Non-circular gear transmission represents an innovative approach in gear technology, enabling variable transmission ratios through specialized pitch curve functions for applications requiring non-uniform motion, periodic movements, or reciprocating operations. These gears are vital in low-speed, high-torque systems like printing machines, packaging equipment, pumping unit reversing mechanisms, gear pumps, hydraulic motors, and flow meters. Despite their advantages, challenges in modeling transmission errors and quantifying backlash limit their precision in critical applications. This study addresses these gaps by developing comprehensive models for bidirectional transmission errors and backlash under eccentricity errors, validated through rigorous experiments.
Modeling Transmission Error and Backlash
Traditional transmission error models for constant-ratio gears are inadequate for non-circular variants due to their time-varying kinematics. Our approach adapts the meshing line increment method to non-circular gears, incorporating eccentricity errors to establish bidirectional transmission error and backlash models. For counterclockwise rotation of the driving gear (Fig. 2), meshing line increments are:
$$
\Delta L_1^+ = e_1 \sin(\theta_1 + \varphi_1 + \alpha)
$$
$$
\Delta L_2^+ = e_2 \sin(\theta_2 + \varphi_2 – \alpha)
$$
For clockwise rotation (Fig. 3):
$$
\Delta L_1^- = e_1 \sin(\theta_1 + \varphi_1 – \alpha)
$$
$$
\Delta L_2^- = e_2 \sin(\theta_2 + \varphi_2 + \alpha)
$$
where \(e_1\) and \(e_2\) denote eccentricity errors, \(\theta_1\) and \(\theta_2\) are rotation angles, \(\varphi_1\) and \(\varphi_2\) are initial phases, and \(\alpha\) is the pressure angle. Transmission errors for counterclockwise (\(\Delta \phi_n\)) and clockwise (\(\Delta \phi_s\)) rotations are:
$$
\Delta \phi_n = \frac{180}{\pi r_b} \left[ e_1 \sin(\theta_1 + \varphi_1 + \alpha) + e_2 \sin(\theta_2 + \varphi_2 – \alpha) – e_1 \sin(\varphi_1 + \alpha) – e_2 \sin(\varphi_2 – \alpha) \right]
$$
$$
\Delta \phi_s = \frac{180}{\pi r_b} \left[ e_1 \sin(\theta_1 + \varphi_1 – \alpha) + e_2 \sin(\theta_2 + \varphi_2 + \alpha) – e_1 \sin(\varphi_1 – \alpha) – e_2 \sin(\varphi_2 + \alpha) \right]
$$
where \(r_b = r \cos\alpha\) is the base circle radius. Backlash (\(b_v\)) combines constant and time-varying components:
$$
b_v = \frac{180 \times 2 \tan \alpha}{\pi} \left( e_2 \cos(\theta_2 + \varphi_2) – e_1 \cos(\theta_1 + \varphi_1) \right)
$$
Parametric Analysis of Transmission Error and Backlash
Using the parameters in Table 1, we analyzed the influence of eccentricity ratio (\(k\)) and eccentricity error (\(e\)) on transmission performance.
| Parameter | Value |
|---|---|
| Module \(m\) (mm) | 3 |
| Center Distance \(a\) (mm) | 150 |
| Tip Clearance Coefficient \(C^*\) | 0.25 |
| Eccentricity Ratio \(k\) | 0.2 |
| Number of Teeth \(Z\) | 47 |
| Addendum Coefficient \(h_a^*\) | 1 |
| Face Width \(B\) (mm) | 30 |
| Pitch Curve Equation | \( r = \frac{64.667}{1 \pm 0.3287 \cos \theta} \) |
Eccentricity Ratio Effects
Figure 4 illustrates transmission error sensitivity to \(k\). Both \(\Delta \phi_n\) and \(\Delta \phi_s\) increase with \(k\), exhibiting periodic fluctuations. Higher \(k\) amplifies error magnitudes due to enhanced geometric asymmetry, a critical consideration in gear technology design.
Eccentricity Error Effects
Figure 5 shows transmission errors under varying \(e\). At \(k=0.3\), errors grow with \(e\), confirming manufacturing precision as a key factor in minimizing transmission inaccuracies in advanced gear technology systems.
Backlash Dynamics
Backlash derives from bidirectional transmission errors: \(b_v = \Delta \phi_n – \Delta \phi_s\). Figure 6 validates this relationship, while Figure 7 demonstrates how initial phase (\(\varphi\)) affects backlash. Optimal phases minimizing backlash are:
$$
\varphi_1 = \pi n – \alpha, \quad \varphi_2 = \pi n + \alpha
$$
Post-optimization backlash reductions exceed 40% (Fig. 8–9), underscoring phase tuning as a vital gear technology calibration step.
Experimental Validation
We developed a test rig (Fig. 10) with torque sensors, optical encoders, and programmable loads. Key components included:
- Torque sensors (YH-502, ±0.1% accuracy)
- Optical encoders (K-100, 48,000 PPR resolution)
- AC servo motors (MSME504G, 15.9 N·m torque)
- Data acquisition (NI-6351, 1.25 MS/s sampling rate)
Transmission Error Measurements
Figure 11 compares theoretical and experimental transmission errors. Both directions exhibit periodicity, with minor deviations attributed to friction and dynamic effects. Bidirectional error curves enable backlash extraction via \(b_v = \Delta \phi_n – \Delta \phi_s\).
Load Influence
Under constant speed (5 rpm), increasing torque from 5 N·m to 15 N·m amplifies transmission errors (Fig. 12) and backlash (Fig. 13). Elastic deformations cause backlash accumulation between cycles, a phenomenon mitigated at higher loads as tooth compliance dominates:
$$
\Delta b_{\text{accum}} = \frac{F_t}{k_m} \left( \frac{1}{\cos\alpha} – \mu \sin\alpha \right)
$$
where \(F_t\) is tangential load, \(k_m\) is mesh stiffness, and \(\mu\) is friction coefficient.
Conclusions
- Bidirectional transmission errors directly quantify backlash: \(b_v = \Delta \phi_n – \Delta \phi_s\).
- Transmission errors increase with eccentricity ratio (\(k\)), eccentricity error (\(e\)), and load (\(T\)), while backlash accumulation arises from elastic deformations.
- Initial phase optimization reduces backlash by >40%, enhancing precision in non-circular gear technology applications.
- Experimental validation confirms model accuracy, providing a foundation for high-precision non-circular gear design.
These insights advance gear technology for applications demanding variable transmission ratios with minimal error, establishing methodologies for precision manufacturing and calibration.