Harmonic gear drives, a critical advancement in modern gear technology, are extensively utilized in aerospace, robotics, CNC machine tools, and automated control systems due to their high precision and compact design. In flexspline-free configurations, severe surface wear on the cam wave generator presents a significant challenge. Addressing this, a novel cam wave generator design was proposed, with computational fluid dynamics (CFD) identifying an optimal elliptical clearance ratio of 3 for superior lubrication performance in elliptical sliding bearings. This study advances this work by optimizing biomimetic anti-wear micro-dimple textures on the cam surface, inspired by natural adaptations observed in lotus leaves, tree frogs, shark skin, and geckos.
1. Analytical Model
The fluid domain between the cam wave generator and flexspline is modeled as shown in Figure 2. The cam rotates counterclockwise at speed \(n_1\), opposing the flexspline’s rotation. The oil film thickness \(h(\theta)\) varies from a maximum \(h_{\text{max}}\) at \(\theta = 0^\circ\) to a minimum \(h_{\text{min}}\) at \(\theta = 90^\circ\). Micro-dimples with spherical-cap geometry are engineered on the cam’s outer surface along the z-axis (width direction). Texture parameters include distribution location, depth (\(h_p\)), shape (characterized by \(r_p / h_p\)), and density (\(\phi = \frac{N \pi r_p^2}{B L}\)), where \(B\) and \(L\) define the elliptical clearance dimensions, \(B_p\) relates to distribution position, \(r_p\) is the dimple radius, and \(N\) is the dimple count.
2. Results and Analysis
2.1 Influence of Texture Distribution Location
Three distribution schemes were evaluated: Full-region (uniform coverage), Convergent-region (textures only in convergent zones of the cam-flexspline gap), and Divergent-region (textures only in divergent zones). Under conditions \(z=0\), \(\delta=3\), \(h_{\text{min}}=0.1 \text{mm}\), \(\omega=2000 \text{rpm}\), \(h_p=r_p=1 \text{mm}\):
The oil film pressure \(p(\theta)\) on the flexspline inner wall (Figure 4) reveals:
- Divergent-region distribution yields smoother pressure profiles than convergent-region placement.
- Convergent-region distribution exhibits pressure curves nearly identical to full-region in the positive-pressure zone but lacks symmetry in negative-pressure regions.
- Full-region distribution generates localized pressure spikes at dimple sites, enhancing hydrodynamic lift. This configuration balances load capacity and friction optimally.
Quantitative comparisons show:
| Distribution | Load Capacity \(F_L\) (N) | Friction Force \(F_f\) (N) | Lubrication Efficiency \(\eta = F_L / F_f\) |
|---|---|---|---|
| Full-Region | 485 | 8.2 | 59.15 |
| Convergent-Region | 480 | 12.1 | 39.67 |
| Divergent-Region | 510 | 11.8 | 43.22 |
Full-region distribution maximizes \(\eta\), critical for durable gear technology applications.
2.2 Influence of Texture Depth (\(h_p\))
Varying \(h_p\) (0.05–0.20 mm) at fixed \(r_p = 6 \text{mm}\), \(\phi = 15\%\), and \(h_{\text{min}} = 0.1 \text{mm}\) yields:
$$ F_L = k_1 \cdot e^{-\alpha (h_p – h_{\text{min}})^2} \quad ; \quad F_f = k_2 \cdot h_p + c $$
where \(k_1\), \(k_2\), \(\alpha\), and \(c\) are model constants. Optimal performance occurs at \(h_p = 0.1 \text{mm}\) (Figure 5):
| \(h_p\) (mm) | \(F_L\) (N) | \(F_f\) (N) |
|---|---|---|
| 0.05 | 285 | 8.2 |
| 0.10 | 320 | 7.5 |
| 0.15 | 260 | 9.1 |
| 0.20 | 225 | 10.3 |
Exceeding \(h_{\text{min}}\) drastically reduces \(F_L\) due to disrupted hydrodynamic pressure buildup.
2.3 Influence of Texture Shape (\(r_p / h_p\))
With \(h_p = 0.1 \text{mm}\) and \(\phi = 15\%\), \(r_p\) was varied (2–10 mm). Dimple shape is characterized by aspect ratio \(\lambda = r_p / h_p\). Load and friction responses (Figure 6) follow:
$$ \frac{\partial F_L}{\partial r_p} = \beta r_p e^{-\gamma r_p} \quad ; \quad F_f \propto \frac{1}{r_p} $$
Results confirm:
| \(r_p\) (mm) | \(\lambda\) | \(F_L\) (N) | \(F_f\) (N) |
|---|---|---|---|
| 2 | 20 | 290 | 9.0 |
| 4 | 40 | 305 | 8.0 |
| 6 | 60 | 320 | 7.5 |
| 8 | 80 | 318 | 7.6 |
| 10 | 100 | 315 | 7.7 |
Peak \(F_L\) and minimal \(F_f\) occur at \(r_p = 6 \text{mm}\) (\(\lambda = 60\)), beyond which effects saturate.
2.4 Influence of Texture Density (\(\phi\))
Varying \(\phi\) (5%–25%) with \(h_p = 0.1 \text{mm}\), \(r_p = 6 \text{mm}\) demonstrates strong positive scaling (Figure 7):
$$ F_L = \kappa_1 \phi + \kappa_2 \quad ; \quad F_f = \kappa_3 / \phi + \kappa_4 $$
where \(\kappa_i\) are fitting parameters. Higher density enhances hydrodynamic lift accumulation:
| \(\phi\) (%) | \(F_L\) (N) | \(F_f\) (N) | \(\eta\) |
|---|---|---|---|
| 5 | 250 | 10.5 | 23.8 |
| 10 | 290 | 8.8 | 32.9 |
| 15 | 320 | 7.5 | 42.7 |
| 20 | 340 | 6.9 | 49.3 |
| 25 | 355 | 6.5 | 54.6 |
Maximum \(\phi\) minimizes friction and maximizes load capacity, crucial for high-efficiency gear technology.
3. Conclusion
This study establishes design principles for biomimetic micro-dimple textures on flexspline-free harmonic drive cam wave generators, significantly enhancing gear technology reliability:
- Full-region texture distribution optimizes lubrication symmetry and hydrodynamic lift, outperforming localized convergent/divergent patterns.
- Texture depth (\(h_p\)) must approximate the minimum oil film thickness (\(h_{\text{min}}\)). An optimal \(h_p = 0.1 \text{mm}\) maximizes \(F_L\) and minimizes \(F_f\).
- Dimple shape characterized by \(\lambda = r_p / h_p = 60\) (i.e., \(r_p = 6 \text{mm}\) at \(h_p = 0.1 \text{mm}\)) balances curvature and pressure generation.
- Higher texture density (\(\phi\)) linearly enhances load capacity and reduces friction via cumulative hydrodynamic effects. Densities >20% are recommended.
These findings provide a foundation for surface engineering in next-generation harmonic gear technology, directly addressing wear challenges in demanding applications like robotics and aerospace systems. Future work will explore multi-scale textures and dynamic wear modeling.