My research focuses on a novel type of power transmission system. This article details my investigation into the contact mechanics of this specific gear geometry. Traditional worm drives often face limitations in load capacity and efficiency due to high sliding friction and limited contact lines. The design I studied, characterized by its unique generation principle, aims to overcome these limitations. By utilizing rollers as the teeth of the worm wheel, the meshing action transforms sliding friction into a combination of rolling and sliding, theoretically reducing wear and power loss. Furthermore, the enveloping generation process of the worm thread is designed to promote a greater number of simultaneous contact lines compared to conventional designs. The primary objective of this study was to verify the contact pattern and analyze the stress distribution under load for this roller-enveloping end face meshing screw gear set using the Finite Element Method (FEM). Establishing a reliable finite element model and understanding its stress behavior is a foundational step for developing high-load-capacity screw gear transmissions.
The geometry of the meshing surfaces is fundamental to any contact analysis. For the roller-enveloping screw gear, the worm surface is not a simple helicoid but a complex envelope generated by the family of roller surfaces in relative motion. To mathematically describe this, I established a series of coordinate systems, as shown conceptually in the figure below. A fixed coordinate system is attached to the worm, and another to the gear. Moving coordinate systems are linked to each rotating body. The key is the coordinate system attached to an individual roller. The worm thread surface is derived as the envelope of the roller family as it moves according to the kinematic relationship between the worm and the gear.
Let $\Sigma_1$ and $\Sigma_2$ be the fixed coordinate systems for the worm and gear, respectively, and $\Sigma_1’$ and $\Sigma_2’$ be their rotating counterparts. The roller geometry is defined in its local system $\Sigma_0$. The worm rotates with angular velocity $\omega_1$ and the gear with $\omega_2$, where their ratio defines the transmission ratio $i_{12} = \omega_1 / \omega_2 = z_2 / z_1$, with $z_1$ and $z_2$ being the number of worm threads and gear teeth. The rotation angles are $\phi_1$ and $\phi_2$, related by $\phi_2 = i_{21} \phi_1$ (where $i_{21} = 1/i_{12}$). The center distance is denoted by $a$.
The surface of a single roller in $\Sigma_0$ is given by a simple cylindrical equation:
$$ \mathbf{r}_0 = x_0 \mathbf{i}_0 + y_0 \mathbf{j}_0 + z_0 \mathbf{k}_0 $$
with the constraint $x_0^2 + y_0^2 = R^2$, where $R$ is the roller radius.
Through coordinate transformations based on the spatial kinematics of the screw gear set, the family of roller surfaces in the worm coordinate system $\Sigma_1’$ is obtained. According to gear meshing theory, the worm tooth surface is the envelope of this family, satisfying both the family equation and the equation of meshing. The derived worm surface $\mathbf{r}_1’$ in $\Sigma_1’$ is:
$$
\mathbf{r}_1′ = x_1 \mathbf{i}_1′ + y_1 \mathbf{j}_1′ + z_1 \mathbf{k}_1′
$$
where the components are:
$$
\begin{aligned}
x_1 &= -\cos \phi_1 \cos \phi_2 (a_2 – z_0) + \cos \phi_1 \sin \phi_2 \, x_0 – y_0 \sin \phi_1 + a \cos \phi_1, \\
y_1 &= \sin \phi_1 \cos \phi_2 (a_2 – z_0) – \sin \phi_1 \sin \phi_2 \, x_0 – y_0 \cos \phi_1 – a \sin \phi_1, \\
z_1 &= -\sin \phi_2 (a_2 – z_0) – \cos \phi_2 \, x_0.
\end{aligned}
$$
Here, $(a_2, b_2, c_2)$ defines the position of the roller center in the gear coordinate system, and the meshing condition provides the relation $\phi_2 = i_{21} \phi_1$. This mathematical model is essential for generating accurate CAD geometry for the subsequent finite element analysis of these complex screw gears.
To analyze the real contact behavior under load, theoretical Hertzian contact formulas are insufficient due to the complex, multi-tooth engagement and the spatial curvature of the contact lines. Therefore, I employed a nonlinear finite element contact analysis. The basic geometric parameters of the studied screw gear pair are summarized in Table 1.
| Parameter | Symbol | Value | Unit |
|---|---|---|---|
| Center Distance | $a$ | 125 | mm |
| Number of Worm Threads | $z_1$ | 1 | – |
| Number of Gear Teeth | $z_2$ | 25 | – |
| Roller Radius | $R$ | 11 | mm |
| Face Width Coefficient | $k_1$ | 0.3 | – |
| Addendum Coefficient | $h_a^*$ | 0.8 | – |
| Dedendum Coefficient | $h_f^*$ | 0.8 | – |
| Clearance Coefficient | $c^*$ | 0.2 | – |
The first step was to create a volumetric 3D model. I generated the spatial curve data for the worm surface using computational software (MATLAB) based on the derived equations. This data was then imported into a CAD system (Creo Parametric) to create a solid model of both the worm and the gear assembly with its multiple rollers. This assembly was crucial for studying the interaction in this multi-contact screw gear system.
For the finite element analysis, I used ANSYS Workbench. To manage computational resources while capturing the essential multi-tooth contact phenomenon, I created a partial model. This model consisted of one segment of the worm (approximately one axial pitch) and the section of the gear containing the five rollers that were in simultaneous meshing contact with that segment, as predicted by theory. The material properties assigned to both the worm and gear rollers were those of AISI 1045 steel, a common gear material. The properties are listed in Table 2.
| Property | SI Value | Simulation Units |
|---|---|---|
| Young’s Modulus, $E$ | 2.0 × 1011 Pa | 2.0 × 105 MPa |
| Poisson’s Ratio, $\nu$ | 0.3 | 0.3 |
| Density, $\rho$ | 7850 kg/m³ | 7850 kg/m³ |
The contact between the worm threads and the roller surfaces was defined as surface-to-surface contact. Since the primary focus was on contact pressure and stress distribution, and the coefficient of friction is relatively low in well-lubricated screw gears, I initially modeled the contact as “frictionless” to simplify the nonlinear convergence. The meshing process is dynamic, so I selected a Transient Structural analysis module. This allows for the simulation of the engagement process over a short period. A key aspect was the application of constraints and loads. The worm’s inner cylindrical surface was assigned a remote displacement constraint, allowing only rotation about its axis. A rotational velocity of 157 rad/s (approximately 1500 rpm) was applied to this surface. Conversely, the gear’s bore was also constrained to allow only rotation about its axis, and a resisting torque was applied to it. The magnitude of the torque was varied to simulate different load conditions. The contact forces are generated automatically by the solver based on the defined interactions and constraints, simulating the power transmission through the meshing screw gear teeth.
A high-quality mesh is critical for accurate contact results. I used a predominantly hexahedral mesh for better numerical stability. The mesh was refined specifically in the contact regions where high stress gradients were expected. After several convergence studies, the final model contained approximately 133,658 elements and 225,151 nodes, with an average element quality above 0.8, which is considered suitable for such an analysis. The boundary conditions and the meshed model are conceptually critical for understanding the loading state of the screw gear assembly.
I performed analyses for several input torque levels: 15.60 N·m, 31.90 N·m, 38.99 N·m, and 46.76 N·m. The primary results extracted were the equivalent (von Mises) stress, total deformation, and contact pressure. One of the most significant findings was the confirmation of the multi-pair contact. The results clearly showed that five roller teeth were sharing the load simultaneously during the simulated mesh cycle for this single-thread screw gear. The stress distribution was not uniform among these five pairs. Typically, the middle pair (Pair 3) carried the highest load, although under the highest torque, the entering pair (Pair 1) sometimes showed the highest stress. This load-sharing behavior is a desirable characteristic of this enveloping screw gear design, contributing to its potential for high load capacity. A summary of the maximum equivalent stress on the worm and gear for different torques is shown in Table 3.
| Input Torque (N·m) | Max Equivalent Stress – Worm (MPa) | Max Equivalent Stress – Gear Roller (MPa) |
|---|---|---|
| 15.60 | 38.58 | 22.14 |
| 31.90 | 64.12 | 36.32 |
| 38.99 | 82.05 | 46.45 |
| 46.76 | 100.20 | 56.80 |
As expected, the stress in both components increased with the applied torque. The worm consistently experienced higher stress than the gear rollers, indicating it is the critical component for strength analysis in this configuration. Under the nominal design torque of 31.90 N·m, the maximum worm stress was 64.12 MPa. The deformation followed a similar trend, with maximum deformation occurring at the outer edges of the worm segment. The total deformation increased by about 22% when the torque increased from 15.60 N·m to 46.76 N·m.
The contact pressure distribution along the contact lines provided deeper insight. For a given torque, the maximum contact pressure on each successive roller tooth, from the entering side (Pair 1) to the exiting side (Pair 5), generally increased. For example, at 46.76 N·m, the maximum contact pressure on Pair 5 was about 151% higher than on Pair 1. Furthermore, the location of the maximum pressure on each roller shifted from near the tooth tip (addendum) towards the tooth root (dedendum) as the meshing progressed from entry to exit. Under the nominal torque, the peak contact pressure was found to be 186.07 MPa. Analyzing the stress-time history for the most stressed node revealed a consistent pattern across loads: the stress rapidly increased during the initial engagement phase (first 0.1-0.3 seconds of simulated time) and then stabilized to a nearly constant value for the remainder of the analysis, indicating steady-state meshing conditions.
To understand the load distribution on a single tooth flank, I examined the equivalent stress along the predicted contact line on the third (middle) roller tooth for four load levels. The stress distribution consistently exhibited a “U”-shaped profile, with higher stresses at the ends of the contact line and lower stress in the middle. This is characteristic of edge effects and slight misalignment or deflection under load. Crucially, as the load increased, the stress level increased proportionally, but the “U”-shape profile was maintained, suggesting the contact remained stable. The highest stress on the roller flank was consistently found near the root region, identifying it as a potential critical area for fatigue in these screw gears.
| Input Torque (N·m) | Stress Increase on Pair 5 vs. Pair 1 |
|---|---|
| 15.60 | 122.4% |
| 31.90 | 143.2% |
| 38.99 | 147.9% |
| 46.76 | 151.1% |
In conclusion, my finite element contact analysis of the roller-enveloping end face meshing screw gear has yielded several important insights. First and foremost, the model successfully demonstrated the fundamental advantage of this design: multiple simultaneous lines of contact. The analysis confirmed that five pairs of teeth were actively sharing the load for the single-thread worm configuration studied. This multi-pair engagement is a direct contributor to the high potential load capacity of this type of screw gear drive. The contact pattern and stress distribution derived from the mathematical model were validated by the FEM results, showing line contact along the predicted spatial curves.
Second, the stress analysis under varying loads provided quantitative data. The relationship between applied torque and resulting stress was linear in the studied range. The worm was identified as the more critically stressed component. The contact pressure distribution showed a progressive load sharing from the entering to the exiting tooth pairs, with the middle and exiting pairs carrying higher loads. The “U”-shaped stress distribution along the contact line and the root region as the point of maximum stress on the roller provide critical guidance for future design optimization, such as proposing slight crown modifications to the rollers or optimizing the worm profile to achieve a more uniform pressure distribution.
The successful establishment and execution of this finite element model lay a solid foundation for further research. Future work will include incorporating friction into the model to estimate efficiency losses, performing dynamic analysis to study vibration characteristics, and conducting fatigue life prediction based on the contact stress results obtained. This comprehensive approach is essential for advancing the development of high-performance, high-load-capacity roller-enveloping screw gear drives for demanding industrial applications.