In our work on a military-trade unmanned aerial vehicle (UAV) system, we encountered a critical practical challenge during the engineering development of the rope-hook recovery device. The telescopic arm slider clearance, if improperly set, frequently led to severe failures such as steel cable detachment from pulleys and pulley damage, directly threatening the safety of UAV recovery. Addressing this issue required a multi-dimensional modeling approach that integrates thermodynamics, tribology, and dynamics to define a theoretical clearance range, and then a comprehensive evaluation model to identify the optimal clearance. This article presents our methodology and engineering validation, offering a reusable framework for similar recovery systems in drone technology.
Rope-hook recovery remains one of the classic methods for small fixed-wing UAVs. During recovery, the UAV’s wingtip hook engages a rope suspended between two horizontal bars mounted on a telescopic boom structure. The boom, composed of four sections, employs a system of steel cables and pulleys to adjust the height. When the UAV impacts the rope, significant lateral oscillations occur. If the slider clearance between telescopic sections is not optimized, the resulting impact forces can cause cables to jump out of their pulleys or pulleys to crack. Figure below illustrates the typical structural arrangement of the recovery device we studied.
Our objective was to determine the slider clearance that minimizes these failures while maximizing energy absorption, response speed, and fatigue life, and achieving the best cost-performance balance. The approach began with establishing a theoretical clearance interval [δ_min, δ_max] through three boundary analyses.
1. Theoretical Clearance Interval
1.1 Thermodynamic Boundary
Thermal deformation due to temperature changes can cause the slider to jam if the clearance is too small. The total thermal deformation ΔL of the inner arm relative to the outer arm is given by:
$$ \Delta L = \alpha \cdot L \cdot \Delta T $$
where α = 12×10⁻⁶ /°C (steel), L = 3020 mm (extension length), and ΔT = 50 °C (maximum working temperature range). Computing ΔL = 1.812 mm. The minimum clearance δ₁ required to accommodate half of this deformation with a safety factor η = 1.2 is:
$$ \delta_1 = \frac{\Delta L}{2} \cdot \eta = 1.087 \text{ mm} $$
Considering manufacturing tolerances and assembly errors, we applied an assembly margin M_total using ISO 286-1:2010 standards. After including a margin of 8%, the adjusted thermal clearance minimum became 1.0 mm. This ensures no thermal jamming under extreme conditions.
1.2 Tribological Boundary
From a lubrication and wear perspective, the clearance must be sufficient to maintain an oil film of adequate thickness to prevent direct metal-to-metal contact. The minimum oil film thickness h_min is computed based on surface roughness:
$$ h_{min} = k \cdot \sqrt{R_{q1}^2 + R_{q2}^2} $$
with k = 3.0, R_q1 = 2 μm (slider), R_q2 = 1 μm (guide rail), giving h_min ≈ 6.71 μm. Including a 20% engineering safety margin, the threshold oil film thickness was set to 8 μm. The relationship between clearance δ₂ and oil film thickness h_oil was empirically modeled as:
$$ h_{oil} = 38.07 \, \delta_2^{0.7} \mu^{0.5} \nu^{0.2} $$
where μ = 0.15 Pa·s (grease dynamic viscosity), ν = 0.1 m/s (sliding speed). The wear rate W (mm per thousand cycles) is:
$$ W = k \cdot \frac{P L}{H} \cdot \lambda_{oil} \cdot (1.2 – 0.15 \delta_2) $$
with k = 2.5×10⁻⁶ (nylon on steel), P = contact pressure, L = 3.02 m, H = material hardness. Table 1 summarizes computed values for different clearances.
| δ₂ (mm) | h_oil (μm) | W (mm/10³ cycles) |
|---|---|---|
| 1.0 | 9.3 | 0.015 |
| 1.2 | 10.5 | 0.012 |
| 1.5 | 12.3 | 0.011 |
At δ₂ = 1.0 mm, the oil film thickness (9.3 μm) exceeds the 8 μm threshold, and the wear rate is acceptable. Therefore, the minimum clearance from tribology is δ₂ = 1.0 mm as well. Combining both boundaries, we set δ_min = max(δ₁, δ₂) = 1.0 mm.
1.3 Dynamic Boundary
The maximum allowable clearance δ_max is constrained by the need to limit the time response of the slider during impact. The maximum response time t_max is related to clearance by:
$$ t_{max} = \sqrt{\frac{2 \delta_{max}}{a}} $$
With t_max = 0.015 s and a = 25 m/s² (contact acceleration), solving yields δ_max = 2.8125 mm. To maintain a safety margin of about 29% (assembly margin per ISO 286-1), we set δ_max = 2.0 mm. Additionally, we checked torsional compliance: the twist angle θ of the telescopic arm under maximum torque T = 5000 N·m is:
$$ \theta = \frac{T L}{G J} $$
where G = 79×10⁹ Pa, J = 1.2×10⁻⁴ m⁴. The computed θ = 0.091°, well below the ISO 10223 standard limit of 0.15°. Thus, the theoretical clearance interval is [1.0, 2.0] mm.
2. Comprehensive Evaluation Model
2.1 Key Performance Indicators
To pinpoint the optimal clearance within this interval, we evaluated three primary metrics: energy absorption rate E, response time t, and fatigue life F.
Energy absorption rate E is defined as the ratio of energy absorbed by the recovery device to the initial kinetic energy of the UAV. Through experimental data fitting, we obtained a quadratic polynomial:
$$ E(\delta) = 8 + 74\delta – 24\delta^2 \quad (\%) $$
Response time t was measured directly in recovery tests. The shortest t (8.1 ms) occurred at δ = 1.5 mm.
Fatigue life F (number of stress cycles to failure) was calculated using:
$$ F = N_f = \frac{C}{(\Delta \sigma_{eff})^m} $$
where C = 2.5×10¹², m = 3.2, and Δσ_eff is the effective stress amplitude given by:
$$ \Delta \sigma_{eff} = \Delta \sigma_{nom} [1 + q (K_t – 1)] $$
with q = 0.85 (fatigue sensitivity), and the stress concentration factor K_t:
$$ K_t = 1 + 0.75 \left(\frac{\delta}{b}\right)^{-0.5} e^{-0.9\delta} $$
and Δσ_nom = T / (2μ A_c). For δ = 1.5 mm, b = 15 mm, μ = 0.12, A_c = 0.0015 m², we computed:
$$ \Delta\sigma_{nom} = \frac{5000}{2 \times 0.12 \times 0.0015} \approx 13.89 \text{ MPa} $$
$$ K_t = 1 + 0.75 \left(\frac{1.5}{15}\right)^{-0.5} e^{-0.9 \times 1.5} \approx 1.616 $$
$$ \Delta\sigma_{eff} = 13.89 [1 + 0.85(1.616 – 1)] \approx 21.17 \text{ MPa} $$
$$ F = \frac{2.5 \times 10^{12}}{21.17^{3.2}} \approx 1.365 \times 10^8 \text{ cycles} $$
All computed fatigue lives exceeded the design threshold of 5×10⁷ cycles.
2.2 Normalized Scoring and Cost-Performance Model
To combine these three metrics with different units, we normalized each to a 0–1 scale. For energy absorption rate:
$$ S_E = \frac{E – E_{min}}{E_{max} – E_{min}} $$
with E_min = 58% (δ=1.0 mm), E_max = 65% (δ=1.5 mm). For response time (where smaller is better):
$$ S_R = 1 – \frac{R – R_{min}}{R_{max} – R_{min}} $$
with R_min = 8.1 ms, R_max = 18.0 ms. For fatigue life:
$$ S_F = \frac{F – F_{min}}{F_{max} – F_{min}} $$
with F_min = 0.5×10⁷ cycles (design minimum) and F_max = 24.5×10⁷ cycles (maximum computed).
The overall performance score P was derived via principal component analysis (PCA) with regression weights:
$$ P = 1.953 S_E – 0.844 S_R + 0.287 S_F $$
Note: The negative sign for S_R indicates that faster response (smaller R) yields a higher P, consistent with the normalization. The weights reflect the relative importance: energy absorption is most critical, response time second, and fatigue life third.
To account for cost, we introduced a cost index C based on relative manufacturing and maintenance costs (with δ=2.0 mm as baseline C=3.00). The comprehensive cost-performance index S is:
$$ S = 1.0498 P – 0.02924 C + 0.0928 $$
Table 2 presents the computed S for clearances from 1.0 to 2.0 mm.
| δ (mm) | E (%) | P | C | S |
|---|---|---|---|---|
| 1.0 | 58.00 | 0.89 | 3.65 | 0.92 |
| 1.1 | 60.36 | 0.91 | 3.50 | 0.95 |
| 1.2 | 62.24 | 0.92 | 3.40 | 0.96 |
| 1.3 | 63.64 | 0.93 | 3.30 | 0.97 |
| 1.4 | 64.56 | 0.94 | 3.20 | 0.99 |
| 1.5 | 65.00 | 0.95 | 3.14 | 1.00 |
| 1.6 | 64.96 | 0.94 | 3.10 | 0.99 |
| 1.7 | 64.44 | 0.93 | 3.05 | 0.98 |
| 1.8 | 63.44 | 0.92 | 3.02 | 0.97 |
| 1.9 | 61.96 | 0.91 | 3.01 | 0.96 |
| 2.0 | 60.00 | 0.91 | 3.00 | 0.96 |
2.3 Optimal Clearance Identification
From Table 2, the optimal performance zone lies in the interval [1.4, 1.7] mm, where energy absorption stays above 64.4% and S exceeds 0.98. The absolute optimum occurs at δ = 1.5 mm, achieving the highest performance score (P=0.95), the highest comprehensive index (S=1.00), and peak energy absorption (65%). Response time at 1.5 mm is the fastest (8.1 ms). This point represents the classic “sweet spot” balancing performance, cost, and manufacturability in drone technology.
3. Key Indicator Verification and Engineering Validation
3.1 Multi-Physics Comparison at Three Selected Clearances
We performed detailed comparisons at δ = 1.0, 1.5, and 2.0 mm, as summarized in Table 3.
| Metric | δ=1.0 mm | δ=1.5 mm | δ=2.0 mm | Threshold | Advantage at 1.5 mm |
|---|---|---|---|---|---|
| Thermal stress (MPa) | 35 | 28 | 22 | ≤44 | Balanced; lower than 1.0 mm, sufficient margin |
| Response time (ms) | 10.2 | 8.1 | 12.0 | ≤15 | Fastest; 20% quicker than 1.0 mm, 48% quicker than 2.0 mm |
| Oil film thickness (μm) | 9.3 | 12.3 | 15.1 | ≥8 | Sufficient and efficient; good dynamic damping |
| Fatigue life (10⁷ cycles) | 5.22 | 13.65 | 24.50 | ≥0.5 | Far exceeds threshold; acceptable trade-off |
| Cost index | 3.65 | 3.14 | 3.00 | ≤4.0 | Best cost-performance ratio |
| Manufacturing yield (%) | 98 | 98 | 98 | ≥95 | All equally good; no fabrication drawback |
3.2 Field Validation Results
Using δ = 0.8 mm as a failure baseline (previously causing 18 thermal jamming events per thousand hours), we conducted engineering validation on the actual recovery device. Table 4 presents the results.
| Parameter | δ=1.0 mm | δ=1.5 mm | δ=2.0 mm | δ=0.8 mm | Advantage at 1.5 mm |
|---|---|---|---|---|---|
| Thermal jamming rate (events/10³ h) | 0 | 0 | 0 | 18 | Zero failures; 0.8 mm demonstrates disaster of too‑small gap |
| Slider operating life (h) | 8100 | 9500 | 8700 | 3200 | Longest life; ~1000 h more than alternatives |
| Maintenance cost ($/year) | 4,800 | 4,500 | 4,600 | 9,200 | Lowest cost due to long life and zero failures |
| Energy absorption rate (%) | 58 | 69 | 68 | 25 | Highest; 7% better than 1.0 mm, 5% better than 2.0 mm |
| Comprehensive cost‑performance index | 0.92 | 1.00 | 0.96 | 0.35 | Global optimum; benchmark for all other values |
The field tests confirmed that δ = 1.5 mm eliminates cable detachment and pulley damage entirely, while achieving the lowest maintenance cost and longest slider life. The energy absorption rate (69%) closely matched the fitted value (65%), validating the model. This optimal clearance has been permanently adopted in the production of our drone technology recovery systems.
4. Conclusion
Through systematic multi-physics modeling and comprehensive cost-performance evaluation, we established that the telescopic slider clearance for a UAV rope-hook recovery device should be set to 1.5 mm. This value lies within the theoretical interval [1.0, 2.0] mm derived from thermodynamic, tribological, and dynamic constraints. The 1.5 mm clearance achieves the highest energy absorption rate (65%), fastest response time (8.1 ms), longest slider life (9500 h), and the best comprehensive cost-performance index (S=1.00). The method has been successfully applied in military-trade UAV systems, completely solving the previously catastrophic failures of cable detachment and pulley damage. Our approach provides a reusable technical framework for the fine design of similar recovery systems in drone technology, contributing to safer, more reliable, and more economical UAV operations.