This work presents the comprehensive design, structural analysis, and optimization of a critical tilting mechanism for a 30 kg class rotary wing unmanned aerial vehicle (RWUAV). Combining the vertical take-off and landing (VTOL) capabilities of rotorcraft with the efficient high-speed cruise performance of fixed-wing aircraft, this rotary wing unmanned aerial vehicle configuration offers significant operational flexibility. The core innovation lies in the tilting mechanism enabling the rapid and synchronized transition of the forward rotors between helicopter and airplane modes. The design employs a worm gear mechanism driving a torsion bar connected to both rotor nacelles, ensuring precise synchronization. Structural integrity verification and subsequent lightweight optimization using advanced response surface methodologies are detailed, resulting in a significant mass reduction while maintaining operational safety.
1 Introduction
The rotary wing unmanned aerial vehicle, specifically the tiltrotor variant, represents a transformative advancement in unmanned aerial systems (UAS). It eliminates the need for runways, enabling VTOL operations like traditional rotorcraft, while achieving the higher speeds, extended range, and broader flight envelopes characteristic of fixed-wing aircraft during cruise. This dual capability makes the rotary wing unmanned aerial vehicle exceptionally suited for diverse missions, including military logistics, reconnaissance, medical evacuation, cargo transport, surveying, and aerial photography. The pivotal component enabling this mode transition is the tilting mechanism responsible for rotating the rotor nacelles and their associated propulsion units.
Existing tilting mechanisms often utilize various actuation principles – servo motors with gear trains, chain drives, linkages, or worm gears. While functional, common drawbacks include relatively long transition times (typically several seconds), complex aerodynamic flow fields during transition challenging to analyze and control, susceptibility to reversal forces, and weight inefficiencies. The primary objective of this work was to design a robust, lightweight tilting mechanism capable of an extremely rapid 90-degree tilt (0.5 seconds) for a 30 kg RWUAV. This rapid transition minimizes the impact of complex transitional aerodynamics and control issues. Furthermore, a significant focus was placed on the structural validation and subsequent mass optimization of the torsion bar, identified as a critical and highly loaded component within the mechanism.
2 RWUAV Overall Design
2.1 Configuration Layout
The designed rotary wing unmanned aerial vehicle features a tailless, blended-wing-body configuration. This integrated design enhances the overall aerodynamic efficiency by creating a streamlined shape where the entire fuselage contributes to lift, resulting in a high lift-to-drag ratio and reduced drag during cruise. The landing gear is housed within the inboard wing sections. Propulsion is provided by four rotors:
- Two forward rotors are tiltrotors, mounted on nacelles capable of rotating through 90 degrees.
- Two aft rotors are fixed, providing propulsion solely in fixed-wing mode.
The forward rotor nacelles are rigidly connected to opposite ends of a central torsion bar. Lift forces generated by these rotors are transmitted directly to the airframe through this torsion bar. Crucially, the tilting torque applied by the actuation system rotates the torsion bar, which simultaneously and synchronously tilts both forward nacelles. The overall layout is conceptualized as a streamlined, tailless flying wing with integrated propulsion units.
2.2 Propulsion System Design
The propulsion system comprises brushless DC motors (BLDC), electronic speed controllers (ESC), and propellers. Each motor drives a propeller to generate thrust perpendicular to its plane of rotation. The ESC regulates the power delivered to each motor based on commands from the flight control system. Key propulsion specifications are summarized below:
- Takeoff Mass: 30 kg
- Required Lift per Rotor (Hover): Minimum 75 N (considering maneuvering and wind resistance, maximum lift requirement per rotor is 150 N).
- Tiltrotor Motors: Scorpion SII-6530-180 kV BLDC Motors (Power: 4.8 kW, Nominal Voltage: 44.4 V, KV: 180, Nominal RPM: 7992 rpm).
- Fixed Rotor Motors: T-MOTOR MN805S BLDC Motors (KV: 170).
- Propellers: 2-blade, Diameter: 26 inches, Pitch: 10 inches.
- Battery: 12S (44.4V nominal), 30 Ah Lithium Polymer (LiPo).
- Operation Modes:
- Rotor Mode (VTOL/Hover): All four motors active.
- Fixed-Wing Mode (Cruise): Only the two forward tiltrotor motors active; aft fixed motors stopped. The forward motors are tilted to 0 degrees by the tilting mechanism.
The thrust TT generated by a propeller is empirically modeled as:T=0.00025⋅d⋅L⋅N2⋅PT=0.00025⋅d⋅L⋅N2⋅P
where:
- dd is the propeller diameter (in),
- LL is the propeller pitch (in),
- NN is the rotational speed (RPM),
- PP is the atmospheric pressure (Pa),
- 0.000250.00025 is an empirical constant.
Thrust testing confirmed the model’s validity, showing measured thrust values within 10% of calculated values across the operational RPM range. At the motor’s rated power, sufficient thrust is generated to meet the maximum 150 N requirement per rotor with a safety margin.
3 Tilting Mechanism Design
3.1 Dynamic Load Analysis
The primary function of the tilting mechanism is to rotate the forward rotor nacelles and their attached motors/propellers through 90 degrees in 0.5 seconds. A critical challenge is managing the gyroscopic torque generated by the spinning propellers during this rapid angular acceleration. Treating the propeller as a rigid body undergoing combined rotation (spin ω1ω1) and precession (tilt rate ω2ω2) around the torsion bar axis, the absolute angular velocity θθ is:θ=ω1+ω2θ=ω1+ω2
The angular momentum LL of the propeller about the torsion bar pivot point is:L=IzθL=Izθ
where IzIz is the propeller assembly’s moment of inertia about its spin axis. According to the theorem of angular momentum, the external torque MM acting on the propeller is:M=dLdt=ω2×L=ω2×(Izθ)M=dtdL=ω2×L=ω2×(Izθ)
Given that the spin axis ω1ω1 and the precession axis ω2ω2 are perpendicular during the tilt initiation, the magnitude of the gyroscopic torque TgTg opposing the tilting motion is:Tg=∣M∣=Izω1ω2Tg=∣M∣=Izω1ω2
Using CATIA modeling, the moment of inertia IzIz of the propeller assembly (motor + propeller + nacelle structure) was determined to be 0.025 kg·m². The required tilt angular velocity ω2ω2 is:ω2=π/2 rad0.5 s=π rad/s≈3.1416 rad/sω2=0.5 sπ/2 rad=π rad/s≈3.1416 rad/s
The nominal propeller spin speed ω1ω1 is:ω1=2π×799260≈837.0 rad/sω1=602π×7992≈837.0 rad/s
Therefore, the gyroscopic torque TgTg that the tilting mechanism must overcome for each rotor is:
This torque is a significant load on the mechanism, particularly the torsion bar.
3.2 Actuation System Selection and Design
The key requirements driving the actuation system design were:
- Ability to generate sufficient torque to overcome inertial loads and gyroscopic effects.
- Prevention of reverse rotation due to aerodynamic or gyroscopic forces acting on the rotors (self-locking).
- Precision and synchronization for both rotors.
- Minimization of mass.
Worm gear drives offer inherent self-locking capability when the lead angle is sufficiently small, preventing unwanted reverse rotation of the rotors – a critical safety feature. This made it the preferred choice over other transmission types like spur gears or linkages, despite potential efficiency trade-offs.
- Motor Selection: A 57-series stepper motor (holding torque: 2.4 N·m) was chosen. Applying a safety factor of 2, the usable output torque is 1.2 N·m.
- Worm Gear Design: A high reduction ratio is needed to amplify the motor torque. A 4-start worm meshing with a 32-tooth worm gear provides a reduction ratio of i=32/4=8i=32/4=8. The torque output from the worm gear is:Tworm gear=Tmotor×i×ηTworm gear=Tmotor×i×ηAssuming an efficiency ηη of ~70% for a multi-start worm, the output torque is approximately . This torque is sufficient to drive the torsion bar and overcome the gyroscopic loads distributed through it.
- Worm Gear Implementation: To minimize mass and limit the rotation range precisely to 90 degrees, the worm gear segment was designed as a 180-degree arc. Lightening holes were machined into the segment, and limit screws were installed on the gear face to physically restrict its travel relative to the worm.
- Torsion Bar Integration: The worm gear segment is centrally mounted on the torsion bar. The torsion bar rotates within bearings supported by rigid mounts attached to the airframe structure. The ends of the torsion bar are connected directly to the rotor nacelle supports. The stepper motor, driving the worm via a coupling, is mounted on the airframe. Rotating the worm causes the worm gear segment (and thus the torsion bar) to pivot, synchronously tilting both connected rotor nacelles. The core components include the mounting plate, bearing blocks, bearings, worm, worm gear segment, torsion bar supports, torsion bar, motor mount, stepper motor, and couplings.
4 Torsion Bar Structural Analysis
The torsion bar is the most critically loaded and complex component within the tilting mechanism (“vulnerable part”). It simultaneously transmits:
- Lift forces (FL1,FL2FL1,FL2) from both rotors.
- Reaction torques (TR1,TR2TR1,TR2) from the rotors resisting spin.
- Drive torque (TDTD) from the worm gear.
- Gyroscopic torques (TG1,TG2TG1,TG2) during tilting.
The maximum combined stress state occurs at the initiation of the tilting maneuver.
4.1 Material Selection
T300 Carbon Fiber Reinforced Polymer (CFRP) was chosen for the torsion bar due to its exceptional specific strength (strength-to-density ratio) and specific stiffness (stiffness-to-density ratio), crucial for lightweight aerospace structures. Its material properties are listed below:
Table 1: T300 CFRP Material Properties
| Property | Symbol | Value | Unit |
|---|---|---|---|
| Poisson’s Ratio | ν | 0.33 | – |
| Density | ρ | 1.79 | g/cm³ |
| Elastic Modulus | E | 120 | GPa |
| Yield Strength (Tensile) | σ_y | 493 | MPa |
4.2 Finite Element Analysis (FEA)
Static structural analysis was performed using the ANSYS Workbench Static Structural module. The torsion bar geometry was imported, meshed with suitable elements (tetrahedral or hex-dominant), and T300 CFRP properties assigned. Boundary conditions simulated the physical constraints:
- The regions where the torsion bar passes through the bearing supports were modeled as fixed supports (cylindrical or fixed geometry), representing the rigid connection to the airframe via the bearings.
- All external loads (Lift forces, Reaction Torques, Drive Torque, Gyroscopic Torques) calculated for the worst-case scenario (tilt initiation) were applied to their respective attachment points on the torsion bar model.
The FEA solved for deformation and stress distribution under these loads. Key results for the initial torsion bar design were:
- Maximum Deformation: 0.583 mm.
- Maximum Equivalent (von Mises) Stress: 55.783 MPa.
4.3 Structural Assessment
The maximum equivalent stress (55.783 MPa) was significantly lower than the material’s yield strength (493 MPa). The calculated safety factor SFSF is:SF=σyσmax=49355.783≈8.84SF=σmaxσy=55.783493≈8.84
This high safety factor confirms the initial torsion bar design possesses substantial structural strength reserves. The maximum deformation (0.583 mm) was also deemed acceptable, unlikely to interfere with the precise meshing of the worm gear mechanism or the overall tilting kinematics. While structurally sound, the significant safety margin indicated a clear opportunity for mass reduction through structural optimization.
5 Torsion Bar Optimization
The primary objective was to minimize the mass of the torsion bar while ensuring structural integrity under operational loads. The constraints were maintaining maximum equivalent stress below the material’s allowable stress (conservatively set based on yield strength) and limiting maximum deformation below a permissible threshold to ensure mechanism functionality. Response Surface Methodology (RSM), integrated within ANSYS Workbench, was employed for this optimization task.
5.1 Optimization Problem Formulation
The optimization was mathematically defined as:minimizem(do,di)subject toσeqv, max(do,di)≤σallowδmax(do,di)≤δallowdoLB≤do≤doUBdiLB≤di≤diUBdi<dominimizesubject tom(do,di)σeqv, max(do,di)≤σallowδmax(do,di)≤δallowdoLB≤do≤doUBdiLB≤di≤diUBdi<do
Where:
- mm is the mass of the torsion bar (objective function).
- σeqv, maxσeqv, max is the maximum equivalent (von Mises) stress.
- δmaxδmax is the maximum deformation.
- σallowσallow is the allowable stress (derived from material yield strength with safety factor).
- δallowδallow is the allowable deformation.
- dodo is the outer diameter of the torsion bar (design variable).
- didi is the inner diameter of the torsion bar (design variable, defining a hollow tube).
- doLB,doUB,diLB,diUBdoLB,doUB,diLB,diUB define the lower and upper bounds for the design variables.
5.2 Design of Experiments (DoE)
Building an accurate Response Surface Model (RSM) requires strategically sampling the design space defined by dodo and didi. The Central Composite Design (CCD) method was chosen. CCD efficiently samples the space using:
- Factorial Points (Cube Points): Points at the corners of the design variable range (2^2 = 4 points).
- Axial Points (Star Points): Points along the axes outside the factorial range (2*2 = 4 points).
- Center Point: Point at the center of the design space (1 point).
This combination provides good coverage for fitting a quadratic response surface model. ANSYS Workbench automatically generated the required 9 design points based on the defined variable bounds.
5.3 Response Surface Model (RSM) Construction
For each design point generated by the CCD, ANSYS Workbench automatically executed an FEA simulation to compute the mass, maximum stress, and maximum deformation. These results were used to construct surrogate models (RSM) predicting these outputs for any combination of dodo and didi within the bounds. Two common RSM types were evaluated:
- Standard Quadratic Polynomial (SQP): Models the response y^y^ as:y^=β0+β1do+β2di+β3do2+β4di2+β5dodiy^=β0+β1do+β2di+β3do2+β4di2+β5dodiwhere β0,β1,…,β5β0,β1,…,β5 are regression coefficients.
- Kriging Model: A more sophisticated interpolation method providing a global approximation plus localized deviations:y^(x)=f(x)Tβ+Z(x)y^(x)=f(x)Tβ+Z(x)where f(x)Tβf(x)Tβ is a global trend model (often polynomial), and Z(x)Z(x) is a Gaussian process representing local deviations based on spatial correlation.
5.4 RSM Validation and Selection
The accuracy of the fitted RSMs was assessed using the Coefficient of Determination (R2R2). R2R2 values close to 1.0 indicate the model explains almost all the variation in the FEA data.
- SQP Model: Achieved R2R2 values > 0.98 for mass, stress, and deformation, indicating a good fit.
- Kriging Model: Achieved R2R2 = 1.0 for all three responses, indicating a perfect fit to the sampled FEA data points.
Given its superior fit accuracy, the Kriging model was selected for the subsequent optimization phase. A plot of Kriging model predictions versus actual FEA results for all responses confirmed the points lay almost perfectly on the diagonal line, validating the model’s predictive capability.
5.5 Optimization Execution and Results
Using the Kriging RSM as a computationally inexpensive surrogate for the actual FEA, a multi-objective genetic algorithm (MOGA) within ANSYS Workbench was employed to search the design space for the optimum (do,di)(do,di) pair minimizing mass while satisfying the stress and deformation constraints. The MOGA efficiently explored numerous potential designs based on the RSM predictions.
The optimization process yielded a candidate design point. Minor adjustments were made to this point to conform to standard manufacturing sizes (e.g., rounding diameters to nearest 0.5 mm or 1.0 mm), resulting in the final optimized dimensions. A final FEA verification was performed on this manufacturable design. The results comparing the initial and optimized torsion bars are presented below:
Table 2: Torsion Bar Initial vs. Optimized Design Comparison
| Parameter | Initial Design | Candidate Design | Final Optimized Design | Unit | Change (%) |
|---|---|---|---|---|---|
| Outer Diameter (dodo) | 40.000 | 36.836 | 37.000 | mm | -7.5% |
| Inner Diameter (didi) | 25.000 | 26.944 | 27.000 | mm | +8.0% |
| Max. Equiv. Stress | 55.783 | 102.380 | 103.610 | MPa | +85.7% |
| Max. Deformation | 0.583 | 1.401 | 1.409 | mm | +141.7% |
| Mass | 1.4529 | 0.9453 | 0.9686 | kg | -33.3% |
Key Optimization Outcomes:
- Mass Reduction: Achieved a significant 33.3% mass reduction (from 1.4529 kg to 0.9686 kg).
- Stress: The maximum equivalent stress increased to 103.61 MPa but remains well below the T300 CFRP yield strength (493 MPa), maintaining a high safety factor (SF≈493/103.61≈4.76SF≈493/103.61≈4.76).
- Deformation: The maximum deformation increased to 1.409 mm but was confirmed to be within the permissible limit (δallowδallow) defined to ensure no interference with the worm gear operation or overall mechanism function.
- Design Change: The optimization effectively increased the inner diameter (making the tube thinner-walled) and slightly decreased the outer diameter, significantly reducing cross-sectional area and thus mass, while still meeting strength and stiffness requirements under the applied loads.
Final FEA stress and deformation contour plots for the optimized torsion bar confirmed the predictions: stress was highest near the bearing support regions and central worm gear attachment point, and deformation was highest at the ends (nacelle attachment points), with minimal deflection in the central section critical for worm gear engagement.
6 Conclusions
This work detailed the complete design and engineering process for a high-performance tilting mechanism enabling rapid transition in a 30 kg rotary wing unmanned aerial vehicle. The key achievements and findings are:
- Rapid and Synchronized Tilting: A novel mechanism utilizing a worm gear drive coupled with a central torsion bar was successfully designed and engineered. The worm gear provides essential self-locking to prevent unintended rotor reversal. The torsion bar inherently ensures perfect synchronization of both forward rotor nacelles during the 90-degree tilt. The mechanism achieves an exceptionally fast transition time of 0.5 seconds, effectively mitigating the complexities associated with transitional aerodynamics and control.
- Robust Structural Design: Comprehensive structural analysis using ANSYS Workbench confirmed the initial design, particularly the critical torsion bar component manufactured from T300 CFRP, possessed substantial strength reserves (Safety Factor > 8) under the demanding combined loads (lift, reaction torque, drive torque, gyroscopic torque). Deformation was also within acceptable limits for mechanism functionality.
- Effective Lightweight Optimization: Leveraging the identified structural safety margin, a rigorous optimization process was implemented. Employing Central Composite Design (CCD) for sampling and Kriging interpolation to build highly accurate Response Surface Models (RSM), a Multi-Objective Genetic Algorithm (MOGA) successfully identified an optimized torsion bar design. The final design, verified by FEA, achieved a 33.3% mass reduction (from 1.4529 kg to 0.9686 kg) while maintaining maximum stress (103.61 MPa) safely below the material yield strength and keeping deformation within the permissible functional limit. This optimization significantly enhances the rotary wing unmanned aerial vehicle‘s payload capacity and flight performance.
- Validated Methodology: The integration of parametric CAD, FEA, Design of Experiments, Kriging RSM, and MOGA optimization within the ANSYS Workbench environment proved highly effective for the structural design and mass minimization of this complex aerospace component.
The developed tilting mechanism provides a robust, lightweight, and high-performance solution enabling the unique VTOL-to-cruise transition capability of the tiltrotor rotary wing unmanned aerial vehicle. The successful application of advanced structural optimization techniques demonstrates a pathway for further mass savings in critical aerospace structures. Future work could explore topology optimization of support brackets, advanced composite layup optimization for the torsion bar, dynamic analysis of the full tilt transient, and experimental validation of the mechanism’s performance under load. This rotary wing unmanned aerial vehicle platform, enabled by this optimized tilting mechanism, holds significant potential for expanding the operational scope of unmanned aerial systems.