In the domain of gear dynamics, the precise calculation of time-varying meshing stiffness (TVMS) is fundamental for predicting vibration, noise, and the overall operational stability of transmission systems. Among various gear types, the helical gear is widely favored for its smooth engagement and high load-carrying capacity. However, the inherent helix angle introduces an axial force component, which is often neglected in conventional stiffness models. In my research, I propose an improved analytical algorithm for the TVMS of helical gear pairs, which integrates the axial force effect into the Potential Energy Method (PEM). This work aims to enhance the accuracy of stiffness predictions and systematically investigate the influence of key geometric parameters.
Traditional PEM, as established by Yang and Lin, decomposes the total elastic potential energy into Hertzian, bending, and axial compressive components. Tian later incorporated shear energy. However, for helical gears, the presence of a helix angle \(\beta\) results in a significant axial force \(F_a\). As illustrated in the force analysis, the normal load \(F_n\) is decomposed into radial \(F_r\), tangential \(F_t\), and axial \(F_a\) components. The fundamental equations governing this decomposition are:
\[
\begin{aligned}
F_r &= F_n \sin \alpha_1 \\
F_t &= F_n \cos \alpha_1 \cos \beta \\
F_a &= F_n \cos \alpha_1 \sin \beta
\end{aligned}
\]
In my improved model, each lamina of a sliced helical gear is treated as a variable cross-section cantilever beam. The core innovation lies in accounting for the axial bending deformation caused by \(F_a\), which introduces an additional potential energy term \(U_{af}\). The total potential energy of a single tooth pair is thus extended to six components:
\[
U = U_h + U_{b1} + U_{s1} + U_{a1} + U_{f1} + U_{af1} + U_{b2} + U_{s2} + U_{a2} + U_{f2} + U_{af2}
\]
Consequently, the comprehensive mesh stiffness \(k\) for a single tooth pair is:
\[
k = 1 / \left( \frac{1}{k_h} + \frac{1}{k_{b1}} + \frac{1}{k_{s1}} + \frac{1}{k_{a1}} + \frac{1}{k_{f1}} + \frac{1}{k_{af1}} + \frac{1}{k_{b2}} + \frac{1}{k_{s2}} + \frac{1}{k_{a2}} + \frac{1}{k_{f2}} + \frac{1}{k_{af2}} \right)
\]
The bending and shear stiffness formulas were also modified to reflect the influence of the axial force. The improved bending stiffness \(k_{ib1}\) for the \(i\)-th slice, considering the helix angle, is expressed as:
\[
\frac{1}{k_{ib1}} = \int_{-\alpha_1}^{\alpha_2} \frac{3(\cos \beta – 1) (\alpha_1 + \alpha_2) \sin \alpha_1 \cos \alpha_1 + \sin^2 \alpha_1}{2Ey\left[ \sin \alpha + (\alpha_2 – \alpha) \cos \alpha \right]^3} \, d\alpha + \int_{-\alpha_1}^{\alpha_2} \frac{\left[ \cos \alpha_1 – \cos \alpha + (\alpha_2 – \alpha) \sin \alpha \right] \cos \alpha_1 \cos \beta}{2Ey\left[ \sin \alpha + (\alpha_2 – \alpha) \cos \alpha \right]^3} \, d\alpha
\]
The improved shear stiffness \(k_{is1}\) is:
\[
\frac{1}{k_{is1}} = \int_{-\alpha_1}^{\alpha_2} \frac{1.2 (1 + \nu) (\alpha_2 – \alpha) \cos \alpha \cos \beta \cos^2 \alpha_1}{Ey\left[ \sin \alpha + (\alpha_2 – \alpha) \cos \alpha \right]} \, d\alpha
\]
The newly introduced axial bending stiffness \(k_{iaf}\) for a single slice was derived from the axial bending potential energy. The moment \(M_{af}\) induced by the axial force is \(M_{af} = F_a (y_\alpha – y_1)\), leading to the following formula:
\[
\frac{1}{k_{iaf}} = \int_0^{r_f} \frac{6 r_b \cos^2 \alpha_1 \sin^2 \beta \left[ (\alpha_1 + \alpha_2) \sin \alpha_1 + \cos \alpha_1 \right]^2 – y^2}{E y^3 \sqrt{r_f^2 – y’^2}} \, dy’
\]
The total time-varying meshing stiffness \(k_p\) for the entire helical gear pair is then obtained by integrating the stiffness of all \(n\) slices:
\[
k_p = \sum_{i=1}^{n} k_i^p = \sum_{i=1}^{n} \left[ 1 / \left( \frac{1}{k_{ih}} + \frac{1}{k_{ib1}} + \frac{1}{k_{is1}} + \frac{1}{k_{ia1}} + \frac{1}{k_{if1}} + \frac{1}{k_{ib2}} + \frac{1}{k_{is2}} + \frac{1}{k_{ia2}} + \frac{1}{k_{if2}} + \frac{1}{k_{iaf1}} + \frac{1}{k_{iaf2}} \right) \right]
\]
To validate my proposed algorithm, I compared the maximum single-tooth mesh stiffness \(k_{max}\) against results from the ISO 6336-1:2019 standard, the finite element method (FEM), and a previous slice-PEM method. The benchmark gear parameters used for validation are shown in the table below. The results demonstrate that my method yields a stiffness value very close to the ISO standard, with an error comparable to other advanced methods, thereby confirming the feasibility and accuracy of the approach.
| Method | \(k_{max}\) (×10⁸ N/m) | Error (%) |
|---|---|---|
| ISO 6336-1:2019 | 3.61 | 0 |
| Finite Element Method | 3.55 | 1.66 |
| Previous Slice-PEM [12] | 3.65 | 1.11 |
| My Improved Algorithm | 3.67 | 1.66 |
Using the verified algorithm, I conducted a systematic parametric study to understand the influence of key design variables on the TVMS of helical gears. The baseline parameters for this study are listed in the table below.
| Parameter | Symbol | Value |
|---|---|---|
| Normal Module | \(m_n\) | 4 mm |
| Helix Angle | \(\beta\) | 20° |
| Pressure Angle | \(\alpha_0\) | 20° |
| Number of Teeth (Pinion/Gear) | \(z_1(z_2)\) | 40 (40) |
| Face Width | \(b\) | 30 mm |
| Young’s Modulus | \(E\) | 2.06×10¹¹ Pa |
| Poisson’s Ratio | \(\nu\) | 0.3 |
Effect of Helix Angle \(\beta\)
I varied the helix angle from 0° to 20° in increments of 5°. The results clearly show that for small helix angles (e.g., \(\beta=5^\circ\)), the axial force is minimal, and my improved algorithm yields stiffness values nearly identical to the traditional PEM. As \(\beta\) increases, the axial force and axial overlap ratio \(\varepsilon_\beta\) increase significantly. This leads to a greater contribution from the axial bending stiffness and a significant divergence from the traditional method. A larger helix angle prolongs the multi-tooth contact region and reduces the fluctuation of the mesh stiffness, making the transmission smoother and quieter. The improved algorithm is therefore essential for accurately predicting the behavior of helical gears with large helix angles.
Effect of Pressure Angle \(\alpha_0\)
I investigated pressure angles of 14.5°, 20°, and 25°. An increase in the pressure angle thickens the tooth root, enhancing bending strength. The analysis shows that while the single-tooth stiffness peak increases, the total contact ratio decreases. Consequently, the mean value of the combined mesh stiffness shows a slight overall decrease. However, the influence on the stiffness fluctuation amplitude is relatively minor, suggesting that the pressure angle is not a primary driver for altering the fluctuation characteristics of TVMS in helical gears.
Effect of Number of Teeth \(z\) and Module \(m_n\)
Keeping the center distance constant, I varied the number of teeth from 20 to 70, with the module adjusting accordingly. The results indicate that the single-tooth stiffness peak decreases with an increasing number of teeth. However, the mean combined stiffness does not follow a simple linear trend. Importantly, the fluctuation of the mesh stiffness, defined as the difference between the maximum and minimum values, is minimized when the number of teeth is around 50. This highlights the complex interaction between the tooth count, module, and the resulting contact ratio.
| Number of Teeth \(z\) | Module \(m_n\) (mm) | Mean Stiffness (×10⁸ N/m) | Stiffness Fluctuation (×10⁸ N/m) |
|---|---|---|---|
| 20 | 8.00 | 15.2 | 3.8 |
| 30 | 5.33 | 14.8 | 2.1 |
| 40 | 4.00 | 14.5 | 1.5 |
| 50 | 3.20 | 14.6 | 1.1 |
| 60 | 2.67 | 14.7 | 1.4 |
| 70 | 2.29 | 14.4 | 1.7 |
Effect of Face Width \(b\)
I increased the face width from 20 mm to 60 mm. The results demonstrate a near-linear increase in the mean mesh stiffness with increasing face width. This is intuitive, as a wider helical gear provides more contact lines and greater structural rigidity. Furthermore, the axial overlap ratio \(\varepsilon_\beta\) is directly proportional to the face width. The study revealed a critical design insight: the fluctuation of the TVMS is minimized when the axial overlap ratio is close to an integer value. In this specific case, the smallest fluctuation was observed at a face width of 40 mm, which corresponds to an \(\varepsilon_\beta\) of 1.088, very close to 1.0.
| Face Width \(b\) (mm) | Axial Overlap Ratio \(\varepsilon_\beta\) | Mean Stiffness (×10⁸ N/m) | Stiffness Fluctuation (×10⁸ N/m) |
|---|---|---|---|
| 20 | 0.544 | 10.1 | 2.5 |
| 30 | 0.816 | 14.5 | 1.5 |
| 40 | 1.088 | 18.8 | 0.8 |
| 50 | 1.360 | 23.2 | 1.1 |
| 60 | 1.632 | 27.5 | 1.6 |
In conclusion, my research presents a significant enhancement to the analytical calculation of TVMS for helical gears. By rigorously incorporating the axial force component, the proposed algorithm offers superior accuracy, particularly for helical gears with substantial helix angles. The parametric study provides valuable guidelines for helical gear design, highlighting that selecting a helix angle to maximize smoothness, choosing a face width that aligns the axial overlap ratio with an integer, and optimizing the tooth count can effectively reduce mesh stiffness fluctuations and improve the dynamic performance of helical gear transmissions.