Intake Design for an Atmosphere-Breathing Electric Propulsion System (ABEP)

An atmosphere-breathing electric propulsion system (ABEP), see Fig. 1, is composed of two main components: the intake and the electric thruster. The system is designed for satellites orbiting at very low altitude altitudes, for example in very low Earth orbit (VLEO), defined for altitudes $h<~$450\text{\,}\mathrm{km}$$ [1]. The ABEP system collects the residual atmospheric particles encountered by the satellite through the intake and uses them as propellant for the electric thruster. The system is theoretically applicable to any planetary body with an atmosphere, and can drastically reduce the on-board propellant storage requirement while extending the mission’s lifetime [2]. Many ABEP concepts have been investigated in the past based on radio frequency ion thrusters (RIT) [3, 4, 5], ECR-based thruster [6, 7, 8, 9, 10], Hall-effect thrusters (HET) [11, 12, 13, 14, 15, 16, 17], and plasma thrusters [18]. The only laboratory tested ABEP systems to date are the ABIE, composed of an annular intake and an ECR-based thruster into one device [6, 7, 8, 9], and the RAM-HET system, comprised of the intake and a HET assembled into one device [13, 14, 15, 16, 17].

The intake is the device of an ABEP system that collects and delivers the atmospheric particles to the electric thruster, the general design is shown in Fig. 2. The intake must ensure efficient collection of atmospheric particles and provide the required density, pressure, and mass flow for the thruster’s operation, while minimizing the required frontal area.

The number density is $n$ and $\dot{N}$ is the particle flux. A key parameter to evaluate the intake performance, is the collection efficiency $\eta_{c}(h)$ defined in Eq. 1, where $\dot{N}_{out}$ is the collected particle flux and $\dot{N}_{in}$ is the incoming one, the subscript $i$ refers to each different particle species $i=[1,N_{s}]$, where $N_{s}$ is the total number of species found in the inflow environment. As the VLEO composition is highly variable, the task to efficiently collect each species in a variable density range is challenging.

A further key parameter is the pressure at the intake chamber section $p_{ch}$, which is an input for the thruster. By applying the ideal gas assumption $p_{ch}$ is derived as in Eq. 2, where $n_{ch}$ and $T_{ch}$ are the particle number density and temperature at the chamber section, and $k_{b}=$1.380\,649\text{\times}{10}^{-23}\text{\,}\mathrm{J}\mathrm{/}{\mathrm{K}}$$ is the Boltzmann constant.

Finally, the output mass flow to the thruster $\dot{m}_{thr}$ is obtained as in Eq. 3, where $m_{p}$ is the particle mass.

Within this study, hexagonal and circular geometries, for the whole intake but also for the ducts are investigated. To differentiate between them, the aspect ratio ($AR$) is defined in Eq. 4 and used, where $S$ is the hexagonal duct short diagonal, $D=2R$ is the circle diameter, and $L$ the intake length.

Over the last decades, several research groups have approached the intake design for an ABEP system [20]. Most of those are based on diffuse scattering materials, and implement a front section of small ducts to reduce the backflow, acting as molecular trap, to achieve higher $\eta_{c}<0.5$. The intake designed by Nishiyama and JAXA [6, 7, 8, 9, 10], uses long ducts packed in a ring-shaped section surrounding the SC body and ending on a conical reflector at the back, $\eta_{c}<0.46$. There they are ionized by electron cyclotron resonance (ECR) and accelerated by grids to produce thrust. Busek Co. Inc. [11, 12] designed the MABHET spacecraft for Mars orbit equipped with an ABEP system based on HET technology. It has a long and slender cylindrical intake in front of the SC with a honeycomb duct section at the front to operate as a molecular trap delivering $\eta_{c}=0.2-0.4$. The intake designs developed at IRS [21, 22, 23] are based on a long slender cylindrical intake with a honeycomb duct section in the front optimized for both Earth and Mars atmosphere that can achieve $\eta_{c}=0.43$. The Central Aerohydrodynamic Institute TsAGI [24, 25, 26, 27, 28, 29, 30] developed a similar concept, but the honeycomb is changed for a squared duct section delivering $\eta_{c}=0.33-0.34$. SITAEL and the von Karman Institute for Fluid Dynamics, Aeronautics and Aerospace [13, 14, 15, 16, 17, 31], refined the intake design that started with the 2007 ESA RAM-EP study [3]. The intake features several coaxial cylinders connected by plates which form divisions of different AR between each cylinder delivering $\eta_{c}=0.28-0.32$. This has been tested in combination with a modified HET for ABEP operation and is the only ABEP system of intake and thruster in one device that has been tested for continuos operation, to date, in the laboratory [13, 14, 15, 16, 17, 31]. The Lanzhou Institute of Space Technology and Physics [32], designed an active intake with a two stage system, a passive multi-hole plate followed by an active compression with a turbo pump system that can achieve $\eta_{c}=0.42-0.58$. Finally, the University of Colorado [33] investigated specular scattering and finally designed an intake using the optical proprieties of a parabola leading to an estimated $\eta_{c}>0.9$. A summary of these intake studies and their main features and performances are shown in Tab. 1.

Finally, each intake design always needs to be tailored based on the thruster applied and its requirements and also on the mission profile: expected operational orbits (altitudes and inclinations) and duration (i.e. to account for solar activity variation).

Upon particle impact with a surface an exchange of momentum and energy takes place, and the velocity of the incoming molecule is altered by the specific local properties of the impacted surface. The Maxwell model is a simplified approach that uses two types of surface-particle interactions: specular and diffuse reflection. The schematics are shown in Fig. 3 in which the velocity vector of the particles $v$ is represented by the black arrows with a thinner tip, $\hat{n}$ and $\hat{t}$ are the normal and tangential unit vectors, and $\delta_{i}$ and $\delta_{r}$ the incident and re-emission angles.

In specular reflections, particles rebound when hitting the surface without energy exchange, the angle of reflection equals the incident angle $\delta_{r}=\delta_{i}$, and the tangential component of particle velocity $\mathbb{v}\hat{t}=v_{t}$ remains constant. Only the normal velocity component $\mathbb{v}\hat{n}=v_{n}$ is affected undergoing a complete inversion $v_{t}\rightarrow-v^{\prime}_{t}$. In diffuse reflection, instead, the velocity of reflected molecules $\mathbb{v}^{\prime}$ is independent of the incident one. Particles reach thermal equilibrium with the surface and are reflected according to a Rayleigh distribution corresponding to the wall temperature $T_{wall}$. The Maxwell model defines the accommodation coefficient $\sigma_{B}$, to represent the fraction between diffuse and specular reflection, and it is highly dependent on the surface material properties. The accommodation coefficient can be implemented in both the momentum and energy equations [39] and is defined in Eq. 5, where the energy of the incident particle is $E$, and $E^{\prime}$ the energy of the reflected particle, while $E_{W}$ is the energy of the particle which it would have at a full accommodation to $T_{wall}$.

For realistic free molecular flows (FMF) and most surface materials, a full accommodation $\sigma_{B}=1$, corresponding to diffuse reflection, provides good simulation results on surface interactions  [34, Ch. 5, pp. 118]. Instead, specular reflection is for $\sigma_{B}=0$.

The simulation domain of the intake is discretized using unstructured hexahedral cells. Due to the symmetry of the intake design, see Fig. 2, and for an incoming flow parallel to $z$, symmetry planes normal to $x$ and $y$ are used to reduce the mesh size and therefore optimize the simulation efficiency. Additionally, more optimized concepts are explored with two different mesh setups and simulation times to verify the results [45]. The ABEP VLEO altitude range of $h=150-$250\text{\,}\mathrm{km}$$ corresponds to the orbital velocity $v_{SC}\sim$7.8\text{\,}\mathrm{k}\mathrm{m}\mathrm{/}{\mathrm{s}}$$ [2]. The free stream conditions at VLEO are in the free molecular flow regime [23, 2], and the particle flow can be treated as hyperthermal for the selected $AR$ in this study [32, 23]. This is valid at low temperatures when the random thermal motion of the incoming particles is negligible compared to the SC velocity $v_{SC}$. Therefore, the incoming flow is collimated entering the intake with a single free stream bulk velocity $v_{in}(h)=v_{SC}(h)$ at a specific incoming angle $\alpha$. Assuming a circular orbit, $v_{SC}(h)$ is calculated as in Eq. 6, where $\mu_{E}=$3.986\text{\times}{10}^{15}\text{\,}\mathrm{k}\mathrm{m}\mathrm{/}{\mathrm{s}^{2}}$$ is the Earth’s gravitational parameter, and its average radius $R_{E}=$6371\text{\,}\mathrm{km}$$.

Finally, the temperature of the intake walls $T_{wall}$ is set to that of a SC in LEO $T_{wall}=$300\text{\,}\mathrm{K}$$ [21, 2]. Investigation based on variation of $T_{wall}$ in the intake has been performed previously, but have shown no significant influence on intake performance [23]. The flow parameters used as input in PICLas are shown in Tab. 2, and are extracted from the NRLMSISE-00 atmospheric model as specified in Sec. 3. Finally, all the species shown in Fig. 4, are accounted in the simulations with their respective $n_{i}$.

The intake concepts based on diffuse reflecting materials are represented by MomentumACC$=1.0$, while those based on specular reflecting materials by MomentumACC$=0.0$. In the hybrid concepts, each surface boundary condition (B.C.) is singularly defined as either diffuse or specular. As a final remark, all the intakes are based on an $A_{out}$ that is defined by the selected reference thruster [2].

The design improvement process is iterative. The starting case is set for an altitude $h=$150\text{\,}\mathrm{k}\mathrm{m}$$ at an incoming angle $\alpha=$0\text{\,}\mathrm{\SIUnitSymbolDegree}$$. Based on this different (intake) cases are designed, simulated, and evaluated for better understating the scattering dynamics inside the hexagonal ducts in terms of $\eta_{c}$. The first iteration affects the variation of overall intake size and chamber angles, with a fixed duct $AR$ based on previous studies [21, 22, 23]. Then, the designs providing $\eta_{c}\geq 0.3$ are further adapted to the hexagonal outer shape. In the second iteration step, different duct $AR$ are simulated, while keeping chamber angle, intake length, and intake diameter constant. Finally, the concepts with the highest $\eta_{c}$ are selected for further investigation and the respective collectible $\dot{m}_{thr}$ and pressure distribution of $p_{ch}$ are taken into account.

The three selected designs are named Diffuse-Nominal (D-N), Diffuse-Multiple Duct AR (D-MDAR), and Diffuse-Maximum Filling Ratio (D-MFR). The respective performance and settings are shown in Tab. 3. The D-N case is derived directly from the enhanced funnel design (EFD) [21, 22, 23], adapted to an hexagonal shape, and hereby optimized for the RF Helicon-based plasma thruster. This case is the baseline for all the upcoming concepts.

The D-MDAR design features multiple duct $AR$ to increase the effective intake area $A_{in}$ directly in front of the thruster’s discharge channel with larger $AR$ ducts, while smaller ducts are at the outer perimeter, as this region is expected to have larger backflow. The schematics displaying the frontal view of the D-MDAR is shown in Fig. 6.

The D-MFR case is designed to maximize $A_{in}$, while still taking advantage of the backflow mitigation of the duct section. The number of ducts on the intake is minimized and their $AR$ equal, finally reducing the cross-sectional area of the duct geometry while keeping the same wall thickness.

As listed in Tab. 3, the D-MDAR case provides the highest performance in terms of both $\dot{m}_{thr}$ and $\eta_{c}$. Although, the D-MFR also provides $\eta_{c}>0.4$ with only half frontal area $A_{Intake}$, this intake design provides a $10\%$ lower $\dot{m}_{thr}$ when compared with D-MDAR case. Different incoming flow angles $\alpha$, see Fig. 7, are simulated to assess D-MDAR performance variations at conditions in which the SC may not be able to maintain flow-oriented attitude, or due to the presence of strong not-aligned atmospheric winds as seen by GOCE [46].

In Tab. 4, the results of this analysis are presented, displaying the relative $\Delta\eta_{c}$ variation of $\eta_{c}$ vs $\alpha$.

The design improvement process is iterative. The base case is set for $h=$150\text{\,}\mathrm{km}$$ at an incoming angle $\alpha=$0\text{\,}\mathrm{\SIUnitSymbolDegree}$$. Each case is tailored to the given $A_{out}$ of the reference thruster [19], and has different geometric parameters which affect the focal point location of the parabola targeting the maximization of $\eta_{c}$. Furthermore, the cases providing the highest $\eta_{c}$ are the investigated aiming to maximizing $\dot{m}_{thr}$ and $p_{ch}$. Concerning the S-MS intake concept, different geometrical stage combinations are simulated, based on the improved purely specular designs, to maximize $\dot{m}_{thr}$ and $p_{ch}$ while maintaining $\eta_{c}>0.3$.

The highest performance cases for the intake based on specular reflecting materials are shown in Tab. 5. The naming of the cases refers to the focal point position of the parabola in respect to the thruster discharge channel, the schematics is shown in Fig. 10.

• 1.

  S-FE: focus at the boundary between the intake and discharge channel;

• 2.

  S-FT: focus located inside the discharge channel of the thruster;

• 3.

  S-FI: focus within the intake itself.

Finally, the S-FT-D case includes a frontal duct section to take advantage of both the parabolic optical properties, and the low backflow transmission probabilities of the ducts structure. The parabolic geometry is shared with the S-FT case, but, due to the duct section, the $A_{in}$ of the S-FT-D case is smaller.

Furthermore, the results for the highest performance cases of the S-MS concept is shown in Tab. 6. The S-MS-F and S-MS-B cases are both applied to the S-FT design described above. In the S-MS-F, $A_{Intake}\sim A_{in}$ is increased by adding a second stage at the front of the S-FT case parabola, while in the S-MS-B case, the second stage is added at the back of the S-FT parabola.

The designs providing the highest $\eta_{c}$, the S-FT and the S-FT-D, are further investigated to evaluate the performance sensitivity to flow misalignment $\alpha$. Based on the results of the diffuse duct simulations, it is expected that the duct geometry could prove useful on mitigating the effects of $\alpha>$0\text{\,}\mathrm{\SIUnitSymbolDegree}$$, as particles could be captured by the duct geometry and redirected to the parabolic section. The collection efficiency $\eta_{c}$ and the output mass flow $\dot{m}_{thr}$ at different $\alpha$ are shown in Fig. 11, where the S-FT case is examined in a closer manner between $$10\text{\,}\mathrm{\SIUnitSymbolDegree}$>\alpha>$15\text{\,}\mathrm{\SIUnitSymbolDegree}$$ to better understand the influence of $\alpha$ to the parabolic scattering dynamics.

The hybrid intake concepts are based on the already improved designs presented in Sec. 5, and Sec. 6. Therefore, the design improvement process is focused only on the overall intake performance effect due to different combinations of GSI properties.

At first, the geometry represented by the S-FT case is simulated with all of its surfaces that can have both diffuse and specular scattering. This is achieved by varying $\sigma_{B}$, therefore by changing the MomentumACC in the PICLas parameter file. Finally, it is assumed that for MomentumACC $=0.1$ there is a $10\%$ probability of diffuse reflection, thus, the surface is assumed to have $90\%$ specular surface properties.

The results of the hybrid intake simulations are shown in Tab. 7. The acronym SP indicates that the B.C. applied to the Intake or/and Chamber section is simulated with specular scattering while DI indicates diffuse scattering. The HI case refers to the hexagonal intake with a duct structure similar to the one in the baseline diffuse intake design of Fig. 5. The S-FT-D and S-FI share the same geometric parameters of their pure specular versions, but each B.C. is simulated according to Tab. 7.

Additionally, the S-FT case is simulated with different level $\%$ of material specular reflection over VLEO altitudes, the respective results for $\eta_{c}$ are shown in Fig. 13. Such simulation aims to resemble the effect of different GSI properties for the parabolic intake concept providing the highest $\eta_{c}$. In this set of simulations, each intake section shares the same B.C. GSI property.

The diffuse intake improvement resulted in two designs to be the most effective for propellant collection: the D-MDAR, and the D-MFR. Both provide the highest $\eta_{c}=0.437-0.458$ and $\dot{m}_{thr}=0.0212-$0.0240\text{\,}\mathrm{m}\mathrm{g}\mathrm{/}{\mathrm{s}}$$, while the second has a maximized $FR$ therefore requiring less $A_{Intake}$ which might be positive for reducing the SC frontal area, and therefore the drag. The addition of a frontal section of small ducts enhances $\eta_{c}$ by operating as molecular trap. The sensitivity analysis on the flow misalignment represented by the angle $\alpha$ shows that a $<50\%$ reduction of $\eta_{c}$ is achieved for $\alpha=$10\text{\,}\mathrm{\SIUnitSymbolDegree}$$, while this reduction is below $20\%$ for $\alpha<$5\text{\,}\mathrm{\SIUnitSymbolDegree}$$.

The specular intake improvement resulted in very efficient designs, especially the S-FT and the S-FT-D with a respective $\eta_{c}=0.943$ and $\eta_{c}=0.887$. Both are based on parabolic shapes with their focus located inside the thruster’s discharge channel. The S-FT-D includes a front section of hexagonal ducts to reduce the backflow. In particular, the S-FT-D design resulted more sensitive to the misalignment with the flow $\alpha$ which sees a steeper drop of $\dot{m}_{thr}$ and $\eta_{c}$, see Fig. 11. The S-MS designs provide high $\dot{m}_{thr}$ but lower $\eta_{c}<0.334$ compared to the S-FT and S-FT-D designs. For comparison, the $p_{ch}$ distribution of the S-FT case and the S-FE are shown in Fig. 16 and in Fig. 17. The different $p_{ch}$ distribution compared to diffuse-based intakes is due to the optical properties on which the particle collection is based. The S-FT, see Fig. 16, achieves $p_{ch}\sim$0.3\text{\,}\mathrm{Pa}$$, while the S-FE achieves a larger pressure $p_{ch}>$1\text{\,}\mathrm{Pa}$$ in the central region while also resulting in a greater larger-pressure region, see Fig. 17. However, it delivers lower $\eta_{c}=0.88$. It is also observed that the S-FT case provides an almost constant $\eta_{c}$ until $\alpha\sim$10\text{\,}\mathrm{\SIUnitSymbolDegree}$$. For $\alpha>$10\text{\,}\mathrm{\SIUnitSymbolDegree}$$ the impact on $\eta_{c}$ becomes larger. In particular, by increasing $\alpha=10\rightarrow$11\text{\,}\mathrm{\SIUnitSymbolDegree}$$ there is a further decrease of $\eta_{c}$ by $\sim 10\%$. Such reduction continues until $\alpha=$20\text{\,}\mathrm{\SIUnitSymbolDegree}$$ at which $\eta_{c}\sim 0.13$. A $\alpha>$15\text{\,}\mathrm{\SIUnitSymbolDegree}$$, $\eta_{c}<0.6$ and $\dot{m}_{thr}>$0.1\text{\,}\mathrm{m}\mathrm{g}\mathrm{/}{\mathrm{s}}$$.

The investigation on the combination of diffuse and specular reflecting materials on hybrid concepts differed from the expected result. It was initially thought that by having a diffuse chamber section, the intake performance would increase. However, lower $\eta_{c}$ compared to their pure specular counterpart are achieved. Furthermore, changing the percentage of diffuse versus specular scattering on the S-FT case provided insightful results for the understanding of the behaviour of the parabolic shape with different material properties. The $\eta_{c}$ is maintained $\eta_{c}>0.6$ for all cases with at least $70\%$ specular reflection. Although a specific relation of the $\sigma_{B}$ parameter to the scattering dynamics cannot yet be concluded, the results suggest an almost linear decrease of $\eta_{c}$ as the material becomes less specular and more diffuse, see Fig. 13. Such results are very valuable, as intake designs based on partial specular reflecting materials can provide $\eta_{c}>0.6$. Additionally, the results show how intake performance could change due to possible material degradation over the mission lifetime, for example by the action of AO in VLEO [2], and thus affecting the $\eta_{c}$.

In this section the final diffuse intake is presented. Based on the analysis of the intake concepts using diffuse scattering, the D-MDAR case is selected. Compared to the other best candidate, the D-MFR, provides a more homogenous pressure distribution, $p_{ch}\sim$0.27\text{\,}\mathrm{Pa}$$, in front of the thruster’s discharge channel as shown in Fig. 14, while providing $\eta_{c}>0.4$.

The intake has an hexagonal shape and it features a front section of hexagonal ducts with different $AR$:

• 1.

  Larger ducts in front of the thruster’s discharge channel $\Rightarrow$ increase $A_{in}$;

• 2.

  Narrower ducts in the outer intake region $\Rightarrow$ reduce the backflow transmission probability, molecular trap operation.

At the rear there is a conical section with a chamber angle of $45\text{\,}\mathrm{\SIUnitSymbolDegree}$ terminating on the thruster’s discharge channel diameter.

Although the expected $\dot{m}_{thr}$ is low compared to the currently applied $\dot{m}_{thr}$ to the thruster’s laboratory model [47, 48], 1) the thruster has still a large margin to be optimized in both size and operational conditions (therefore influencing intake geometry as well), 2) the input density is based on low solar activity, therefore for stronger solar activities, higher input density yields larger $\dot{m}_{thr}$ with limited variations of $\eta_{c}$. The pressure distribution within the diffuse intake is shown in Fig. 14.

In Tab. 8, the respective intake performance in the selected VLEO altitude range are presented, while in Fig. 15 the isometric render view of the diffuse intake is shown.

In this section the final specular intake is presented. Based on the analysis based on specular scattering, the S-FT case is selected: a parabolic scoop with its focus inside the thruster discharge channel providing $\eta_{c}=0.94$, and $\dot{m}_{thr}=$0.23\text{\,}\mathrm{m}\mathrm{g}\mathrm{/}{\mathrm{s}}$$ at $h=$150\text{\,}\mathrm{km}$$. Different simulation times, verifying runs, solar activity maximum-minimum, bow shock analysis, and different orbital altitudes that are not reported here for the sake of brevity all confirmed the same performance values. Such design outperforms the diffuse intake also in terms of sensitivity to flow misalignment leading to a limit of $\alpha=$15\text{\,}\mathrm{\SIUnitSymbolDegree}$$ resulting in $\eta_{c}<0.6$. The pressure achieved at the back of the intake is $p_{ch}=$0.3\text{\,}\mathrm{Pa}$$.

The S-FT case is selected as the best specular intake candidate as it provides $\eta_{c}=0.94$ and $\dot{m}_{thr}\sim$0.23\text{\,}\mathrm{m}\mathrm{g}\mathrm{/}{\mathrm{s}}$$.

Although it falls under the estimated value for the current laboratory model thruster’s discharge channel operating pressure $p_{ch}=1-$7\text{\,}\mathrm{Pa}$$ [47, 48], it has been experimentally shown that helicon plasma thrusters can operate as low as $p_{ch}=$0.266\text{\,}\mathrm{Pa}$$ [49]. The specular intake render is shown in Fig. 18, while the performances are shown in Tab. 9.

To assess the ABEP system performance, a preliminary thrust vs drag analysis is hereby performed. The drag $F_{D}$ is calculated assuming the frontal area of the intake (specular and diffuse final designs) to be that of the spacecraft $A_{f}=A_{Intake}$, a drag coefficient $C_{D}=2.2$ is assumed [50]. The SC velocity calculated as in Eq. 6, and the respective density from the NRLMSISE-00 model, see Eq. 7.

The thrust $F_{T}$ is calculated as in Eq. 8.

Where $c_{e}$ is the exhaust velocity that, by applying the condition of full drag compensation $F_{D}/F_{T}=1$, the required $c_{e}$ can be extracted. Finally, as $c_{e}$ is linearly dependent with $v(h)^{2}$, and $v(h)$ changes only slightly in the considered VLEO $h$ range, $c_{e}=9.1-$9.2\text{\,}\mathrm{k}\mathrm{m}\mathrm{/}{\mathrm{s}}$$ for the “Specular Intake”, and $c_{e}=35.8-$39.8\text{\,}\mathrm{k}\mathrm{m}\mathrm{/}{\mathrm{s}}$$ for the “Diffuse Intake”. This highlights the greater effectiveness of the “Specular Intake” compared to the “Diffuse Intake”, for example by maintaining the intake geometry but increasing $A_{f}$ such that $A_{f}>A_{in}$, to stay within the range achieved by the state-of-the-art conventional EP thrusters [51], an $A_{f}=$0.1\text{\,}\mathrm{m}^{2}$$ would require for the “Specular Intake” $c_{e}\sim$48\text{\,}\mathrm{k}\mathrm{m}\mathrm{/}{\mathrm{s}}$$, while for the “Diffuse Intake” $c_{e}\sim 190-$210\text{\,}\mathrm{k}\mathrm{m}\mathrm{/}{\mathrm{s}}$$, way higher than what available with current EP technologies [51]. Finally, the “Diffuse Intake” does not necessarily be discarded, as its performance can be still tuned for delivering higher $\dot{m}_{thr}$, and could benefit greatly from more accurate studies in low-drag SC design, and experimental results on the plasma thruster development. Moreover, it can be combined with a specular-based design to realize a hybrid intake.