Electric VTOL Configurations Comparison

: In the last ten years, different concepts of electric vertical take-off and landing aircrafts (eVTOLs) have been tested. This article addresses the problem of the choice of the best conﬁguration. VTOLs built since the ﬁfties are presented and their advantages, disadvantages, and problems are discussed. Three representative eVTOLs, one for each main conﬁguration, are compared on ﬁve main parameters and three reference missions. The parameters are disk loading, total hover time, cruise speed, practical range, and ﬂight time. The performance of the eVTOLs on the urban, extra-urban, and long-range mission is evaluated computing the time and energy required. The results show that the best conﬁguration depends on the mission. The multirotor is more efﬁcient in hover. The vectored thrust jet is more efﬁcient in cruise and has a higher range. The lift + cruise is a compromise.


Introduction
Electric vertical take-off and landing aircrafts (eVTOLs) are being built and tested, and their configurations vary from hover bikes to electric ducted fans.In 2010, Moore [1] presented the NASA Puffin electric tailsitter VTOL concept and highlighted the potential of electric propulsion to enable cheap, quiet, and reliable short-range VTOLs.That same year the company ZeeAero, now Kitty Hawk [2], was founded by Kroo with the aim of building an eVTOL flying car.From that moment on, many researchers, companies, and startups started to work on eVTOLs.Now, most of the major aircraft companies are directly developing their own electric VTOL or have subsidiaries doing it.More than 130 electric VTOL concepts have been proposed [3] and venture capitalists have invested more than 1 billion dollars into promising eVTOL startups [4].Moore and his colleagues have worked on the idea of on-demand air mobility [5,6], hybrid eVTOLs [7], the advantages of electric propulsion compared to internal combustion and gas turbines [8], and the distributed electric propulsion of the X-57 Sceptor [9].McDonald has worked on electric propulsion modeling for conceptual design [10] and developed the OpenVSP design tool.
Most of the research has been conducted by private companies.Uber has hired both Moore and McDonald and is trying to build, with its program Uber Elevate, the infrastructure for eVTOLs [11].Kitty Hawk, Lilium, Joby Aviation, and E-Hang are four of the startups developing electric VTOLs.Kitty Hawk has developed and is now testing two vehicles: Cora, the lift + cruise air taxi and the Flyer, a hoverbike [12].Lilium is a German startup that is building an electric ducted fan eVTOL.They have flown many prototypes including a two-seater jet and are now developing a five-seater air taxi [13].Joby Aviation has performed tests on electric propulsion and is building an eVTOL prototype [14].E-Hang is a Chinese company manufacturing quadrotor UAVs that has built and tested, with humans on board, the E-Hang 184 passenger drone [15].This article tries to understand which is the best eVTOL design, presenting and discussing all the different configurations, from the first developed in the fifties and sixties to the present eVTOL a fan behind the cockpit which provides, with the nozzle of the main engine swiveled, the thrust required for hover.The Rayan XV-5A had two fans-in-wing powered by the exhaust gases of its two turbojets.The two prototypes crashed during transition because of the slow control response and narrow transition corridor [17].The last VTOL category is the compound helicopter which uses a rotor to hover and has a propeller for forward flight.
In recent years many companies and startups have started developing and testing different electric VTOLs.The website Electric VTOL News [22], published by the Vertical Flight Society, classifies eVTOLs in the following categories: The vectored thrust eVTOLs have a wing for an efficient cruise and use the same propulsion system for both hover and cruise.The Lilium Jet, the Aurora LightningStrike, and the Joby S2 and S4 are in this category (Figure 1).The Lilium Jet is a tilt duct able to increase the lift coefficient of the wing, during the transition, sucking air from the upper surface of the wing and pushing it down with the electric jets.The Aurora LightningStrike is a tilt-wing with fans-in-wing.The Joby S2 is a tilt prop.The main difference between eVTOLs in this category is whether they have fans or propellers.The wingless eVTOLs are multirotors.They have large disk actuator surface which makes them efficient in hover, but they do not have a wing for an efficient cruise.These vehicles are suited for short-range operations in cities where they can fly over traffic jams.Two VTOLs in this class are already in the certification phase: The E-Hang 184 and the Volocopter 2X (Figure 3).Hoverbikes are multirotors that can be flown like a motorbike.The pilot sits on a saddle or is standing.An example is the prototype built and flown by Kitty Hawk (Figure 4).eHelos are electrical conventional helicopters.An example is the Aquinea Volta (Figure 5).The E-Hang 184, the Kitty Hawk Cora, and the Lilium Jet have been chosen as the reference for the comparison of the three main eVTOL categories.Their performances have been computed using analytical methods [32] such as the disk actuator theory [33] to evaluate hover performances, Breguet's equation for electric flight [34] to compute the theoretical range, empirical methods to evaluate the drag of the eVTOL from the Hoerner's book [35] and standard drag, power, efficiency considerations.
The comparison has been performed evaluating five parameters and computing the energy and time required to perform three reference missions.The five parameters are disk loading, total hover time, cruise speed, practical range, and flight time.The reference missions are: -7 km urban mission -30 km extra-urban mission -100 km long-distance mission

Wingless Multirotor Configuration
To present the performances of the wingless multirotor configuration, the e-Hang 184 data [15] has been used and its performance has been evaluated.Figure 6 shows the E-Hang 184.Data of the e-Hang 184 configuration, found in their website [15], is presented in Table 1.The estimated values of geometry, battery, mass balance, and hover performances are presented in Table 2.The battery mass has been computed as total energy divided by energy specific density.The energy specific density and specific power have been assumed equal to the values of the Tesla Model S battery pack [36].This assumption has been made because these batteries are used in a consumer product in a high-power application.They have demonstrated the ability to work after years and hundreds of life cycles, in harsh environments like Norway.Li-ion batteries for power applications have specific energy ranging from 100 to 250 Wh/kg [34,37] and specific power from 700 to 1300 W/kg [37].The assumed values are at pack level, they consider the additional weight of casing, connections, and thermal management system.The energy density and specific power of the batteries used for eVTOLs might be better than the assumed values.This conservative assumption means that the batteries will be able to provide enough power for takeoff and landing even after years of utilization.As batteries age, the energy they can store and the maximum power they can provide decrease [37,38].For electric cars, this means that the total range decreases.For eVTOLs, designed to be able to take off with maximum battery power at the beginning of the life of the battery, this might mean not having enough power to take off after a few years of service.
The average power consumption has been computed dividing the total energy by the total time of flight.The power required to hover, P, has been computed with the disk actuator theory modified for coaxial rotors [39]: where T is the thrust or the weight of the vehicle, ρ is the air density at sea level (1.225 kg/m 3 ), A is the disk actuator area of the vertical thrust system, in this case, the area of the four coaxial rotors, and k int is the interference factor.k int varies from 1 for zero interference to √ 2 for maximum interference corresponding to the case of two rotors corotating in the same plane.The value selected is 1.26 for rotors operated at balanced torque with the lower rotor operating in the fully developed wake of the upper rotor [39].
The power required to hover, 42.7 kW, is different from the average power consumption specified by the producer, 34.6 kW (Table 1), because the power required in cruise is less than in hover due to the lift produced by the vehicle and to the reduction in induced drag [39][40][41].The total hover time found is 20.5 min.
The electric motors have been sized to ensure the safety of the vehicle in case of failure.Each coaxial rotor couple is a failure redundant system.The two rotors are driven by two different motors.When one motor fails, the propeller connected to it stops and the propeller connected to the undamaged motor of the couple provides the entire thrust normally provided by the coaxial rotor couple.To evaluate the power required for this contingency scenario, the standard disk actuator theory has been used [32,33]: where T is the thrust and A is the disk actuator area of the single operative propeller.The thrust considered is a quarter of the weight of the vehicle multiplied by a 1.3 margin factor for maneuver.The power found is 17.5 kW per motor, while the maximum power per motor specified by E-Hang's website [15] is 19 kW.This value has been found by dividing the total power, 152 kW in Table 1, by 8, the total number of motors.

Lift + Cruise Configuration
To present the performances of the lift + cruise configuration, the Kitty Hawk Cora's data has been used (Figure 7 and Table 3).
The mass and battery data of the Kitty Hawk eVTOL are listed in Table 5.The total mass of the vehicle is 1224 kg as specified in [42].The battery mass has been estimated computing the minimum power required to hover P.This is given by the disk actuator Equation ( 2), where the thrust, T, is the weight of the vehicle, ρ is the air density at sea level (1.225 kg/m 3 ), and A is the disk actuator area of the vertical thrust system.The hover performances are listed in Table 6.The gravity acceleration g = 9.8 m/s 2 has been used.The energy required to hover for one minute is computed multiplying the power required to hover by 60 s and the total hover time is computed dividing the total energy available by the power required to hover.Comparing it to the E-Hang 184, the Kitty Hawk Cora requires over four times the energy for one minute of hover, has double the disk loading, and has a lower total hover time.The aerodynamic properties of the Cora vehicle have been estimated using the lifting line theory, adding the additional resistance of the pylons and propellers for the vertical takeoff.For the lifting line procedure, the airfoil's lift slope coefficient and zero lift angle have been assumed 5.34 rad −1 and −3.26 • .These values have been found selecting the airfoil NLF(1)-0115 [43,44] and using the software Xfoil [45] for the computations.The aerodynamic drag of the wing and the horizontal tail are computed integrating the airfoil sections contribution.The drag of the other components is estimated using the parameters listed in Table 7.The fuselage drag is computed using [46]: where C f is given by: for turbulent flow, and by: for laminar flow, F is the form factor given by: and Q is the interference factor set at 1.The flow is assumed to be 20% laminar and 80% turbulent.
The same procedure has been followed for the vertical tail, computing the form factor, F, by F = (F * − 1)cos 2 ∆ 0.5c + 1 (7) where F * = 1 + 3.52(t/c) and ∆ 0.5c is the sweep angle at 50% of the chord.The interference factor, Q, is set at 1.2 for the vertical tail.The drag of the pylons supporting the vertical lift propellers and the drag of the vertical lift propellers has been computed as the base drag of a 3D body [35] (pp.[3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19].The drag coefficient is computed as: then it is scaled to the reference surface, corresponding to the wing surface: Equations ( 8) and ( 9) have been applied for both the pylons supporting the vertical lift propellers and for the vertical lift propellers using their different geometries and different friction coefficients.The drag of the landing gear is computed supposing a C D0 of 0.25 as suggested in [35] and scaling it from the wheel surface to the reference surface.
The interference drag between the wing and fuselage has been added using the following equation [35]: The drag polar of the Cora vehicle computed is: The speed of maximum L/D and the maximum L/D are given by [47]: where C D0 and k are the parameters of the drag polar.This gives a speed of maximum L/D of 145 km/h and a maximum L/D of 13.9.The theoretical range, given by [32]: is 200 km.Limiting the depth of discharge to 70%, improving the cruise speed to save time to 180 km/h, and considering takeoff and landing the range decreases to 107 km.The flight time is 36 min.These results are in accordance with the performances specified by the producer of 180 km/h cruise speed and 19 min flight time plus 10 min of reserves (Table 3).The depth of discharge of the battery has been limited to 70% because Li-ion batteries lifetime depends on the depth of discharge at which they are subjected [37].A 70% depth of discharge gives a good amount of energy preserving the lifetime of the battery.The energy required for takeoff, landing, and transition is 6.3 kWh, which corresponds to 1 min and 40 s of hover.The power required for the cruise is given by: where D is the aerodynamic drag, D = 1 2 ρ ∞ S v 2 C D , v is the flight speed and η is the total efficiency of the power electronics and propeller which has been assumed 75%.
The angle of attack in cruise is 8 degrees.This seems reasonable because, as seen in the drawings, the angle between the wing and the fuselage is 12 degrees.This feature allows a comfortable cruise at high angles of attack that reduces the required wing surface also reducing the aerodynamic drag of the wing.Flight at high angles of attack with low induced drag is possible because Cora has a very high aspect ratio wing.The angle between the wing and the fuselage is also beneficial because turning on the VTOL propellers in flight produces a thrust which has a component opposed to the flight direction, allowing a smooth deceleration, and transition between cruise and vertical landing.

Vectored Thrust
To evaluate the vectored thrust eVTOL category, the Lilium Jet has been selected.The geometric dimensions have been estimated from the image of the first flight test of the prototype (Figure 8 and Table 8).
The total mass has been assumed 490 kg because this is less than the maximum takeoff weight for the ultralight aircraft category in Europe (450 + 45 kg) [48,49].The battery mass has been estimated at 240 kg, with a battery mass to total mass ratio of 49%.These results are presented in Table 9.The hover performances have been computed using the disk actuator theory modified for ducted fans [50].The power required to hover is: where Ti = 1.26 is the thrust increase for ducted fans, T is the thrust required or the weight of the vehicle and A is the disk actuator area of the vertical thrust system.The sea level air density ρ = 1.225 kg/m 3 and gravity acceleration g = 9.8 m/s 2 have been used.The results are listed in Table 10.The power required found is 187 kW and the maximum power available with 240 kg of batteries and a specific power of 735 W/kg is 176 kW.This means that, with the assumptions made, the Tesla batteries considered are not enough to power the Lilium jet.It requires batteries with a higher specific power.Prandtl's lifting line theory was used to compute the wing's lift and drag.The drag produced by the fuselage and by the forward fans has then been added.The resulting drag polar is: Applying Equations ( 11) to ( 14), the speed of maximum L/D is 230 km/h and the maximum L/D is 16.3.The theoretical range, computed with Equation (13) [32], is 380 km.Limiting the depth of discharge to 70%, improving the cruise speed to save time to 250 km/h, and considering takeoff and landing the range decreases to 203 km.The flight time is 48 min.
Lilium is now developing a five-seater version of its eVTOL (Figure 9).Performing the same procedure for this vehicle, the estimated values are 12 m wingspan, 0.28 m fans diameter, 1700 kg total mass, and 900 kg battery mass.The disk loading of this vehicle is the same as the two-seater jet, the practical range with 70% depth of discharge and a cruise speed of 290 km/h is 245 km, and the total flight time is 55 min.

Reference Mission Performance
The time and energy required by the three eVTOLs to perform the three reference missions have been computed.Each mission consists of: -15 s of takeoff at hover power; -acceleration at 2 m/s 2 from zero forward speed to cruise speed at hover power; -cruise flight; -deceleration at −2 m/s 2 from cruise speed to zero forward speed at hover power; -15 s of landing at hover power.
The results are presented in Tables 11-15.

Discussion
The three configurations examined have been compared.Their hover and cruise flight parameters are presented in Table 16 and their performances for the three reference missions are presented in Table 17, Figures 10 and 11.The urban mission comparison shows that multirotors require less energy for short-range missions.In the extra-urban mission, the cruise phase is as important as the hover phase, and the energy required by the three configurations is comparable.In the long-range mission, cruise efficiency is more important.E-hang's range is insufficient to complete it.Cora's parasitic drag caused by the pylons and vertical thrust propellers increases the power required in cruise.Its cruise speed is less than Lilium's and the energy required is more.
Lilium's hover is so power demanding that it requires batteries with higher specific power than the Tesla batteries considered for the computations.This means that the aerodynamic advantages of this configuration are balanced by higher demands on the batteries and on the power electronics.
More practical reasons might influence which eVTOL configuration will be adopted more rapidly in the future.The multirotor configuration seems to be closer to the market and less complex than the lift + cruise and the electric jet.However, the range advantage of the latter two enables missions impossible to the multirotor configuration.

Conclusions
Different configurations of turboshaft powered VTOLs tested in the fifties and sixties have been discussed in the introduction detailing advantages, disadvantages, and problems of each one.The recent eVTOL prototypes categories have then been presented under the classification proposed by the American Helicopter Society.The performances of the three main eVTOL configurations have been evaluated estimating five main parameters, the energy and the time required to complete three reference missions.The performances of the E-Hang 184 have been estimated for the multirotor configuration, the Kitty Hawk Cora was selected for the lift plus cruise configuration and the Lilium jet represented the electric jet configuration.This analysis showed that the best eVTOL configuration depends on the mission.Short-range missions are best performed by multirotors because they have better hover performances.Long-range missions cannot be accomplished by multirotors because their range is not enough.

Figure 1 .
Figure 1.Vectored thrust electric vertical take-off and landing aircrafts (eVTOLs): Lilium Jet [13], Aurora Lightning Strike [23], Joby S2 [24].The lift + cruise eVTOLs have a wing for an efficient cruise, like vectored thrust eVTOLs, but they use two different propulsion systems for hover and cruise flight.The ZeeAero Z-P2, the Kitty Hawk Cora, and the Aurora Flight Sciences eVTOL are in this category (Figure2).

Figure 6 .
Figure 6.E-Hang 184 specs and dimensions, from E-Hang website [15].Dimensions in the top view are in millimeters.

Figure 10 .
Figure 10.Energy required for the three reference missions.

Figure 11 .
Figure 11.Time required for the three reference missions.Table 16 shows that the multirotor configuration represented by the E-Hang 184 is the best suited to hover flight while Lilium is the best suited to cruise flight.The lift + cruise Cora is a compromise.It has less range and flight speed than Lilium but good hover performances comparable to the wingless multirotor configuration.The 7 km urban mission is completed in 4.9 min by E-Hang, 3.3 min by Cora, and 2.8 min by Lilium.E-Hang requires 3 kWh, Cora 7.1 kWh, and Lilium 5.7 kWh.The 30 km extra-urban mission is completed in 18.7 min by E-Hang, 10.9 min by Cora, and 8.2 min by Lilium.E-Hang requires 10.9 kWh, Cora 15.2 kWh, Lilium 8.3 kWh.The 100 km long-range mission cannot be completed by E-Hang and is almost Cora's computed maximum range.It is completed in 34.3 min by Cora and 24.9 min by Lilium.Cora requires 39.7 kWh and Lilium 16.1.Theurban mission comparison shows that multirotors require less energy for short-range missions.In the extra-urban mission, the cruise phase is as important as the hover phase, and the energy required

Table 3
shows the Kitty Hawk Cora specifications.With this data, the propeller area has been computed and is presented in Table4.

Table 5 .
Kitty Hawk Cora mass and battery data.

Table 7 .
Kitty HawkCora geometry data used to compute the drag.

Table 9 .
Lilium Jet mass and battery data.

Table 11 .
Data used in the computations for the reference mission performance.

Table 17 .
Summary of the energy and time required for the three reference missions.