The application of hyperboloid gears, also known as hypoid gears, as the primary reduction drive in automotive axles is extensive due to their significant speed reduction ratio and compact design. The annual domestic demand for such gears in China exceeds 10 million pairs. However, the predominant manufacturing technology relies on Gleason cutting processes. This machining method inevitably severs the metal’s fibrous flow lines, leaves tool marks on the tooth flanks and roots, and introduces stress concentrations at the fillet, critically compromising the gear’s bending fatigue strength. Consequently, developing an anti-fatigue manufacturing strategy suitable for domestic industrial conditions is imperative. Within this context, the cold rotary forging process for the large gear (driven gear) of a hyperboloid pair represents a pivotal step in the post-forging finishing of tooth flanks and root fillets, fulfilling the core requirements of anti-fatigue manufacturing.
Rotary forging, a pressworking technology, became commercially viable internationally in the 1960s and 70s and has since seen widespread research and application. While existing studies detail the principles of cold rotary forging for spiral bevel gears, these conventional approaches often employ complex die cavities with full-surface contact. This complexity can lead to surface defects, stress concentrations during forming, and ultimately fail to achieve the desired anti-fatigue properties. Furthermore, the required forging force is substantial. To address these critical limitations in the manufacturing of hyperboloid gears, I propose a fundamentally novel cold rotary forging scheme. This innovative strategy simplifies the die structure dramatically and utilizes localized line-contact continuous plastic deformation, offering a targeted and technologically superior solution. Given the inherent design of hyperboloid gears with their offset, the pinion (small gear) typically possesses higher bending fatigue strength, with failures more frequently occurring in the larger driven gear. Therefore, this proposed cold rotary forging technology is specifically architected for the large gear.
Conventional Rotary Forging Principles and Their Drawbacks
The established cold rolling principle for spiral bevel gears involves an upper die, typically a conical body, whose central axis is inclined at a tilt angle $\gamma$ relative to the machine spindle’s central axis. During the process, this upper die continuously nutates around the machine axis while simultaneously rotating about its own axis. Concurrently, the lower die pushes the workpiece upward at a constant velocity, applying axial pressure. Through this compounded motion of nutation and feed, the workpiece is incrementally and repeatedly compressed until the desired tooth form is achieved.
The primary disadvantage of this method lies in the intricate geometry of the die cavity. This complexity makes it susceptible to causing defects on the tooth flank and introducing unfavorable stress states during forming, often undermining the goal of anti-fatigue manufacturing. The full-surface contact also results in relatively high forming forces. My proposed scheme directly confronts these issues.
The Novel Cold Rotary Forging Scheme for Large Hyperboloid Gears
My innovative approach is conceptually inspired by the formate (non-generated) method of machining large hyperboloid gears. It substitutes the complex cutting head with a simple trapezoidal die and replaces material removal with localized plastic deformation. The core of the scheme is a trapezoidal die with an extremely simple profile, where the cross-section at the contact zone with the tooth space is identical to the normal section view of the final tooth space at that location. This die is embedded into a conical摆动头 (pendulum head). Crucially, the rotational axis of this die assembly is inclined at a specific angle relative to the machine’s central axis. This inclination enables instantaneous line contact between the forming die and the ideal tooth space while preventing interference at all other positions.
The kinematic motion during cold forging is similar to the conventional process: the pendulum head nutates about the machine axis. However, as it completes one full nutation cycle, the envelope surface formed by the contact line between the trapezoidal die and the ideal tooth space perfectly coincides with the grinding wheel shape required for formate grinding of the same gear. This congruence is the fundamental guarantee of forming accuracy and quality. After one tooth is fully formed, an indexing mechanism rotates the blank by one tooth pitch to begin forming the next.
The forging process is mathematically described using a series of coordinated coordinate systems. For a right-hand large gear, the machine spindle axis lies in the first quadrant of the plane $Z_0 = 0$.
- Machine Coordinate System $S_0(O_0-X_0Y_0Z_0)$: Origin $O_0$ at the machine center; plane $Z_0=0$ is the machine plane.
- Virtual Grinding Wheel System $S_c(O_c-X_cY_cZ_c)$: Defined by the adjustment parameters from formate gear machining. Its origin $O_c$ is offset from $O_0$ by $H_2$ in the X-direction and $V_2$ in the Y-direction.
- Die/Pendulum Head System $S_d(O_d-X_dY_dZ_d)$: Origin $O_d$ is the apex of the conical pendulum head. Its position and orientation are derived from $S_c$ to ensure the contact line trajectory matches the grinding wheel surface. It is offset from $O_c$ by $-L$ along $X_c$ and $-N$ along $Z_c$, with its axis $Z_d$ tilted clockwise by angle $\gamma$ relative to $Z_c$.
- Workpiece (Gear) System $S_2(O_2-X_2Y_2Z_2)$: The gear center $O_2$ is located at a distance $\Delta X_2$ from $O_0$, with the angle between $O_0X_0$ and $O_2X_2$ being the gear root angle $\gamma_2$.
The transformation between these systems uses homogeneous coordinate matrices. The transformation from $S_d$ to $S_c$ is given by:
$$
\mathbf{M}_{cd} = \begin{bmatrix}
\cos\gamma & 0 & -\sin\gamma & -L \\
0 & 1 & 0 & 0 \\
\sin\gamma & 0 & \cos\gamma & -N \\
0 & 0 & 0 & 1
\end{bmatrix}
$$
And the inverse transformation from $S_c$ to $S_d$ is:
$$
\mathbf{M}_{dc} = \begin{bmatrix}
\cos\gamma & 0 & \sin\gamma & L\cos\gamma + N\sin\gamma \\
0 & 1 & 0 & 0 \\
-\sin\gamma & 0 & \cos\gamma & -L\sin\gamma + N\cos\gamma \\
0 & 0 & 0 & 1
\end{bmatrix}
$$
This novel scheme offers significant advantages over traditional cold rotary forging for hyperboloid gears: targeted application, simplified die structure, reduced forming force due to line contact, and improved fiber orientation perpendicular to the principal stress direction, thereby enhancing bending fatigue strength. The finished tooth surface requires no subsequent machining.
Interference Checking Between the Die and Virtual Grinding Wheel
The accuracy of this new forging scheme hinges on a critical condition: the envelope surface generated by the contact line between the trapezoidal die and the theoretical tooth surface during one nutation cycle must perfectly coincide with the grinding wheel surface used in formate grinding. Any deviation—separation or intersection—would result in an incomplete or damaged tooth form. Therefore, rigorous interference checking between the die’s forming surface and the virtual grinding wheel surface is essential.
I employ a distance-based collision detection algorithm. The fundamental principle is to calculate the distance from sample points on the die’s forming surface to the axis of the virtual grinding wheel and compare it with the local radius of the grinding wheel at that axial height. Interference occurs if this distance is less than the local wheel radius for points that should not be in contact.
For a point on the die’s conical forming surface parameterized in $S_d$ as:
$$
\mathbf{r}_d(s_d, \theta_d) = \begin{bmatrix}
(r_d – s_d \sin\alpha_d) \cos\theta_d \\
(r_d – s_d \sin\alpha_d) \sin\theta_d \\
z_d – s_d \cos\alpha_d \\
1
\end{bmatrix}
$$
where $r_d$ and $z_d$ are the coordinates of the cone’s apex point (A or C in Fig. 5 of original context), $s_d$ and $\theta_d$ are surface parameters, and $\alpha_d$ is the die pressure angle (positive for inner cone forming the convex flank, negative for outer cone forming the concave flank). Its coordinates in the grinding wheel system $S_c$ are:
$$
\mathbf{r}_c^{(d)}(s_d, \theta_d) = \mathbf{M}_{cd} \cdot \mathbf{r}_d(s_d, \theta_d)
$$
The distance $r$ from this transformed point to the grinding wheel axis ($Z_c$) is:
$$
r = \sqrt{x_c^2(s_d, \theta_d) + y_c^2(s_d, \theta_d)}
$$
The local radius $r_d$ of the grinding wheel (a conical surface) at the same height $z_c$ is:
$$
r_d = r_{c2} + z_c \cdot \tan\alpha_c
$$
where $r_{c2}$ is the grinding point radius and $\alpha_c$ is the wheel pressure angle.
The non-interference condition must be checked only in regions where contact is possible. For points on the inner die cone (corresponding to the gear’s convex flank), which should only contact along the designated line (e.g., line AB), the condition to avoid unwanted interference elsewhere is $r > r_d$. Conversely, for points on the outer die cone (corresponding to the concave flank), the condition is $r < r_d$ outside the designated contact line (e.g., line CD).
Applying this algorithm to a specific case with the adjustment parameters below confirms the feasibility. The calculated surfaces show perfect line contact without collision, proving the kinematic accuracy of the proposed scheme for manufacturing hyperboloid gears.
| Formate Wheel Radius (mm) | Inner Flank Pressure Angle $\alpha_{cn}$ (°) | Outer Flank Pressure Angle $\alpha_{cw}$ (°) | Wheel Point Width (mm) | Tilt Angle $\gamma$ (°) | Horizontal Offset $H_2$ (mm) | Vertical Offset $V_2$ (mm) |
|---|---|---|---|---|---|---|
| 95.25 | 17 | -24 | 2.286 | 2 | 41.11 | 83.66 |
| Axial Workpiece Correction (mm) | Gear Axis Angle $\gamma_2$ (°) | Pendulum Apex Offset $L$ (mm) | Pendulum Apex Offset $N$ (mm) | |||
| -1.36 | 68.1333 | 0 | -12 | |||
The die pressure angles are adjusted due to the tilt: $\alpha_{dn} = \alpha_{cn} + \gamma$ and $\alpha_{dw} = \alpha_{cw} + \gamma$.
Simulation Validation of the Proposed Scheme
To further validate the proposed cold rotary forging strategy for hyperboloid gears, comprehensive simulation studies were conducted. A three-dimensional model was first constructed in UG NX 6.0 based on the geometric parameters of a typical automotive drive axle large hyperboloid gear.
| Number of Teeth | Face Width (mm) | Outer Cone Distance (mm) | Whole Depth (mm) | Pitch Apex to Crossing Point (mm) | Face Apex to Crossing Point (mm) |
|---|---|---|---|---|---|
| 41 | 28 | 101.26 | 9.73 | -3.05 | -3.60 |
| Root Apex to Crossing Point (mm) | Pitch Angle (°) | Face Angle (°) | Root Angle $\gamma_2$ (°) | ||
| -1.36 | 73.70 | 74.8333 | 68.1333 | ||
This model was subsequently imported into ADAMS software to perform kinematics and dynamics analysis of the mechanism. The virtual prototype simulation confirmed the precise and interference-free motion of the pendulum head and die assembly relative to the workpiece, as theorized.
For the forming process analysis, the model was imported into DEFORM-3D via the STL interface. The finite element model consisted of the trapezoidal die (as a rigid body) and the workpiece (AISI 1045 steel) modeled as a nonlinear strain-hardening plastic material. The initial workpiece was meshed with approximately 288,386 elements and 66,489 nodes using an absolute meshing method. Key simulation parameters are summarized below:
| Parameter | Value |
|---|---|
| Workpiece Temperature | 20 °C (Room Temperature) |
| Pendulum Head Speed | 10 rpm |
| Tilt Angle ($\gamma$) | 2° |
| Friction Model | Shear Friction |
| Friction Coefficient | 0.12 |
| Simulation Step Size | 0.002 |
The simulation results were analyzed from multiple perspectives: the shape of the contact zone between the workpiece and die, the metal flow velocity field, and the forces acting on the die. The analysis clearly demonstrated the characteristic localized line contact inherent to the proposed method. Most notably, the steady-state maximum force $F_z$ on the die in the axial (Z) direction was found to be in the range of 101–126 kN. This represents a reduction of approximately 90% compared to the forces typically required in traditional full-contact cold rotary forging processes for similar components. This drastic reduction in forming force is a direct and significant advantage of the line-contact strategy, leading to lower energy consumption, reduced die wear, and the potential for using smaller, more economical forging equipment. The successful outcome of these multi-domain simulations—kinematic, dynamic, and plastic deformation—provides strong and conclusive evidence for the technical feasibility and practical advantages of the novel cold rotary forging scheme for manufacturing high-performance, fatigue-resistant hyperboloid gears.
Conclusion
1. Addressing the requirements of anti-fatigue manufacturing for automotive hyperboloid gears, a novel cold rotary forging scheme for the large driven gear has been proposed. This method fundamentally overcomes the drawbacks of complex die structure and high forming force associated with conventional cold rotary forging by employing a simplified trapezoidal die and localized line-contact plastic deformation.
2. The kinematic accuracy of this scheme is predicated on the precise coincidence between the envelope of the die-tooth contact line and the virtual grinding wheel surface. A rigorous distance-based interference checking algorithm confirmed that the die and virtual grinding wheel maintain the required instantaneous line contact without collision, validating the scheme’s foundational principle.
3. Comprehensive numerical simulations encompassing kinematics (ADAMS), dynamics, and the plastic forming process (DEFORM-3D) were successfully conducted. The results demonstrated the correct motion trajectory and, critically, revealed a drastic reduction (around 90%) in the required axial forging force compared to traditional methods. These simulations provide robust and multi-faceted verification of the feasibility, precision, and significant practical benefits of the proposed cold rotary forging strategy for enhancing the manufacturing quality and performance of hyperboloid gears.