) was connected only with the energy harvesters num- bered 5,7,8 and 9 and was tested in free space. For this experiment, the tags were connected to the oscilloscope and were held in front of the reader antenna. The DC output for different distances from the reader was captured. The energy harvesters were designed to output a DC voltage of 0.8- 1.2 V. The distance from the reader antenna at which the DC output voltage starts falling below 0.8V is shown in Table 2.3. The tag at the experiment (B) consists of the dipole antenna connected both with the custom- made energy harvesters and the commercial EM4324 chip. The antenna was wire-bonded with 18 Figure 2.9: The dipole-like antenna connected with the oscilloscope. Chip Number Measured detection Analytically calculated expected (cm) for 0.8-1.2V (IN AIR) detection distance (cm) (IN AIR) Number 5 90 77 Number 7 100 70 Number 8 125 77 Number 9 92 450 Table 2.3: Performance of tag with custom made energy harvester, experiment (A). Chip Number Measured detection Analytically calculated expected distance (cm)(IN AIR) detection distance (cm) (IN AIR) Number 6 77 80 Number 7 81 100 Table 2.4: Performance of tag with custom made energy harvester and EM4324, experiment (B). both the energy harvester and the commercial EM4324 in parallel. The commercial EM4324 is able to send, through the designed antenna, a signal with an identification number back to the reader. No oscilloscope is needed in this case. The maximum distance from the reader antenna that the tag can be detected was measured and is shown in Table 2.4 19 Chip Number Measured detection Analytically calculated expected Chip Number distance (cm)(IN MIX) detection distance (cm) (IN MIX) Number 5 57 57 Number 7 30 9 Number 8 20 13 Number 9 50 60 Table 2.5: Performance of tag with custom made energy harvester in mix, experiment (A). Antenna performance in human muscle model According to [37] the absorption loss within human muscle, at 916.5MHz is given by LA = 206.81 × z (dB), where z is the traveled distance of the signal in meters. The loss due to impedance mismatch between the EM4124 chip and the loop antenna is 1.81 dB. The distance between the mix containing the tag and the reader antenna, as calculated using the FSPL equation is 54 cm. The maximum detection distance between the mix, when the tag is in the bottom of it and the reader antenna above the mix, is 90 cm. When the reader antenna was placed horizontal on the table and the mix was on the table with the tag antenna submerged in its bottom, in the middle of the cylindrical bowl, the maximum detection distance was 51 cm. Using the same methodology, the experiment (A) of the dipole antenna connected with the chips number 5,7,8 and 9 was performed in the mix mimicking the human muscle.The reader an- tenna was held above the mix and the DC output for different distances from the mix was measured. The distance from the reader antenna at which the DC output voltage starts falling below 0.8 V is shown in Table 2.5. The experimental and analytically calculated detection distance of the dipole antenna connected with the custom-made energy harvester and the commercial EM4324 is indicated in Table 2.6. The maximum analytically calculated detection distance is 5.5 cm which is considered low. When the two chips are connected in parallel, the total impedance is equal to the parallel impedance of the energy harvester and the EM4324 chip. The chip sensitivity is affected by the impedance value and as a result when the impedance changes, the sensitivity decreases causing the maximum detection distance to reduce. 20 Chip Number distance (cm)(IN MIX) Analytically calculated expected distance (cm)(IN MIX) detection distance (cm) (IN MIX) Number 6 0 5 Number 7 0 5.5 Table 2.6: Performance of tag with custom made energy harvester and EM4324 in mix, experiment (B). 2.5 Conclusion An antenna connected to commercial and in-house implemented energy harvesters is developed. The performance of the antenna in the air and a mix mimicking the human muscle is measured. Unlike conventional RFID tags, the proposed antenna is tolerant to environmental changes, thus, capable of operating in these two extreme environments. The antenna developed through this research operates in two extreme environments and compared to other RFID antennas in literature, operating at 915 MHz, the proposed antenna offers the best detection distance compared to its size, to the best of my knowledge. The combination of the implemented antenna and the low sensitivity of the in-house implemented energy harvester offers large detection distance to this RFID tag when it is submerged to the lossiest part of human and when left outside the human body. This RFID tag can be developed in large production lines and used commercially by healthcare to avoid incidents of gossypiboma. 21 CHAPTER 3 AUTOMOTIVE ANTENNAS FOR CELLULAR AND C-V2X APPLICATIONS 3.1 Background The study about automotive antennas started in 2019. The Cellular Vehicle To Everything (C-V2X) bands have gained great attention the last few years for enabling vehicles to communicate with their surroundings at higher rates. Some potential applications of the C-V2X technology includes communication between a vehicle and the network (V2N), a vehicle and the infrastructure (V2I), a vehicle and a pedestrian (V2P) and two vehicles (V2V). The C-V2X is a novel technology based on cellular modem. Before developing the C-V2X band, the dedicated short range communication (DSRC) bands were utilized to connect a vehicle with other vehicles. C-V2X was introduced by 3GPP in 2017 as an extension of LTE. As a novel technology, it requires new installations on the network infrastructure. For this reason, the automotive antennas designed during this study are desired to cover frequencies within the cellular bands and take advantage of the already deployed cellular network infrastructure. As stated in chapter 1, vehicular communications demand the use of numerous antennas but the metal nature of the vehicle body restricts the places where an antenna can operate normally. This is the main issue we tried to resolve through this research by implementing multi-band antennas and antennas on a flexible substrate that can be adjusted on vehicle plastic parts, occupying limited space. Most of the cellular antennas in literature are implemented on a rigid substrate and are placed inside a shark fin on a rooftop of a vehicle, taking advantage of the metal rooftop that acts as a ground plane, boosting the antenna performance. Automotive industry, has started looking for solutions to hide the numerous antennas while keeping their performance. The main issue of antenna mounting is observed on convertible vehicle that do not have a steady rooftop. The antenna mounting positions are really limited on these vehicles. For this reason, researchers started investigating, other parts of the vehicle to house an antenna. Koch et al., in [39] developed a WLAN 22 antenna on a thin flexible 35µm copper foil with adhesive and placed it into the plastic spoiler of a hatchback. Abbas et al., in [40], evaluated the use of diversity antennas for Vehicle to Vehicle (V2V) applications by placing antennas on the rooftop, side mirror, windscreen and bumper. The side mirror is an attractive place for housing antennas since it is not that low to interfere with the ground and it also provides the plastic cover that can protect the antenna. In [41], Imai et al., mounted a half wavelength dipole antenna on a right side mirror and investigated numerically the effects of the car body on the radiation of an antenna used for inter-vehicle communications. Marantis et al., placed a monopole inside a truck mirror and tested it in the lab without the vehicle’s body in [42]. Hasnain et al., in [43] placed a vehicle in an anechoic chamber and positioned two sets of antennas at the left and right side mirrors in order to test the efficacy of the antenna sub- arrays for robust satellite navigation. Virothu et al., implemented circularly polarized antennas covering the 2.5-2.57 GHz, 2.62-2.69 GHz and 3.4- 3.8 GHz bands and measured them on a planar and curved surface, imitating the side mirror cover in [44]. In [45], Asghar et al., investigated the effect of the antenna position on its radiation. An antenna operating at 700 MHz, 2600MHz and 5870 MHz was mounted on different positions on a vehicle’s rooftop. Measurements on the various locations showed that the effect of the mounting location is negligible at lower frequencies, but as the frequency tends to increase, it becomes significant. Singh et al., in [46], addressed the influence of the plastic embedding and the door frame on the performance of the antenna. The Acrylonitrile Butadiene Styrene (ABS) on the door frame affects the antenna frequency of operation by causing a shift to it. Through this research a suitable design for dual-band antenna operating at two frequencies with 1 a wavelength ratio of 7 was identified. The antenna was fabricated on a flexible substrate using lithography process and was mounted on different automotive plastic parts to study the impact of the different shaped parts on the radiation of the antenna at the cellular and C-V2X bands. As a second experiment, the antennas were tested on a real convertible vehicle and the impact of the ground, of the vehicle surroundings and of the car body on signal transmission and reception on the dual-band microwave antenna was investigated. This study can be used for deciding embedded 23 antenna locations in future vehicles. For the first experiment, two different antenna designs, im- plemented on the flexible Rogers 3850 (Liquid Crystalline Polymer) substrate were mounted on a plastic trunk lid, a mirror cover and a curved plastic retrieved from the inside of a real vehicle. The radiation patterns, efficiency and realized gain of the antennas mounted on the automotive parts were measured in Satimo SG32 near field system. The performance of the antennas on the various plastic parts was compared to the performance of the antennas in literature in Table 3.1 from [28]. The maximum realized gain values reported in Table 3.1 of this work refer to the measurements of the antennas alone, without any plastic below them and when they are mounted on the inside surface of the side mirror cover. The term free under the column of antenna position in vehicle in Table 2.2 refers to the antenna measurements when they are measured alone, meaning that there is no plastic below them. The term curved in the same column of the table refers to the placement of the antenna on the inside surface of the side mirror. Compared to the antennas in literature, the antennas presented here operate at two frequencies with a wavelength ratio of 17 . At 800 MHz the antennas have omni-directional radiation pattern whereas the antennas in [39] have directional radiation pattern. When comparing the antenna measurements in free space, where the antennas lie flat, and on the curved surface, the values of the maximum realized gain slightly increase for antenna 1, as in [44]. At 5.9 GHz both antennas presented here have a slightly lower maximum value of realized gain compared to the [42]. Namely the authors of [42] state a maximum value of gain around 5 dBi but the antenna 1 and 2 measured here have a gain of 4.6 and 4.2 dBi when measured alone in the Satimo near field system. When placing the antenna 2 on the trunk lid which is a flat plastic, similar to the door frame of [46] the maximum realized gain at 800 MHz is 0.43 dBi and at 5.9 GHz around 5.5 dBi. The gain of the antenna in [46] on the flat plastic is around 3.5 dBi at 2.45 GHz. To the best of my knowledge, this is the only research on antenna performance when mounted on different surfaces of automotive plastics. For the second set of experiments, a single antenna design was implemented on the same flexi- ble substrate that was used during the first experiment. The antenna was measured on a convertible 24 Ref. Antenna Resonance Antenna Maximum Dimensions frequencies position Realized (mm3 ) (GHz) on vehicle Gain (dBi) [39] 2∗λ 2.45 spoiler 12 [44] 44*42*0.1 3.575 planar 3.77 3.45 curved 3.87 [42] 26*18*0.815 5.9 flat side mirror 5 [46] 73.9*36.9*1.6 2.45 door frame 3.5 This work 120*70*0.1 0.8 free 1.68 Antenna1 5.9 free 4.6 0.85 curved 2.5 5.9 curved 4.8 This work 100*70*0.1 0.8 free 1.6 Antenna2 5.9 free 4.2 0.8 curved 0.187 5.9 curved 1.54 Table 3.1: Comparison of this work to the literature. From [28] @ 2022 IET. 2018 Mustang, provided by Ford Motor Company at the outdoor antenna range facilities of Oak- land University. The antenna was mounted initially on a stanchion and was measured on a metallic turn table. Next, the vehicle was brought on the turn table and the antenna was placed inside the mirror cover and finally on the trunk lid of the vehicle. Antenna positions on a vehicle, similar to the mounting that is presented in this research is used in [47]. Senega et al. tested two dislocated antennas for satellite communications as a diversity antenna inside a vehicle’s side mirror. In [48] a Sirius-XM antenna operating at 2326 MHz is placed under the trunk lid and its radiation patterns as calculated by Mathematical Absorber Reflection Suppression (MARS) algorithm is compared with the measured radiation pattern at a far field setup at ATC GmbH premises. Testing facilities for automotive antenna measurements similar to the ones used in this experiment to investigate the effect of the ground and the vehicle body on the antenna signal have been employed in [49] where a Car-to-X antenna is placed on the rooftop of a vehicle. Hasturkouglu et al., in [50], tested a Nefer MIMO system on the rooftop of a car in an antenna radome with a car turntable inside. The difference between the literature and our research is that the radome used in the literature is closed and not affected by external environmental factors. Tazi et al., in [51] investigated how and when it is important to consider the ground and the turn table effect in the simulation models for 25 Ref. Antenna Dimensions Resonance Frequencies Maximum Realized (mm3 ) (GHz) Gain (dBi) [47] 32*32*11 (set of 3 antennas) 2.332 5 [41] 25.4 (λ/2 dipole) 5.9 5 (only simulated) [42] 8.8*1.65*0.035 5.9 3 800 1.68 [27] 120*70*0.1 5.9 4.6 0.975 0.3 This work 91.5*30*0.1 5.4 1.6 Table 3.2: Comparison of this work to the literature. From [29] @ 2021 IEEE. the evaluation and development of car antenna systems, the influence of the body shell and the influence of the glass curvature by testing an antenna operating within 50-300 MHz on a rooftop of a commercial vehicle. An evaluation method of LTE-car antennas is presented in [52] by mea- suring the Antenna Under Test (AUT) on a rooftop of a commercial vehicle, in a closed radome. The antenna performance at the second set of experiments is compared to the literature in Table 3.2 from [29]. To the best of my knowledge, this study presents for the first time the influence of the ground and the vehicle body on a flexible dual band antenna with a wavelength ratio of 17 , mounted on a mirror cover and a trunk lid, for C-V2X communications. 3.2 Design 3.2.1 Antennas tested on automotive plastics As stated, two antenna designs are implemented and their performance is measured during the first set of experiments. Both antenna designs consist of a bow-tie which extends to meander lines and are fed by a strip-line with an SMA connector. The logic behind this design is that the antenna is dual band and resonates at the cellular 4G frequencies (below 1GHz) and at the high C-V2X 5.9 GHz. The bow-tie is resonating at 5.9 GHz and the meander lines provide the resonance at the lower cellular frequencies. At the lower cellular frequency the antenna has a traverse magnetic (TM) resonant mode, like an electric dipole but in the higher C-V2X frequency the antenna has a 26 Antenna Design Variable Dimensions (mm) a1 120 b1 70 c1 77.4 d1 22.8 e1 16.6 Design 1 f1 30.8 g1 9.8 h1 10.1 j1 26 k1 6.5 Design 2 a2 100 b2 70 c2 75.5 d2 30.5 e2 20 f2 13.5 g2 10 h2 16.5 j2 33 Table 3.3: Antennas A and B dimensions. From [28] @ 2022 IET. (a) (b) Figure 3.1: (a)Simulated; (b) and implemented antenna A. From [28] @ 2022 IET. hybrid traverse electric/traverse magnetic (TE/TM) mode due to the size of the meander lines and the feed-line. Both the antenna designs are indicated in Fig.3.1 and Fig.3.2 and their dimensions in Table 3.3. Both of them are implemented on the flexible Rogers 3850 (LCP) using lithography. The LCP has a permittivity of ϵr = 2.9 at 10GHz, tanδ = 0.0025 and thickness t=0.1 mm, as stated in [53]. 27 (a) (b) Figure 3.2: (a)Simulated; (b) and implemented antenna B. From [28] @ 2022 IET. 3.2.2 Antennas tested on real vehicle The antenna design used in the second set of experiments was initially designed using Ansys EM (HFSS) software. The dual-band antenna consists of a bow-tie and two meander lines extending the bow-tie. It is fed by a slot-line, on which an SMA connector is soldered for measurements. The design is indicated in Fig.3.3 and it follows the same logic behind the antennas tested on the automotive plastics. The substrate of the antenna is the same Rogers 3850. In cellular frequencies the simulated antenna resonates at 968.8 MHz and for C-V2X applications, when the antenna is simulated alone, it resonates at 5.5 GHz as shown in Fig.3.4. The simulated co- and cross- polarization radiation patterns of the antenna, without the ABS, at both frequencies, are indicated in Fig.3.5. When an ABS part is added below the substrate, the measured resonance around 7.2 GHz shifts to 5.9 GHz, as indicated in Fig.3.6. The resonance to a lower frequency when there is an ABS part below the antenna is expected since the permittivity of the material on which the antenna is mounted increases. As a result the effective wavelength λef f minimizes but the antenna length remains the same. So the antenna resonates to a lower frequency. The antenna was designed to operate with the ABS below the Rogers substrate. However, during the standalone antenna measurements this resonance shifts to 7.2 GHz and another resonance is created around 5.5 GHz. The required bandwidth for C-V2X applications is only 30 MHz extending from 5.895-5.925 GHz. 28 The simulated antenna with the ABS below the Rogers 3850 substrate is indicated in Fig.3.7a. It is obvious that the antenna is linearly polarized at the cellular frequencies but the cross-pol level is increased at 5.5 GHz and 5.9 GHz. At the lower frequency the antenna’s dominant mode is TM since the antenna lies on xy plane in HFSS and the radiated field is mainly on the theta direction. However, at 5.5 GHz the antenna has a hybrid TE/TM mode. At 5.5 GHz and 5.9 GHz the length of the meander line as well as the size of the slot-line are not negligible when compared to the wavelength. As a result the antenna cross-pol level is increased by 15 dB when compared to the lower frequencies. The antenna has an omni-directional radiation pattern at 968.8 MHz as shown in Fig.3.8a but loses the omni-directionality at 5.5 GHz due to the meander and the slot-line, as indicated in Fig.3.8b. At Fig.3.9 the bow-tie antenna is simulated without the feed-line and the meander lines resulting in an omni-directional radiation pattern and linear polarization at 5.5 GHz, which proves that the meander and the feed-line of the antenna affect the antenna behavior at the C-V2X bands. The maximum realized gain of the simulated antenna with the meander at 968.8 MHz is 1.6 dBi and the efficiency 95.4%, as shown in graphs of Fig. 3.10. At 5.5 GHz the efficiency of the full antenna with the meander is 89.2% and the value of the maximum realized gain is 1 dBi, as indicated in graphs of Fig. 3.11. At the Fig. 3.10a and Fig. 3.11a the realized gain on H-plane is plotted because the Linear Average Gain (LAG) requirements by automotive refer to the gain on H-plane. When the ABS is added below the antenna, the efficiency at 5.9 GHz is 99.7% and the maximum realized gain around 4.2 dBi. After designing the antenna in HFSS, it is fabricated using lithography at Michigan State University. In order to measure the fabricated antenna, an SMA connector is soldered at the pads of the feed-line. The final fabricated antenna with its dimensions is shown in Fig. 3.3 and Table 3.4. 29 (a) (b) Figure 3.3: (a)Simulated and (b)implemented antenna. 0 HFSS simulated S11 -5 -10 X 7.644 S11 in dB Y -8.829 -15 X 5.506 Y -13.49 -20 X 0.9688 Y -23.14 -25 0 1 2 3 4 5 6 7 8 Frequency in GHz Figure 3.4: Simulated S-parameters in HFSS. From [29] @ 2021 IEEE. 3.3 Experiments 3.3.1 Antennas measured on automotive plastics without the presence of a vehicle The antennas are initially simulated in Ansys EM (HFSS) software. They are designed on the LCP 30 substrate with an SMA connector at the edge of the slotline. Both the antenna designs A and B 90 90 120 60 120 60 0 0 150 30 150 30 -20 -20 -40 -40 180 0 180 0 210 330 210 330 240 300 240 300 Co-pol (along) of antenna in HFSS 270 Co-pol of antenna in HFSS 270 Cross-pol (vertical to) of antenna in HFSS Cross-pol of antenna in HFSS (a) (b) Figure 3.5: Simulated co- and cross- polarization radiation patterns (a) at 968.8 MHz and (b) at 5.5 GHz. From [29] @ 2021 IEEE. S11 antenna with ABS: PNA measurements S11 antenna with ABS: HFSS Simulation -2 -4 -6 -8 Magnitude (dB) -10 X 5.9 -12 Y -13.06 -14 -16 X 5.915 -18 Y -15.85 -20 -22 4.5 5 5.5 6 6.5 7 7.5 8 Frequency (GHz) Figure 3.6: S-parameters as simulated in HFSS and measured in Satimo when there is ABS below the antenna. From [29] @ 2021 IEEE. 31 90 120 60 0 150 30 -20 -40 180 0 210 330 240 300 Co-pol of antenna in HFSS 270of antenna in HFSS Cross-pol (a) (b) Figure 3.7: (a)Simulated antenna with ABS piece below the Rogers 3850 substrate; (b)Simulated co- and cross- polarization radiation patterns at 5.9 GHz. From [29] @ 2021 IEEE. 90 90 120 60 120 60 0 0 150 30 150 30 -20 -20 -40 -40 180 0 180 0 210 330 210 330 240 300 240 300 E-plane of antenna in HFSS E-plane of antenna in HFSS 270 270 H-plane of antenna in HFSS H-plane of antenna in HFSS (a) (b) Figure 3.8: (a)Simulated E and H plane radiation patterns at 968.8 MHz and (b) 5.5 GHz. From [29] @ 2021 IEEE. consist of a bowtie resonating at 5.9 GHz and its arms are extended to meander lines, introducing a resonance at cellular bands. The differences between the two designs are the length of the slotline and the size of the bowtie. The antenna design A, at cellular bands, resonates at 817.6 MHz and has an efficiency of 89.8%. The maximum realized gain, according to the simulations is 0.99 dBi. 32 90 120 60 0 -20 150 30 -40 -60 180 -80 0 210 330 240 300 Co-pol of270 bowtie in HFSS Cross-pol of bowtie in HFSS (a) (b) Figure 3.9: (a)Bow-tie antenna without the meander line and the slot-line (b) Co and Cross- polarization of the bow-tie on E-plane at 5.5 GHz. From [29] @ 2021 IEEE. 2 100 0 90 Realized Gain on H-plane (dBi) Radiation Efficiency % -2 80 -4 70 -6 60 -8 50 -10 40 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 frequency in GHz frequency in GHz (a) (b) Figure 3.10: Graph of the simulated (a) realized gain on H plane and (b) efficiency versus the frequency at cellular bands. From [29] @ 2021 IEEE. At 5.9 GHz according to Ansys EM, the S11 is around -10 dB and the maximum realized gain around 2.79 dBi. The efficiency is 98.8%. The antenna A has an efficiency of 99.7% at C-V2X frequency and the S11 at 817.6 MHz is around -9 dB but at 5.9 GHz is around -10 dB. In addition conductive loss is higher due to the lower skin depth over metal thickness ratio. The antenna design B resonates at 886 MHz and at 5.9 GHz has an S11 of -17 dB. At cellular frequencies it has an efficiency of 94.6% and a maximum realized gain of 0.8 dBi. At C-V2X frequencies the simulations give a 100% efficiency and a maximum realized gain of 2.96 dBi. The S11 at 5.9 GHz 33 5 90 0 80 Realized Gain on H-plane (dBi) -5 Radiation Efficiency % -10 70 -15 60 -20 -25 50 -30 40 -35 -40 30 5 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 6 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 6 frequency in GHz frequency in GHz (a) (b) Figure 3.11: Graph of the simulated (a) realized gain on H plane and (b) efficiency versus the frequency close to C-V2X bands. From [29] @ 2021 IEEE. Variable Dimension (mm) a 91.5 b 16.5 c 9.5 d 6 e 4 f 10 g 0.5 h 18 j 10 Table 3.4: Antenna dimensions in mm. From [29] @ 2021 IEEE. is -16.7 dB. The antennas are implemented using lithography and an SMA connector is soldered at their feed-line. A balanced to unbalanced (balUn) device is not used as simulations showed that omit- ting a balUn does not have an impact on the antenna performance at the frequencies of interest. The S-parameters of the manufactured antennas are measured using the M5227 PNA by Keysight Technologies. Then the antennas are placed inside the Satimo system and six different measure- ment setups are tested: (a) the antenna A alone, without any automotive plastic part below it; (b) the antenna A mounted inside the mirror cover of a vehicle, after having removed the metallic parts that were inside the mirror; (c) the antenna B alone, without the presence of any automotive plastic part; (d) the antenna B mounted outside of a mirror cover, after having removed the metallic parts 34 0 0 S11 measured with VNA S11 measured with VNA S11 Simuated in HFSS S11 Simuated in HFSS -2 -5 X 5.905 Magnitude (dB) Magnitude (dB) -4 Y -9.926 -10 -6 -15 -8 X 5.904 Y -14.84 X 0.7717 -20 -10 Y -8.478 X 0.8344 -12 Y -10.17 -25 0.7 0.8 0.9 1 5.5 6 6.5 7 Frequency (GHz) Frequency (GHz) (a) (b) Figure 3.12: S-parameters of simulated and implemented antenna A (a) in cellular frequencies; (b) in C-V2X. From [28] @ 2022 IET. from the inside of the cover; (e) the antenna B mounted on a flat trunk lid; (f) the antenna B placed on a curved plastic from the A-pillar of a vehicle. Stand alone antennas measurements in SATIMO SG32 near field system The implemented antenna design A resonates at 771.7 MHz and at 5.9 GHz has an S11 of -15 dB, as shown in Fig. 3.12. Compared to the simulations, there is a resonance shift of only 46 MHz. The E-and H-plane of the antenna at cellular and C-V2X frequencies as measured in Satimo near field system, without the presence of any automotive plastic, are indicated in Fig. 3.13, 3.14 and [27] with more details. At cellular 771.7 MHz the measured efficiency of antenna A is 70.6% and the value of the maximum realized gain is 1.68 dBi. At 5.9 GHz antenna A has an efficiency of 84.9% and the maximum value of the realized gain as measured in Satimo near field system is 4.6 dBi. The implemented antenna design B resonates at 830 MHz and at 5.9 GHz has a measured S11 of -12.2 dB. When compared to the simulations there is a shift of 50 MHz to the resonance frequency at the cellular frequencies. This small shift in resonance frequency can be explained by fabrication tolerance. At 830 MHz it has an efficiency of 83% and the maximum realized gain as measured in Satimo system is 1.6 dB. At 5.9 GHz the efficiency of the antenna is measured at 81.2% and the maximum realized gain at 4.2 dBi. The S11 parameters of the implemented antenna are given in Fig. 3.15 for both frequency bands. The radiation patterns of the antenna without the 35 (a) (b) Figure 3.13: The simulated and measured radiation patterns of antenna A at cellular frequencies, without any automotive plastic (a) on E-plane and (b) on H-plane. From [28] @ 2022 IET. (a) (b) Figure 3.14: The simulated and measured radiation patterns of antenna A at C-V2X frequencies, without any automotive plastic (a) on E-plane and (b) on H-plane. From [28] @ 2022 IET. presence of an automotive plastic, E- and H-planes are given in Fig.3.16 and Fig.3.17 for cellular and C-V2X frequencies. At C-V2X there is a resonance shift of 394 MHz. This can be explained by the size of the soldering of the SMA connector. In simulations the soldering is represented by flat rectangles, extending the feeding pads and connecting them to the SMA, with a thickness of 1mil. In reality the soldering consists of two bulbs covering the whole feeding-pads and extending up to the connector. At 5.9 GHz the wavelength is only 50mm which means that the thickness of the soldering can affect the resonance of the antenna. 36 0 0 S11 measured in VNA S11 measured in VNA S11 Simuated in HFSS -5 S11 Simuated in HFSS X 5.9 Magnitude (dB) Magnitude (dB) -5 -10 Y -12.24 X 5.902 -15 Y -16.77 -10 X 0.83 -20 Y -9.481 -25 -15 -30 0.7 0.75 0.8 X0.85 0.886 0.9 0.95 1 5.5 6 6.5 7 Y -14.53 Frequency (GHz) Frequency (GHz) (a) (b) Figure 3.15: S-parameters of (a) simulated and (b) implemented antenna B. From [28] @ 2022 IET. (a) (b) Figure 3.16: The simulated and measured radiation patterns of antenna B at cellular frequencies, without any automotive plastic (a) on E-plane and (b) on H-plane. From [28] @ 2022 IET. Antenna measurements when mounted on mirror cover in SATIMO SG32 near field system At a second step the antennas are mounted on a mirror cover that is empty from the inside and placed inside the Satimo system to measure their radiation patterns and efficiency. The mirror cover is made by Acrylonitrile Butadiene Styrene (ABS) which has a permittivity of ϵr = 2.5 and tanδ = 0.004 as measured by Agilent E4991A RF Impedance Material Analyzer in the lab. Antenna design A is placed initially inside the mirror cover and then on the outer surface of the same mirror cover. Electrical tape is used to stabilize the antenna and the coaxial cable inside the 37 (a) (b) Figure 3.17: The simulated and measured radiation patterns of antenna B at C-V2X frequencies, without any automotive plastic on (a) E-plane and (b) on H-plane. From [28] @ 2022 IET. (a) (b) Figure 3.18: Antenna A inside the mirror. From [28] @ 2022 IET. cover. Then the cover is placed back on the mirror skeleton as indicated on Fig. 3.18, when the antenna is mounted inside the mirror cover. The antenna realized gain and efficiency are measured at 800 MHz and 5.9 GHz. According to HFSS simulations, when a flat ABS part is placed below the Rogers 3850 substrate the resonance shifts from 816 MHz to 800MHz and from 5.9 GHz to 5.3 GHz, for antenna A, as shown in Fig. 3.19. This is expected since adding a material with high permittivity below the antenna substrate, 38 0 0 S11 Simuated in HFSS S11 Simuated in HFSS -2 -5 -4 -10 S 11 (dB) -6 S 11 (dB) -15 -8 -20 -10 -25X 5.32 Y -28.47 -12 X 0.7984 -30 0.7 Y -11.25 0.8 0.9 1 5 5.5 6 6.5 7 Frequency (GHz) Frequency (GHz) (a) (b) (c) Figure 3.19: S-parameters of Antenna A with ABS below the Rogers3850 substrate (a) at cellular and (b) C-V2X frequency bands and (c) the antenna simulation. From [28] @ 2022 IET. the effective wavelength decreases but the size of the antenna is the same so the antenna now will resonate to lower frequency. The antenna has a measured efficiency of 55.5% at 800 MHz and the maximum realized gain is 2.5 dBi for θ = 114o , as shown in Fig. 3.20. The realized gain is measured at 5.9 GHz since this is the frequency of interest for C-V2X applications. The antenna has a measured efficiency of 74.7% and a measured maximum realized gain of 4.8 dBi. The radiation patterns of the antenna A in this placement are given in Fig. 3.21. At a second step, the antenna A is mounted on the outer surface of the same mirror cover. The radiation patterns of the antenna A at this placement are demonstrated in Fig. 3.22 and 3.23 at the cellular and C-V2X frequencies respectively. The efficiency of the antenna A at cellular frequency 39 (a) (b) Figure 3.20: Radiation patterns of antenna A when mounted inside the mirror cover at 800 MHz. From [28] @ 2022 IET. (a) (b) Figure 3.21: Radiation patterns of antenna A when mounted inside the mirror cover at 5.9 GHz. From [28] @ 2022 IET. when mounted on the outer surface of the mirror is 28% and has a maximum realized gain of -0.34 dBi. At C-V2X frequencies its measured efficiency is 63.4% and the maximum realized gain is 4dBi. The antenna B is mounted on the outside surface of a mirror cover, made of ABS, as indicated in Fig. 3.25. Tape and a styrofoam are used to stabilize the antenna with the coaxial cable on the 40 (a) (b) Figure 3.22: Radiation patterns of antenna A when mounted on the outside surface of the mirror cover at 800 MHz. From [28] @ 2022 IET. (a) (b) Figure 3.23: Radiation patterns of antenna A when mounted on the outside surface of the mirror cover at 5.9 GHz. From [28] @ 2022 IET. mirror cover in the Satimo system. The antenna B radiation pattern is measured at 800 MHz when it is mounted on ABS part because according to S11 measurements of antenna B on a piece of the trunk lid which is a flat ABS surface, below the Rogers 3850, the antenna B resonates at 800 MHz for the cellular frequencies, as indicated in Fig. 3.24. In simulations the antenna resonance has shifted to 768 MHz. At 800 MHz the antenna has a measured efficiency of 43.87% when mounted 41 -2 -2 X 5.9 -4 -4 Y -4.943 -6 -6 Magnitude (dB) Magnitude (dB) Measured -8 HFSS Simulation -8 -10 Measured HFSS Simulation -10 -12 X 0.7672 -12 -14 Y -15.71 -16 X 0.79 -14 X 5.9 Y -17.92 -18 -16 Y -13.61 0.7 0.75 0.8 0.85 0.9 0.95 1 5 5.5 6 6.5 Frequency (GHz) Frequency (GHz) (a) (b) Figure 3.24: S-parameters of antenna B when mounted on the trunk lid of a vehicle. From [28] @ 2022 IET. Figure 3.25: Antenna B mounted on the outer surface of a mirror cover in the Satimo system. From [28] @ 2022 IET. on the curved ABS surface of the mirror cover and a maximum realized gain of 0.187 dBi at this placement. At C-V2X frequencies (5.9 GHz) the antenna has a measured S11 of -9 dB when it is mounted on the flat ABS surface of the trunk lid. The efficiency of the antenna B at 5.9 GHz when is mounted on the curved surface of the mirror cover is 31% and the maximum realized gain 1.54 dBi. The radiation patterns of the antenna B mounted on the outer surface of the mirror cover are given in Fig. 3.26 and Fig. 3.27 at 800 MHz and 5.9 GHz, respectively. 42 (a) (b) Figure 3.26: Radiation patterns of antenna B when mounted on the outer surface of the mirror cover at 800 MHz (a) on E-plane and (b) on H-plane. From [28] @ 2022 IET. (a) (b) Figure 3.27: Radiation patterns of antenna B when mounted on the outer surface of the mirror cover at 5.9 GHz (a) on E-plane and (b) on H-plane. From [28] @ 2022 IET. Antenna measurements when mounted on trunk lid inside the SATIMO SG32 near field system The antenna B is mounted on a plastic trunk lid of a convertible vehicle using tape and transferred in Satimo system to measure its radiation patterns. In Ansys EM software the trunk lid is simulated as a flat ABS surface with thickness t=2.8 mm. The setup in Satimo is indicated in Fig. 3.28. In Ansys EM software the antenna resonates around 768.4 MHz and has an S11 of -5 dB at 5.9 GHz. 43 Figure 3.28: Antenna B mounted on the trunk lid of a real vehicle in the Satimo system. From [28] @ 2022 IET. The implemented antenna when mounted on the real trunk lid resonates around 800 MHz and at 5.9 GHz has an S11 of -13.6 dB, as shown in Fig. 3.24. The efficiency of the simulated antenna B at 768.4 MHz is 93% and of the implemented is 51.4% at 800 MHz. At 5.9 GHz the antenna in simulations has an efficiency of 89.5% when it is placed on the plastic ABS surface and in the Satimo system the efficiency is measured to be 71% at 5.9 GHz. The maximum realized gain of the antenna on the flat ABS according to Ansys EM software is 0.7 dBi at 768.4 MHz and 2.87 dBi at 5.9 GHz. The implemented antenna has a maximum realized gain of 0.4 dBi at 800 MHz and 2.5 dBi at 5.9 GHz when it is mounted on the trunk lid of the vehicle. The radiation patterns of the antenna B in this setup are given in Fig. 3.29 and Fig. 3.30. At cellular frequency the radiation patterns of the simulations match well with the radiation patterns of the measured antenna. The difference in the radiation patterns of the simulated to the manufactured antenna at C-V2X frequency can be explained by the size of the soldering. At 5.9 GHz the wavelength is around 50mm and the soldering of the SMA connector on the implemented antenna consists of two big bulbs covering the whole feeding pads of the antenna. Whereas at the simulations, the soldering is represented by two flat rectangles extending the feeding lines and connecting them to the SMA connector. The simulated soldering has a size of 1mm×0.25mm×0.018mm. Antenna measurements when mounted on plastic retrieved from A-pillar of vehicle inside the SATIMO SG32 near field system The last experimental setup includes the antenna B mounted on a curved plastic retrieved from the A-pillar of a vehicle. The surface of this plastic is coarse and is not coated when compared to the 44 (a) (b) Figure 3.29: Radiation patterns of antenna B when mounted on the trunk lid of a vehicle at cellular frequencies (a) on E-plane and (b) on H-plane. From [28] @ 2022 IET. (a) (b) Figure 3.30: Radiation patterns of antenna B when mounted on the trunk lid of a vehicle at C-V2X frequencies (a) on E-plane and (b) on H-plane. From [28] @ 2022 IET. ABS of the mirror cover and the trunk lid surfaces. The antenna B mounted on this surface and set in Satimo system is indicated in Fig. 3.31. The antenna has an efficiency of 36.75% at 800 MHz and a realized gain of 0.3 dBi at the same frequency. At 5.9 GHz the antenna has an efficiency of 22% and a maximum realized gain of 1.9 dBi. The radiation patterns of the antenna on the curved plastic are shown in Fig.3.32 and 3.33. 45 Figure 3.31: Antenna B mounted on the A-pillar plastic of a real vehicle in the Satimo system. From [28] @ 2022 IET. (a) (b) Figure 3.32: Radiation patterns of antenna B when mounted on a plastic from inside of the vehicle at cellular frequencies (a) on E-plane and (b) on H-plane. From [28] @ 2022 IET. 3.3.2 Antenna measured on real vehicle The antenna is measured under four different experimental setups. Initially it is tested in: (a) a SATIMO SG32 near field system at Michigan State University where the far field of the antenna is extracted through the near field spherical measurements and then transferred to Oakland University in order to be measured in their outdoor antenna range facilities as shown in Fig. 3.35. At Oakland University the antenna is tested; (b) on a stanchion, without the presence of any vehicle; (c) inside the side-mirror cover of a convertible vehicle; and (d) on the trunk lid of the same vehicle. The side mirror as well as the trunk lid are made of ABS material. The SATIMO SG32 system is covered with absorbers and there are no interventions or barriers 46 (a) (b) Figure 3.33: Radiation patterns of antenna B when mounted on a plastic from inside of the vehicle at C-V2X frequencies (a) on E-plane and (b) on H-plane. From [28] @ 2022 IET. to disrupt the antenna’s radiation. To measure the antenna at the outdoor facilities without the presence of a vehicle, the antenna is placed on a stanchion on the turntable of Fig. 3.35. The stanchion is made of non conductive material. Later, a vehicle is added on the turntable of the outdoor antenna range facilities and the antenna is placed inside the side mirror of the co-driver and then on the trunk lid of the vehicle. Measurements for all the mentioned setups are captured and the radiation patterns of the antenna are compared in order to understand how the real environment and the presence of a vehicle affect the antenna characteristics at cellular bands and frequency close to C-V2X band. The maximum values of the co-polarization patterns and the values of the cross-polarization at the same point are indicated in Table 3.5. Stand alone antenna measurements in SATIMO SG32 near field system Before testing the antenna’s radiation patterns, the S11 of the implemented antenna is measured using the M5227 PNA Network Analyzer by Keysight Technologies the measurement is indicated in Fig. 3.36. The antenna is placed on a styrofoam inside the Satimo near field system. The measured co-polarization and cross-polarization patterns of the antenna at the cellular (975 MHz) and around the C-V2X frequencies (5.488GHz), where the antenna resonates, are plotted using 47 Figure 3.34: Outdoor antenna range measurement facility at Oakland University. [54] Matlab code and are indicated in Fig. 3.37. These measurements as well as the simulations refer to the realized gain of the antenna. The maximum realized gain of the antenna as measured in SATIMO at 975 MHz is 1.4 dBi and the efficiency is 80%. The linear average gain (LAG) of the antenna is 1.13 dBi at 975 MHz. At 5.48 GHz the antenna has a measured efficiency of 70% and a maximum realized gain of 1.2 dBi. At C-V2X frequencies the designed antenna has a LAG of -8 dBi. By introducing an amplifier in the structure the transmitted signal at C-V2X frequency can be boosted. The suggested antenna does not have a ground plane and can be mounted on any automotive plastic, which facilitates the use of amplifiers. When comparing the radiation patterns of the simulated antenna (Fig. 3.38) to the measure- ments, there is a reasonable agreement between them. The simulated and measured co-polarization values are almost the same. Namely, in HFSS the highest value of the co-polarization is 1.6 dBi at θ = 90o and in Satimo it is measured to be 1.4dBi at the same angle at cellular frequencies. How- ever the maximum value of the cross-polarization is higher in the SATIMO measurements (-11.5 dBi) compared to the simulations (-22.25 dBi) around 975 MHz at the same θ. The difference in 48 Figure 3.35: Antenna placed at the outdoor far field measurement setup of Oakland University without any vehicle. From [29] @ 2021 IEEE. the averaged cross-polarization values of the simulated and measured antenna is around 7 dBi. At the C-V2X frequencies the measured radiation pattern of the antenna matches the pattern in the simulations and the differences can be explained by the soldering of the SMA connector which cannot be precisely simulated in Ansys EM software, but its size is comparable to the wavelength at 5.5 GHz. In Ansys EM software the soldering was simulated by planar lines, at the edge of the feeding pads connecting the antenna to the SMA connector where in reality the soldering consists of two big bulbs covering the whole area of the feeding pads and extending above the connector. As mentioned in [55] the value of the cross-polarization is affected by the length of the feeding line and the placement of the connector. At 5.5 GHz the simulated co-polarization on E-plane has a maximum value of -7.8 dBi at θ = 90o and the cross-polarization on E-plane has a value of -17 dBi at same θ. The maximum value of co-polarization as measured in Satimo near field system reaches 49 0 dB(S11) -2 -4 -6 Magnitude (dB) -8 -10 X 5.488 -12 X 0.975 Y -13.16 Y -10.83 -14 -16 -18 -20 0 1 2 3 4 5 6 7 8 Frequency (GHz) Figure 3.36: PNA measured S-parameters. From [29] @ 2021 IEEE. Figure 3.37: Measured co- and cross- polarization gain of implemented antenna (a) at 975 MHz and (b) at 5.5 GHz in Satimo near-field system. From [29] @ 2021 IEEE. the -8 dBi at same θ value and the cross-polarization is -14.2 dBi at the same angle. In order to avoid the radiation pattern deterioration at C-V2X frequency due to the size of the feed-line, an absorbent material can be used to cover the feed-line and eliminate any radiation by it. 50 Figure 3.38: Simulated in HFSS and measured in Satimo (a) co- and (b) cross- polarization gain of antenna around 975 MHz. From [29] @ 2021 IEEE. Figure 3.39: Simulated in HFSS and measured in Satimo (a) co- and (b) cross- polarization gain of antenna around 5.5 GHz. From [29] @ 2021 IEEE. Stand alone antenna measurements at far-field, outdoor antenna range facilities For the second set of measurements, the antenna is measured at the outdoor range facilities of Oakland University without the presence of a vehicle, as indicated in Fig. 3.35. Initially, the antenna is fastened on a stanchion and its co- and cross-polarization radiation patterns on E-plane are measured. The stanchion is placed on a turntable which is conductive, acting as a ground plane. The distance between the antenna and the turntable is 132cm and the 51 source antenna of Fig. 3.34 is opposite to the measured antenna, at θ = 90o . The distance between the antenna and the ground is chosen to be 132cm because this is equal to the height of the vehicle side mirror from the ground. The co- and cross- polarization of the fabricated antenna are shown in Fig. 3.40 for the two frequency bands. The antenna patterns are different than what measured in Satimo and simulated in Ansys EM, where there was no ground plane. This is a result of the ground effect. Simulations with the antenna at a distance of 132 cm above an infinite ground plane are performed in HFSS and indicated in Fig. 3.41. The presence of the ground plane causes an increase in the cross-pol level of the antenna and deteriorates its radiation pattern due to the reflections of the signal on the ground plane. The radiation patterns at the higher frequencies are affected more than in lower frequencies since, as explained before, the wavelength at higher frequencies is smaller so the effect of the ground and any small obstacles around the antenna becomes more obvious at 5.5 GHz rather than in 960 MHz. For the measurements at cellular frequencies the radiation pattern of the antenna was measured from 800 MHz to 960 MHz with a step of 20 MHz, this is the reason why the measurements at the outdoor facility of Oakland university are plotted at 960 MHz. The co-polarization has a maximum value of 1.5 dBi at ϕ = 10o and the cross-polarization is equal to -7.5 dBi at the same angle at 960 MHz. At 5.5 GHz the co-polarization has a maximum value of 4.4 dBi at ϕ = 334o and the cross-polarization at the same position drops to -8 dBi. At the cellular band, simulations in HFSS with an infinite ground plane below the antenna at a distance of 132 cm are performed. The setup in the HFSS is indicated in Fig. 3.41a. The co and cross polarization radiation patterns of the antenna at 984.8 MHz as well as the radiation patterns of the antenna are shown in Fig. 3.41b. The simulations are run only for the lower frequency band because the antenna has a linear polarization when there is no ground. The antenna is not linearly polarized at 5.5 GHz even when the ground is missing due to the size of the meander line and of the feed-line. As is demonstrated in Fig. 3.41b, the ground has a significant effect. 52 90 90 120 60 120 60 0 0 150 30 150 30 -20 -20 -40 -40 180 0 180 0 210 330 210 330 240 300 240of antenna at Oakland University without Co-pol (along) 300 Vehicle Co-pol (along) of antenna at Oakland University without Vehicle Cross-pol (vertical to) of antenna at Oakland University without Vehicle 270 at Oakland University without Vehicle Cross-pol (vertical to) of antenna 270 (a) (b) Figure 3.40: Measured co- and cross- polarization gain patterns of implemented antenna (a) at 960 MHz and (b) at 5.45 GHz at outdoor antenna range facility of Oakland University, without the presence of a vehicle. From [29] @ 2021 IEEE. 90 120 60 0 150 30 -100 180 -200 0 210 330 240 300 Co-pol (along) of antenna in HFSS when there is infinite ground 270 Cross-pol (vertical to) of antenna in HFSS when there is infinite ground (a) (b) Figure 3.41: (a)Antenna with infinite ground plane at 132 cm below it in HFSS. (b) Simulated co- and cross- polarization gain patterns of antenna at 984.8 MHz, when there is an infinite ground plane at 132cm below the antenna. From [29] @ 2021 IEEE. Antenna measurements when mounted inside the vehicle’s side mirror on a real vehicle at far-field, outdoor antenna range facilities After measuring the antenna characteristics without a vehicle, a convertible car is added on the turntable and the antenna is positioned inside the side mirror of the passenger, as indicated in Fig. 3.42. The antenna is placed inside the side mirror cover and the mirror cover is set back on its original position. The mirror cover is made of ABS and is a good candidate to print the antenna on 53 (a) (b) Figure 3.42: Antenna placed inside the side-mirror of vehicle. From [29] @ 2021 IEEE. it at a next step. Also by positioning the antenna inside the mirror cover, the antenna radiation can cover the sides and part of the front of the vehicle. The co- and cross-polarization patterns of the antenna for the different frequency bands can be found in Fig. 3.43. The co- and cross-polarization of the antenna are plotted for θ = 90o which means that the transmitter/source antenna of Fig. 3.34 is at the same level with the car. During our experiments, in contrast to Fig. 3.34, the Antenna Under Test (AUT) is not placed on the rooftop of the car but inside the side-mirror cover and then on the trunk lid of the vehicle. At cellular frequencies, when the antenna is placed inside the side mirror, the maximum value of co-polarization is 0.8 dBi at ϕ = 48o . The cross-polarization at same position is measured around -2 dBi at 960 MHz. At 5.9 GHz the max value of the co-polarization is -0.75 dBi at ϕ = 350o and the cross-polarization at the same angle is -3.15 dBi. Antenna measurements when mounted on the vehicle’s trunk lid on a real vehicle at far-field, outdoor antenna range facilities For the last set of experiments the antenna is taped on the trunk lid of the vehicle as shown in Fig.3.44. The cross- and co-polarization of the AUT for θ = 90o of the source antenna, at 960 MHz and 5.9 GHz are plotted and shown in Fig.3.45. As indicated, at 960 MHz the maximum co- polarization value is -2 dBi at ϕ = 90o and the cross-polarization at same angle is -5 dBi. At 5.9 54 (a) (b) Figure 3.43: Measured co- and cross- polarization of implemented antenna (a) at 960 MHz and (b) at 5.9 GHz at the outdoor antenna range facility of Oakland University, when the antenna is inside the side-mirror cover of a vehicle. From [29] @ 2021 IEEE. Figure 3.44: Antenna placed on the trunk lid of the vehicle. From [29] @ 2021 IEEE. GHz the co-polarization has a maximum value of 0.3 dBi at ϕ = 124o where the cross-polarization is 1.04 dBi. 55 (a) (b) Figure 3.45: Measured co- and cross- polarization of implemented antenna (a) at 960 MHz and (b) at 5.9 GHz at outdoor antenna range facility of Oakland University, when the antenna is placed on the trunk lid of the vehicle. From [29] @ 2021 IEEE. Cellular C-V2X Max Co-Pol Cross-Pol Max Co-Pol Cross-Pol Value (dBi) Value (dBi) Value (dBi) Value (dBi) HFSS 1.6 -22.25 -7.8 -17 SATIMO 1.4 -11.5 -8 -14.2 Oakland No vehicle 1.5 -7.5 4.4 -8.2 Vehicle Mirror 0.8 -2 -0.75 -3.15 Vehicle Trunk -2.1 -5 0.3 1.04 Table 3.5: Comparison of polarization values at the different setups. From [29] @ 2021 IEEE. 3.4 Results 3.4.1 Results of antennas measured on automotive plastics without the presence of a vehicle The efficiency and maximum realized gain values of the various experimental setups are summa- rized in the table 3.6. By comparing the results of the two antenna designs when they are measured alone, it is obvi- ous that both the antennas have similar values of maximum realized gain. Their radiation patterns at cellular frequencies are similar. Both the antennas have an omni-directional radiation pattern at 56 Experimental Setup Efficiency Max Realized Gain (dBi) Cellular C-V2X Cellular C-V2X Antenna A alone 50% 84.8% 1.64 4.6 Antenna B alone 82.9% 81.2% 1.6 4.2 Antenna A in mirror cover 40.3% 76% 1.38 4.6 Antenna A on mirror cover 26% 63.4% -0.34 4 Antenna B on mirror cover 44% 31% 0.187 1.54 Antenna B on trunk lid 51.4% 71% 0.43 2.5 Antenna B on A-pillar 36.75% 22% 0.34 1.9 Table 3.6: Comparison of antenna properties at the different setups. From [28] @ 2022 IET. 800 MHz. At C-V2X their radiation patterns are different because antenna A has a longer slotline than antenna B and antenna B has longer meander lines. At 5.9 GHz the wavelength in free space is λ = 50.8mm and the size of the slotline as well as the meander lines is comparable to the wave- length. As a result there is radiation coming from the meander lines and the slotline at 5.9 GHz, which results to loss of the omnidirectionality. When placing the antennas on the ABS we expect that the resonance of the antennas will change. Namely, since the antenna size remains the same but the effective wavelength changes to λ= √c , where ϵr refers to the electrical permittivity of ABS. The new effective wavelength is f ϵr smaller but the antenna size remains the same, as a result the antenna resonates to lower frequencies when placed on ABS automotive parts, compared to when measured alone. This is also obvious in Fig. 3.15 and Fig. 3.24 where the resonance of antenna B drops from 830 MHz to 800 MHz. When mounting the antennas on the curved mirror cover there is a shift at the value of the maximum realized gains. Both antennas A and B have a decrease of the maximum realized gain value. The maximum realized gain of antenna A is 0.22 dBi higher at 800 MHz when the antenna is mounted on the inside surface of the mirror than when it is mounted on the outside surface of the same mirror cover. At 5.9 GHz the maximum realized gain is 0.6 dBi higher. When mounting the antenna A on the inner surface of the mirror cover, the radiation pattern on the H-plane becomes more directional compared to the radiation pattern of the same antenna when measured alone and on the outer surface of the same mirror cover at cellular frequency. When antenna A is mounted at the inner part of the mirror cover, it gets bent along the bowtie and meander direction, as a result 57 the radiation patterns are modified. The radiated power gets concentrated in the concave direction as also proved in [56]. Kellomaki et al. study the effects of bending circular and linear polarized GPS antennas. According to their research the polarization of the tested patch antennas changed from circular to linear. Also the radiated power of the inverted-F antenna increased in the vertical and decreased in the horizontal polarization, compared to the unbent antenna. When the antenna Ais mounted on the outer surface of the same mirror cover it gets bent on both directions, across and along the bowtie and meander lines. The radiation pattern on H-plane compared to the pattern of the same antenna when measured alone is more directional but compared to the patter of the antenna mounted on the inner surface of the mirror cover, it is less directional. The efficiency of the antenna and the maximum realized gain values, at both cellular and C-V2X frequency bands, decrease. When antenna B is mounted on the outer surface of the mirror cover the maximum value of the realized gain drops almost 1.4 dBi at cellular frequencies, when compared to the measurements of the same antenna alone in the Satimo system and 2.66 dBi at C-V2X. The place where the antenna B is mounted on the mirror has a smaller bending radius than when mounting the antenna A on the inner surface of the mirror cover. In [57] is proved that antenna bending broadens the radiation pattern in the bending plane which results to a drop of gain. In [58] the authors suggest that the larger the bending of the surface the more the decrease in the maximum realized gain value of the antenna. This is proved also here when mounting the antenna to the plastic part retrieved from the A-pillar of the vehicle. This plastic has a smaller radius than the mirror cover and as a result is more curved. The efficiency and the value of the maximum realized gain of antenna B are decreased as stated in the Table 3.6. The radiation patterns of the antenna are more affected when the antenna is placed on the plastic part from the A-pillar of the vehicle. 3.4.2 Results of antennas measured on automotive plastics on a real vehicle To compare the co- and cross-polarization values at the different experimental setups the data indicated in Fig. 3.37, 3.38, 3.40, 3.43 and 3.45 are gathered, separated to co-polarization and 58 cross-polarization radiation patterns and are presented in Fig. 3.46, 3.47, 3.48 and 3.49. When comparing the co- and cross-polarization gain patterns of the antenna at the different experimental setups in cellular frequency bands, it is obvious that although the antenna has a linear polarization when simulated and measured in the Satimo system, without a ground plane, the cross-polarization level increases when exposed to open space, with or without a vehicle due to the ground effect. As a result the direction of polarization has been shifted. Fig. 3.4 and Fig. 3.36 show the resonance frequencies of the simulated and the implemented antennas as measured in the lab using the M5227 PNA Network Analyzer of Keysight Technolo- gies. The simulated and the implemented antennas resonate at 984.8 MHZ and 975 MHz respec- tively for the cellular bands. Also, both of them have S11 < -10 dB at 5.5 GHz. The resonance frequencies of the simulated and the implemented antennas are the same. In Fig. 3.38 and Fig.3.39 the co and cross- polarization radiation patterns of the simulated antenna and the implemented one as measured in Satimo system are gathered in the same plot. In these figures it is obvious that the radiation patterns of the simulated and measured antennas show reasonable agreement. The average cross-polarization of the implemented antenna at the cellular bands is 7 dB higher than the simulated. At 5.5 GHz the cross-polarization is high because the size of the meander and the feeding line are not negligible when compared to the wavelength. Each automotive company has set its own standards on the required antenna gain. To improve the deterioration of the antenna radiation pattern at C-V2X frequency, an absorbing material can be used to cover the feed-line. Moreover, further investigation can be performed in identifying a suitable feed method for this antenna that will not interfere with its performance at the higher frequency band. At cellular frequencies the maximum co-polarization values when the antenna is measured in the Satimo system and at the Oakland outdoor range facility, presented in Fig. 3.40, are close to the simulated value of 1.6 dBi with a deviation less than 0.2 dBi, as indicated in Table 3.5. The radiation patterns of the antenna when measured alone at the outdoor measurement setup of Oakland are different compared to the radiation patterns of the antenna when measured in Satimo 59 near field system. The difference in the patterns is caused by the metallic turntable on which the antenna is placed and acts as a ground plane. Simulations with the antenna 132 cm above a round ground plane are indicating that the patterns of the antenna are affected when there is ground below it, as shown in Fig. 3.41. As stated in [59] the diffracted fields of the ground affect the pattern and the values of the cross-polarization. The simulated maximum value of the co-polarization at 984.8 MHz when there is ground below it, is 7 dBi but it has many lobes due to reflections. The antenna loses its linearity and the cross-polarization increases to -13 dBi at 180o , close to the co-polarization values. When the antenna is mounted on the trunk lid of the vehicle, we get the minimum value of the co-polarization radiation pattern on E-plane at cellular frequencies. The closest the an- tenna is on the ground, the least is the value of the realized gain and the higher the value of the cross-polarization because as stated in [60] and [59] the surface waves propagating along the ground affect the level of cross polarization. The difference between the co-polarization and cross- polarization values decreases when the vehicle is added in the experimental setup. When the antenna is measured on the car, there are huge metal surfaces around it because of the vehicle’s body which creates many reflections and deformities to the antenna radiation pattern as indicated in Fig. 3.43 and Fig. 3.45. When comparing the antenna radiation pattern of the four measurement setups, it is obvious that the reflections due to the outdoor space, presence of ground plane and vehicle body disturb the antenna polarization resulting in loss of antenna linearity, at cellular frequencies as indicated in Fig. 3.46. When placing the antenna on the vehicle’s trunk lid, we get similar results as indicated in Fig. 3.45, 3.47 and 3.49. The difference between placing the antenna on the trunk lid and placing it inside the side-mirror cover is that the trunk lid is closer to the turn-table and ground. As a result there are reflections affecting the antenna’s radiation pattern and the values of the co- and cross- polarization patterns are deviating more from the simulated values when compared with the other experimental setups. At 5.9 GHz the maximum value of the cross-polarization of the antenna 60 (a) (b) Figure 3.46: Comparison of antenna (a)co- and (b)cross- polarization as simulated in HFSS, as measured in Satimo system and as at the outdoor facility without and with a vehicle, when antenna is placed inside the mirror cover, at cellular band. From [29] @ 2021 IEEE. (a) (b) Figure 3.47: Comparison of antenna (a)co- and (b)cross- polarization patterns as simulated in Ansys HFSS, as measured in Satimo system and at the outdoor facility without and with a vehicle, when antenna is placed on the trunk lid at cellular bands. From [29] @ 2021 IEEE. exceeds the maximum value of the co-polarization. The antenna’s radiation pattern is distorted at the higher frequency bands. The antenna at 5.9 GHz is not designed to be linearly polarized due to the SMA connector, the size of the feedline 61 (a) (b) Figure 3.48: Comparison of antenna (a)co- and (b)cross- polarization patterns as simulated in HFSS, as measured in Satimo system and at the outdoor facility without and with a vehicle, when antenna is placed inside the side-mirror cover, at C-V2X band. From [29] @ 2021 IEEE. and the meander lines, which are comparable to the wavelength since the cross-polarization is created due to the transverse currents of the higher order modes. In Fig. 3.48 and Fig. 3.49 the co- and cross-polarization of the antenna as simulated in Ansys EM (HFSS) and measured during the various experimental setups is indicated. The measured maximum values of the co-polarization are higher than the simulated. The reflections created by the obstacles in the real environment, the ground and the vehicle’s body modify the antenna’s radiation patterns and add more to its maximum radiation pattern values. The smallest value of the cross-polarization at the angle where we have the maximum co-polarization value at 5.9 GHz, when the vehicle is present, is measured to be -3.15 dBi when the antenna is placed inside the mirror’s cover. At that placement the antenna is higher from the ground, compared to the trunk lid position, but the conductive body of the vehicle is closer to the antenna. As stated in [60] and [59] the presence of ground causes the generation of higher level cross-polarization due to surface waves propagating along the ground. For this reason, the values of the cross-polarization increase when the vehicle is added, compared to the measurements on the stanchion and in the Satimo system. 62 (a) (b) Figure 3.49: Comparison of antenna (a)co- and (b)cross- polarization patterns as simulated in HFSS, as measured in Satimo system and at the outdoor facility without and with a vehicle, when antenna is placed on its trunk lid at C-V2X band. From [29] @ 2021 IEEE. 3.5 Conclusion Comparing the measured results of antenna B when placed on the automotive plastic parts it is understood that the antenna B has the maximum value of the realized gain at both frequency bands when it is mounted on the flat ABS of the trunk lid. The measured radiation patterns at both frequencies for this placement, agree well with the simulated results. When mounting the antenna on the curved ABS surfaces of the mirror cover and the A-pillar plastic, the value of the maximum realized gain decreases. An increase in the bending of the surface results in a decrease in the value of the maximum realized gain and deterioration of the radiation patterns. Moreover, at 5.9GHz the efficiency of the antenna B on the curved mirror cover and the A-pillar plastic is 31% and 22% respectively. This low efficiency value can also be a result of loss of the resonance at this frequency. At both [57] and [58] the antennas lost the resonance at some frequencies when bending them. The effects of the shape and surface of automotive plastic parts on the efficiency, gain and radiation patterns of two flexible dual-band antennas were investigated for the first time, to the best of my knowledge. From the experiments and analysis of measurements, the mirror cover is a 63 preferred location to mount the C-V2X antenna. The curvature of the mirror cover compared to the antenna size does not impede the antenna performance. The trunk lid which is a flat surface is also compatible with C-V2X antennas. The antenna tested in this research resembles a dipole, which means that the current flows through the bowtie and the meander lines. By mounting the antenna at the inside of the mirror cover, the axis on which the current flows is not bent significantly and as a result the current flow is not impeded. The antenna performance is not severely affected. Moreover, the effect of the ground on the antenna performance at cellular and C-V2X commu- nications was evaluated for the first time to the best of my knowledge at an outdoor far field system, incorporating a real vehicle. This research can benefit automotive companies that struggle with the increasing amount of antennas and the limited mounting positions on a vehicle. It identifies how candidate positions like the plastic automotive parts and their distance form the ground can affect the antenna performance. 64 CHAPTER 4 ADDITIVE MANUFACTURED AUTOMOTIVE ANTENNAS ON ABS 4.1 Background After implementing the dual band antennas on the thin and flexible Rogers 3850 substrate in chap- ter 3, I slightly modified the design of the antenna that was tested on the real vehicle, to develop it on automotive ABS using additive manufacturing techniques. The design implemented on ABS is slightly modified compared to the one presented in chapter 3, to accommodate the dielectric prop- erties of the new substrate. ABS is widely used in automotive industry in parts like the bumper, the mirror cover, the trunk lid, the instrument panel, the lighting, the dashboard, the interior trim, the door handles and the seating. Other automotive plastics are developed by polypropylene (PP), polyurethene (PU), polyethylen (PE), and polyvynil chloride (PVC), as stated in [61]. A common characteristic of these materials is that they all have a melting temperature lower than 164o C. As a consequence, when printing silver nanoparticles on these materials, thermal sintering is not a viable option for the substrate. Different sintering methods have to be developed. IPL was used to sinter the silver nanoparticles ink on the automotive ABS in order to develop the dual-band anten- nas. It was investigated how the wide range of processing parameters of IPL affect the sintering performance of not-pre-sintered silver nanoparticles on the ABS substrate. Researchers have been testing photonic sintering methods to develop electrically conductive devices without affecting the substrate. Lines using a silver paste are created in [62] on a clear ABS substrate. Ultraviolet (UV) sintering for more than 1 minute results in a slightly higher conductivity compared to the conduc- tivity achieved through IPL. In contrast to the literature, the antennas tested through my research were created using silver nano-particles and were sintered for a few milliseconds. The authors of [22] sintered silver resonators on polyimide using IPL. The resonators had a conductivity of 5× 106 S/m before IPL sintering. Their conductivity increased to 11.9× 106 S/m after the application of IPL. The devices I developed were not conductive before IPL sintering. To the best of my knowledge, this is the first study investigating how the wide range of process- 65 Ref. Material Substrate Sintering Method Conductivity (S/m) [62] silver paste clear ABS UV light 2.6 × 106 [22] inkjet printed silver polyimide IPL 11.9 × 106 This work silver nanoparticles clear ABS IPL 2.13 × 106 Table 4.1: Comparison of this work to literature. From [30] @ 2022 IEEE. ing parameters of IPL affect the sintering performance of not-pre-sintered silver nanoparticles on ABS substrate widely used in automotive parts. In Table 4.1, from [30], a comparison is presented between the work of this research to the literature. 4.2 Design The antennas are designed to operate at the cellular and C-V2X frequencies when implemented on ABS. The antenna design is the same with the one demonstrated in chapter 3. The bowtie is designed to resonate at 5.9 GHz and its arms extend to meander lines in order to achieve resonance at 857 MHz. The overall dimensions of the antenna are 91.5×30mm2 . The implemented antennas are presented in Fig.4.2a. 4.3 Fabrication Process The aqueous based Metalon JS-A221E silver nano-particle ink is deposited on clear ABS substrate using the Ylingsu handheld airbrush, as indicated in Fig.4.1. The airbrush is used as an alternative fast implementation method when creating big structures that do not require the development of fine features. The AJP was not used during this study as the fine feature printhead of the AJP required 1 hour to print one layer of this antenna design, providing around 0.8µm of silver thick- ness. The wide feature printhead of AJP was not tested yet during the development of this antenna. The ABS has a permittivity ϵr of 2.5, thickness of 2.8mm and tanδ of 0.004 as measured using the Agilent E4991A RF Impedance Material Analyzer. After the deposition of the silver on the substrate, the ink is dried in an oven at 80o C for an hour and then the structures are sintered using the PulseForge Invent IPL system. IPL has a xenon flashlight lamp. This lamp emits light within 250nm-1150nm with a pulse- 66 Ant. Impl. Sintering Method Thickness Conductivity section profile (µm) (S/m) A IPL (25, 30Hz, 450V, 700µs) 1.63 ∗ 106 B IPL (25, 30Hz, 450V, 650 µs) 0.88 ∗ 106 C IPL (25, 30Hz, 450V, 600 µs) 0.6 ∗ 106 D IPL (25, 40Hz, 450V, 650 µs) 0.836 ∗ 106 Abs1 airbrush 8 E IPL (25, 50Hz, 450V, 650 µs) 1.4165 ∗ 106 F IPL (25, 55Hz, 450V, 650 µs) 1.523 ∗ 106 G IPL (40, 55Hz, 450V, 650 µs) 1.78 ∗ 106 H IPL (60, 55Hz, 450V, 650 µs) 2.13 ∗ 106 Abs2 airbrush thermal, 24 hours at 80o C 5 0.23 106 Table 4.2: Thickness and conductivity values of antennas. From [30] @ 2022 IEEE. length of 0.1ms-10ms. The peak radiated power that can be delivered is 35 kW/cm2 , as stated in [63]. The pulse length can be adjusted with an increment of 1µs and the pulse spacing with an increment of 20 µs. The light emitted by the lamp is absorbed by the dark silver lines and reflected by the white-colored substrate. The light absorbed by the silver, induces a rise at its temperature which dissolves the agents around the silver nanoparticles, allowing them to connect and create paths that the current can pass through, making the device conductive. During this research, different settings were used on IPL in order to identify the sintering profile that will make the antenna conductive, without affecting the substrate. The thickness of the silver is measured using the NanoMap 500LS system by AEP technology. The silver conductivity is exported by the Lucas Pro4-4400 Signatone QuadPro Resistivity sys- tem. The S-parameters of the antennas are measured using the Keysight PNA N5227A Network Analyzer and the radiation characteristics are measured in the Satimo near field system. 4.4 Measurements The antennas are designed to operate at the cellular and C-V2X frequencies. To identify the best IPL sintering profile for the silver on ABS, the antenna Abs1 is separated in sections A-H which are sintered using different length and frequency of pulses. Table 4.2 summarizes the values of these variables as well as the thermal sintering profile of the antenna Abs2. The S-parameters of the antennas are presented in Fig. 4.2b. 67 Figure 4.1: Antenna development using airbrush. 0 -5 X 5.897 Y -8.583 -10 X 5.897 X 0.8575 Y -9.615 -15 Magnitude (dB) Y -12.37 -20X 0.8575 Y -17.42 -25 -30 -35 -40 ABS1 ABS2 -45 0 1 2 3 4 5 6 7 Frequency (GHz) Figure 4.2: (a) Antennas created on ABS; (b) S-parameters of antennas on ABS. From [30] @ 2022 IEEE. The antennas were mounted on a stanchion inside the Satimo near field measurement system. The antenna radiation patterns on H and E plane at cellular and C-V2X frequency bands , as measured in the Satimo, are given in Fig. 4.4 and Fig.4.5. Next, the antenna Abs2 was transferred to Drive-ability Test Facility of Ford Motor Company and was measured on a mirror cover of a Bronco vehicle inside the MVG 3000 near field system. The experimental setup is indicated in Fig.4.3. The antenna is centered under the top sensor of the MVG 3000 near field system and the mirror of the cover is aligned with the ϕ = 0o angle of the system. 68 Figure 4.3: Antenna Abs2 mounted on a side view mirror cover of a Bronco at DTF to be measured using the MVG 3000 near field system. The measured radiation patterns of antenna Abs2 at DTF, when the antenna was mounted on the mirror cover of the Bronco are presented in Fig.4.6. The Fig.4.6a presents the total gain, Eθ and Eϕ of the antenna at θ = 90o at f=857.5 MHz and the Fig.4.6b indicates the total gain, Eθ and Eϕ of the antenna at θ = 90o at f=5.9 GHz. 4.5 Results To sinter the antenna on ABS using IPL, the voltage cap bank was set to 450V for all the sections of the antenna Abs1 whereas the number of pulses, their frequency and length were varied. By keeping the number of pulses and frequency of pulses steady and decreasing the duration of the pulses the conductivity decreases as indicated by the sections A, B and C of antenna Abs1. This is also verified in [62] with UV irradiation. An increase in the frequency of the IPL pulses, with the number of pulses and duration held constant, results in an increase of the conductivity as indicated when comparing the areas D, E and F of antenna Abs1. This is expected and as explained in [63], using higher pulse frequency the sample heats up more. Moreover, from the areas G and H it 69 Antenna Frequency Efficiency Maximum Realized Maximum Realized (GHz) (%) Gain (dBi) as measured Gain (dBi) as measured alone in Satimo on mirror cover on Bronco 0.8575 51.56 0.81 - Abs1 5.9 57.6 7 - 0.8575 47.4 0.14 2.9 Abs2 5.9 32.4 5.2 5.14 Table 4.3: Antenna properties as measured in Satimo near field system. From [30] @ 2022 IEEE. is apparent that increasing the number of pulses creates an increase in the conductivity as also observed in [64]. The conductivity of antenna Abs1 is higher than the conductivity of the Abs2 antenna that was thermally sintered for 24 hours at 80o C, even though the silver of Abs1 is thicker than the silver of Abs2. The S-parameters of the antennas are measured from 700 MHz to 3 GHz and from 5 GHz to 7 GHz since these bands carry the cellular and C-V2X frequencies that are of interest for the specific application. The antennas resonate at 857.5 MHz, 5.3 GHz and have an S11 around -10 dB at 5.9 GHz. The S-parameters of Abs2 are lower than those of Abs1 as shown in Fig. 3.6 because Abs2 has more losses due to lower conductivity than Abs1 antenna. Table 4.3 presents the efficiency and the maximum realized gain values of the antennas at 857.5 MHz and 5.9 GHz. The antennas with lower conductivity values have lower efficiency and as a result lower maximum realized gain values. When the antenna Abs2 is measured on the mirror cover of the vehicle, the radiation pattern of the antenna becomes more directive which results in an increase at the maximum realized gain of the antenna. The metallic body of the vehicle is affecting the radiation pattern of the antenna, and this effect is more prominent at the lower frequency of the dual band antenna where the wavelength is larger. The distance between the metallic vehicle body and the antenna is comparable to the wavelength at 857.5 MHz where λ = 349 mm. At 5.9GHz, the wavelength is only λ = 50mm. The distance between the mirror cover and the metallic vehicle body is around mm. 70 Figure 4.4: (a) Measured gain on E plane at 857.5 MHz of antennas implemented on ABS; (b) Measured gain on H plane at 857.5 MHz of antennas implemented on ABS when they are measured in Satimo near field system. Figure 4.5: (a) Measured gain on E plane at 5.9 GHz of antennas implemented on ABS; (b) Mea- sured gain on H plane at 5.9 GHz of antennas implemented on ABS when they are measured in Satimo near field system. 4.6 Conclusion Through this research it is verified that IPL sinters silver nano-particles in milliseconds without destroying the temperature sensitive clear ABS substrate, making it suitable for developing 3D printed low cost vehicular antennas on it. The resulting conductivity depends on the power, the 71 Figure 4.6: Antenna Abs2 mounted on a side view mirror cover of a Bronco at DTF to be measured using the MVG 3000 near field system. duration, the number and the frequency of the pulses of the light. The maximum conductivity of silver nano-particles was obtained by using 60 pulses with 450V power and a duration of 650µs. This is the maximum number of pulses tested on ABS in this research. The thickness of the silver impacts thermal conduction to the substrate. Thin structures are sintered faster and demand less power to be sintered. The gain and efficiency of the antennas achieved through IPL sintering are sufficient for use in automotive communications. The performance of a 3D printed antenna on an automotive part made of ABS, developed us- ing an airbrush and IPL is presented for the first time to the best of my knowledge. This research suggests the incorporation of AM techniques in the development of automotive antennas on auto- motive ABS. IPL is a fast and low cost method that can be incorporated in large production lines, reducing the manufacturing cost and time. 72 CHAPTER 5 ADDITIVE MANUFACTURED CO-PLANAR WAVEGUIDE (CPW) LINES ON ABS WITH DIFFERENT SILVER THICKNESS 5.1 Background To investigate further the effectiveness of IPL on the silver nano-particle conductivity, co-planar waveguide lines were printed on the same ABS substrate that was used in chapter 4, and sintered using IPL. Different silver thickness was printed on the ABS, using the AJP. The measured con- ductivity of the silver devices is close to thermally sintered silver traces and the measured insertion loss at 20GHz and 40GHz is comparable to thermally sintered 3-D printed CPW lines on a different substrate, found in literature. In [65], LeBlanc et al. tested how photonic sintering can improve the conductivity of already sintered silver paste on Poly-ether-ether-ketone (PEEK) substrate. CPW samples were thermally sintered in the oven at 160o C for 5 minutes and then processed with a NovaCentrix PulseForge photonic sintering system. The photonically sintered samples were exposed to the lamp for 500 µs, with a peak power of 1700 W/cm2 . The photonic sintering resulted in a reduction of 2.58 dB in dissipative loss at 40 GHz. Hawasli et al. in [66] increased the line conductivity of 3D printed silver paste by 806% using a commercial PulseForge system. Silver paste was printed and sintered using UV, NIR and laser sintering in [62]. The conductivity of silver paste reached 3.6 × 106 S/m by using UV annealing on ABS substrate after flashing the silver trace for 65 seconds. During my experiments, the silver was only sintered for 10 ms and a conductivity of 2.67 × 106 S/m was achieved. In [67], Firat et al. 3-D printed suspended CPW using ABS and CB028 conductive silver paste. The ABS was created using Fused Deposition Modeling (FDM) and the CPW line loss was measured around 0.24 dB/mm at 20 GHz. The insertion loss of one 4 µm CPW line implemented in our study was 0.16 dB/mm at the same frequency. The conductivity of the suspended CPW in [67], which is created using the CB028 paste and sintered thermally at 90o C is 2 × 106 S/m. In our work, a 2.67 × 106 S/m conductivity of a CPW line is achieved by AJP and photonic sintering. 73 Ref. Conductive Material Substrate Sintering Conductivity Method (S/m) [62] silver paste clear ABS UV light 3.6 ∗ 106 [67] silver paste FDM printed ABS bridges thermal 2 ∗ 106 This work silver nanoparticles clear ABS IPL 2.67 ∗ 106 Table 5.1: Comparison of this work conductivity to literature. From [30] @ 2022 IEEE. The ABS substrate used in [67] has different dielectric properties from the commercial automotive ABS used in my study. Moreover the CPW lines were created using micro-dispensed method on silver paste and not AJP. CPW lines were printed on a diamond substrate in [68] by He et al. The silver CPW lines were thermally sintered from 150o C-200o C and the reported insertion loss at 20 GHz and 40 GHz of the CPW lines was a little higher than the insertion loss measured in our study where the CPW lines were sintered photonically using the IPL. The exact values are reported in Table 5.3. Delage et al., in [69], 3D printed CPW lines using silver nano-particle ink on ceramic substrate and sintered it thermally. The reported value of the insertion loss at 40GHz was almost double of the insertion loss achieved in my study at the same frequency. Conductivity values of the CPW lines developed in this research are compared to the literature in Table 5.1 , from [30]. In Table 5.3 the insertion loss (dB/mm) of the CPW lines implemented in this study, sintered using IPL is compared to the insertion loss of other 3D-printed CPW lines found in literature, sintered thermally. To print the CPW lines on the ABS the Aerosol Jet Printer (AJP) by Optomec was used. The AJP uses aerodynamic focusing to deposit electronic inks onto substrates, as described in [70] and chapter 1. The sintering method employed for this research is IPL. IPL makes use of a high inten- sity xenon flash lamp emitting light from 250 nm to 1150 nm, as described in chapter 4. Photonic sintering is preferable for thermally sensitive and light colored substrates as they generally absorb less of the light energy from the flash lamp than the targeted nano-particle ink. The flash lamp is driven by the discharge of a capacitor bank through an inductor in what is referred to as a pulse forming network (PFN). The purpose of the inductor is to slow down the rush of current from the capacitor bank. Changing the inductance value changes the characteristic “pulse length” of the 74 (a) (b) Figure 5.1: (a) AJP by Optomec; (b) Diagam indicating how the ink is ’mixed’ with the sheath to form the mist that is printed on the substrate. From [30] @ 2022 IEEE. discharge, as described in [71]. 5.2 Design The ABS used in this study, is the same used in chapter 4 that has a thickness of t=2.8 mm, permittivity of ϵr = 2.5, and tanδ=0.004 as measured by the Agilent E4991A RF Impedance Material Analyzer. It has a white/clear color. The light colors do not easily absorb the light by the flash lamp. As a result, the white/clear substrate is less affected by photonic sintering. We created two different CPW designs. The first one can be measured using probe tips with 250 µm pitch and the second CPW design can be measured using probes with 500 µm pitch. The first CPW design has dimensions of 6.25×1.8 mm2 and the second design of 10.86×2.6 mm2 . The CPW lines with their dimensions are indicated in Fig. 5.2a and 5.2b. The width of the signal line in design 1 is 0.3 mm and the spacing between the signal line and the ground planes is 0.025 mm. The length of the line is 1.8 mm. The width of the signal line of design 2 is 0.8 mm and the spacing between the signal line and the ground planes is 0.05 mm. The line of design 2 is 2.6 mm long. Design 1 and design 2 of the implemented CPW lines after being sintered are indicated in Fig. 5.3a and Fig. 5.3b. 75 (a) (b) Figure 5.2: CPW designs: (a) design 1; (b) design 2. From [30] @ 2022 IEEE. (a) (b) Figure 5.3: Implemented CPW designs: (a) design 1; (b) design 2. From [30] @ 2022 IEEE. 5.3 Fabrication Aerosol Jet Printer (AJP) by Optomec is used to print CPW lines on clear ABS. The aqueous based Metalon JS-A221AE silver ink by NovaCentrix is employed to 3D-print the transmission lines. A nozzle with diameter of 150 µm is used to print the CPW lines. The printer’s process speed is set at 2 mm/sec and the rapid speed at 8 mm/sec. The sheath gas pressure is set at 0.85-1.23 PSI and the Ultrasonic Atomizer pressure is varied between 1.98-2.19 PSI. The CPW lines with a thickness of 12-15 µm were created by printing 10 times over the same design. The lines with a silver thickness of 6-8µm were developed by having 5 passes over the same device and the lines with a thickness around 3-5 µm were created by having 3 printing runs over the same design. Each run had a duration of 3mins and 35 secs. The NanoMap 500LS profilometer by AEP technologies is used to determine the silver thickness of each CPW line and the Lucas Pro 4-400 Signatone Quad Pro Resistivity system for conductivity measurements. The thickness of each CPW line and the corresponding conductivity values are indicated in Table 5.2. After implementing the CPW lines on the ABS, the silver ink is dried out in an oven at 80o C for 1 hour. The sintering of the silver 76 X 39.9 -14 -0.2 0.55 Y 0.525017 -16 0.5 -0.4 0.45 -18 -0.6 0.4 -20 X 20.2 Insertion Loss dB/mm Y 0.342017 0.35 S 11 and S22 S 21 and S12 -0.8 -22 0.3 -24 -1 0.25 -26 dB(S11 ) -1.2 0.2 dB(S22 ) -28 dB(S12 ) 0.15 -1.4 -30 dB(S21 ) 0.1 -32 -1.6 0.05 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 Frequency (GHz) Frequency (GHz) (a) (b) Figure 5.4: CPW design 2, 14um thickness : (a) S-parameters; (b) Insertion Loss (dB/mm). From [30] @ 2022 IEEE. ink is performed by the Intense Pulsed Light system of PulseForge Inc. The sintering profile of PulseForge used on all the presented CPW lines varies from: 250 Volts to 300 Volts and has 10 ms pulse duration. 5.4 Measurements The Keysight PNA Network Analyzer N5227A combined with the GGB Industries GSG-type CPW probes 40A-GSG-250-C, 40A-GSG-500-C and the MPI-TS150 Probe Station are used to measure the S-parameters of the implemented CPW lines. The CS-9 calibration kit and the Line- Reflect-Reflect-Match (LRRM) method are employed to calibrate the station. The measured S- parameters for each CPW line with their insertion loss graphs are indicated in Fig.5.4a, and Fig. 5.4b, in Fig. 5.5a, and Fig. 5.5b, in Fig. 5.6a, and Fig. 5.6b, in Fig. 5.7a, and Fig. 5.7b, in Fig. 5.8a, and Fig. 5.8b. The measured thickness, conductivity and insertion loss (dB/mm) of each CPW are shown in Table 5.2. 5.5 Results The results of the performed experiments indicate that direct 3-D printing and photonic sinter- ing of conductive films on automotive ABS is a feasible application. Through the conductivity measurements it is proved that IPL sintering of the silver nano-particle ink is more effective on thinner silver traces. The 4 µm thick CPW lines have the highest conductivity values among the 77 -5 0 0.6 -0.1 X 20.1 -10 0.5 Y 0.476161 -0.2 X 39.9 dB(S11 ) Y 0.492399 -15 -0.3 dB(S22 ) 0.4 Insertion Loss dB/mm dB(S12 ) -0.4 S 11 and S22 S 21 and S12 -20 dB(S21 ) -0.5 0.3 -25 -0.6 0.2 -30 -0.7 -0.8 0.1 -35 -0.9 -40 -1 0 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 Frequency (GHz) Frequency (GHz) (a) (b) Figure 5.5: CPW design 1, 12um thickness : (a) S-parameters; (b) Insertion Loss (dB/mm). From [30] @ 2022 IEEE. CPW Thickness Conductivity Sintering IL (dB/mm) IL (dB/mm) Design (µm) (S/m) Profile @ 20GHz @ 40GHz Design 2 14 0.31 ∗ 106 250V, 10ms 0.34 0.55 Design 1 12 0.73 ∗ 106 250V, 10ms 0.47 0.48 Design 1 7.5 0.9 ∗ 106 250V, 10ms 0.33 0.4 Design 1 4 0.984 ∗ 106 250V, 10ms 0.35 0.47 Design 2 4 2.67 ∗ 106 300V, 10ms 0.16 0.37 Table 5.2: CPW fabrication and measurement characteristics. From [30] @ 2022 IEEE. implemented lines. The 14 µm thick CPW has the lowest conductivity value of 0.34×10^6 S/m. However, thicker traces are necessary to avoid losses caused by skin effect, especially at the lower frequency bands. Another process to be investigated in order to increase the conductivity of the thicker silver traces is to print a thin layer of silver ink, sinter it photonically and then print on top of it another silver trace by repeating the same sintering process. By repeating this process we might be able to produce thicker and more conductive silver lines. The measurements of the insertion loss prove that the 4 µm thick CPW line of "design 2", which has the highest conductivity, demonstrates the least losses at 20 GHz and 40 GHz. The "4 µm thick design 2" CPW line was sintered using a 300 Volts pulse by keeping the duration of the pulse steady at 10 ms. The silver of this line started peeling off at the edges. For this reason, the voltage of the pulse was lowered to 250 Volts and the duration of the pulse was kept the same at 10 ms for the sintering of the rest of the lines. 78 Ref. Substrate Sintering Printing IL (dB/mm) IL (dB/mm) Method Method @ 20GHz @ 40GHz [67] ABS Thermal (90o C) Micro-dispensed 0.24 - [68] Diamond Thermal (200o C) AJP 0.28 0.46 [69] Ceramic Thermal (300o C) AJP - 0.7 This work ABS IPL AJP 0.16 0.37 Table 5.3: Comparison of CPW lines to literature. From [30] @ 2022 IEEE. -5 0 0.45 -0.1 0.4 -10 X 20.1 X 40 0.35 Y 0.328879 Y 0.401283 -0.2 -15 0.3 Insertion Loss dB/mm -0.3 dB(S11 ) S 11 and S22 S 21 and S12 -20 0.25 dB(S22 ) -0.4 dB(S12 ) 0.2 -25 dB(S21 ) -0.5 0.15 -30 -0.6 0.1 -35 -0.7 0.05 -40 -0.8 0 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 Frequency (GHz) Frequency (GHz) (a) (b) Figure 5.6: CPW design 1, 7.5µm thickness : (a) S-parameters; (b) Insertion Loss (dB/mm). From [30] @ 2022 IEEE. When comparing the insertion loss of the "design 1" CPW lines, which are all sintered using the same energy, it is verified that the thinner lines are more conductive and present less loss. The "design 1" CPW lines with a thickness of 4 µm and 7.5 µm have similar conductivity values, around 0.9-0.98×10^6 S/m and similar insertion losses at both 20 GHz and 40 GHz even though they have 3 µm difference in thickness. The conductivity decreases when the line thickness in- creases to 12 µm causing an increase to the insertion loss. As the conductivity values decrease, the loss increases, as expected. 5.6 Conclusion The effectiveness of AM techniques on automotive plastics has been investigated in this study. CPW lines have been aerosol jet printed on automotive ABS directly and sintered using IPL. The obtained conductivity and insertion loss values of the CPW lines at 20 GHz and 40 GHz are overall 79 -5 0 0.5 -0.1 0.45 X 40 -10 Y 0.477942 0.4 -0.2 -15 0.35 dB(S11 ) -0.3 X 20.1 Insertion Loss dB/mm dB(S22 ) 0.3 Y 0.350509 S 11 and S22 S 21 and S12 -20 -0.4 dB(S12 ) 0.25 dB(S21 ) -0.5 -25 0.2 -0.6 -30 0.15 -0.7 0.1 -35 -0.8 0.05 -40 -0.9 0 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 Frequency (GHz) Frequency (GHz) (a) (b) Figure 5.7: CPW design 1, 4µm thickness : (a) S-parameters; (b) Insertion Loss (dB/mm). From [30] @ 2022 IEEE. -10 0.2 0.4 -15 0.35 0 X 39.8 Y 0.356062 -20 0.3 -0.2 dB(S11 ) Insertion Loss dB/mm -25 0.25 dB(S22 ) S 11 and S22 S 21 and S12 -0.4 dB(S12 ) -30 0.2 X 20.1 dB(S21 ) -0.6 Y 0.157752 -35 0.15 -0.8 -40 0.1 -1 -45 0.05 -50 -1.2 0 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 Frequency (GHz) Frequency (GHz) (a) (b) Figure 5.8: CPW design 2, 4µm thickness : (a) S-parameters; (b) Insertion Loss (dB/mm). From [30] @ 2022 IEEE. better than the thermally sintered CPW lines found in the presented literature. IPL is a suitable alternative sintering method of silver nano-particle ink when using temperature sensitive substrates. The obtained experimental results demonstrate that aerosol jet printing directly on the auto- motive plastics and photonically sintering using IPL has potential as a non-destructive, low-cost process, making it suitable for automotive production lines. This is the first time, to the best of my knowledge that AJP combined with IPL are used to im- plement electronic devices on an automotive ABS substrate. This study evaluates the effectiveness of AJP in the implementation of small structures on automotive ABS and presents how the IPL affects the conductivity of different printed silver thickness. 80 CHAPTER 6 HELICAL ANTENNA FOR C-V2X APPLICATIONS 6.1 Background The aim of our research is to mitigate the number of automotive antennas on the rooftop of the ve- hicles. For this reason we researched on antenna designs for automotive C-V2X communications that can be hidden inside automotive plastic parts, maintain adequate performance for the applica- tion and not impede the performance of the other electrical elements around them. An electrically small, spherical helix antenna, in the normal mode of operation was fabricated and tested inside a car high mount stop light (CHMSL) of a Ford SUV vehicle. The CHMSL is also known as the third stop light that is in the rear of the vehicle. The suggested antenna is mounted behind the LED light and tested inside a near field MVG 3000 system while mounted on a real SUV vehicle. Most papers in literature suggest antennas that can be mounted on the rooftop of the vehicle for C-V2X communications. In [72] Jouini et al. compared the performance of different patch antennas that can be mounted on the rooftop for C-V2X applications, through simulations. The researchers of this study suggest that the rectangular patch antenna is the most directive with a gain of 10.1dBi. Zhou et al., in [73] simulated a low profile antenna, with a height of 4mm, vertical polarization, omnidirectional radiation pattern and a maximum simulated gain of 4.5dBi. In [74] Alavis et al., studied the near field and far field data of antenna measurements for C-V2X commu- nications by using a monocone antenna on the rooftop of a vehicle, through simulations. In [75], Shen et al. suggest a flush mount compact vehicle antenna system that supports 4-channel MIMO operation of LTE, 5G Sub 6, WLAN, and C-V2X communications. The antenna has a diameter of 15.2cm and a height of 2cm and is mounted on the rooftop of a vehicle during the simulations. The Table 6.1 sums up the comparison between the antenna developed for this research and the C-V2X antennas proposed in literature. Compared to the literature, the antenna developed for our research, is an electrically small antenna that can fit inside a Car High Mount Stop Light (CHMSL), without requiring a rooftop to serve as a ground plane. Our antenna radiation is more omni-directional than 81 Ref. Vehicle Mounting Point Antenna Size (mm3 ) Simulated Gain (dBi) [72] rooftop 24.4×31.5×2 10.1 [73] rooftop 1963.5 × 4 4.5 [75] rooftop 18146×20×90 10 This work CHMSL 50×37.5×5 6.2 Table 6.1: Comparison of helical antenna performance to C-V2X antennas from literature. the antenna design proposed by Jouini et al., more compact the the one designed by Shen et al. and has a better simulated realized gain than the antenna proposed by Zhou et al., even though it is not mounted on a ground plane during the simulations. To the best of my knowledge, the antenna presented in this dissertation is the only C-V2X antenna that is tested while mounted inside a CHMSL of a real vehicle and measurement data prove that its performance is comparable to antennas mounted on a rooftop of a vehicle. 6.2 Design The antenna design is inspired by Best in [76]. It consists of three helical arms, wrapped around a hemisphere with radius of 4.8mm and are mounted over a ground plane. Two of its arms are soldered on the ground plane and the third one is connected to a pigtail, that is on the back of the ground plane, to feed the antenna. The three arms are connected at the top of the hemisphere. The antenna operates in a monopole configuration and has a vertical polarization, as it is mounted over a PEC ground, the vertical components of current within the helical arms are in phase so they add constructively, whereas the parallel components cancel each other. The circumference of the spherical helix antenna is 30.144mm which is less than a wavelength at 5.9 GHz and k × α = 2π 0.5 where k = λ and α is the radius of a sphere encompassing the antenna. According to the definition of [76], the suggested antenna is an electrically small antenna. The antenna, as shown in Fig.6.1a consists of a hemisphere with radius 4.8mm adjusted on a ground plane with a size of 50×37.5 mm2 . The Rogers 3850 which is copper plated on both sides is used as a ground plane. It has a thickness of 1mil, ϵr =2.9 and tanδ=0.0025. The radiating elements of the antenna are the three helical arms wrapped around the hemisphere. The helical 82 (a) (b) Figure 6.1: Antenna design as simulated in Ansys HFSS and the simulated S-parameters. (a) (b) Figure 6.2: Simulated antenna co- and cross- polarization on E and H planes. arms have a radius of 0.127mm. The two of them are soldered on the ground plane. A hole is opened on a ground plane and the third arm is driven to the other side of the ground plane where it is connected to a pigtail, serving as the feed for the antenna. The three arms are connected on the top of the hemisphere. A reflector is adjusted λ/4 away from the one side of the antenna, to help boost its performance and protect it from the reflections caused by the metallic body of the vehicle. According to the simulations, the antenna has a maximum realized gain of 6.2dBi at 5.9 GHz and θ = 90o . The simulated S11 is -20dB and the simulated efficiency of the antenna at this frequency is 94.2%. The simulated resonance and radiation patterns of the antenna are shown in Fig.6.2. 83 (a) (b) Figure 6.3: Fabricated antenna with pigtail and reflector. 6.3 Fabrication To fabricate the antenna a 50×37.5 mm2 part of the Rogers 3850 is used as a ground plane and a 50×25 mm2 part of the same material is used as a reflector. A hemisphere made by vero-white is 3D printed using the Connex 3D polyjet printer. The vero-white is an opaque polyjet resin that is used to create 3D printed objects. Its base is a photosensitive polymer liquid. It has a permittivity of ϵr =2.97 and tanδ=0.0285, as stated in [77] , and a heat deflection temperature of 50o C, [78]. For the radiating elements of the antenna, a 30 AWG wire was cut and soldered on the ground and on the top of the vero-white hemisphere. A pigtail was used to feed the antenna. One of the arms was soldered with the pigtail. Photos of the fabricated antennas are shown in Fig.6.3. 6.4 Measurements After fabricating three similar antennas, the S-parameters were measured using the M5227 PNA by Keysight Technologies. The S-parameters were measured while the antennas were alone, not mounted inside the CHMSL. The antennas were transferred at the Drive-ability Test Facility (DTF) of Ford Motor Company. The MVG3000M near field system was used to measure the antennas performance. Four different experiment setups were performed at the DTF; a) the antennas were mounted one by one, on a foam pillar, b) the antennas were mounted on a circular ground plane with a diameter of 1m, absorbers were adjusted below the ground plane to remove any unwanted 84 Figure 6.4: Experimental setup B: Ground plane inside MVG 3000 at DTF. reflection from the turn table, c) the antennas were mounted one by one inside a CHMSL of a Ford SUV vehicle, behind the LED light while the cables of the LED light were removed from inside the CHMSL, and d) the antennas were mounted one by one inside a CHMSL of a Bronco vehicle, behind the LED light and with all the cables of the LED light connected. Again, absorbers were adjusted around the vehicle and on the walls of the room, as indicated in Fig.6.5a to isolate any reflection of the surroundings on the antenna radiation pattern. The two different setups B and D are indicated in Fig.6.4 and Fig.6.5a. The measured S-parameters when the antennas were alone, not mounted on the ground or the vehicle are indicated in Fig.6.6. The radiation properties of the three fabricated antennas when mounted on the foam pillar, during experimental setup A, are presented in Fig.6.7. The radiation performance of the antennas when mounted on the ground plane, during experimental setup B, is indicated in Fig.6.8. The radiation performance of the three antennas when mounted inside the CHMSL of the Bronco, behind the LED light, without the cables of the LED light, during experi- mental setup C, is shown in Fig.6.9, and the radiation performance of the antennas as measured at DTF, during experimental setup D is indicated in Fig.6.10. 85 (a) (b) Figure 6.5: Experimental setup C and D: a)Bronco inside MVG 3000 at DTF and b)antenna inside the CHMSL of the Bronco. (a) (b) (c) Figure 6.6: Measured S-parameters: a) of antenna I; b) of antenna II; c) of antenna III. 6.5 Results All the antennas are measured at 5.9GHz. The measurement results of the experimental setup A where the antennas are mounted on the foam pillar are presented in Tables 6.2 and 6.3. The measured results of the experimental setup B, when the antennas are placed on the ground plane are presented in Tables 6.4 and 6.5. The measurement results of the experimental setup C where the antennas are mounted inside the CHMSL of the vehicle, with the LED cables removed, are 86 (a) (b) (c) Figure 6.7: Measured radiation properties of a) antenna I, b) antenna II , and c) antenna III, when the antennas are mounted on the pillar of foams during experimental setup A. (a) (b) (c) Figure 6.8: Measured radiation properties of a) antenna I, b) antenna II , and c) antenna III, when the antennas are mounted on the ground plane during experimental setup B. presented in Tables 6.6 and 6.7, and the measurement results of the experimental setup D where the antennas are mounted inside the CHMSL of the SUV, with the LED cables present are gathered in Tables 6.8 and 6.9. The data of the antennas measurements on the pillar of foams and the ground are averaged in Tables 6.2 and 6.4 over the angles 0o ≤ ϕ ≤ 60o and 300o ≤ ϕ ≤ 360o on the azimuth range. Because the antenna has a reflector and the antenna is centered under the sensor of the MVG 3000, its maximum gain is expected at ϕ = 0o . In Tables 6.3 and 6.5 the gain of the antenna is averaged over the azimuth range of 60o ≤ ϕ ≤ 90o and 270o ≤ ϕ ≤ 300o . Again, the reason for averaging over this range is that the antenna is centered under the top sensor of the MVG 3000 system and the expected maximum value of realized gain is at ϕ = 0o . During 87 (a) (b) (c) Figure 6.9: Measured radiation properties of a) antenna I, b) antenna II , and c) antenna III, when the antennas are mounted inside the CHMSL, with the LED light on, after removing the LED cables, during experimental setup C. (a) (b) (c) Figure 6.10: Measured radiation properties of a) antenna I, b) antenna II , and c) antenna III, when the antennas are mounted inside the CHMSL, with the LED light and the LED cables on, during experimental setup D. the experiments C and D, when the antenna is placed inside the CHMSL of the SUV vehicle, the CHMSL is at the back of the vehicle, and the top sensor of the MVG 3000 system is centered over the top of the antenna, so the maximum of the antenna is expected at ϕ = 180o . This is the reason the gain of the antenna is averaged over 120o ≤ ϕ ≤ 240o , 90o ≤ ϕ ≤ 120o and 240o ≤ ϕ ≤ 270o in Tables 6.6, 6.7, 6.8, and 6.9. Since the antenna is placed in the center back of the vehicle, where ϕ = 180o , it makes sense to test the antenna radiation performance in the azimuth range of 90o ≤ ϕ ≤ 270o . 88 When comparing the performance of the three antennas on the foam pillar and the ground plane in the range of 0o ≤ ϕ ≤ 60o and 300o ≤ ϕ ≤ 360o , in Tables 6.2 and 6.4, antenna II has the highest values of average realized gain compared to antennas I and III. This is expected since antenna II has a better resonance at 5.9 GHz compared to antennas I and III, as indicated in Fig.6.6a, 6.6b , and 6.6c. When the antennas are placed inside the CHMSL of the vehicle, as indicated in Tables 6.6 and 6.8, antenna II has the highest values of the averaged realized gain compared to antennas I and III. This can be explained again by the better matching of the antenna at 5.9 GHz. The average gain of the antenna II is lower when the antenna is placed inside the CHMSL compared to the values of the average realized gain of the antenna on the foam pillar and the ground plane for θ ≤ 90o . For θ ≥ 90o the antenna performance is affected by the CHMSL. The antenna during experimental setups C and D is placed behind a LED light, inside the plastic cover of the CHMSL. There is a spare tire, right in front and below the antenna position. As a result, there are reflections created by the vehicle metallic body, losses due to the LED light, mounted in front of the antenna and the cables of the LED light that are placed in close proximity to the antenna. From Table 6.10, where the maximum measured realized gain of the three antennas at the different experimental setups is gathered, we can see that antenna II has very similar maximum realized gain values at all the setups. However antennas I and III have a better performance when mounted inside the CHMSL. This can be explained by the resonance of these two antennas. When mounted inside the CHMSL, the antennas are placed on an Acrylonitrile Butadiene Styrene (ABS) vehicular part with an ϵr = 2.5, as a result, the resonance shifts to lower frequencies. Antennas I and III resonate around 6 GHz when measured alone, not mounted on any part. When the antennas are mounted on a part with ϵr = 2.5 the resonance shifts closer to 5.9GHz where the measurements are made. When comparing the radiation performance of the three antennas on the foam pillar and the infinite ground, as indicated in Fig.6.7 and 6.8, the antennas cross-polarization level is lower during the experimental setup B. Moreover the maximum realized gain increases when there is an infinite ground plane. However, the radiation patterns of all three antennas are not as smooth as 89 Measured Average Realized Gain (dBi) Azimuth Range Polar Range Antenna I Antenna II Antenna III 90o ≤ θ ≤ 96o -5.91 -6.1 -3.83 0o ≤ ϕ ≤ 60o 89o ≤ θ ≤ 91o -5.91 -3.59 -4.16 84o ≤ θ ≤ 90o -5.16 -2.16 -3.13 and 80o ≤ θ ≤ 84o -5.16 -1.21 -3.17 70o ≤ θ ≤ 80o -7 -2.12 -4.16 300o ≤ ϕ ≤ 360o 50o ≤ θ ≤ 70o -2.47 -0.77 -1.26 Table 6.2: Measurement results of antenna on the pillar of foams at DTF (experiment A). Measured Average Realized Gain (dBi) Azimuth Range Polar Range Antenna I Antenna II Antenna III 90o ≤ θ ≤ 96o -14 -12.58 -14.17 60o ≤ ϕ ≤ 90o 89o ≤ θ ≤ 91o -6.2 -3.6 -4.16 84o ≤ θ ≤ 90o -13.6 -14.5 -12.51 and 80o ≤ θ ≤ 84o -10.65 -13.03 -10.46 70o ≤ θ ≤ 80o -9.2 -9.65 -9.95 270o ≤ ϕ ≤ 300o 50o ≤ θ ≤ 70o -9.3 -9.07 -7.87 Table 6.3: Measurement results of antennas on the pillar of foams at DTF (experiment A). when they are measured on the pillar of foams. During the experimental setups C and D where the antennas are measured in the CHMSL area of the vehicle, the cross-polarization level increases and the radiation pattern of the antennas develops some lobes. At this position the antennas have no infinite ground plane and there are reflections created by the vehicle body, the spare tire adjusted below the mounting position of the antennas and other vehicular metal parts above and behind the antenna. From Table 6.10 it is obvious that the presence of the cable in experimental setup D, results in lower maximum realized gain values, compared to the experimental setup C, where the antennas are mounted in the exact same position, but we removed the LED cable. However, the average realized gain values, measured during experimental setups C and D are similar. The overall cross- polarization levels of the antennas are lower when the LED cable is absent compared to the cross- polarization patterns of experimental setup D, when the cables of the LED light are present. 90 Measured Average Realized Gain (dBi) Azimuth Range Polar Range Antenna I Antenna II Antenna III 90o ≤ θ ≤ 96o -5.9393 -5.1258 -7.2103 0o ≤ ϕ ≤ 60o 89o ≤ θ ≤ 91o -3.5267 -3.1339 -5.4965 84o ≤ θ ≤ 90o -1.3826 -1.7866 -3.8436 and 80o ≤ θ ≤ 84o 1.144 0.3584 -1.4932 70o ≤ θ ≤ 80o 2.3142 1.4574 -0.1062 300o ≤ ϕ ≤ 360o 50o ≤ θ ≤ 70o -1.0159 -1.7891 -3.2391 Table 6.4: Measurement results of antenna on the ground at DTF (experiment B). Measured Average Realized Gain (dBi) Azimuth Range Polar Range Antenna I Antenna II Antenna III 90o ≤ θ ≤ 96o -12.8622 -12.4697 -19.6434 60o ≤ ϕ ≤ 90o 89o ≤ θ ≤ 91o -9.9245 -9.6169 -11.0198 84o ≤ θ ≤ 90o -7.3841 -9.6481 -7.0814 and 80o ≤ θ ≤ 84o -4.5053 -7.7627 -4.2692 70o ≤ θ ≤ 80o -2.9833 -6.9139 -2.609 270o ≤ ϕ ≤ 300o 50o ≤ θ ≤ 70o -3.2349 -11.2414 -5.8874 Table 6.5: Measurement results of antenna on the ground at DTF (experiment B). Measured Average Realized Gain (dBi) Azimuth Range Polar Range Antenna I Antenna II Antenna III 90o ≤ θ ≤ 96o -4.83 -3.77 -4.66 89o ≤ θ ≤ 91o -4.34 -3.72 -4.56 120o ≤ ϕ ≤ 240o 84o ≤ θ ≤ 90o -4.66 -2.6 -4.36 80o ≤ θ ≤ 84o -3.66 -3 -2.72 70o ≤ θ ≤ 80o -3.43 -1.69 -2.52 50o ≤ θ ≤ 70o -2.46 -0.95 -2.32 Table 6.6: Measurement results of antenna in the CHMSL, LED cables removed (experiment C). Measured Average Realized Gain (dBi) Azimuth Range Polar Range Antenna I Antenna II Antenna III 90o ≤ θ ≤ 96o -9.66 -8.77 -7.57 90o ≤ ϕ ≤ 120o 89o ≤ θ ≤ 91o -9.67 -9.13 -7.2 84o ≤ θ ≤ 90o -9.88 -10 -10.12 and 80o ≤ θ ≤ 84o -9.93 -10.48 -6.97 70o ≤ θ ≤ 80o -7.67 -9.29 -7.5 270o ≤ ϕ ≤ 300o 50o ≤ θ ≤ 70o -7.83 -10.72 -8.85 Table 6.7: Measurement results of antenna in the CHMSL, LED cables removed (experiment C). 6.6 Conclusion A small helical antenna implemented on the 3D printed vero white material has been tested inside a CHMSL of a real vehicle, resulting in adequate performance for C-V2X communications. This 91 Measured Average Realized Gain (dBi) Azimuth Range Polar Range Antenna I Antenna II Antenna III 90o ≤ θ ≤ 96o -6.0623 -3.7845 -4.4395 89o ≤ θ ≤ 91o -5.9841 -3.4955 -4.2427 120o ≤ ϕ ≤ 240o 84o ≤ θ ≤ 90o -5.981 -2.7928 -3.7405 80o ≤ θ ≤ 84o -5.435 -2.4521 -3.1144 70o ≤ θ ≤ 80o -3.7635 -2.9239 -3.1614 50o ≤ θ ≤ 70o -5.4809 -4.4388 -5.2846 Table 6.8: Measurement results of antenna inside CHMSL. LED cables present (experiment D). Measured Average Realized Gain (dBi) Azimuth Range Polar Range Antenna I Antenna II Antenna III 90o ≤ θ ≤ 96o -7.931 -7.0116 -5.327 90o ≤ ϕ ≤ 120o 89o ≤ θ ≤ 91o -7.3589 -6.4735 -5.0446 84o ≤ θ ≤ 90o -6.4447 -5.9346 -5.2053 and 80o ≤ θ ≤ 84o -5.3016 -4.9558 -5.0956 70o ≤ θ ≤ 80o -7.0815 -5.9862 -6.5918 240o ≤ ϕ ≤ 270o 50o ≤ θ ≤ 70o -10.3142 -8.7928 -10.5426 Table 6.9: Measurement results of antenna inside CHMSL. LED cables present (experiment D). Max Measured Realized Gain (dBi) Experimental Setup Antenna I Antenna II Antenna III A - antennas on the pillar of foams -1.46 2.4 0.67 B - antennas on the ground plane 1.42 2.5 0.55 C - antennas mounted on the rear area of 1.77 4.12 1.62 Bronco, while LED cables are removed D - antennas mounted on the rear area of 2.1 2.4 1.28 Bronco , while LED cables are present Table 6.10: Max measured realized gain of antennas at the four different experimental setups. research introduces for the first time the CHMSL area of an SUV vehicle as candidate placement of automotive antennas. The small size of the presented antenna permits positioning it behind the third stop light, contributing to the mitigation of the number of antennas from the rooftop, which results in optimization of the overall aesthetics of the vehicle. The measured performance of the antenna is efficient for C-V2X communications and comparable to the performance of the antenna when mounted on an infinite ground plane, like the rooftop. However, a second antenna should be placed in the front part of the vehicle to cover the 360o of the horizon. Compared to published research on automotive C-V2X antennas, this is the first time to the best of my knowledge, that an antenna 92 has been tested inside the CHMSL of a real vehicle and proves to have efficient performance. 93 CHAPTER 7 FULLY 3D PRINTED HELICAL ANTENNA 7.1 Background The electrically small helical antenna for C-V2X communications, presented in previous chapter is developed using AM techniques. Introducing AM techniques in the development of automotive antennas can reduce the cost and time to mass produce automotive antennas. The antenna substrate and helical arms are fully 3D printed using AJP and silver nanoparticle ink. Silver epoxy is used for antenna connectorization. To the best of our knowledge, this is the first helical antenna implemented by AJP. In Table 7.1 the implementation method of the antenna suggested in this dissertation is compared to 3D printed techniques for the implementation of similar antennas in literature. In [79], Yan et al. developed a hemispherical helix GPS antenna with similar design, by selective metallization technique. Kong et al. implemented a magnetic monopole through three-dimensional printing (3DP) technology, resonating around 224 MHz as indicated in [80]. A helix operating in axial mode at 4.72 GHz-5.32 GHz is fabricated using photo- curable 3-D printer by Song et al., in [81]. Compared to the literature, the proposed helical antenna, consists of 3 arms connected at the top. The metallic part of the antenna is implemented using AJP and the substrate is 3D printed through Connex 3D printer. The helix has an omnidirectional radiation pattern, linear polarization, and resonates around 5.9GHz. Ref. Implementation Antenna Size (mm3 ) Method [79] Selective Metallization 70×77×40 [80] 3DP h= 0.0016λ,ka=0.276 [81] Photo-curable 3D printer 22.4×22.4×90 This work AJP and photo-curable 50×50×4.6 3D printer Table 7.1: Comparison of AM techniques for the implementation of a helix antenna. 94 (a) (b) Figure 7.1: 3D printed antenna design a) as simulated in Ansys Electronics Desktop; b)implemented using AJP and Connex 3D printer. 7.2 Design The antenna design is the same to the antenna implemented in chapter 6. It consists of three spher- ical helical arms, wrapped around a hemisphere and mounted over a ground plane. As explained before, the antenna operates in a monopole configuration, providing omni-directional radiation pattern and vertical polarization. The antenna design is indicated in Fig.7.1. The simulated S- parameters and co-polarization of the 3D-printed antenna on E and H-plane are shown in Fig.7.4. The maximum simulated realized gain of the antenna at 5.9GHz and θ=90o is -0.59dBi. 7.3 Fabrication The antenna, as shown in Fig.7.1 consists of a vero-white hemisphere with radius 4.6mm mounted on a ground plane with a size of 50×50 mm2 . The vero-white hemisphere is 3D printed using the Connex 3D polyjet printer as in chapter 6. The radiating elements of the antenna are three helical arms printed on the hemisphere using the AJP. The helical arms have a rectangular cross section with a width of 0.255mm and are printed using the reactive silver nano-particle ink El-615 by Electroninks, [82]. The two of its arms are soldered on the ground plane using the Epo-Tek H20E silver epoxy, and one of them is connected to an SMA pigtail, serving as the feed for the 95 Figure 7.2: Pattern of helical arms to be printed. antenna. The three arms are connected on the top of the hemisphere. The pattern that was uploaded on AJP software and printed on the vero-white, is indicated in Fig.7.2. The pattern to be printed is planar. The stage of the printer is kept planar and the printer nozzle tip is steady on the top of the dome at a distance of 6mm from the print stage. In total, 8 layers of silver were printed, providing a total silver thickness of 0.8um. After printing the helical arms, keeping the same printer settings and same amount of printed layers, a 5mm × 5mm rectangle was printed on a flat vero-white. The rectangle was sintered using the same steps as for the helical arms. The rectangle was used to estimate the conductivity and thickness of the printed helical arms. After printing, the antenna and the 5mm × 5mm rectangle were placed in the oven at 180o C for half an hour. Using a commercially available tool that utilizes IPL technology, the antenna arms and the rectangle were flashed. The IPL tool has a repetition rate of 0.5-0.9 seconds. The max optical output is 6 J/cm2 and has a pulse length of 2.6msec. The wavelength of the light is between 510-2000nm and the power line input is 100-240V, 50-60Hz, 1.7-0.91A. The head of the tool through which the light was applied on the antenna and the rectangle has a surface of 1.5 cm2 . The antenna and the rectangle were kept at a distance of 1cm from the device head. In total, 8 flashes were applied on top of the antenna dome and around the hemisphere by turning the antenna every 45o angles. 96 0 dB(S11 ) -5 -10 Magnitude (dB) X 5.64 Y -10.1436 -15 -20 X 5.89333 -25 Y -23.564 1 2 3 4 5 6 7 Frequency (GHz) Figure 7.3: 3D printed antenna measured and simulated S-parameters, . 7.4 Results The conductivity of the printed silver rectangle was measured using the Lucas Pro4-4400 Signatone QuadPro Resistivity system and was found to be 9.66 × 106 S/m. As a result there is a value of resistance that causes a shift to the antenna resonance towards lower frequencies as indicated in the measured S-parameters of Fig.7.3. The realized gain of the implemented antenna was measured on a foam pillar at the far field measurement system of Michigan State University. The maximum measured realized gain of the antenna inside the far field measurement system, at 5.9GHz, for θ = 90o is -2dBi and at 5.6GHz where the fabricated antenna presents a better resonance, the measured maximum realized gain is -1.88dBi. The measured antenna radiation pattern is omni- directional, as indicated in Fig.7.4. The measured radiation patterns of the antenna at θ = 90o have some variances. This can be explained by the measurement setup. The Rogers 3850 is a flexible material and the antenna was mounted on a foamy stanchion using tape during measurements. The mechanical strengths of the cables during measurements caused some curves on the Rogers 3850. The gain of the implemented antenna at 5.9 GHz, is 2dB lower than the simulated antenna. This is expected as the 3D printed antenna is not fully conductive at 5.9GHz and has a low S11 , impeding its performance. 97 (a) (b) Figure 7.4: 3D printed antenna: a)measured and simulated radiation pattern on H-plane , b)measured and simulated radiation pattern on E-plane. 7.5 Conclusion A fully 3D printed helical antenna on a hemisphere, developed using AJP, polyjet printers and photonic sintering is presented. The antenna was implemented using reactive silver nanoparticle ink on a hemisphere with a radius of 4.6mm by AJP. The hemisphere was printed using vero- white by the Connex 3D polyjet printer and photonic sintering was applied on the module. This research proves that AJP, Connex 3D polyjet printer and IPL are viable methods for manufacturing automotive antennas on curved, temperature sensitive surfaces. By introducing AM techniques in automotive industry for the development of vehicular antennas, the time and cost of manufacturing can be decreased. 98 CHAPTER 8 CONCLUSION Part of the presented research was to identify an antenna design that can operate in two extreme environments and be detected by another antenna in the far field, using RFID technology. The antenna designed and connected to the custom energy harvesters has the ability to be detected in a distance of 57cm from the reader antenna when implanted in a human muscle which is the lossiest part of the human body and in a distance of 125 cm when measured outside the human muscle, in free space. Compared to other RFID rectennas from literature, operating at 915 MHz, the developed rectenna has the best performance when submerged in the mix mimicking the human muscle while having a size of only 38×12.7 mm2 . For the second part of the presented research, an additively manufactured antenna that can be mounted on an automotive plastic has been developed and tested on real vehicles. The issue that arises from the multiple antennas and the limited available mounting positions, especially for vehicles with no permanent rooftop, can benefit from this research. New candidate positions for antenna mounting on a vehicle have been identified through this research and the performance of different antennas has been tested, pinpointing factors that affect it. As a result, there is better understanding of the solution to the issue that arises because of the numerous antennas and the min- imum available placements on a vehicle. The suggested antenna designs can be used for cellular and C-V2X automotive communications. Moreover, this research suggests additive manufacturing techniques for the implementation of large and small structures on thermal sensitive substrates, using photonic sintering. The results retrieved by the studies made so far, indicate that AJP and IPL are promising techniques for the manufacture of antennas of different shapes on light colored, thermal sensitive substrates. AM techniques are preferred for large scale production lines. The minimum time to manufacture the desired electronic devices as well as the detailed features that can be printed combined with the conductivity values that IPL offers show that the AM techniques are effective for the production of the desired devices. 99 APPENDIX A: FULLY 3D PRINTED ANTENNA FOR APPLICATIONS AT 28 GHZ The fully 3D printed antenna presented in Chapter 7 is modified to resonate at 28GHz. A fourth helical arm is added to the radiated elements to increase the antenna resistance. The vero-white hemisphere that is used as a substrate has a radius of 2.9mm, the Rogers 3850 is used as a ground plane for the hemisphere and has dimensions of 35mm×35mm×0.1mm. The helical arms have a rectangular cross section with a width of 0.2mm. The simulated antenna, and the simulated results are presented in Fig.8.1 and 8.2. 0 S11 measured -5 -10 Magnitude (dB) -15 -20 -25 24 25 26 27 28 29 30 31 32 Frequency (GHz) (a) (b) Figure 8.1: Simulated antenna for 28GHz applications a) antenna design; b)simulated S- parameters. To fabricate this antenna, the vero-white hemisphere is created using the Connex 3D printer and the helical arms are printed on the vero-white using the Optomec AJP and the El-1403 silver ink by Electron Inks. The four arms of the fabricated antenna are connected to the ground using Epo-Tek H20E silver epoxy and the fourth arm is connected to a 2.92mm to SMA connector to feed the antenna, as indicated in Fig.8.3. The measured S-parameters of the fabricated antenna are presented in Fig.8.3c. The S-parameters of the fabricated antenna were measured using the Keysight Network Analyzer. Compared to the simulations, the fabricated antenna resonates around 32 GHz, instead of 28 GHz. This is caused most probably because of human error during the connectorization of the antenna. As explained, the fourth arm of the antenna is connected to a 2.92mm connector using silver epoxy. The small size of the antenna makes it difficult to precisely connect the fourth arm to 100 (a) (b) Figure 8.2: Simulated antenna for 28GHz applications co and cross-polarization a) on E-plane; b)on H-plane. (a) (b) (c) Figure 8.3: 3D printed antenna for 28GHz applications a) antenna design top view; b)antenna design back view; c)measured S-parameters of the fabricated antenna.. the feeding pin of the connector. Placing the feeding pin of the connector slightly higher than the end of the helical arm caused the shift of resonance to higher frequency because the length of the fourth arm is reduced. More similar antennas will be fabricated, to optimize the fabrication process. To the best of my knowledge, this is the first fully 3D printed helical antenna, using Connex 3D printer and AJP for applications at 28 GHz. 101 APPENDIX B: AEROSOL JET PRINTING MATERIALS The additively manufactured antennas presented in Chapter 4 and the CPW lines presented in Chapter 5 were developed using the Metalon JS-A221AE silver ink. It is a water based silver nano-particle with silver content of 50 wt % as stated in [83]. The viscosity of this ink is 10-20cP which makes it suitable for ultrasonic atomization. According to the technical specifications of this ink, if it is sintered at 200o C for 1 hour, it can reach a resistivity equal to 5.8 times of the bulk silver. While researching materials that can be sintered to become conductive in low frequencies, the PSPI-1000 silver nanoparticle ink was tested. The PSPI-1000 is an aqueous based ink, with silver nanoparticles of 50-100nm in diameter and 40wt.% silver. It requires only 15 minutes heating at 70o C to become conductive or 1 minute at 120o C but has a very sensitive pH and requires constant setting. For this reason it was not used for the implementation of antennas that will be manufactured in a large production line. Details on the handling and the specifications of this ink can be found in [84]. Silver Sintering Conductivity Thick. Rough. Ink Process (S/m) (µm) µm) Clariant EXPT UV sintering and Prelect TPS 50G2 Thermal at 150o C for 1 hour 1.04 × 10^6 1.5 0.1207 Metalon JS-A221 AE Thermal at 150o C for 1 hour 0.23 × 106 2.5 0.0991 PSPI-1000 Thermal at 75o C for 15 minutes 2.09 × 106 20 - Table 8.1: Comparison of silver conductivity on dark ABS. At the beginning of this research the Clariant EXPT Prelect TPS 50G2 silver ink was printed on black ABS. The Clariant ink has a viscosity of 15cP and reaches a a resistivity 2 times of the bulk silver when sintered at 200o C for 1 hour, as stated in [85]. As an experiment, to test the performance of these three silver inks on ABS, rectangles with a surface of 10 × 10 mm2 were printed on dark ABS and sintered thermally and photonically. Two of these rectangles are showed in Fig.8.4. The results of this experiment are indicated in Table 8.1. For the additively manufactured helical antenna, the El-615 reactive silver ink by Electron Inks 102 (a) (b) Figure 8.4: Rectangles printed on dark ABS: a)using Metalon JS-A221 AE silver ink, b)using PSPSI-1000 silver ink. was used. When this ink is sintered at 120o C for 1 hour, it can reach a resistivity equal to 6 times the resistivity of bulk silver. 103 APPENDIX C: AEROSOL JET PROCEDURES The rectangles, using the Clariant EXPT Prelect TPS 50G2 and the Metalon JS-A221AE silver inks were printed on the dark ABS using the ultrasonic atomizer of the AJP. The rectangle of PSPI-100 silver ink was printed using the pneumatic atomizer of the AJP. The settings of the AJP while printing the rectangles are indicated in Table 8.2. An antenna similar to the one presented in Chapter 4 was printed using the Metalon JS-A221 AE silver ink in the pneumatic atomizer of the AJP and the wide feature printhead. When the fine feature printhead of the AJP was used to print this antenna, it would take 1 hour and 30 minutes to print one layer. While using the wide feature printhead, one layer of this antenna would be printed in 19 minutes. The settings of the wide feature printhead while printing the antenna are showed in Table 8.3. This printing was performed as a test to verify the performance of AJP in production lines for the implementation of large devices. Silver Nozzle D Sheath UA PA Exhaust Ink (µm) (SCCM) (SCCM) (SCCM) (SCCM) Clariant EXPT Prelect TPS 50G2 300 110 20 - - Metalon JS-A221 AE 300 80 22 - - PSPI-1000 300 80 - 690 540 Table 8.2: AJP printing settings when printing the rectangles of different silver inks. Sheath (SCCM) PA (SCCM) Exhaust (SCCM) 500 40 300 Table 8.3: AJP wide-feature printhead settings. The CPW lines, presented in Chapter 5 were developed using the Metalon JS-A221 AE silver ink in the ultrasonic atomizer of AJP and the fine feature printhead. The settings of the printer to develop the presented CPW lines are indicated in Table8.4. Nozzle D (µm) Sheath (SCCM) UA (SCCM) 150 65 20 Table 8.4: AJP printing settings when printing the CPW lines. 104 BIBLIOGRAPHY [1] Xiaoyu Liu, J Lee Berger, Ajay Ogirala, and Marlin H Mickle. A touch probe method of operating an implantable rfid tag for orthopedic implant identification. IEEE Transactions on Biomedical Circuits and systems, 7(3):236–242, 2012. [2] Zhibin Xiao, Xi Tan, Xianliang Chen, Sizheng Chen, Zijian Zhang, Hualei Zhang, Junyu Wang, Yue Huang, Peng Zhang, Lirong Zheng, et al. An implantable rfid sensor tag to- ward continuous glucose monitoring. IEEE journal of biomedical and health informatics, 19(3):910–919, 2015. [3] Wen Yao, Chao-Hsien Chu, and Zang Li. The use of rfid in healthcare: Benefits and barriers. In 2010 IEEE International Conference on RFID-Technology and Applications, pages 128– 134. IEEE, 2010. [4] Md Mohiuddin Soliman, Muhammad EH Chowdhury, Amith Khandakar, Moham- mad Tariqul Islam, Yazan Qiblawey, Farayi Musharavati, and Erfan Zal Nezhad. Review on medical implantable antenna technology and imminent research challenges. Sensors, 21(9):3163, 2021. [5] Niels Koch and J Carmo. Antennas for automobiles. New Advances in Vehicular Technology and Automotive Engineering, page 410, 2012. [6] Markforged. Additive manufacturing history: From the 1980’s to now. page 1, 2000. [7] twiglobal. Additive manufacturing history. page 1, 2000. [8] 3DVERKSTAN. History of additive manufacturing. page 1. [9] A Shastri, B Sanz-Izquierdo, and S Gao. Manufacturing considerations for the development of reconfigurable antennas using inexpensive inkjet printing. IEEE Transactions on Compo- nents, Packaging and Manufacturing Technology, 2022. [10] J Heirons, S Jun, A Shastri, Benito Sanz-Izquierdo, David Bird, L Winchester, L Evans, and A McClelland. Inkjet printed gps antenna on a 3d printed substrate using low-cost machines. In 2016 Loughborough Antennas & Propagation Conference (LAPC), pages 1–4. IEEE, 2016. [11] Yong Li, Chunting Wang, Hongwei Yuan, Ning Liu, Hualong Zhao, and Xiali Li. A 5g mimo antenna manufactured by 3-d printing method. IEEE Antennas and Wireless Propagation Letters, 16:657–660, 2016. [12] Sung Yun Jun, Benito Sanz-Izquierdo, Edward A Parker, David Bird, and Alan McClelland. Manufacturing considerations in the 3-d printing of fractal antennas. IEEE Transactions on Components, Packaging and Manufacturing Technology, 7(11):1891–1898, 2017. [13] Sung Yun Jun, Ahmed Elibiary, Benito Sanz-Izquierdo, L Winchester, D Bird, and A Mc- Cleland. 3-d printing of conformal antennas for diversity wrist worn applications. IEEE Transactions on Components, Packaging and Manufacturing Technology, 8(12):2227–2235, 2018. 105 [14] Rui Xu, Steven Gao, Benito Sanz Izquierdo, Chao Gu, Patrick Reynaert, Alexander Standaert, Gregory J Gibbons, Isakov Dmitry, Wolfgang Bösch, and Michael Ernst Gadringer. 140 ghz additive manufacturing low-cost and high-gain fabry-perot resonator antenna. In 2020 International Workshop on Antenna Technology (iWAT), pages 1–4. IEEE, 2020. [15] Vincens Gjokaj, John Papapolymerou, John D Albrecht, Brian Wright, and Premjeet Chahal. A compact receive module in 3-d printed vivaldi antenna. IEEE Transactions on Components, Packaging and Manufacturing Technology, 10(2):343–346, 2019. [16] Sung Yun Jun, Anshuman Shastri, Benito Sanz-Izquierdo, David Bird, and Alan McClel- land. Investigation of antennas integrated into disposable unmanned aerial vehicles. IEEE Transactions on Vehicular Technology, 68(1):604–612, 2018. [17] Lixin Mo, Zhenxin Guo, Li Yang, Qingqing Zhang, Yi Fang, Zhiqing Xin, Zheng Chen, Kun Hu, Lu Han, and Luhai Li. Silver nanoparticles based ink with moderate sintering in flexible and printed electronics. International journal of molecular sciences, 20(9):2124, 2019. [18] Kate Black, Jetinder Singh, Danielle Mehta, Sarah Sung, Christopher Sutcliffe, Paul Chalker, et al. Silver ink formulations for sinter-free printing of conductive films. Scientific reports, 6(1):1–7, 2016. [19] Lixin Mo, Zhenxin Guo, Zhenguo Wang, Li Yang, Yi Fang, Zhiqing Xin, Xiu Li, Yinjie Chen, Meijuan Cao, Qingqing Zhang, et al. Nano-silver ink of high conductivity and low sintering temperature for paper electronics. Nanoscale research letters, 14(1):1–11, 2019. [20] Amit J Lopes, In Hwan Lee, Eric MacDonald, Rolando Quintana, and Ryan Wicker. Laser curing of silver-based conductive inks for in situ 3d structural electronics fabrication in stere- olithography. Journal of Materials Processing Technology, 214(9):1935–1945, 2014. [21] Wendong Yang, Felix Hermerschmidt, Florian Mathies, and Emil JW List-Kratochvil. Com- paring low-temperature thermal and plasma sintering processes of a tailored silver particle- free ink. Journal of Materials Science: Materials in Electronics, 32(5):6312–6322, 2021. [22] Chaimaa El Hajjaji, Nicolas Delhote, Serge Verdeyme, Malgorzata Piechowiak, and Olivier Durand. Optimization of the conductivity of microwave components printed by inkjet on polymeric substrates by photonic sintering. In 2020 50th European Microwave Conference (EuMC), pages 380–383. IEEE, 2021. [23] Omnexus. Material selection platform, abs glass transition temperature. [24] stratasys. Vero material properties. page 1, 2000. [25] Adamantia Chletsou, John Papapolymerou, Ahmet Cagri Ulusoy, and Ibrahim Kagan Ak- soyak. Uhf rectenna for implanted and free space communications. In 2019 IEEE Interna- tional Symposium on Antennas and Propagation and USNC-URSI Radio Science Meeting, pages 641–642. IEEE, 2019. [26] Ibrahim Kagan Aksoyak, Adamantia Chletsou, John Papapolymerou, and Ahmet Cagri Ulu- soy. A high sensitivity rf energy harvester for diverse environments. In 2020 50th European Microwave Conference (EuMC), pages 444–447. IEEE, 2021. 106 [27] Adamantia Chletsou, Yuxiao He, John F Locke, and John Papapolymerou. Multi-band, flex- ible, lightweight antenna on lcp for automotive applications. In 2020 IEEE International Symposium on Antennas and Propagation and North American Radio Science Meeting, pages 1507–1508. IEEE, 2020. [28] Adamantia Chletsou, John F Locke, and John Papapolymerou. Effects of plastic vehicular covers on radiation characteristics of lightweight, dual-band antenna for vehicular communi- cations. The Journal of Engineering, 2022(10):1017–1031, 2022. [29] Adamantia Chletsou, John F Locke, and John Papapolymerou. Vehicle platform effects on performance of flexible, lightweight, and dual-band antenna for vehicular communications. IEEE Journal of Microwaves, 2(1):123–133, 2021. [30] Adamantia Chletsou, John F Locke, and John Papapolymerou. Additive manufactured anten- nas on automotive abs using intense pulsed light sintering. In 2022 IEEE International Sym- posium on Antennas and Propagation and USNC-URSI Radio Science Meeting (AP-S/URSI), pages 1–2. IEEE, 2022. [31] Adamantia Chletsou, Alexandra Bannon, Xenophon Konstantinou, Lauren Reimnitz, John Locke, and John Papapolymerou. Additive manufactured cpw lines cured by intense pulse light for automotive microwave applications. In 2022 52nd European Microwave Conference (EuMC), pages 377–380. IEEE, 2022. [32] Anastasis C Polycarpou, Antonis Dimitriou, Aggelos Bletsas, Panayiotis C Polycarpou, Loizos Papaloizou, George Gregoriou, and John N Sahalos. On the design, installation, and evaluation of a radio-frequency identification system for healthcare applications [wireless corner]. IEEE Antennas and Propagation Magazine, 54(4):255–271, 2012. [33] Gianmario Scotti, Sue-Yuan Fan, Chun-Hsiang Liao, and Yi Chiu. Body-implantable rfid tags based on ormocer printed circuit board technology. IEEE Sensors Letters, 4(8):1–4, 2020. [34] Shubin Ma, Lauri Sydänheimo, Leena Ukkonen, and Toni Björninen. Split-ring resonator antenna system with cortical implant and head-worn parts for effective far-field implant com- munications. IEEE Antennas and Wireless Propagation Letters, 17(4):710–713, 2018. [35] Rula S Alrawashdeh, Yi Huang, Muayad Kod, and Aznida Abu Bakar Sajak. A broadband flexible implantable loop antenna with complementary split ring resonators. IEEE antennas and wireless propagation letters, 14:1506–1509, 2015. [36] Fu-Jhuan Huang, Chien-Ming Lee, Chia-Lin Chang, Liang-Kai Chen, Tzong-Chee Yo, and Ching-Hsing Luo. Rectenna application of miniaturized implantable antenna design for triple-band biotelemetry communication. IEEE Transactions on Antennas and Propagation, 59(7):2646–2653, 2011. [37] Ilka Dove. Analysis of radio propagation inside the human body for in-body localization purposes. Master’s thesis, University of Twente, 2014. 107 [38] Junji ITO, Hiroshi MORIYAMA, Seiichiro INOKUCHI, and Noboru GOTO. Human lower limb muscles: an evaluation of weight and fiber size. Okajimas folia anatomica Japonica, 80(2.3):47–56, 2003. [39] Niels Koch. Vehicular spoiler antenna for high data rate wlan. In 2015 Loughborough An- tennas & Propagation Conference (LAPC), pages 1–4. IEEE, 2015. [40] Taimoor Abbas, Johan Karedal, and Fredrik Tufvesson. Measurement-based analysis: The effect of complementary antennas and diversity on vehicle-to-vehicle communication. IEEE Antennas and Wireless Propagation Letters, 12:309–312, 2013. [41] Suguru Imai, Kenji Taguchi, Tatsuya Kashiwa, and Takeshi Kawamura. Effects of car body on radiation pattern of car antenna mounted on side mirror for inter-vehicle communications. In 2014 IEEE Antennas and Propagation Society International Symposium (APSURSI), pages 601–602. IEEE, 2014. [42] Leonidas Marantis, Anastasios Paraskevopoulos, Dimitrios Rongas, Athanasios Kanatas, Christos Oikonomopoulos-Zachos, and Sebastian Voell. A printed monopole espar antenna for truck-to-truck communications. In 2017 International Workshop on Antenna Technol- ogy: Small Antennas, Innovative Structures, and Applications (iWAT), pages 239–242. IEEE, 2017. [43] Syed N Hasnain, Ralf Stephan, Marius Brachvogely, Michael Meureryz, and Matthias A Hein. Interference mitigation for robust automotive satellite navigation achieved with com- pact distributed antenna sub-arrays. In 2020 14th European Conference on Antennas and Propagation (EuCAP), pages 1–5. IEEE, 2020. [44] Sateesh Virothu and M Satya Anuradha. Flexible circularly polarized antenna for lte-v2x/5g vehicular safety applications. In 2021 International Conference on Computing, Communica- tion, and Intelligent Systems (ICCCIS), pages 814–821. IEEE, 2021. [45] Muhammad Ehtisham Asghar, Frank Wollenschläger, Aidin Asgharzadeh, Jasmeet Singh, and Matthias A Hein. Influence of antenna mounting location on the radiation pattern of an automotive antenna. In 12th European Conference on Antennas and Propagation (EuCAP 2018), pages 1–5. IET, 2018. [46] Jasmeet Singh, Andreas Neumann, Thomas Wack, Tristan Koppe, Ralf Stephan, and Matthias A Hein. Novel conformai automotive di-patch antenna verified through car door frame measurements. In Antennas and Propagation Conference 2019 (APC-2019), pages 1–5. IET, 2019. [47] S Senega, J Kammerer, and S Lindenmeier. Scan-phase antenna diversity for digital satellite radio (sdars) in a single automotive side mirror. In The 8th European Conference on Antennas and Propagation (EuCAP 2014), pages 3255–3259. IEEE, 2014. [48] Dieter Pototzki and Andreas Griesche. Exploring radiating and scattering sources of part of vehicles by means of hemi-spherical-near-field antenna measurements. In 2012 6th European Conference on Antennas and Propagation (EUCAP), pages 2249–2251. IEEE, 2012. 108 [49] Miguel Bueno Diez, Wilfrid Pascher, and Stefan Lindenmeier. Electromagnetic characteri- zation of automotive sunroofs for car-to-x applications. In 2016 46th European Microwave Conference (EuMC), pages 1323–1326. IEEE, 2016. [50] Sertan Hastürkoğlu, Mahmoud Almarashli, and Stefan Lindenmeier. A compact wideband terrestial mimo-antenna set for 4g, 5g, wlan and v2x and evaluation of its lte-performance in an urban region. In 2019 13th European Conference on Antennas and Propagation (EUCAP), pages 1–5. IEEE, 2019. [51] Hicham Tazi, Christoph Ullrich, and Thomas F Eibert. Virtual development of automotive antennas printed on glass using a hybrid of method of moments and method of auxiliary sources. In Asia-Pacific Microwave Conference 2011, pages 614–617. IEEE, 2011. [52] Mahmoud Almarashli and Stefan Lindenmeier. A new method for evaluation of lte mimo antennas in automotive application. In 2017 11th European Conference on Antennas and Propagation (EUCAP), pages 2450–2453. IEEE, 2017. [53] Rogers. Datasheet of rogers3850 dielectric. pages 1–4, 2003. [54] Daniel N Aloi and Ehab Abdul-Rahman. Modeling of a far-field automotive antenna range using computational electromagnetic tools. In 2020 14th European Conference on Antennas and Propagation (EuCAP), pages 1–5. IEEE, 2020. [55] Huang Jingjian, Zhang Xiaofa, Xie Shaoyi, Wu Weiwei, and Yuan Naichang. Suppression of cross-polarization of the microstrip integrated balun-fed printed dipole antenna. International Journal of Antennas and Propagation, 2014, 2014. [56] Tiiti Kellomaki, Jouko Heikkinen, and Markku Kivikoski. Effects of bending gps antennas. In 2006 Asia-Pacific Microwave Conference, pages 1597–1600. IEEE, 2006. [57] Pekka Salonen, Mikko Keskilammi, and Yahya Rahmat-Samii. Textile antennas: Effect of antenna bending on radiation pattern and efficiency. In 2008 IEEE Antennas and Propagation Society International Symposium, pages 1–4. IEEE, 2008. [58] Luigi Vallozzi and Hendrik Rogier. Effects of bending on the radiation characteristics of a textile patch antenna. In 8th UGent PhD symposium, 2007. [59] Jorge L Salazar, Nafati Aboserwal, José D Díaz, Javier A Ortiz, and Caleb Fulton. Edge diffractions impact on the cross polarization performance of active phased array antennas. In 2016 IEEE International Symposium on Phased Array Systems and Technology (PAST), pages 1–5. IEEE, 2016. [60] AB Smolders, RMC Mestrom, ACF Reniers, and A Zamanifekri. Effect of a finite ground plane on the axial ratio of circularly-polarized microstrip arrays. In 2013 European Mi- crowave Conference, pages 203–206. IEEE, 2013. [61] ResourceInnovations. Automotive plastics. page 1, 2000. 109 [62] Simone Neermann, Michael Steindl, Joerg Franke, Edgar Mayer, and Michael Schmidt. Qual- ification of nir, uv and laser irradiation as alternative photonic sintering methods for printed electronics. In 2020 Pan Pacific Microelectronics Symposium (Pan Pacific), pages 1–8. IEEE, 2020. [63] Juha Niittynen, Enrico Sowade, Hyunkyoo Kang, Reinhard R Baumann, and Matti Män- tysalo. Comparison of laser and intense pulsed light sintering (ipl) for inkjet-printed copper nanoparticle layers. Scientific reports, 5(1):1–10, 2015. [64] Dana Weise, Kalyan Yoti Mitra, Enrico Sowade, and Reinhard R Baumann. Intense pulsed light sintering of inkjet printed silver nanoparticle ink: Influence of flashing parameters and substrate. MRS Online Proceedings Library, 1761(1):1–6, 2015. [65] Sam LeBlanc, Kenneth Church, and Eduardo A Rojas-Nastrucci. Photonic curing of mm- wave coplanar waveguides for conductor loss enhancement. In 2021 IEEE 21st Annual Wire- less and Microwave Technology Conference (WAMICON), pages 1–3. IEEE, 2021. [66] S Hawasli, H Tsang, N Lazarus, G Smith, and E Forsythe. Improving conductivity of 3d printed conductive pastes for rf & high performance electronics. In 2018 IEEE MTT-S Inter- national Microwave Workshop Series on Advanced Materials and Processes for RF and THz Applications (IMWS-AMP), pages 1–3. IEEE, 2018. [67] Omer F Firat, Mohamed M Abdin, Jing Wang, and Thomas M Weller. Low-loss suspended crossover interconnects using laser enhanced direct print additive manufacturing. In 2019 IEEE 20th Wireless and Microwave Technology Conference (WAMICON), pages 1–4. IEEE, 2019. [68] Y He, M Becker, T Grotjohn, A Hardy, M Muehle, T Schuelke, and J Papapolymerou. Rf characterization of coplanar waveguide (cpw) transmission lines on single-crystalline dia- mond platform for integrated high power rf electronic systems. In 2017 IEEE MTT-S Inter- national Microwave Symposium (IMS), pages 517–520. IEEE, 2017. [69] Anthony Delage, Nicolas Delhote, Serge Verdeyme, Barbara Bonnet, Ludovic Carpentier, Cindy Schick, Thierry Chartier, and Christophe Chaput. Aerosol jet printing of millimeter wave transmission lines on 3d ceramic substrates made by additive manufacturing. In 2018 IEEE/MTT-S International Microwave Symposium-IMS, pages 1557–1560. IEEE, 2018. [70] Aerosol jet printer, 2015. [71] The evolution of photonic curing, 2015. [72] Anis Jouini, Salem Hasnaoui, and Adnane Cherif. Comparative study of patch antennas for c-v2x communication. In 2022 5th International Conference on Advanced Systems and Emergent Technologies (IC_ASET), pages 61–66. IEEE, 2022. [73] Yi Zhou, Ge Zhao, Yun Jing Zhang, and Mei Song Tong. An omnidirectional vertical- polarized c-v2x antenna with high gain and low profile. In 2021 IEEE International Sym- posium on Antennas and Propagation and USNC-URSI Radio Science Meeting (APS/URSI), pages 1956–1957. IEEE, 2021. 110 [74] Rezvan Rafiee Alavi, Ahmed Harb, and Daniel N Aloi. Comparison of nf and ff automotive antenna measurement in dsrc and c-v2x applications. In 2022 IEEE International Symposium on Antennas and Propagation and USNC-URSI Radio Science Meeting (AP-S/URSI), pages 780–781. IEEE, 2022. [75] Haoran Shen and Chi-Chih Chen. A flush mount automobile 4-channel universal antenna design. In 2022 IEEE International Symposium on Antennas and Propagation and USNC- URSI Radio Science Meeting (AP-S/URSI), pages 475–476. IEEE, 2022. [76] Steven R Best. The radiation properties of electrically small folded spherical helix antennas. IEEE Transactions on antennas and propagation, 52(4):953–960, 2004. [77] Richard L Dumene, Paul Kennedy, Christopher B Williams, Dennis Sweeney, and Greg Earle. Creating embedded radiofrequency structures using polyjet material jetting. In 2015 Interna- tional Solid Freeform Fabrication Symposium. University of Texas at Austin, 2015. [78] veroProperties. Polyjet resin. page 1. [79] Yonghui Yan, Qingjiang Nan, Chunsheng Liu, Yan Zhang, and Ji Li. 3d-printed hemispheri- cal helix gps antenna with stable phase center. In 2019 International Applied Computational Electromagnetics Society Symposium-China (ACES), volume 1, pages 1–2. IEEE, 2019. [80] Myeongjun Kong, Su-Hyeon Lee, Geonyeong Shin, Junghyo Nah, and Ick-Jae Yoon. Inves- tigation of 3-d printed, electrically small, and thin magnetic dipole antenna. IEEE Antennas and Wireless Propagation Letters, 17(4):654–657, 2018. [81] Zhiwei Song, Hongxing Zheng, Yongjian Li, and Erping Li. Investigation of axial mode dielectric helical antenna. IEEE Transactions on Antennas and Propagation, 70(5):3806– 3811, 2021. [82] ElectronInks. El615 by electroninks. [83] NovaCentrix. Js-a221ae silver ink. page 1. [84] NovaCentrix. Pspi silver ink. page 1. [85] Optomec. Clariant silver ink. page 1. 111