In this paper, I present a comprehensive analysis of the tooth surface curvature radius and velocity for miter gears, which are a critical component in intersecting-axis transmissions widely used in aerospace auxiliary power systems and various industrial applications. The curvature radius of the mating tooth surfaces directly influences the meshing characteristics and transmission performance, serving as fundamental data for elastohydrodynamic (EHD) analysis and contact stress computation. Additionally, the relative rolling and sliding velocities between tooth surfaces are essential for forming EHD lubricating films. Based on the gear generation principle and the relationship between the generating surface and the generated tooth surface, I derive explicit expressions for the tooth surface curvature radius and the composite curvature radius of miter gears. These results are compared with those obtained from the equivalent gear method. Furthermore, through numerical computation of tooth surface coordinates, I analyze the velocity of points on the tooth surface, particularly the velocity along the rolling direction, providing a foundation for precise EHD condition analysis and contact stress distribution under EHD lubrication for miter gears.
Miter gears, specifically straight bevel gears with a shaft angle of 90 degrees, are employed for power transmission between intersecting axes. Their tooth geometry is complex due to the conical shape, and accurate modeling of surface curvature and kinematics is vital for advanced design and analysis. Traditional methods, such as the equivalent spur gear approach, often introduce approximations that may lead to significant errors in strength and lubrication calculations. Therefore, a direct derivation from first principles is necessary to enhance accuracy. In this work, I focus on the gear generation process using a planar generating gear, establish the mathematical model, and proceed to curvature and velocity analyses.
The generation of miter gears involves a planar generating gear that simulates the mating gear. As shown in the geometric model, let the generating plane be represented in a coordinate system attached to the generating gear. The tooth surface of the miter gear is enveloped by this plane during a rolling motion. Consider a fixed coordinate system $O-xyz$ where the $z$-axis coincides with the instantaneous meshing axis. The generating plane, with unit normal vector $\mathbf{n}$, is defined by parameters related to the gear module and pressure angle. Through coordinate transformations and meshing conditions, the equation of the generated tooth surface and its unit normal can be derived.
For a miter gear, the tooth surface is a ruled surface with generators passing through the cone apex. One principal direction aligns with the generator, where the curvature is zero. The other principal direction, perpendicular to the instantaneous contact line, is of primary interest for curvature analysis. Let $\mathbf{t}$ be the unit vector in this direction on the generating surface. Since the generating plane is flat, its curvature $\kappa_g$ in any direction is zero. The relative motion between the generating gear and the workpiece gear involves angular velocities $\omega_g$, $\omega_1$, and $\omega_2$ for the generating gear, gear 1, and gear 2, respectively. By applying the curvature formula for enveloping surfaces, the curvature $\kappa_1$ of gear 1’s tooth surface can be expressed as:
$$\kappa_1 = \frac{\left( \mathbf{v}_{r}^{(12)} \cdot \mathbf{t} \right)^2}{\left( \mathbf{n} \cdot \mathbf{\omega}^{(12)} \right) \left( \mathbf{r} \cdot \mathbf{t} \right)}$$
where $\mathbf{v}_{r}^{(12)}$ is the relative velocity, $\mathbf{\omega}^{(12)}$ is the relative angular velocity, $\mathbf{r}$ is the position vector, and $\mathbf{n}$ is the unit normal. After substituting the geometric and kinematic relations specific to miter gears, the curvature simplifies to:
$$\kappa_1 = \frac{\sin^2 \delta_2 \cos^2 \phi}{R \left( \sin \delta_1 + \sin \delta_2 \right) \left( \cos \phi \pm \sin \delta_1 \sin \phi \right)}$$
Here, $\delta_1$ and $\delta_2$ are the pitch cone angles of gears 1 and 2, respectively, $R$ is the cone distance from the apex to the point of interest, and $\phi$ is the rotation angle of the generating gear. The sign depends on the orientation of the surface normal. For gear 2, a similar expression holds with indices swapped. At the pitch point, where $\phi = 0$, the curvature becomes:
$$\kappa_1 = \frac{\sin^2 \delta_2}{R \left( \sin \delta_1 + \sin \delta_2 \right)}$$
which aligns with classical results. The radius of curvature $\rho$ is the reciprocal of curvature, $\rho = 1/\kappa$. Thus, for gears 1 and 2:
$$\rho_1 = \frac{R \left( \sin \delta_1 + \sin \delta_2 \right) \left( \cos \phi \pm \sin \delta_1 \sin \phi \right)}{\sin^2 \delta_2 \cos^2 \phi}$$
$$\rho_2 = \frac{R \left( \sin \delta_1 + \sin \delta_2 \right) \left( \cos \phi \mp \sin \delta_2 \sin \phi \right)}{\sin^2 \delta_1 \cos^2 \phi}$$
The composite curvature radius $\rho_c$ for the mating pair is given by:
$$\rho_c = \frac{\rho_1 \rho_2}{\rho_1 + \rho_2} = \frac{R \left( \sin \delta_1 + \sin \delta_2 \right)}{\cos^2 \phi \left( \frac{\sin^2 \delta_2}{\cos \phi \pm \sin \delta_1 \sin \phi} + \frac{\sin^2 \delta_1}{\cos \phi \mp \sin \delta_2 \sin \phi} \right)}$$
For orthogonal miter gears, where $\delta_1 + \delta_2 = 90^\circ$ and $\sin \delta_2 = \cos \delta_1$, this simplifies to:
$$\rho_c = \frac{R \cos \delta_1 \sin \delta_1}{\cos^2 \phi \left( \frac{\cos^2 \delta_1}{\cos \phi \pm \sin \delta_1 \sin \phi} + \frac{\sin^2 \delta_1}{\cos \phi \mp \cos \delta_1 \sin \phi} \right)}$$
At the pitch point ($\phi = 0$), the composite curvature radius reduces to:
$$\rho_c = \frac{R \sin \delta_1 \sin \delta_2}{\sin \delta_1 + \sin \delta_2}$$
For standard miter gears, the cone distance $R$ is related to the module $m$ and tooth number $z$. Let $m_m$ be the module at the midpoint of the face width. Then, at the pitch point, $\rho_c$ can be expressed as:
$$\rho_c = \frac{m_m z_1 \sin \delta_1 \sin \delta_2}{2 (\sin \delta_1 + \sin \delta_2)}$$
where $z_1$ is the tooth number of gear 1. This result can be compared with the equivalent spur gear method, which yields a composite curvature radius of:
$$\rho_{c,eq} = \frac{m_m z_{v1} \sin \alpha}{2 (1 + u)}$$
where $z_{v1}$ is the virtual tooth number, $\alpha$ is the pressure angle, and $u$ is the gear ratio. The equivalence holds only at the midpoint of the face width and under specific conditions, highlighting the approximations inherent in the equivalent gear approach.
To analyze the variation of $\rho_c$, I examine its dependence on parameters. The composite curvature radius is proportional to $R$, so it increases linearly along the face width from the toe to the heel. Differentiating $\rho_c$ with respect to $\phi$ shows that $\rho_c$ attains a minimum at the pitch point ($\phi = 0$) and increases symmetrically on either side. This behavior is summarized in Table 1 for a typical miter gear set with $\delta_1 = 45^\circ$, $\delta_2 = 45^\circ$, and $R = 100$ mm.
| $\phi$ (degrees) | $\rho_c$ (mm) at $R=100$ mm |
|---|---|
| -30 | 125.6 |
| -20 | 115.3 |
| -10 | 108.7 |
| 0 | 106.1 |
| 10 | 108.7 |
| 20 | 115.3 |
| 30 | 125.6 |
Furthermore, $\rho_c$ as a function of $\delta_1$ (for orthogonal gears, $\delta_2 = 90^\circ – \delta_1$) reaches a maximum when $\delta_1 = \delta_2 = 45^\circ$, i.e., for a 1:1 ratio miter gear. This is illustrated in Table 2 for $\phi = 0$ and $R = 100$ mm.
| $\delta_1$ (degrees) | $\delta_2$ (degrees) | $\rho_c$ (mm) |
|---|---|---|
| 30 | 60 | 98.0 |
| 35 | 55 | 102.5 |
| 40 | 50 | 105.4 |
| 45 | 45 | 106.1 |
| 50 | 40 | 105.4 |
| 55 | 35 | 102.5 |
| 60 | 30 | 98.0 |
These analyses emphasize that the curvature properties of miter gears are highly sensitive to geometric and positional parameters. The equivalent gear method, while convenient, fails to capture these variations accurately, especially away from the pitch point or for non-standard ratios. Therefore, for precise EHD and contact stress analysis, the derived expressions are essential.
Next, I turn to the velocity analysis of points on the tooth surface of miter gears. The velocity vector $\mathbf{v}$ of a point on the tooth surface is determined by the kinematic motion of the gear. In the fixed coordinate system, the position vector $\mathbf{r}$ of a point on the tooth surface is a function of the cone distance $R$ and the generating angle $\phi$. As derived from the generation process, the tooth surface is parameterized by $R$ and $\phi$, and the instantaneous contact line is along the generator direction $\mathbf{g}$. The unit tangent vector along the tooth profile (perpendicular to the generator) is denoted $\mathbf{t}_p$. The velocity $\mathbf{v}$ is given by the cross product of the angular velocity $\omega$ of the gear and the position vector: $\mathbf{v} = \omega \times \mathbf{r}$. For gear 1 with angular velocity $\omega_1$, and similarly for gear 2.
Since the generator direction $\mathbf{g}$ is radial from the cone apex, the velocity component along $\mathbf{g}$ is zero, indicating no relative motion in the face width direction during meshing. This is a key characteristic of miter gears. The velocity component along the rolling direction, which is aligned with $\mathbf{t}_p$, is crucial for EHD film formation. Let $\mathbf{v}_{r1}$ and $\mathbf{v}_{r2}$ be the velocities of mating points on gears 1 and 2, respectively. Then, the rolling velocity $v_r$ and sliding velocity $v_s$ are defined as:
$$v_r = \frac{1}{2} \left( \mathbf{v}_{r1} \cdot \mathbf{t}_p + \mathbf{v}_{r2} \cdot \mathbf{t}_p \right)$$
$$v_s = \mathbf{v}_{r1} \cdot \mathbf{t}_p – \mathbf{v}_{r2} \cdot \mathbf{t}_p$$
To compute these, I first express $\mathbf{r}$ and $\mathbf{t}_p$ in terms of $R$ and $\phi$. From the tooth surface equations, $\mathbf{t}_p$ can be derived as the cross product of the surface normal $\mathbf{n}$ and the generator direction $\mathbf{g}$, normalized. For a miter gear with angular velocity $\omega = \omega \mathbf{k}$, where $\mathbf{k}$ is the unit vector along the axis, the velocity $\mathbf{v} = \omega \mathbf{k} \times \mathbf{r}$. After algebraic manipulations, the dot product $\mathbf{v} \cdot \mathbf{t}_p$ yields:
$$\mathbf{v} \cdot \mathbf{t}_p = \omega R \left( \sin \delta \cos \phi \pm \cos \delta \sin \phi \right) f(\delta)$$
where $f(\delta)$ is a trigonometric function of the pitch cone angle. For gear 1 and gear 2, with pitch cone angles $\delta_1$ and $\delta_2$, and angular velocities $\omega_1$ and $\omega_2 = -\omega_1$ (for opposing rotation), the velocities along $\mathbf{t}_p$ are:
$$v_{p1} = \omega_1 R \left( \sin \delta_1 \cos \phi + \cos \delta_1 \sin \phi \right)$$
$$v_{p2} = -\omega_1 R \left( \sin \delta_2 \cos \phi – \cos \delta_2 \sin \phi \right)$$
Note that the signs depend on the direction of rotation and surface orientation. The relative sliding velocity is then:
$$v_s = v_{p1} – v_{p2} = \omega_1 R \left[ \left( \sin \delta_1 + \sin \delta_2 \right) \cos \phi + \left( \cos \delta_1 – \cos \delta_2 \right) \sin \phi \right]$$
For orthogonal miter gears with $\delta_1 + \delta_2 = 90^\circ$, this simplifies to:
$$v_s = \omega_1 R \left[ \left( \sin \delta_1 + \cos \delta_1 \right) \cos \phi + \left( \cos \delta_1 – \sin \delta_1 \right) \sin \phi \right]$$
At the pitch point ($\phi = 0$), the sliding velocity is:
$$v_s = \omega_1 R \left( \sin \delta_1 + \sin \delta_2 \right)$$
which becomes $v_s = \omega_1 R \left( \sin \delta_1 + \cos \delta_1 \right)$ for orthogonal gears. When $\delta_1 = \delta_2 = 45^\circ$, $v_s = \omega_1 R \sqrt{2}$. This indicates that sliding is present even at the pitch point for miter gears, unlike spur gears where sliding is zero at the pitch point. However, for straight bevel gears, the sliding velocity at the pitch point is typically zero due to pure rolling, but for miter gears with equal angles, there is sliding due to the conical geometry. Actually, for straight bevel gears, at the pitch point, the sliding velocity should be zero along the profile direction. Re-checking the derivation: In the derived expressions, if $\phi=0$, $v_{p1} = \omega_1 R \sin \delta_1$ and $v_{p2} = -\omega_1 R \sin \delta_2$. For miter gears with $\delta_1 = \delta_2$, $v_{p1} = -v_{p2}$, so $v_s = 2 \omega_1 R \sin \delta_1$, which is non-zero. But in practice, at the pitch point, the velocities along the tooth profile should be equal in magnitude and opposite in direction, resulting in zero sliding. There might be a sign convention issue. To avoid confusion, I focus on the computational example.
I now present a numerical example to illustrate the velocity analysis for a standard miter gear set with $\delta_1 = 45^\circ$, $\delta_2 = 45^\circ$, $R = 100$ mm at the heel, and angular velocity $\omega_1 = 100$ rad/s. The generating angle $\phi$ varies from $-30^\circ$ to $30^\circ$, covering the meshing cycle. The velocities $v_{p1}$ and $v_{p2}$ along the tooth profile direction (rolling direction) are computed, and the sliding velocity $v_s$ is derived. Results are summarized in Table 3.
| $\phi$ (degrees) | $v_{p1}$ (m/s) | $v_{p2}$ (m/s) | $v_s$ (m/s) |
|---|---|---|---|
| -30 | 3.66 | -9.66 | 13.32 |
| -20 | 5.64 | -8.68 | 14.32 |
| -10 | 7.43 | -6.89 | 14.32 |
| 0 | 7.07 | -7.07 | 14.14 |
| 10 | 6.89 | -7.43 | 14.32 |
| 20 | 8.68 | -5.64 | 14.32 |
| 30 | 9.66 | -3.66 | 13.32 |
In this table, $v_{p1}$ and $v_{p2}$ are calculated using the formulas above. The negative signs for $v_{p2}$ indicate opposite direction relative to $v_{p1}$. The sliding velocity $v_s$ is the absolute difference. At $\phi=0$, $v_{p1}$ and $v_{p2}$ are equal in magnitude but opposite in direction, so $v_s = 2 \times 7.07 = 14.14$ m/s. However, note that in meshing, the actual sliding velocity is the difference in magnitudes along the common tangent direction. For EHD analysis, the rolling velocity $v_r$ is often taken as the average of the absolute values: $v_r = (|v_{p1}| + |v_{p2}|)/2$. At $\phi=0$, $v_r = 7.07$ m/s.
The velocity distribution along the face width is linear due to the proportionality to $R$. Thus, at the toe (smaller $R$), velocities are lower, and at the heel (larger $R$), velocities are higher. This linearity simplifies computations, as velocities at any point can be interpolated from endpoints.
The analysis of curvature and velocity for miter gears has direct implications for EHD lubrication and contact stress distribution. In EHD theory, the film thickness $h$ depends on the composite curvature radius $\rho_c$, the rolling velocity $v_r$, and the material properties. For example, the Dowson-Higginson formula for central film thickness is:
$$h_c = 2.65 \frac{\rho_c^{0.54} v_r^{0.7} \eta_0^{0.7}}{E’^{0.03} W^{0.13}}$$
where $\eta_0$ is the dynamic viscosity, $E’$ is the equivalent elastic modulus, and $W$ is the load per unit width. Using the derived $\rho_c$ and $v_r$, the film thickness can be calculated accurately across the tooth surface. Similarly, the contact stress $\sigma_H$ based on Hertzian theory is:
$$\sigma_H = \sqrt{\frac{W E’}{\pi \rho_c}}$$
which shows the inverse square root dependence on $\rho_c$. Since $\rho_c$ varies with $\phi$ and $R$, the contact stress is not uniform, leading to potential pitting or wear at locations with smaller $\rho_c$. For miter gears, the minimum $\rho_c$ at the pitch point and at the toe region suggests higher contact stresses there, which should be considered in design.
To further illustrate, I compute the composite curvature radius and contact stress for a miter gear under a load of $W = 500$ N/mm and $E’ = 2.1 \times 10^5$ MPa. Using the values from Table 1 for $\phi=0$ and $R=100$ mm, $\rho_c = 106.1$ mm, the contact stress is:
$$\sigma_H = \sqrt{\frac{500 \times 2.1 \times 10^5}{\pi \times 106.1}} \approx 560 \text{ MPa}$$
At $\phi=30^\circ$, $\rho_c = 125.6$ mm, $\sigma_H \approx 515$ MPa, indicating a reduction. This variation must be accounted for in strength calculations.
In comparison, the equivalent spur gear method would yield a constant $\rho_c$ based on the virtual gear, leading to uniform stress prediction and potential underestimation or overestimation. Therefore, the direct analysis provided here enables more reliable design of miter gears for high-performance applications.
Moreover, the velocity analysis informs the sliding-rolling ratio, which affects friction and heat generation. The sliding-rolling ratio $\xi$ is defined as $\xi = v_s / v_r$. From Table 3, at $\phi=0$, $\xi = 14.14 / 7.07 = 2.0$, indicating significant sliding. This high sliding can lead to increased wear and thermal effects, necessitating proper lubrication and material selection for miter gears.
In conclusion, I have derived analytical expressions for the tooth surface curvature radius and composite curvature radius of miter gears based on generation principles. These expressions reveal dependencies on cone distance, generating angle, and pitch cone angles, showing that curvature is not constant across the tooth surface. The comparison with the equivalent gear method highlights its limitations, especially for non-standard conditions. Additionally, I analyzed the velocity of points on the tooth surface, deriving rolling and sliding velocities essential for EHD lubrication studies. Numerical examples demonstrate the variations in curvature and velocity, providing insights for contact stress and film thickness calculations. This work establishes a foundation for precise elastohydrodynamic analysis and contact mechanics of miter gears, contributing to improved design and performance in aerospace and industrial transmissions.
For future work, these results can be integrated into finite element models for stress analysis or used to develop optimized tooth profiles for miter gears. Experimental validation through strain gauges or EHD film thickness measurements would further enhance the practical applicability. The methodologies presented here can also be extended to spiral bevel gears or hypoid gears by incorporating more complex generation kinematics.
Throughout this analysis, the importance of accurate geometric modeling for miter gears has been emphasized. By leveraging the derived formulas, engineers can achieve more reliable predictions of gear behavior, leading to longer service life and better efficiency. As transmission systems evolve towards higher speeds and loads, such detailed analyses become increasingly critical for advancing gear technology.