In the field of aero-engine transmission systems, bevel gears play a critical role due to their ability to transmit power between non-parallel shafts. However, the high-speed operational environment of aviation applications imposes severe dynamic loads on these components, leading to vibration issues such as traveling wave resonance. This phenomenon, characterized by diametral modes, can cause excessive vibration stresses, resulting in fatigue failures like tooth breakage or gear cracking, which jeopardize flight safety. For instance, historical incidents have highlighted failures in central drive bevel gears due to three-nodal-diameter traveling wave resonance after limited flight hours. Therefore, implementing effective vibration suppression measures is paramount for enhancing the durability and reliability of aeronautical bevel gears. In this study, we focus on the application of a damping ring, specifically a spiral elastic ring, to mitigate traveling wave resonance in bevel gears. Through a combination of numerical simulations and experimental validation, we analyze the damping effect and verify the feasibility of this approach. The research encompasses modal analysis, friction damping theory, transient dynamic simulations, and vibration testing, with an emphasis on using tables and formulas to summarize key findings. Throughout this work, the term ‘bevel gears’ will be frequently referenced to underscore the central subject of our investigation.
The primary objective is to address traveling wave resonance in bevel gears by integrating a damping ring. We begin by describing the research objects: a pair of central drive bevel gears from an aero-engine, consisting of a cylindrical pinion and a crown-shaped driven gear. The design parameters of these bevel gears are summarized in Table 1. The driven bevel gear, which is the focus of our analysis, operates at a design speed of 20,000 rpm and features a thin web structure that makes it susceptible to vibrational modes. To provide a visual reference, the three-dimensional model of the bevel gears is depicted below, illustrating their geometry and configuration.
| Parameter | Value |
|---|---|
| Number of teeth (pinion) | 47 |
| Number of teeth (driven gear) | 35 |
| Module at large end | 3.875 mm |
| Design speed of driven gear | 20,000 rpm |
| Material density | 7,860 kg/m³ |
| Elastic modulus | 200 GPa |
| Poisson’s ratio | 0.3 |
The damping ring employed is a spiral elastic ring installed in an annular groove on the inner side of the driven bevel gear’s rim. This design leverages elastic deformation to ensure tight contact with the groove, enabling frictional energy dissipation. The parameters of the damping ring are detailed in Table 2. The interaction between the damping ring and the bevel gear is crucial for vibration reduction, as it introduces additional damping into the system.
| Parameter | Value |
|---|---|
| Outer diameter | 115 mm |
| Groove bottom diameter | 114 mm |
| Material density | 7,920 kg/m³ |
| Elastic modulus | 184 GPa |
| Poisson’s ratio | 0.24 |
To understand the damping mechanism, we delve into the theoretical foundation of friction damping. The forced vibration of a multi-degree-of-freedom system can be described by the differential equation:
$$ M\ddot{X} + C\dot{X} + KX = F(t) $$
where \( M \), \( C \), and \( K \) represent the mass, damping, and stiffness matrices, respectively, and \( F(t) \) is the excitation force. For bevel gears, the primary excitation source is the meshing force from gear tooth engagement. When a damping ring is added, the system damping increases, as expressed by:
$$ C = C_0 + C_e $$
Here, \( C_0 \) denotes the structural damping of the bevel gear, and \( C_e \) is the equivalent viscous damping contributed by the damping ring. During resonance, the damping ring dissipates energy through micro-slip friction at the contact interface. The relationship between frictional energy dissipation and equivalent damping is given by:
$$ C_e = \frac{4W_{\text{all}}}{\pi \omega_n A_m^2} $$
where \( W_{\text{all}} \) is the energy dissipated per cycle, \( \omega_n \) is the natural frequency, and \( A_m \) is the amplitude. The energy dissipation \( W_{\text{all}} \) depends on contact pressure \( p \), tangential stiffness \( k_t \), and tangential displacement amplitude \( V_t \). For a circular contact, the energy dissipation can be integrated over the circumference:
$$ W_{\text{all}} = \int_0^{2\pi} 4 \left[ p \mu V_t – \frac{(p \mu)^2}{k_t} \right] ds $$
where \( \mu \) is the friction coefficient. This model indicates that sliding friction occurs when \( k_t V_t > p \mu \), leading to energy dissipation. Thus, the damping ring effectively reduces vibration amplitudes in bevel gears by converting mechanical energy into heat through friction.
We proceed with numerical simulations to analyze the dynamic behavior of the bevel gear. First, modal analysis is conducted using finite element methods. The bevel gear model is constrained appropriately: radial and axial displacements are fixed at the bearing fit surface, and axial displacement is constrained at the spline. The mesh is generated, and natural frequencies are computed. The traveling wave resonance modes for the first four nodal diameters are identified, as summarized in Table 3. These modes are critical because they can be excited within the operating speed range of the bevel gears.
| Nodal Diameter | Frequency (Hz) | Forward Wave Speed (rpm) | Backward Wave Speed (rpm) |
|---|---|---|---|
| 1 | 3,148 | 4,837 | 4,568 |
| 2 | 3,779 | 6,805 | 6,069 |
| 3 | 7,028 | 13,361 | 11,252 |
| 4 | 11,567 | 22,740 | 18,075 |
The resonance speeds are derived from the relationship:
$$ N = \frac{60 f_d}{Z \pm n} $$
where \( N \) is the resonance speed, \( f_d \) is the frequency, \( Z \) is the number of teeth, and \( n \) is the nodal diameter number. The plus sign corresponds to backward traveling waves, and the minus sign to forward traveling waves. A Campbell diagram is constructed to visualize these resonances, showing that the 4-nodal-diameter backward wave occurs at 18,075 rpm, which is 90.4% of the design speed, marking it as a dangerous mode for the bevel gears.
Next, we analyze the contact pressure between the damping ring and the bevel gear. This pressure consists of assembly pressure due to elastic deformation and centrifugal pressure during rotation. At the 4-nodal-diameter backward wave resonance speed, the contact pressure is computed to be 4,697 N. The elastic deformation of the spiral ring is also evaluated, confirming its tight fit. The meshing force, as the excitation source, is calculated through loaded tooth contact analysis. By incorporating tooth modifications, the contact pattern is optimized to center on the tooth flank. The dynamic meshing load, including effects of mesh impact and transmission error, is derived for the driven bevel gear. The continuous meshing load over multiple cycles is represented by a time-domain function, which is used as input for transient dynamics simulations.
For vibration stress analysis, transient dynamic simulations are performed using the modal superposition method. The meshing force is applied sequentially to tooth flanks, and the damping effect is modeled by increasing the modal damping ratio from 0.07% (without damping ring) to 0.30% (with damping ring), assuming a friction coefficient of 0.1. The resonance stress at 18,075 rpm is computed, focusing on the 4-nodal-diameter backward wave mode. The results show that without the damping ring, the maximum vibration stress at a monitored node (near strain gauge location) is 177.3 MPa. With the damping ring, this stress reduces to 104.3 MPa. A frequency-domain analysis via Fourier transform reveals that the 39th harmonic, corresponding to the 4-nodal-diameter mode, has an amplitude of 69.9 MPa without damping and 32.3 MPa with damping, indicating a reduction of 53.8%.
To validate these simulations, we conduct experimental vibration testing on the bevel gears. The test rig includes a lubrication system, loading system, and telemetry for data acquisition. Twelve strain gauges are attached tangentially along the rim of the driven bevel gear, close to the tooth roots, to measure vibration strain. The data collection process involves transmitting signals from rotating strain gauges via telemetry to a data acquisition system. A sweep test is performed by gradually increasing the speed from 0 to beyond the resonance range. The waterfall plot obtained from the test clearly shows peaks at the 3- and 4-nodal-diameter traveling wave resonances. The measured resonance speeds are compared with calculations in Table 4, demonstrating minimal deviation and confirming the accuracy of the modal analysis for bevel gears.
| Resonance Mode | Measured Speed (rpm) | Calculated Speed (rpm) | Error (%) |
|---|---|---|---|
| 3-nodal-diameter backward wave | 11,356 | 11,252 | -0.92 |
| 3-nodal-diameter forward wave | 13,320 | 13,361 | 0.31 |
| 4-nodal-diameter backward wave | 18,084 | 18,075 | -0.05 |
| 4-nodal-diameter forward wave | 22,510 | 22,740 | 1.02 |
The effect of the damping ring on resonance speeds is negligible, with a maximum change of 2.52%, as shown in Table 5. This implies that the damping ring primarily affects vibration amplitude rather than altering the modal characteristics of the bevel gears.
| Resonance Mode | Frequency without Ring (Hz) | Frequency with Ring (Hz) | Change (%) |
|---|---|---|---|
| 3-nodal-diameter backward wave | 7,192 | 7,118 | -1.03 |
| 3-nodal-diameter forward wave | 7,104 | 6,925 | -2.52 |
| 4-nodal-diameter backward wave | 11,755 | 11,649 | -0.90 |
| 4-nodal-diameter forward wave | 11,630 | 11,460 | -1.46 |
For vibration stress comparison, we extract the 39th harmonic amplitude from the experimental data. Without the damping ring, the maximum measured strain is \(356 \times 10^{-6}\), corresponding to a stress of 71.2 MPa (using Hooke’s law: \( \sigma = E \epsilon \), with \( E = 200 \) GPa). With the damping ring, the strain reduces to \(151 \times 10^{-6}\), or 30.2 MPa, indicating a damping effectiveness of 57.6%. The simulation results for the same harmonic are 69.9 MPa (without ring) and 32.3 MPa (with ring), yielding errors of -1.8% and 7.0%, respectively, as summarized in Table 6. This close agreement validates the numerical simulation methodology for predicting vibration stresses in bevel gears with damping rings.
| Condition | Measured Stress (MPa) | Calculated Stress (MPa) | Error (%) |
|---|---|---|---|
| Without damping ring | 71.2 | 69.9 | -1.8 |
| With damping ring | 30.2 | 32.3 | 7.0 |
The damping performance can be further analyzed using a dimensionless parameter, the damping ratio \( \zeta \), which is related to the equivalent viscous damping. For the bevel gear system, the damping ratio with the ring is estimated from the stress reduction. The logarithmic decrement \( \delta \) can be derived from the amplitude ratio:
$$ \delta = \ln\left(\frac{A_1}{A_2}\right) $$
where \( A_1 \) and \( A_2 \) are amplitudes before and after damping. Using the stress values, \( \delta \approx \ln(71.2 / 30.2) = 0.86 \). Then, the damping ratio is approximately \( \zeta = \delta / (2\pi) = 0.137 \), indicating significant energy dissipation. This aligns with the increased modal damping ratio used in simulations.
In conclusion, our study demonstrates the effectiveness of a spiral elastic damping ring in suppressing traveling wave resonance in aeronautical bevel gears. Key findings include: (1) The driven bevel gear exhibits the first four nodal diameter traveling wave modes within its operating range, with the 4-nodal-diameter backward wave resonance at 18,084 rpm posing a high risk. Modal analysis showed excellent accuracy with errors as low as 0.05%. (2) The damping ring has minimal impact on the resonance speeds of bevel gears, with changes under 2.52%, confirming that it does not alter the fundamental dynamics. (3) The damping ring achieves a substantial reduction in vibration stress—57.6% experimentally and 53.8% numerically—for the critical 4-nodal-diameter mode, validating the simulation approach with errors within 7.0%. (4) The friction damping mechanism, based on energy dissipation through micro-slip, is effectively captured by theoretical models and finite element analysis. For future work, we recommend exploring different damping ring materials, geometries, and assembly methods to optimize performance, along with fatigue testing to assess long-term benefits for bevel gears. This research underscores the importance of integrated simulation and testing in advancing vibration control strategies for high-speed bevel gears in aviation applications.
To further elaborate on the theoretical aspects, we can derive the equations of motion for a simplified model of a bevel gear with a damping ring. Consider a single-degree-of-freedom system representing a dominant mode of the bevel gear. The equation is:
$$ m\ddot{x} + c\dot{x} + kx = F_0 \cos(\omega t) $$
where \( m \) is the modal mass, \( c \) is the damping coefficient (including contributions from the ring), \( k \) is the stiffness, and \( F_0 \) is the amplitude of the meshing force. The steady-state amplitude \( X \) is given by:
$$ X = \frac{F_0}{\sqrt{(k – m\omega^2)^2 + (c\omega)^2}} $$
At resonance, \( \omega = \omega_n = \sqrt{k/m} \), so:
$$ X_{\text{res}} = \frac{F_0}{c \omega_n} $$
This shows that increasing \( c \) through the damping ring reduces the resonance amplitude. For the bevel gear, \( c \) is enhanced by the equivalent damping \( C_e \) from the ring, as previously discussed.
Additionally, the contact mechanics between the damping ring and bevel gear can be modeled using Hertzian contact theory. The contact pressure \( p \) for a curved surface is:
$$ p = \frac{2E^*}{\pi} \sqrt{\frac{\delta}{R}} $$
where \( E^* \) is the equivalent elastic modulus, \( \delta \) is the interference fit, and \( R \) is the effective radius. This pressure influences the frictional energy dissipation. For the spiral ring, the interference fit varies with speed due to centrifugal effects, which we accounted for in simulations.
In terms of experimental setup, the strain gauge configuration on the bevel gear allowed for capturing circumferential strain variations. The strain \( \epsilon \) is related to stress \( \sigma \) via \( \sigma = E \epsilon \), and for multi-axial states, we use von Mises stress for fatigue assessment. However, in this study, we focused on principal stresses from vibration modes. The data acquisition system sampled at a rate sufficient to resolve high-frequency components up to the 50th harmonic of the rotational speed, ensuring accurate detection of traveling wave resonances in bevel gears.
The numerical simulations involved extensive finite element modeling. We used quadratic tetrahedral elements for the bevel gear mesh, with refinement near the tooth roots and rim where stress concentrations occur. The damping ring was modeled as a separate body with frictional contact defined by a penalty method. The friction coefficient was assumed constant at 0.1, based on typical values for steel-steel interfaces with lubrication. Transient analysis covered multiple mesh cycles to achieve steady-state response, and results were post-processed to extract stress histories and frequency spectra.
Overall, this comprehensive approach—combining theory, simulation, and experiment—provides a robust framework for analyzing and mitigating vibrations in bevel gears. The success of the damping ring highlights the potential for passive damping solutions in aero-engine transmissions, contributing to enhanced safety and longevity of bevel gears. Future investigations could integrate advanced materials like shape memory alloys for adaptive damping or explore active control methods, but for now, the spiral elastic ring offers a practical and effective means of addressing traveling wave resonance in high-speed bevel gears.