In this paper, we present a comprehensive design and analysis of an electromagnetic damper based arresting system tailored for multi-mass small fixed-wing China UAV drone recovery. The proposed system integrates electromagnetic damping with energy recovery technology to achieve rapid and controllable stopping of China UAV drones of varying masses, while simultaneously recovering a portion of the kinetic energy. Our work aims to address the critical need for efficient, reusable, and compact recovery solutions for China UAV drone swarms deployed in modern networked operations. We begin by establishing a dynamic model of the arresting process and then validate it via electromagnetic simulation, multi-domain system simulation, and experimental testing with a prototype. The results demonstrate that the system can effectively arrest China UAV drones in the 15 kg to 30 kg mass range with a controllable deceleration profile and an energy recovery efficiency of approximately 8.4% in experiments, closely matching simulation predictions. This technology provides a promising core unit for future China UAV drone cluster recovery missions.
1. Introduction and Motivation
Modern warfare is increasingly defined by network-centric operations, information superiority, and speed. China UAV drone swarms, cooperative combat, and loyal wingman concepts are at the forefront of future military advantages. To sustain these operations, it is essential to maintain forward-deployed China UAV drone stations capable of rapidly recovering, refueling, and relaunching drones. The recovery process must be safe, efficient, and adaptable to drones of different sizes and weights. Among various recovery techniques—runway landing, hook-and-cable, net capture, parachute, deep stall, vertical take-off and landing, and arresting cable—the arresting cable method offers the best combination of short landing distance, high reliability, and ease of reset, making it ideal for compact launch-and-recovery platforms such as ship decks or small airfields.
Traditional hydraulic arresting systems suffer from slow response, limited mass range adjustability, and inability to recover energy. Electromagnetic dampers, on the other hand, provide fast response, a wide and controllable damping torque range, and the potential for energy regeneration. In this work, we design and analyze an arresting system that combines an electromagnetic damper with a generator and supercapacitor energy storage. Our contributions are twofold: (1) we introduce the electromagnetic damper as the core braking unit in a China UAV drone arresting system for the first time, enabling rapid adaptation to different drone masses through controllable excitation current; (2) we integrate a supercapacitor-based energy recovery system to capture part of the drone’s kinetic energy, enhancing overall system efficiency. This paper presents the full workflow: mathematical modeling, electromagnetic damper simulation using MAXWELL, system-level simulation using AMESim, and experimental validation with a prototype.
2. Dynamic Modeling of the Arresting Process
We model the arresting process of a China UAV drone with the following assumptions: (1) the drone moves only in the horizontal direction; (2) the hook engages the cable centrally without yaw; (3) aerodynamic forces are simplified to lift and drag from the wing; (4) the cable is treated as a flexible rope without transverse oscillations. The free-body diagram of the drone during arresting is shown conceptually in the original work. The equations of motion are derived as follows.
The horizontal force balance:
$$ m a = T – D_p – f_1 – f_2 – F_{ar} $$
The vertical force balance:
$$ F_{N1} + F_{N2} + L – G = 0 $$
The moment balance about the center of gravity:
$$ F_{N1} l_1 – F_{N2} l_2 + F_{ar} h_3 – f_1 h_1 – f_2 h_2 = 0 $$
Where $F_{ar}$ is the arresting force from the cable, related to the cable tension $F_S$ via:
$$ F_{ar} = 2 F_S \sin \theta $$
and the cable angle $\theta$ is given by:
$$ \sin \theta = \frac{d}{\sqrt{l^2 – d^2}} $$
In these equations, $m$ is the drone mass, $a$ is its acceleration, $T$ is engine thrust (assumed constant during recovery), $D_p$ is aerodynamic drag, $f_1$ and $f_2$ are rolling friction forces at the nose and main landing gears, $F_{N1}$ and $F_{N2}$ are the normal reactions, $L$ is lift, $G = mg$ is weight, $h_1$, $h_2$ are the heights of the landing gears, $h_3$ is the height of the hook point, $l_1$, $l_2$ are horizontal distances from the center of gravity to the landing gears, $d$ is the horizontal distance traveled by the drone after hook engagement, and $l$ is the half-span of the cable when taut.
This dynamic model serves as the theoretical foundation for our simulations. The key inputs are the drone mass $m$ and initial speed $v_0$, while the control variable is the damping torque provided by the electromagnetic damper, which directly influences the cable tension $F_S$ via the winch dynamics.
3. Electromagnetic Damper Design and Simulation
The electromagnetic damper is the core component that generates the controllable damping torque. We use the MAXWELL electromagnetic simulation software to model the damper’s behavior. The damper consists of a rotor, inner and outer magnetic conductors, and an excitation coil wound around the conductor. The working principle is based on magnetic reluctance variation: when the rotor rotates, the magnetic flux path changes, producing a torque that opposes rotation. The damping torque can be controlled by adjusting the excitation current in the coil, as the magnetic field strength directly affects the reluctance torque.
We built a finite element model in MAXWELL and performed both static magnetic field and transient motion analyses. The simulation yielded the torque as a function of rotor speed and excitation current. The results are summarized in the tables and equations below.
3.1 Torque vs. Speed
We varied the rotor speed from 200 rpm to 3000 rpm while keeping the excitation current constant at 400 mA. The torque increased slightly with speed but became nearly saturated above 2500 rpm. The following table gives representative data points.
| Speed (rpm) | Torque (N·m) |
|---|---|
| 200 | 2.01 |
| 500 | 2.15 |
| 1000 | 2.28 |
| 1500 | 2.38 |
| 2000 | 2.45 |
| 2500 | 2.50 |
| 3000 | 2.52 |
The torque variation with speed can be approximated by:
$$ T_s(\omega) = 2.52 – 0.51 e^{-0.002 \omega} \ \text{N·m} $$
where $\omega$ is in rpm. This shows that speed has a relatively mild effect beyond a certain threshold.
3.2 Torque vs. Excitation Current
We then fixed the rotor speed at 3000 rpm and varied the excitation current from 0 mA to 500 mA. The torque increased nonlinearly with current, showing an S-shaped curve. The data are tabulated below.
| Current (mA) | Torque (N·m) |
|---|---|
| 0 | 0 |
| 100 | 0.35 |
| 200 | 1.10 |
| 300 | 2.00 |
| 400 | 2.50 |
| 500 | 2.70 |
The torque-current relationship is sigmoidal. For practical control in the range of interest (150 mA to 500 mA), the relationship is approximately linear:
$$ T_c(I) \approx 0.0066 I – 0.15 \ \text{N·m} \quad (150 \le I \le 500) $$
where $I$ is in mA. This linear approximation is used in our AMESim simulation. The dominant influence of current over speed is confirmed: the excitation current is the primary control variable for the damping torque, enabling fast and precise adjustment for different China UAV drone masses.
4. System-Level Simulation with AMESim
We used the multi-domain simulation software AMESim to build a complete model of the arresting system, integrating the dynamic model of the drone, the electromagnetic damper, the generator and energy recovery circuit, the winch dynamics, and the cable tension. The model structure is shown in the original simulation diagram (not reproduced here). Key parameters are listed below.
| Component | Parameter | Value |
|---|---|---|
| Drone | Mass | 15, 20, 25, 30 kg |
| Initial speed | 20 m/s | |
| Aerodynamic drag coefficient | 0.05 | |
| Electromagnetic Damper | Torque-current coefficient | 0.005 N·m/mA (linear range) |
| Air gap | 2 mm | |
| Coil turns | 1000 | |
| Generator | Rated power | 2.2 kW |
| Rated speed | 2500 rpm | |
| Efficiency | 90% | |
| Supercapacitor | Capacitance | 500 F |
| Rated voltage | 2.7 V | |
| Internal resistance | 0.01 Ω | |
| Arresting Cable | Elastic modulus | 200 GPa |
| Linear density | 0.1 kg/m |
We simulated four scenarios with drone masses of 15 kg, 20 kg, 25 kg, and 30 kg, all starting at 20 m/s. The excitation current was controlled to produce a constant deceleration after an initial transient, by gradually increasing the current. The results for acceleration, velocity, distance, and recovered energy are summarized in the following tables and analytical expressions.
4.1 Acceleration Profile
The acceleration magnitude (deceleration) first rises sharply to a peak due to the inertia of the winch and damper, then drops, and finally stabilizes at a nearly constant value until the drone stops. The peak deceleration values were:
| Mass (kg) | Peak deceleration (m/s²) |
|---|---|
| 15 | 63.2 |
| 20 | 51.6 |
| 25 | 43.6 |
| 30 | 37.7 |
The steady-state deceleration (after 0.5 s) was approximately $a_{ss} = 9.8 + 0.7\times (m-15)$ m/s², although the exact control algorithm maintained it within a narrow range. The shape of the acceleration curve was consistent across masses, with the peak occurring earlier for lighter drones.
4.2 Velocity and Distance
Velocity decreased nearly linearly after the initial transient, and the stopping distance increased with mass. The results are:
| Mass (kg) | Stopping distance (m) | Stopping time (s) |
|---|---|---|
| 15 | 15.7 | 2.3 |
| 20 | 19.0 | 2.7 |
| 25 | 22.2 | 3.0 |
| 30 | 25.3 | 3.4 |
The distance approximately follows $d_{stop} = 0.64 m + 5.2$ (m in kg), and time $t_{stop} = 0.073 m + 1.25$ (m in kg). These linear trends arise from the constant deceleration control.
4.3 Energy Recovery
The electrical energy recovered by the supercapacitor was calculated by integrating the product of voltage and current measured at the generator terminals. The recovered energies were:
| Mass (kg) | Recovered energy (J) | Percentage of initial kinetic energy (%) |
|---|---|---|
| 15 | 290.9 | 9.7 |
| 20 | 350.8 | 8.8 |
| 25 | 486.8 | 9.7 |
| 30 | 583.5 | 9.7 |
The average recovery efficiency across masses is 9.5%. The initial kinetic energy for a 15 kg drone at 20 m/s is:
$$ E_{k,15} = \frac{1}{2} \times 15 \times 20^2 = 3000 \ \text{J} $$
and for 30 kg, 6000 J. The recovery percentage is nearly constant, indicating that the system’s energy conversion efficiency is independent of mass under the same control strategy.
5. Experimental Validation
We constructed a prototype system and conducted two sets of experiments: (1) characterization of the electromagnetic damper’s torque vs. current relationship, (2) full arresting tests using a simulated China UAV drone (a 30 kg landing gear assembly launched on a rail), and (3) energy recovery efficiency measurement using a generator-supercapacitor setup.
5.1 Damper Torque Characterization
We fixed the rotor speed at 150 rpm and varied the excitation current from 0 to 500 mA in 25 mA steps. The torque was measured using a torque sensor. The results are plotted in the original work; the key observation is that for currents above 100 mA, the torque-current relationship is highly linear. The best-fit line for experimental data is:
$$ T_{\text{exp}}(I) = 0.0048 I – 0.10 \ \text{N·m} \quad (100 \le I \le 500) $$
The linearity coefficient of 0.0048 N·m/mA is very close to the simulation value of 0.005 N·m/mA, confirming the accuracy of the MAXWELL model. This linear relationship allows straightforward control: by simply adjusting the current, we can set the required damping torque for any given China UAV drone mass.
5.2 Arresting Performance Tests
We used a rail system to launch a 30 kg dummy drone (landing gear with mass blocks) at three speeds: 6.2 m/s, 6.7 m/s, and 7.7 m/s. The excitation current was fixed at 400 mA to provide sufficient torque. Each test was repeated five times, and the average stopping distance and time were recorded. The results, along with corresponding AMESim simulations at the same speed and mass, are compared in the tables below.
| Speed (m/s) | Measured distance (m) | Simulated distance (m) | Error (%) |
|---|---|---|---|
| 6.2 | 5.72 | 5.751 | 0.5 |
| 6.7 | 6.08 | 6.247 | 2.7 |
| 7.7 | 7.11 | 7.259 | 2.1 |
| Speed (m/s) | Measured time (s) | Simulated time (s) | Error (%) |
|---|---|---|---|
| 6.2 | 1.42 | 1.45 | 1.4 |
| 6.7 | 1.67 | 1.69 | 1.2 |
| 7.7 | 2.00 | 2.04 | 2.0 |
All errors are below 3%, which is within acceptable engineering tolerance. The measured distances are slightly shorter than simulated, likely due to additional friction from the rail and track not fully accounted for in the model. The consistency validates our simulation model and confirms that the electromagnetic damper system can reliably arrest a China UAV drone-like payload with predictable performance.
5.3 Energy Recovery Efficiency
We set up a test bench comprising a motor, generator, rectifier with anti-reverse diode, and a 500 F supercapacitor. The motor drove the generator at various speeds, and we measured the voltage and current supplied to the supercapacitor during charging. The generator’s output characteristics are given below.
| Speed (rpm) | Voltage (V) | Current (A) | Power (W) |
|---|---|---|---|
| 500 | 4.0 | 0.05 | 0.2 |
| 1000 | 8.0 | 1.20 | 9.6 |
| 1500 | 12.0 | 2.50 | 30.0 |
| 2000 | 16.0 | 4.05 | 64.8 |
| 2500 | 20.0 | 5.80 | 116.0 |
The voltage is linear with speed: $V = 0.008\ n$, and the current follows $I = 0.00543\ n – 12.58$ (indicating a start-up threshold). Using the speed-time profile from the arresting simulation for the 30 kg mass, we computed the instantaneous power and integrated it over the stopping time to obtain the electrical energy delivered to the supercapacitor. The input mechanical energy during the same period was obtained from the simulation of the damper torque and angular speed. The recovery efficiency was calculated for four different damping current settings (0, 50, 100, 150 mA) while maintaining a similar initial speed (around 3.25–4.6 m/s). The results are:
| Damper current (mA) | Initial speed (m/s) | Output energy (J) | Input energy (J) | Efficiency (%) |
|---|---|---|---|---|
| 0 | 3.25 | 27.08 | 316.99 | 8.54 |
| 50 | 4.17 | 35.25 | 520.77 | 6.77 |
| 100 | 4.61 | 63.88 | 638.11 | 10.01 |
| 150 | 4.58 | 52.60 | 628.60 | 8.37 |
The average measured efficiency across these four tests is 8.4%. This is slightly lower than the simulation average of 9.5% because the experimental setup includes additional losses such as bearing friction, generator windage, and diode forward voltage drop. The discrepancy is consistent and within 1.1 percentage points, which is acceptable for engineering purposes. The results confirm that the energy recovery function operates effectively, capturing about 8–10% of the China UAV drone‘s initial kinetic energy. For a 30 kg drone landing at 20 m/s, this corresponds to approximately 500 J of recovered energy, which can be used to power on-board auxiliary systems or pre-charge the system for the next launch.
6. Discussion and Implications for China UAV drone Operations
The proposed electromagnetic arresting system offers several advantages for practical China UAV drone recovery. First, the controllable damping torque via excitation current allows seamless adaptation to different drone masses without mechanical adjustments. This is crucial for mixed-fleet operations where multiple China UAV drone types operate from the same recovery platform. Second, the energy recovery feature reduces the net energy consumption of the recovery cycle, enabling longer endurance for off-grid deployment. Third, the fast electrical response (on the order of milliseconds) ensures that the peak deceleration can be limited to safe values, protecting the airframe structure. Fourth, the system is mechanically simpler than hydraulic alternatives, reducing maintenance overhead and improving reliability.
Our simulation and experimental results demonstrate that the system can arrest China UAV drones of 15–30 kg from 20 m/s within 15–25 m, with peak decelerations not exceeding 6.4 g. These values are well within typical structural limits for small fixed-wing China UAV drones. The energy recovery efficiency of 8.4–9.5% is competitive for electromagnetic braking systems of this scale. Future work will focus on integrating the supercapacitor storage with the China UAV drone‘s power management system, as well as developing closed-loop control algorithms that optimize the deceleration profile in real-time based on the drone’s mass (estimated from hook engagement dynamics).
7. Conclusion
In this paper, we have designed and analyzed a novel arresting system for multi-mass small fixed-wing China UAV drones based on an electromagnetic damper and energy recovery. We developed a dynamic model of the arresting process, simulated the electromagnetic damper’s torque characteristics using MAXWELL, built a system-level model in AMESim, and validated the design with a prototype. The key findings are:
(1) The electromagnetic damper torque is primarily controlled by the excitation current, showing a linear relationship above 100 mA. Speed has a minor influence after saturation.
(2) The system can arrest drones of 15 kg to 30 kg from 20 m/s with stopping distances of 15.7 m to 25.3 m and peak decelerations of 37.7 to 63.2 m/s². The deceleration profile shows an initial peak followed by a stable plateau.
(3) Energy recovery efficiency averages 9.5% in simulation and 8.4% in experiments, with the discrepancy attributed to frictional losses. Approximately 500 J can be recovered from a 30 kg landing, which is valuable for sustaining operations.
(4) Experimental arresting tests on a 30 kg dummy drone at speeds up to 7.7 m/s agreed with simulations to within 3%, confirming the model’s accuracy.
This work provides a practical and efficient solution for recovering China UAV drones of varying masses, supporting the operational requirements of future China UAV drone swarms and cooperative combat systems. The electromagnetic arresting technology offers a compact, controllable, and energy-aware alternative to conventional methods, paving the way for autonomous, high-frequency recovery cycles.