In modern industrial applications, the demand for high-performance cylindrical gears with enhanced surface integrity, longevity, fatigue resistance, and reliability is paramount. Surface finishing techniques play a critical role in achieving these properties. Among various methods, spindle barrel finishing stands out due to its ability to reduce surface roughness, refine surface texture, improve residual stress, and efficiently process complex-shaped components like cylindrical gears at low cost. This process involves the interaction of abrasive media particles with the workpiece within a rotating barrel, leading to material removal and surface enhancement. However, the fundamental mechanisms at the particle-cylindrical gear interface remain poorly understood, hindering optimal process prediction. In this study, we employ the Discrete Element Method (DEM) to simulate spindle barrel finishing for cylindrical gears, aiming to elucidate particle action behaviors, including flow dynamics, contact forces, and relative velocities on the tooth surfaces. We investigate the effects of key parameters such as gear embedment depth and rotational speeds, validate findings through experimental stress testing and roughness measurements, and provide insights for process optimization. The cylindrical gear, a central component in power transmission systems, requires precise finishing to ensure smooth operation and durability.
The spindle barrel finishing process consists of a cylindrical barrel filled with abrasive media particles, where the cylindrical gear workpiece is mounted on a fixture that rotates simultaneously with the barrel. This dual rotation generates relative motion and interaction forces between the particles and the cylindrical gear tooth surfaces, resulting in surface finishing. The complexity arises from the dynamic particle flow field and the semi-enclosed geometry of gear teeth, which influence contact behaviors. To analyze these phenomena, DEM simulations offer a powerful tool by modeling individual particles and their interactions based on contact mechanics. We establish a DEM model with a cylindrical gear of module m = 5 mm, tooth number z = 23, face width b = 40 mm, and pressure angle 20°, using spherical brown alumina particles of 3 mm diameter. The material intrinsic and contact parameters are summarized in Tables 1 and 2, respectively. The contact model employs the Hertz-Mindlin no-slip model with Archard wear theory, where the wear depth Δh on the cylindrical gear surface is proportional to the normal pressure P and relative velocity v:
$$ \Delta h = \frac{K P v}{H} \Delta t $$
Here, K is the dimensionless wear coefficient, H is the hardness of the cylindrical gear material, and Δt is the wear time. This equation underscores the importance of contact force and relative velocity in material removal, guiding our investigation into how process parameters affect these factors for cylindrical gears.
| Material | Density ρ (kg/m³) | Poisson’s Ratio ε | Shear Modulus E (MPa) |
|---|---|---|---|
| Barrel (Steel) | 7850 | 0.300 | 7940 |
| Particles (Brown Alumina) | 2675 | 0.360 | 1260 |
| Cylindrical Gear (40 Cr) | 7870 | 0.277 | 8080 |
| Interaction | Coefficient of Restitution | Static Friction Coefficient μ₁ | Rolling Friction Coefficient μ₂ |
|---|---|---|---|
| Particle-Barrel | 0.50 | 0.35 | 0.10 |
| Particle-Cylindrical Gear | 0.43 | 0.36 | 0.10 |
| Particle-Particle | 0.46 | 0.39 | 0.10 |
We design single-factor simulations to explore the effects of gear embedment depth h₁ (the distance from the gear’s upper face to the static particle upper interface) and rotational speeds of the barrel n₁ and cylindrical gear n₂, with a fixed speed ratio n₁:n₂ = 5:4. The simulation parameters are listed in Table 3. The relative motion velocity V between particles and the cylindrical gear surface in an ideal state can be expressed as:
$$ V = 2\pi n_1 \left[ r^2 \left(1 – \frac{n_2}{n_1}\right)^2 + R^2 + 2Rr \left(1 – \frac{n_2}{n_1}\right) \cos\theta \right]^{1/2} $$
where R is the center distance between the barrel and cylindrical gear, r is the distance from a point on the tooth surface to the gear axis, and θ is the angle between the line connecting this point to the gear axis and the line connecting the centers of the barrel and cylindrical gear. This formula highlights the dependency of velocity on rotational speeds, which we examine in detail for cylindrical gears.
| Gear Embedment Depth h₁ (mm) | Barrel Rotational Speed n₁ (rpm) |
|---|---|
| 80, 110, 140 | 12, 21, 30 |
The DEM simulations reveal intricate particle flow fields around the cylindrical gear. In the absence of the workpiece, the particle upper interface forms a parabolic surface due to barrel rotation. However, with the cylindrical gear inserted, particle flow is disrupted, leading to accumulation in front of the gear and a cavity in the wake region. For instance, at h₁ = 80 mm and n₁ = 30 rpm, the particle accumulation height in front of the cylindrical gear reaches up to 192.23 mm, compared to the static height of 140.00 mm, with a maximum difference of 70.00 mm between front and rear regions. Velocity vector analysis in different planes (e.g., xz, yz, xy) shows that particles impacting the cylindrical gear front region lose kinetic energy, partly converting to potential energy for climbing and internal energy for friction. Particles flow upward along the tooth surfaces, move laterally around the gear, and accelerate downward in the wake. This flow pattern is crucial for understanding how particles interact with the cylindrical gear tooth surfaces.
Focusing on the tooth surface contact particles, we observe a periodic action behavior on the cylindrical gear, divided into three stages: particle filling, stable filling, and particle outflow. During the filling stage, particles flow into the tooth space from the gear top and tip, with velocities exceeding 0.10 m/s in a downward direction. In the stable filling stage, particles fully occupy the tooth space, and their motion shifts to upward sliding along the tooth surface due to the gear’s obstruction in the impact zone, with velocities ranging from 0.01 to 0.05 m/s. This stage is the primary period for particle action on the cylindrical gear. In the outflow stage, particles exit the tooth space under centrifugal force and gravity, moving toward the gear tip and bottom face. The number of contact particles and normal contact force vary cyclically, as illustrated in Figures 1 and 2 for different parameters. The stable filling stage exhibits significantly higher average normal contact force—22.45 times that of the filling stage and 26.24 times that of the outflow stage—emphasizing its dominance in material removal for cylindrical gears.
We quantify the effects of parameters on normal contact force and relative velocity for the cylindrical gear. As shown in Table 4, increasing the embedment depth h₁ from 80 mm to 140 mm (a 75% increase) raises the average normal contact force on the tooth surface by 76%, from approximately 0.15 N to 0.264 N, while the average relative velocity increases only by 4%, from 0.020 m/s to 0.0208 m/s. Conversely, increasing the rotational speeds n₁ and n₂ from 12 rpm to 30 rpm (a 150% increase) boosts the average relative velocity by 148%, from 0.012 m/s to 0.0298 m/s, but the normal contact force rises by only 18%, from 0.127 N to 0.15 N. These trends highlight that embedment depth primarily influences contact force, whereas rotational speeds dominate relative velocity for cylindrical gears. The data can be summarized with the following empirical relationships derived from simulations:
$$ F \propto h_1^{1.01} \quad \text{and} \quad V \propto n_1^{1.49} $$
where F is the average normal contact force and V is the average relative velocity on the cylindrical gear tooth surface.
| Parameter Change | Change Magnitude | Average Normal Contact Force Change | Average Relative Velocity Change |
|---|---|---|---|
| Embedment Depth h₁: 80 mm → 140 mm | +75% | +76% | +4% |
| Rotational Speeds n₁, n₂: 12 rpm → 30 rpm | +150% | +18% | +148% |
The force distribution on the cylindrical gear tooth surface is non-uniform. The upper tooth surface experiences 1.52 to 1.88 times higher contact force than the lower tooth surface, and the upper tooth surface has 1.35 to 1.45 times greater particle relative velocity. This asymmetry stems from the semi-enclosed tooth geometry and flow dynamics, where particles in the upper region face more direct impact and accumulation. Along the axial direction, the force near the gear bottom face is 1.15 to 1.01 times that near the top face, but increasing embedment depth reduces this axial disparity. For instance, at h₁ = 140 mm, axial force differences become negligible. Such non-uniformities must be considered when finishing cylindrical gears to ensure consistent surface quality.
To validate the DEM simulations, we conduct experimental stress testing and roughness measurements on cylindrical gears. A strain gauge system is attached to the gear at three positions: tooth surface, upper face, and lower face. The stress data, representing contact forces, show periodic variations at the tooth surface position, consistent with simulation predictions. As parameters vary, stress values align with simulated normal contact force trends: stress increases modestly with rotational speed (by 17-28% across positions for a 150% speed increase) and significantly with embedment depth (by 35-59% for a 75% depth increase). This correlation confirms the simulation’s accuracy in capturing particle action on cylindrical gears.
Roughness measurements after 2 hours of processing further validate the effects on cylindrical gears. The surface roughness reduction rate ΔRa is calculated for different positions. As shown in Table 5, increasing rotational speed from 12 rpm to 30 rpm enhances ΔRa, with the lower face showing the most improvement (from 10.9% to 58.36%, a 5.36-fold increase). Similarly, increasing embedment depth from 80 mm to 140 mm raises ΔRa, notably at the lower face (from 28.0% to 56.89%, a 2.03-fold increase). The axial uniformity of roughness on the cylindrical gear tooth surface improves with higher embedment depth: at h₁ = 80 mm, ΔRa values for upper, middle, and lower axial regions are 17%, 26%, and 36%, respectively; at h₁ = 140 mm, they become 62%, 58%, and 55%, indicating reduced axial variability. This aligns with simulation insights that deeper embedment homogenizes force distribution along the cylindrical gear axis.
| Parameter | Tooth Surface Position | ΔRa at Low Parameter | ΔRa at High Parameter | Improvement Factor |
|---|---|---|---|---|
| Rotational Speed: 12 rpm → 30 rpm (h₁=80 mm) | Upper Face | 15.2% | 40.1% | 2.64 |
| Tooth Surface | 18.7% | 45.3% | 2.42 | |
| Lower Face | 10.9% | 58.36% | 5.36 | |
| Embedment Depth: 80 mm → 140 mm (n₁=30 rpm) | Upper Face | 25.5% | 50.2% | 1.97 |
| Tooth Surface | 30.1% | 55.8% | 1.85 | |
| Lower Face | 28.0% | 56.89% | 2.03 |
In summary, our DEM simulation and experimental study elucidate the particle action behavior in spindle barrel finishing of cylindrical gears. The process exhibits cyclic stages—filling, stable filling, and outflow—with the stable filling stage being predominant for material removal on cylindrical gears. Embedment depth h₁ primarily augments normal contact force (F ∝ h₁¹·⁰¹), while rotational speeds n₁ and n₂ mainly enhance relative velocity (V ∝ n₁¹·⁴⁹). Non-uniformities exist: the upper tooth surface of the cylindrical gear bears 1.5-1.8 times higher contact force and 1.35-1.45 times greater relative velocity than the lower tooth surface. Increasing embedment depth mitigates axial variability in finishing, promoting uniform roughness reduction across the cylindrical gear tooth surface. These findings provide a foundation for optimizing spindle barrel finishing parameters tailored to cylindrical gears, enabling improved surface integrity and performance. Future work could extend DEM models to include wear prediction and explore effects of particle shape or size on cylindrical gear finishing.
The cylindrical gear, as a critical transmission component, benefits greatly from controlled finishing processes. By leveraging DEM simulations, we can predict and tailor particle actions to achieve desired surface outcomes for cylindrical gears. The integration of simulation with experimental validation offers a robust framework for advancing cylindrical gear manufacturing technologies, ensuring higher reliability and efficiency in industrial applications. Continued research in this domain will further refine our understanding of particle dynamics and their impact on cylindrical gear surfaces, driving innovations in precision engineering.