In the realm of precision power transmission, the dynamic performance of spur gears is paramount. As a critical metric for evaluating this performance, the loaded transmission error (LTE) directly correlates with noise, vibration, and harshness (NVH) characteristics. The amplitude of fluctuation in the LTE is of particular concern; a smaller amplitude signifies smoother meshing action and superior dynamic behavior. While extensive research has been devoted to the LTE of external spur gears, studies focusing on the unique configuration of internal spur gears, especially those employing a short-tooth design common in planetary systems, remain relatively scarce. This work aims to address this gap by systematically investigating the influence of modification parameters on the LTE of internal short-tooth spur gears and developing an effective optimization methodology.
The fundamental geometry of spur gears is defined by several key parameters. For the purpose of modeling and analysis in this study, the primary design parameters for the internal spur gear pair are established as follows. The pinion (sun gear) is modeled as a solid component, while the internal gear (ring gear) requires consideration of rim thickness for structural integrity.
| Parameter | Pinion (Sun Gear) | Internal Gear (Ring Gear) |
|---|---|---|
| Number of Teeth, z | 53 | 160 |
| Module, m (mm) | 4.0 | 4.0 |
| Pressure Angle, α (°) | 20 | 20 |
| Addendum Coefficient, ha* | 0.8 | 0.8 |
| Dedendum Coefficient, c* | 0.3 | 0.3 |
| Face Width, b (mm) | 212 | 212 |
| Profile Shift Coefficient, x | +0.5 | 0 |
The loaded transmission error, $\Delta \theta_{LTE}$, can be conceptually defined as the deviation between the actual rotational position of the output gear and its ideal theoretical position under load, expressed as a function of the input rotation $\phi_{in}$:
$$\Delta \theta_{LTE}(\phi_{in}) = \theta_{out,actual}(\phi_{in}) – \theta_{out,ideal}(\phi_{in})$$
The peak-to-peak value of this function, $A_{LTE}$, is the critical performance indicator:
$$A_{LTE} = \max(\Delta \theta_{LTE}) – \min(\Delta \theta_{LTE})$$
The goal of gear modification is to minimize $A_{LTE}$.
Modification Strategies for Spur Gears
To mitigate transmission error excitations, intentional deviations from the ideal involute profile and lead are applied—a process known as gear modification. For the internal spur gear pair studied, modifications are applied solely to the pinion for manufacturing simplicity.
Profile Modification
Profile modification involves altering the tooth profile, typically at the tip and root regions, to compensate for deflections and improve the load transition between tooth pairs. For ease of manufacturing specification, modifications are described along the tooth height direction rather than the line of action. Two primary forms are considered: parabolic and linear relief.
The profile modification parameters are defined as:
– $y_1$: Maximum tip relief amount.
– $y_2$: Length of tip relief along the tooth height.
– $y_3$: Maximum root relief amount.
– $y_4$: Length of root relief along the tooth height.
The relief length is often set proportional to the theoretical length of the double-contact region for optimal effect.
Lead (Helix) Modification
Lead modification compensates for misalignments and edge-loading across the face width. For spur gears, two common types are analyzed:
– End Relief (Linear Taper): Material is removed linearly from both ends of the tooth.
– Crowning (Parabolic): Material is removed in a parabolic pattern, creating a central bulge.
The key parameter is $y_5$, the maximum lead modification amount at the tooth center (for crowning) or the start of relief (for end relief). A secondary parameter $y_6$ defines the length of the end relief zone, typically set to $0.25b$.
Influence of Modification Parameters on LTE Amplitude
A baseline analysis of the unmodified internal spur gear pair yielded an LTE fluctuation amplitude $A_{LTE,0} = 6.69 \mu m$. To understand the parametric influence, a series of simulations were conducted, varying one parameter at a time while holding others at a nominal value.
Effect of Profile Modification Amounts
With lead crowning fixed at $y_5 = 20 \mu m$ and relief lengths set to the double-contact zone, the impact of tip and root relief was isolated.
1. Maximum Tip Relief ($y_1$): As $y_1$ increased, $A_{LTE}$ exhibited a non-linear, concave response for parabolic relief, initially decreasing before rising again. This indicates an optimal value exists.
| Tip Relief $y_1$ ($\mu m$) | LTE Amplitude $A_{LTE}$ ($\mu m$) – Parabolic | LTE Amplitude $A_{LTE}$ ($\mu m$) – Linear |
|---|---|---|
| 10 | 5.42 | 5.85 |
| 15 | 4.98 | 5.32 |
| 20 | 4.87 | 5.10 |
| 25 | 5.15 | 5.25 |
| 30 | 5.61 | 5.68 |
2. Maximum Root Relief ($y_3$): The effect varied significantly with relief type. For parabolic relief, $A_{LTE}$ generally decreased monotonically with increasing $y_3$. For linear relief, a similar concave trend as tip relief was observed, suggesting a less stable influence.
| Root Relief $y_3$ ($\mu m$) | LTE Amplitude $A_{LTE}$ ($\mu m$) – Parabolic | LTE Amplitude $A_{LTE}$ ($\mu m$) – Linear |
|---|---|---|
| 10 | 5.38 | 5.70 |
| 15 | 5.05 | 5.25 |
| 20 | 4.87 | 4.95 |
| 25 | 4.78 | 5.08 |
| 30 | 4.75 | 5.40 |
Effect of Lead Modification Amounts
With profile modifications held constant ($y_1=y_3=20 \mu m$, parabolic), the effect of lead modification type and amount was studied. The amplitude $A_{LTE}$ increased with larger lead modification $y_5$ for both types, attributable to a reduction in overall mesh stiffness. However, parabolic crowning consistently yielded lower $A_{LTE}$ values compared to linear end relief, demonstrating its superiority for optimizing the meshing of these spur gears.
| Lead Relief $y_5$ ($\mu m$) | LTE Amplitude $A_{LTE}$ ($\mu m$) – Parabolic Crowning | LTE Amplitude $A_{LTE}$ ($\mu m$) – Linear End Relief |
|---|---|---|
| 10 | 3.82 | 4.45 |
| 15 | 4.25 | 5.10 |
| 20 | 4.87 | 5.92 |
| 25 | 5.68 | 6.85 |
| 30 | 6.52 | 7.73 |
The comparative analysis confirms that a judicious selection of modification parameters can significantly reduce the LTE amplitude in internal spur gears, with parabolic forms generally offering better potential for optimization.
Optimization of Modification Parameters
Given the complex, non-linear interactions between modification parameters and the LTE amplitude, a systematic optimization approach is necessary to find the global minimum. A Particle Swarm Optimization (PSO) algorithm is employed for this task, chosen for its efficiency and effectiveness in handling non-linear, multi-variable problems without requiring gradient information.
The optimization vector is defined as $\mathbf{Y} = [y_1, y_2, y_3, y_4, y_5]^T$, where all modifications are parabolic. The objective function $f(\mathbf{Y})$ is the peak-to-peak LTE amplitude, $A_{LTE}$, computed via a Loaded Tooth Contact Analysis (LTCA) model that solves for the static equilibrium of the gear mesh under load, considering tooth compliance and contact conditions. The optimization problem is stated as:
$$\text{Minimize: } f(\mathbf{Y}) = A_{LTE}(\mathbf{Y})$$
$$\text{Subject to: } \mathbf{Y}_{lb} \leq \mathbf{Y} \leq \mathbf{Y}_{ub}$$
where $\mathbf{Y}_{lb}$ and $\mathbf{Y}_{ub}$ are practical lower and upper bounds for the modification parameters.
The PSO algorithm operates by maintaining a swarm of particles (potential solutions), each with a position $\mathbf{Y}_i$ and velocity $\mathbf{V}_i$ in the search space. The position represents a specific set of modification parameters, and the velocity guides its movement. Each particle remembers its personal best position ($\mathbf{P}_{best,i}$) and the swarm knows the global best position ($\mathbf{G}_{best}$). The velocity and position are updated iteratively (for generation $k+1$) as:
$$\mathbf{V}_i^{k+1} = \omega \mathbf{V}_i^{k} + c_1 r_1 (\mathbf{P}_{best,i}^{k} – \mathbf{Y}_i^{k}) + c_2 r_2 (\mathbf{G}_{best}^{k} – \mathbf{Y}_i^{k})$$
$$\mathbf{Y}_i^{k+1} = \mathbf{Y}_i^{k} + \mathbf{V}_i^{k+1}$$
where $\omega$ is the inertia weight, $c_1$ and $c_2$ are acceleration coefficients, and $r_1$, $r_2$ are random numbers in [0,1].
The optimization workflow proceeds as follows: 1) Randomly initialize the swarm. 2) Evaluate $f(\mathbf{Y}_i)$ for each particle using LTCA. 3) Update $\mathbf{P}_{best,i}$ and $\mathbf{G}_{best}$. 4) Update velocities and positions. 5) Repeat steps 2-4 until convergence or a maximum generation count is reached.
Results and Comparative Analysis
The PSO algorithm converged to an optimal solution for the internal spur gear pair. To contextualize the result, it is compared against two other modification schemes analyzed within the same simulation framework: a manufacturer-recommended linear profile modification and an optimized design generated by a different algorithm (Genetic Algorithm).
Optimized Modification Parameters from PSO:
– $y_1 = 28.39 \mu m$
– $y_2 = 2.14 mm$
– $y_3 = 28.91 \mu m$
– $y_4 = 1.67 mm$
– $y_5 = 13.52 \mu m$
| Modification Scheme | Description | Optimized LTE Amplitude, $A_{LTE}$ ($\mu m$) | Improvement vs. Baseline |
|---|---|---|---|
| Baseline (Unmodified) | Standard involute profile, no lead modification. | 6.69 | 0% |
| Recommended Linear | Linear tip & root profile relief only ($y_1=y_3=16.08 \mu m$). | 6.45 | 3.6% |
| Alternative Algorithm Optimization | Parabolic profile & lead crowning, optimized via GA. | 3.10 | 53.7% |
| Proposed PSO Optimization | Parabolic profile & lead crowning, optimized via PSO. | 2.86 | 57.2% |
The results clearly demonstrate the efficacy of comprehensive parabolic modification. The simple linear profile modification offered minimal improvement. Both optimization-based approaches delivered substantial reductions in LTE amplitude, with the proposed PSO-based method achieving the lowest value of $2.86 \mu m$, corresponding to a 57.2% reduction from the baseline. This underscores the critical importance of a holistic and optimized modification strategy for enhancing the dynamic performance of high-performance internal spur gears. The optimal parameters suggest a balanced, symmetric approach to profile relief and a relatively mild lead crowning are most effective for this specific gear geometry and loading condition.
Conclusions
This investigation provides a detailed analysis of loaded transmission error in internal short-tooth spur gears, focusing on the impact of modification parameters and optimization. Key conclusions are drawn:
1. The influence of profile modification amounts ($y_1$, $y_3$) on LTE amplitude is non-linear and interdependent with the modification form (linear vs. parabolic). Parabolic relief generally offers more favorable and predictable trends for optimization.
2. Lead modification significantly impacts the meshing behavior of spur gears. Parabolic crowning is superior to linear end relief for minimizing LTE amplitude, although excessive crowning increases amplitude due to stiffness loss.
3. A systematic optimization framework, such as the Particle Swarm Optimization algorithm applied here, is essential for identifying the complex interplay between multiple modification parameters ($y_1$, $y_2$, $y_3$, $y_4$, $y_5$) to achieve a global minimum in LTE fluctuation.
4. The proposed optimization methodology successfully identified a modification set that reduced the LTE amplitude by over 57% compared to the unmodified baseline, significantly outperforming a standard recommended linear modification scheme.
These findings establish a foundational methodology for the design and analysis of high-performance internal spur gear sets, contributing directly to the goal of achieving quieter, smoother, and more reliable planetary and other gear drives utilizing internal spur gears.