3-D Printable Synthetic Metasurface to Realize 2-D Beam-Steering Antenna

This article presents highly radio-frequency (RF) transparent phase gradient synthetic metasurfaces made of sub-wavelength-sized 3D printable meta-atoms with tailored permittivity that cannot be achieved with off-the-shelf, commercially available materials. The synthesized meta-atoms design uses one dielectric block of PREPERM ® ABS 1000 material with air and metallic inclusions to make low- and high-permittivity materials. The inclusions’ size and height are varied to achieve a complete phase range from 0 to 360°, while maintaining transmission magnitudes greater than −3.0 dB. A two-dimensional array of meta-atoms forms a metasurface, which can be used for phase-shifting structures. Such metasurfaces can manipulate RF waves by introducing progressive phase delay into array elements. The proposed meta-atoms are employed to create highly RF transparent phase-gradient metal-dielectric composite metasurfaces (MDCMs) operating at 11 GHz. These MDCMs can be implemented through 3D printed technology using low-cost thermoplastics or polymers with composite filaments and minimal human intervention. A pair of MDCMs are combined with an array of microstrip patches to demonstrate 2D beam steering functionality numerically. The antenna system provides a peak directivity of 19.9 dBi with a maximum conical scanning angle of 114° and a directivity variation of less than 3 dB.


I. INTRODUCTION
C USTOMIZED electromagnetic (EM) response with two-dimensional (2D) metasurfaces has been a topic of considerable interest among the research community. Their ability to modify the phase of the propagating EM fields makes them suitable for beam-steering antennas desired for the next generation of wireless standards [1], [2], [3], [4]. However, the capability of the metasurfaces cannot be fully utilized in physical prototypes with commercial off-theshelf (COTS) materials that come with preset effective permittivity and permeability [5], [6]. Subsequently, the recent development on metasurfaces has seen an alternative approach to artificially achieving the equivalent characteristics through the inclusion of impurities in host materials that offer greater flexibility to incorporate metasurfaces in practical systems [7], [8].
The widely adopted strategy to realize metasurfaces is to first design their fundamental repeating blocks, referred to as "meta-atoms". A few challenges are evident in the fabrication of meta-atoms and, subsequently, metasurfaces, principally the difficulty in fabricating micro-and nanothree-dimensional structures with unrealistic overhangs using traditional machining techniques like lithography and nanoprinting that involve tedious machining and bonding steps, which in turn escalate the component cost [9], [10], [11]. Moreover, because of strong dispersion and high losses, traditional materials are inadequate to design beam-steering metasurfaces with high transmission magnitude [9], [12].
Additive manufacturing (AM), also known as 3D printing technology, is rapidly taking the place of traditional manufacturing because of its many advantages, including low cost, rapid prototyping, automated manufacturing, and repeatability [10], [13], [14], [15]. It can produce extremely customized metasurfaces for state-of-the-art applications that can meet the stringent benchmarks of modern communication systems. Heterogeneous metamaterials that are otherwise difficult to fabricate using traditional micro-machining with modified geometries can be efficiently manufactured by the AM technique. Moreover, because bonding and machining are not involved, the AM techniques make design inherently robust for even complex environments [16].
In this article, we investigate and present highly transparent, low-cost 3D printable meta-atoms to create metasurfaces that can achieve continuous permittivity within a fixed permittivity range. Two separate techniques are presented to achieve low-and high-permittivity meta-atoms by embedding air and metal inclusions in a single dielectric. The effective parameters (permittivity and permeability) of the proposed meta-atoms are calculated using the S-parameter retrieval method. These meta-atoms can be synthesized for any permittivity value between 2.6 and 21.4 using PREPERM R ABS 1000 filament as a base material, which has a base permittivity of 10 [17]. It is to be noted here that for phase-transforming metasurfaces, the meta-atoms must be able to achieve complete 2π phase shifts, which are achieved here by varying the heights of the two meta-atoms.
The meta-atoms are arranged with gradient phase delay to form a 3D printable metal-dielectric composite metasurface (MDCM) for steering the beam of a patch array antenna. With a pair of MDCMs, the broadside beam of a patch array is steered in a conical region having an apex angle of 114 • by the mutual physical rotation of two metasurfaces following the concept of Risley prism [18], [19]. The proposed MDCMs are superior in characteristics compared to previously reported all-dielectric and all-metal near-field phase transforming structures [20], [21], [22], [23], [24], which have their respective limitations pertaining to excessive height, ohmic losses, and fabrication complexity. The maximum height of MDCM is 9.8 mm (≈ 0.36λ 0 , where λ 0 is the wavelength at the operating frequency of 11 GHz), which is 88.7% shorter than the all-dielectric structure reported in [20], which would have a maximum height of 31.9 mm (3.19λ 0 ). The maximum lateral dimension of metal inclusion in each periodic meta-atom is limited to 3.4 mm, which is 43% of the meta-atom's total area, to keep the ohmic losses less than −3.0 dB. The phase-transforming structure reported in [21] needs to be constructed using at least four separate 3D filaments and very accurate micro-milling techniques. Moreover, our proposed MDCM is thinner, polarization-insensitive, and has a broader range of permittivity values as opposed to the 2.4λ c thick and polarization-dependent gradient index lens with limited permittivity values demonstrated in [25] for beam-steering application. In contrast, conventional thin-printed metasurfaces need more than one metal layer (at least three) to achieve a similar level of performance in terms of transmission magnitude, and phase range [3]. The proposed 3D-printed design approach is free from that and also evades multi-layer air-metal structures [22], [23], [24], [26] to attain a higher transmission magnitude.
The rest of the paper is organized as follows. Section II describes the mechanism that is employed to steer the beam in both azimuth and elevation using a pair of MDCMs. The S-parameter retrieval method and design of low-and highpermittivity materials that form the basis of MDCM are explained in Section III. The design of MDCM is explained in Section IV. Results are discussed in Section V, and a conclusion is given in Section VI.

II. BEAM-STEERING MECHANISM
The configuration of the beam-steering antenna system is shown in Fig. 1. It has a fixed-beam static base antenna at the bottom with a large radiating aperture and a pair of independently rotatable MDCMs atop; each MDCM is separated by a fraction of the free-space wavelength (4 mm or 0.15λ 0 ). The closely packed array of patches radiates electric fields with uniform phase distribution in the near-field region, while the MDCMs, which fully cover the antenna aperture, transform the phase of the propagating electric fields. In the classical configuration, MDCMs are designed such that they change the direction of radiated field propagation from broadside to an offset angle δ by introducing a gradient in the phase of the electric fields. Thus, when used separately with the base antenna, each MDCM tilts the antenna beam by an angle of δ. If a single MDCM is placed above the base antenna and rotated, its tilted beam with a fixed elevation angle can be moved to any azimuth angle between 0 • and 360 • . The combination of two MDCMs gives freedom to steer the beam in both elevation and azimuth planes in a conical region having an apex angle of at least 2δ. Such beam-steering mechanisms also known as meta-steering antenna systems, are described in detail in [3], [21], [23], [27], [28]. The steering aspects are, however, briefly explained below with analytical and numerical examples to assist in better understanding.
The beam direction for any rotational combination of MDCMs can be predicted with an analytical method using the design parameters of the MDCMs and antenna system. The beam tilt angles of the MDCM-1 and MDCM-2 are represented by δ 1 and δ 2 , while their respective rotation angles measured in an anticlockwise direction from the x-axis are represented by α 1 and α 2 [27]. The tilt angle of the MDCM is related to the phase delay gradient (p i ) introduced in the phase of the electric field while propagating through the metasurface as follows: where δ i is the tilt angle of the surface and k 0 = 2π/λ 0 is the wavenumber in free space, and λ 0 is free space-wavelength at the operating frequency. The wave propagating in the direction of the antenna beam has two projected wavenumbers: k x and k y in the direction of unit vectorsx andŷ, respectively. The wavenumbers are related to phase-delay gradients, and physical rotation angles of the MDCMs as follows [27].
The resultant azimuth (φ) and elevation angles (θ ) can be calculated using the wavenumbers, given by The equations can be further simplified by substituting (2a) and (2b) in (3a) and (3b).
Using (4) and (5), the elevation and azimuth angles of the output beam can be analytically estimated by appropriately choosing the rotation angles of the two MDCMs while the base antenna is absolutely fixed.  Table 1.
In a special case, when α 1 = α 2 = 0 • and k y = 0, the resultant beam is scanned to the farthest elevation angle, i.e., θ = θ max = 61.16 • and at the azimuth angle of φ = 0 • . With the counter-clockwise rotation of MDCM-2, both projection components change the beam location, and the azimuth angle follows a sinusoidal path. When α 1 = 0 • and α 2 = 180 • , both projection components are nullified, and the beam points in the broadside direction. As a demonstration, beam positions for two different cases where α 2 = 60 • and α 2 = 120 • are shown in Fig. 2. The following section explains the design details of meta-atoms with air, dielectric, and metal composites, which can be arranged with gradient phase delay to develop MDCMs for beam-steering applications.

III. META-ATOM DESIGN AND CHARACTERIZATION
Meta-atoms are the basic building blocks of MDCMs, and they must have a continuous permittivity between 1 and r (max). This is accomplished through the use of two distinct meta-atoms made of dielectric material with an intermediate base permittivity, b . The first meta-atom, referred to as M1 hereafter, covers the lower permittivity range from 1 to b , while M2, the second meta-atom, covers the permittivity range from b to r(max) .
The effective permittivity of meta-atoms, particularly at higher frequencies, can be derived either by the analytical Drude-Lorentz model [29] or by the scattering parameter (S-parameter) retrieval method [30], [31], [32]. For complex, inhomogeneous meta-atoms with multiple material composites, the former is less accurate and unreliable. Contrary to this, the S-parameter retrieval method is more accurate, as it extracts reflections and transmissions through full-wave simulations of actual meta-atoms. This method is used here, where the S-parameters of the meta-atoms are retrieved with two independent commercial EM simulators: the finite integration technique (FIT)-based CST Microwave Studio and the finite element method (FEM)-based high-frequency structure simulator (HFSS).
The meta-atoms are simulated with meta-atom boundary conditions to extract the S-parameter. The meta-atom conditions are realized using periodic boundaries and masterslave boundaries in two lateral directions in CST and HFSS, respectively [33]. They are excited by the plane wave propagating in the longitudinal direction along the +z-axis. The pictorial representation of two meta-atom models is shown in Fig. 3 (a) and (b), respectively. The extracted S-parameters are then used to estimate the effective permittivity and permeability of the meta-atoms over a wider frequency band via numerical methods [33]. First, the effective refractive index (n) and intrinsic impedance (z) are calculated from retrieved S-parameters using (6) and (7).
where, X = 1/2S 21 (1 − S 2 11 + S 2 21 ). Since the metamaterial under consideration is a passive medium, the signs in (6) and (7) are determined based on the following conditions [31]: where (.) and (.) denote the real part and imaginary part of operators, respectively. The value of refractive index n can be determined from (7) as follows: Since we have considered the case of a homogeneous slab for which the boundaries of the slab are well defined along the direction of propagation of a plane wave, the S-parameters are accurately known. So, effective parameters of the slab can be calculated directly from n and z using (10) and (11) [31].
The method is verified with a meta-atom made out of four different dielectrics with known permittivities: Rogers TMM10, Rogers RT5880, rubber, and glass. The permittivity values for four cases are estimated through the S-parameter retrieval method and compared with actual values. The metaatom used for analysis in both simulation environments is a square block with length w = 7.8 mm (λ 0 /3.5) and thickness (t) of 2.5 mm. The estimated effective permittivity and permeability of meta-atoms in the frequency range from 8 to 12 GHz are plotted in Fig. 4. The method is extended to estimate the effective characteristics of M1 and M2, which are explained in the following two subsections.

A. META-ATOM I (M1) WITH AIR INCLUSIONS
The meta-atom, M1 is designed to attain permittivity values between 1 and b . This is achieved using a dielectric material with an air inclusion, as shown in Fig. 5 (a). The width of the VOLUME 4, 2023 509 Authorized licensed use limited to the terms of the applicable license agreement with IEEE. Restrictions apply.  dielectric is w, and that of air inclusion is b. The 3D printable dielectric material used with M1 is PREPERM R ABS 1000, which has a base permittivity ( b ) of 10. The effective permittivity of the meta-atom is reduced below 10 by increasing the size of the air inclusion [34]. It is noted that if traditional machining methods are used to introduce air inclusions, the drill bits available are of discrete sizes, and hence discrete effective permittivity values can only be achieved. To achieve continuous effective permittivity between 2.6 and 10, additive manufacturing can be used instead, which can provide accuracy at the sub-millimeter level to realize M1 with desired permittivity values [8]. Contemporary 3D printing tools equip nozzles with diameters of as little as 0.3 -0.4 mm and printers that can print the layer height of as thin as 0.1 mm. So they offer huge customization capabilities for the design of non-homogeneous materials. Furthermore, square perforations instead of conventional cylindrical perforations are easy to realize using 3D printing [34], as square drill bits are nonexistent. Given the most common printing nozzle size of 0.4 mm, the perforation size in M1 can be adjusted in steps of 0.4 mm between 0 mm and 7.2 mm. S-parameters were computed for each perforation size with full-wave simulations, which are plotted in Fig. 6. As the perforation's size increases, the dielectric volume decreases, resulting in less dielectric losses. Thus, the insertion loss is minimal, which significantly enhances the transmission. The effective permittivity was calculated using (10) and is plotted in Fig. 7. The calculated permittivity values for M1 are compared with the theoretically expected values estimated by the volumetric average of the dielectric and air in the meta-atom structure [34]. The two curves tracked each other closely, and the M1 was able to achieve continuous permittivity values from 2.6 to 10.

B. META-ATOM II (M2) WITH METAL INCLUSIONS
The configuration of M2 is shown in Fig. 5 (b), where a metal block of size d is printed into the host dielectric. The thickness (or height) of the metal block is chosen to be the same as that of the dielectric. The M2 has a continuous effective permittivity between b and r (max). Traditionally, high permittivity artificial materials are engineered by mounting multiple metal and dielectric layers, which are unrealistic to realize using 3D printing as no class of printers can handle both metals and dielectrics simultaneously [8]. The rationale for using a solid metal block in M2 instead of the metal and dielectric layers conventionally used is that such an arrangement can be realized at a low cost with additive manufacturing.
The boundaries of the proposed meta-atom are well defined in the direction of wave propagation (z-axis), so (10) and (11) can be applied directly without sign ambiguities to calculate the effective parameters. This contrasts with the split ring resonator (SRR) meta-atom, which has multiple metal and dielectric layers in the direction of incident wave propagation, resulting in sign ambiguity during the calculation of its effective parameters [31].
In M2, the size of the metal inclusion controls effective permittivity and insertion loss. If the size of the metal inclusion is comparable to the local periodicity of the meta-atom, the reflections are too high to be useful for transmitting  metasurfaces. The M2 was simulated using periodic boundary conditions similar to the M1, with metal inclusion sizes ranging from 0 to 4.0 mm. The S-parameters were recorded at the operating frequency of 11 GHz and plotted in Fig. 8. It is observed that the RF transparency decreases when the size of the metal inclusions increases as a result of the reflections off the metal meta-atoms. The effective permittivity of the meta-atoms is plotted in Fig. 9. As can be seen in the figure, the M2 achieves continuous permittivity values between 10 and 21.4. Because it renders meta-atoms extremely reflective and ideal for transparent metasurfaces, it is important to note that the maximum value of the metal inclusion is limited to 4 mm and not increased beyond that [8], [12]. Using both of the aforementioned methods, we can engineer low as well as high permittivity materials and have proposed a mechanism to devise values from 2.6 to 21.4 using the base material's permittivity of 10. In the next section, the idea of meta-atom design has been extended to develop highly transmitting MDCMs for beam steering applications.

IV. MDCM DESIGN STRATEGY
Instead of the much more time-consuming and computationally intensive process of first simulating all possible combinations of the meta-atoms by varying their physical geometric parameters [23], [35], we used a relatively straightforward method to design MDCM that relied on the effective characteristics of the meta-atoms. In order to do that, a reference database of transmission phases is taken into consideration for all-dielectric unit elements. Then the metaatoms with equivalent phases can be substituted to develop the proposed MDCM. As a result, the first step is to create a reference database that covers the entire phase range from 0 to 360 • , and can be used to compare the phase of proposed meta-atoms. For this purpose, a transmission phase (∠S 21 ) database for ideal dielectrics was created. The overall procedures are further explained in the following subsections.

A. PHASE DELAY OF 2π FROM IDEAL DIELECTRIC
In our design strategy, we first use transmission line theory to predict transmission through ideal dielectric materials. To explain this further, consider an electric field propagation along a direction (+z-axis here). The transmission phase and magnitude between any two planes separated by a distance t in the direction of propagation can be analytically calculated using (12), and (13) derived via transmission line theory [36]. If the space between two virtual planes is occupied by a homogeneous dielectric of length h and air of length (t-h), the total phase delay between these two planes is given by, where, τ 1 and τ 2 are intrinsic reflection coefficients of two mediums (air and dielectric), ρ 1 and ρ 2 intrinsic transmission coefficients of mediums. Further details on this phase delay between two imaginary planes have been reported in previous publications [36], [37] and are not repeated here for brevity. Using (13), we predicted the phase and magnitude of the propagating electric field for various combinations of dielectric materials and their height, h. It is also considered that the spacing between two planes is 9.8 mm (or t = 9.8 mm). The height was varied from 1.4 mm to 9.8 mm while the permittivity was varied between 2.6 and 21.4. The maximum height value is set to a value that ensures a complete 2π phase shift for the given range of permittivity values. The magnitude and phase of the electric field are indicated in a color map plot against the two variables (permittivity and height of dielectric) in Fig. 10 (a) and (b). These two plots of ideal dielectrics are used as a reference database to select the appropriate heights and permittivity of meta-atoms (either M1 or M2) for attaining desired phase-shift through M1 and M2.
The selection of meta-atoms in a metasurface is guided by the desired tilt angle and progressive phase-delay gradient, as discussed in Section II. Considering the periodicity of meta-atom (w = 7.8 mm) and the desired tilt angle of each metasurface (δ = 26 • ), the progressive phase delay of 45 • is required between adjacent meta-atoms [38], [39]. From Fig. 10, we picked eight unique elements with a progressive phase delay of 45 • . The geometric parameters (height and permittivity) of all eight elements of the all-dielectric surface are given in Table 2. It is to be mentioned here that while choosing elements with a progressive phase shift of 45 • , there can be more than one element with different geometric parameters but identical phase shifts. In such cases, our selection is based on using the element having the least insertion loss while achieving the closest to the desired phase gradient. As an example, from Fig. 10, the elements with respective permittivities 5.7 and 8, and heights 9.8 and 7.6 return identical normalized phase of 180 • . However, the loss of the former is −3.88 dB compared to that of −2.007 of the latter. Hence, the meta-atom with the lowest losses was considered. The eight ideal dielectric elements selected for the design of the beam-manipulating surface are not commercially available, so meta-atoms that have equivalent transmission and reflection characteristics to those of these eight dielectrics need to be artificially engineered.

B. META-ATOMS REPLACE IDEAL DIELECTRICS
The prime reason why metasurfaces with artificial dielectrics are superior to all-dielectric surfaces is that they have the ability to offer permittivity and phase values that cannot be offered by natural materials within the limited physical parameters. The incorporation of low-loss artificial dielectrics helps create highly transmitting metasurfaces within the limited permittivity and height range. The next step is to replace meta-atoms of equivalent transmission phases with those of specific all-dielectric unit elements to design MDCM with a high transmission magnitude while maintaining a low height profile. To convert all-dielectric elements to meta-atoms in the MDCMs, the height of all eight meta-atoms will remain the same as that of ideal dielectrics, but the effective permittivity will be achieved by replacing the ideal dielectric blocks with the meta-atoms with air or metal inclusion. If the permittivity value of a particular ideal dielectric element is lower than 10, then the corresponding meta-atom will have air inclusion, and if it exceeds 10 then a meta-atom with metal inclusion will be introduced.
This can be explained with one example. From Table 2, the first meta-atom of the ideal dielectric has a height of 1.4 mm and permittivity 2.6. From Fig. 7, it can be seen that by introducing perforation of 6.8 mm, the same effective permittivity can be achieved. So it can be assumed that an ideal dielectric unit element with permittivity of 2.6 and height of 1.4 mm will have the same reflection and transmission parameters as that of artificial dielectric with host permittivity of 10 and 6.8 mm cubical air inclusion. It was verified by simulating both meta-atoms in CST Microwave Studio. Likewise, other dielectric elements in Table 2 were replaced by meta-atoms. The corresponding heights for both types of meta-atom were the same and the effective permittivity of the ideal dielectric meta-atom was achieved by finding the right inclusion size from Fig. 7 and Fig. 9. This design procedure, thus, did not involve any parametric sweep and hence synthetic metasurface with artificial materials can easily be coined using this technique.
Another way to design MDCM with greater flexibility is to select all height and permittivity values from the all-dielectric database that have identical transmission phases with the progressive phase delay of 45 • . As a result, multiple meta-atoms will have identical ∠S 21 values, and one with the highest transmission can be chosen to incorporate in MDCM. For example, consider three meta-atoms with respective heights of 7.8 mm, 8.8 mm, and 9.6 mm, and perforation sizes of  This is critical for AM process as different models of 3D printers have variable design dimension constraints, and a set of meta-atoms with desirable inclusion sizes can be chosen according to the printer's limitation for MDCM design. Moreover, the freedom to choose from multiple metaatoms with identical transmission phases allows selecting the desired number of meta-atoms with air or metal inclusions because phase overlap exists for both inclusion types. Meta-atoms with air inclusions offer a normalized phase range between 0 • to 258.58 • and those with metal inclusions have a normalized phase range from 77.1 • to 331.7 • . Hence, the phase range of 77.1 • to 258.58 • overlaps, and either meta-atom (with air or metal inclusion) can be selected for MDCM design. This has been depicted as transmission phase vs. insertion loss in Fig. 11 (a) and (b) in the form of a database of precomputed values.

C. ARRANGEMENT OF META-ATOMS
The chosen eight meta-atoms are then strategically arranged with a phase delay gradient to form MDCM, which has a constant phase in the y-axis and a progressive phase in the x-axis. It implies that meta-atoms will be repeated in the y-axis and will have increasing phase delay in the x-axis. A total of twenty-one meta-atoms were arrayed to form the metasurface of size 163.8 mm (6λ 0 ). In this way, low-cost 3D printable synthetic MDCM can be realized using filaments with thermoplastics, polymers, metallic, and ceramic inclusions, as shown in Fig. 12. Metal composites can be printed using Markforged printers, and dielectrics can be printed using polymer printers.

V. RESULTS AND DISCUSSION
To demonstrate the efficacy of the MDCM pair, a beamsteering antenna is developed to steer the broadside beam of a patch array antenna. The overall antenna configuration is shown in Fig. 13. A classical patch array of diameter 163.8 mm (6λ 0 ) is used as a primary feed. The array comprises 90 patches, each having lateral dimensions of 7.96mm×6.96mm. The basic configuration of the patch array with MDCM pair is shown in Fig. 13. The array has an extended ground plane and Rogers Ultralam 2000 ( r = 2.5) substrate of thickness 1.59 mm. The MDCM pair is suspended at 13.8 mm (≈ 0.5λ 0 ) above the feed, and the subwavelength gap between the two MDCMs is 4 mm. It is important to note that such air spacings affect the overall antenna beam-steering performance and should be optimized at the design level. A detailed discussion is already given in the reported article [23] and is not repeated here for brevity. The relative angular positions of MDCMs can dynamically define the beam directions over a large conical space in the upper hemisphere. Controls of azimuth and elevation angles for two different beam steering configurations, namely elevation and azimuth scanning (Case I) and elevation scanning only (Case II), have been demonstrated in the following sub-sections. These configurations are simulated with the full-wave transient solver in CST Microwave Studio. The predicted results are then compared with the analytical results to validate the beam steering method using the proposed synthetic MDCMs.

A. CASE I: SIMULTANEOUS BEAM-STEERING IN THE ELEVATION AND AZIMUTH PLANES
The first configuration demonstrates simultaneous variation in azimuth and elevation angle, for which, MDCM-1 is fixed, and MDCM-2 is rotated in the counter-clockwise direction in steps of 30 • , i.e., α 1 = 0 • and α 2 = 30 • , 60 • , . . . , 180 • . This results in six different orientations of MDCM-2, and the beam peak has a distinct azimuth and elevation angle for each orientation. The radiation patterns for all six orientations are predicted through a full-wave simulation environment. All the elevation cuts, taken at azimuth angles, containing beam peaks are plotted in Fig. 14.
The elevation angle is maximum, i.e., θ = θ max = 55 • when α 1 = 0 • and α 2 = 30 • . With the stepped increase in α 2 from 30 • to 180 • , the beam simultaneously moves in the azimuth and elevation planes and ends up in a broadside direction. A comparison between theoretically estimated and predicted values through full-wave EM simulations of the elevation angles of the beam peaks is given in Table 3. The precise angular positions of the beam peaks are estimated analytically using (10) and (11). Whereas, for a special condition of δ 1 = δ 2 , the azimuth angle (φ) of the beam  peaks can be calculated using (14) [27].
The values of azimuth and elevation angles for all six orientations of the MDCM pair are given in the legend of Fig. 14. The elevation angle (θ ) is the beam peak's position on the figure's horizontal axis. The maximum peak directivity of 19.9 dB is achieved when the MDCM pair is oriented 180 • apart, i.e., α 1 = 0 • and α 2 = 180 • and the beam points in the broadside direction. The minimum peak directivity of 17 dBi is achieved when the beam is scanned furthest in the elevation for α 1 = 0 • and α 2 = 0 • . Such phenomena are very common in the near-field meta-steering antenna systems [3], [21], [23].

B. CASE II: BEAM-STEERING IN THE ELEVATION PLANE ONLY
The two MDCMs are rotated equally in opposite directions for sole elevation angle variation without affecting the azimuth angle. Similar to the prior instance, simulations for six alternative orientations of the MDCM pair are used to confirm the variance in elevation angles. MDCM-1 is rotated clockwise by 15 • from α 1 = 15 • to 90 • , whereas MDCM-2 is rotated anticlockwise by the same steps, from α 2 = −15 • to −90 • .
The elevation cuts for all orientations of the MDCM pair containing the beam peak (taken at φ = 0 • ) are plotted in Fig. 15. The beam peak is in the broadside direction when α 1 = 90 • and α 2 = −90 • and it gradually reduces with the increase in elevation angle. Like the previous case of configurations, the highest peak directivity of 18.9 dBi is exhibited in the broadside direction, and the lowest directivity of 16.1 dBi is observed for α 1 = 15 • and α 2 = −15 • when the beam points to the largest elevation angle of 57 • . The orientation of the MDCMs (relative angular positions) and a comparison between beam peaks estimated using (10) and (11) and predicted by the simulations are given in Table 5. They agree well; the difference is not more than 5 • . Table 5 compares the key features of the proposed antenna with state-of-the-art technologies to assess the MDCM's efficacy. Its footprint is smaller, and its design is straightforward. Table 5 demonstrates that the physical and electrical performance of the patch array with MDCM is comparable and promising. It has similar directivity/gain, better scanning capability and aperture efficiency. The proposed 3D printed technology is an alternative to the PCB technology with rapid prototyping, lower cost and little to no human intervention. In comparison to alternative approaches, such as an all-metal transmitarray [22] or a beam-steering antenna with a metal-only phase gradient metasurface [23], the antenna profile is also significantly reduced. Additionally, the MDCM only needs one filament to build, in contrast to the counter design's requirement for at least four different 3D filaments and extremely precise micro-milling methods [21].

VI. CONCLUSION
Low-cost metal-dielectric composite metasurfaces (MDCMs) have been presented for beam steering applications of a fixed-beam static base antenna. The MDCMs can be realized using filaments with thermoplastics, polymers, ceramics, and metallic composites, where phase transition through MDCMs is locally controlled by varying permittivity values from 2.6 to 21.4. The nearly continuous permittivity range is achieved with synthetic MDCM, which cannot be realized through metasurfaces developed using commercial off-the-shelf standard materials. Such MDCM can be fabricated in a single step, eliminating the tedious machining or bonding typically required in multilayered thin metasurfaces. It should be noted that the properties of 3D printed structures may differ from VOLUME 4, 2023 515 Authorized licensed use limited to the terms of the applicable license agreement with IEEE. Restrictions apply. Engineering, Loughborough University. He is a named Investigator on EPSRC grants totaling in excess of £14m. He has authored approximately 280 peer-reviewed journal and conference papers in topics related to electromagnetic materials, synthetic dielectrics, dielectric measurements, 3-D-printing, wearable antennas, VHF antennas, specific absorption rate, finite difference time domain, specific absorption rate, metamaterials, heterogeneous substrates, embroidered antennas, inkjet printing, electromagnetic compatibility, RFID tags, phantoms, and genetic algorithms.
Prof. Whittow was the Coordinating Chair of the Loughborough Antennas and Propagation Conference from 2007 to 2011. He has served as an Associate Editor for Electronics Letters (IET) and also Microwaves, Antennas and Propagation. He serves on the technical program committees for several IEEE international conferences. He has been asked to give more than 30 invited conference presentations; a four days invited workshop on bioelectromagnetics and teaches about dielectric measurements at the European School of Antennas. In 2017, he won the Women in Engineering Men as Allies Award and he is the inaugural male Associate Fellow of Women's Engineering Society. He is a Senior Fellow of the Higher Education Academy. More than 100 of his academic journal papers can be freely downloaded here: http://publications.lboro.ac.uk/publications/all/collated/elwgw.html.