The pursuit of high-performance power transmission has led to the development of advanced screw gear configurations. Among these, the worm drive family offers unique advantages for achieving high reduction ratios in compact spaces. This analysis focuses on a specific variant: the inclined double-roller enveloping hourglass worm drive. This configuration is engineered to address limitations found in related designs, such as the thin root of the worm in zero-backlash versions or the sensitivity to radial installation errors in tapered-roller designs. The core innovation lies in the worm wheel, which is constructed from two components—a fixed wheel and a movable wheel. Each wheel carries a circumferential array of cylindrical rollers, with their axes intentionally inclined relative to the radial direction of the wheel. The worm thread is generated by the enveloping motion of these rollers, with left and right flanks formed by the rollers on the fixed and movable wheels, respectively. While a single set of rollers inherently possesses backlash, ensuring operational freedom and lubrication, the overall system uses two staggered rows to eliminate kinematic reversal error, thereby enhancing transmission accuracy and smoothness.
The performance and longevity of any screw gear system are fundamentally tied to the lubrication conditions at the contacting tooth surfaces. In high-precision worm drives, the lubricant film thickness is often on the order of micrometers, comparable to the surface roughness resulting from standard manufacturing processes like grinding. Therefore, an analysis assuming perfectly smooth surfaces may be insufficient. The objective here is to investigate the Elasto-Hydrodynamic Lubrication (EHL) characteristics of the conjugate tooth pair in this inclined double-roller worm gear, explicitly accounting for the influence of surface roughness. The analysis is conducted from a first-person research perspective, detailing the model development, numerical solution, and discussion of results for this specific type of screw gear.
The kinematic and geometric model of this worm drive is foundational for the lubrication analysis. The worm wheel tooth surface is essentially the surface of the cylindrical roller with radius \(R_k\). The worm tooth surface is the envelope generated by this roller family. During meshing, multiple tooth pairs are in contact simultaneously, each along a spatially complex contact line. For EHL modeling, the contact at any instant is simplified to a line contact problem between two equivalent cylinders. The equivalent or reduced radius of curvature \(R\) at the contact point is crucial and is derived from the principal curvatures of the worm and wheel surfaces. It is given by the inverse of the induced normal curvature \(k_{\sigma}\):
$$R = \frac{1}{k_{\sigma}}$$
The induced normal curvature is determined through differential geometry and the kinematics of the enveloping process. The velocities of the worm (\(v_g\)) and the roller/wheel (\(v_w\)) at the contact point along the common normal are required to calculate the entrainment or rolling velocity, a key parameter in film formation. These are not trivial and depend on the specific design parameters and instantaneous position. They can be expressed as:
$$v_w = \frac{v_{11} (v_{12}^1 / R_k – \omega_{12}^2) + v_{12} \omega_{12}^1}{\sqrt{(v_{12}^1 / R_k – \omega_{12}^2)^2 + (\omega_{12}^1)^2}}$$
$$v_g = \frac{v_{21} (v_{12}^1 / R_k – \omega_{12}^2) + v_{22} \omega_{12}^1}{\sqrt{(v_{12}^1 / R_k – \omega_{12}^2)^2 + (\omega_{12}^1)^2}}$$
where \(v_{11}, v_{12}, v_{21}, v_{22}, v_{12}^1, \omega_{12}^1, \omega_{12}^2\) are functions of design geometry (center distance \(A\), roller offset \(c_2\), inclination angle \(\gamma\), etc.), transmission ratio \(i_{21}\), worm rotation, and the roller’s generating parameters \((u, \theta)\). The entrainment velocity \(U\) for the EHL model is then:
$$U = \frac{v_w + v_g}{2}$$
The load distribution across the contacting teeth is another critical input. The total transmitted torque is shared among the simultaneously engaged tooth pairs. The load per unit length \(w\) on a specific conjugate pair is:
$$w_i = \frac{F_{n_i}}{L} = \frac{2 K_i T_1}{L d_1 \cos \alpha_n \cos \beta}$$
where \(i\) indexes the contact line, \(K_i\) is the load-sharing factor, \(L\) is the instantaneous length of the contact line, \(\alpha_n\) is the normal pressure angle, \(\beta\) is the lead angle, \(T_1\) is the input torque, and \(d_1\) is the worm pitch diameter. For a typical screw gear like this, the parameters \(R\), \(U\), and \(w\) vary significantly as a tooth pair moves from entry to exit in the mesh zone.
The physical EHL problem is modeled as the contact between an equivalent elastic cylinder of radius \(R\) and a rigid plane, representing the localized interaction between the worm and roller surfaces. The governing equations for isothermal, line-contact EHL with Newtonian fluid behavior are employed. These equations are rendered dimensionless to facilitate numerical solution. The dimensionless parameters are defined as: \(X = x/b\), \(P = p/p_h\), \(H = hR/b^2\), \(\bar{\eta} = \eta/\eta_0\), \(\bar{\rho} = \rho/\rho_0\), \(W = w/(E’ R)\), \(U = \eta_0 u_e/(E’ R)\), \(G = \alpha E’\). Here, \(b\) is the Hertzian half-width, \(p_h\) is the maximum Hertzian pressure, \(E’\) is the effective elastic modulus, and \(\eta_0\) and \(\rho_0\) are the ambient viscosity and density.
The core system of equations includes:
1. The Reynolds Equation:
$$\frac{d}{dX}\left(\epsilon \frac{dP}{dX}\right) = \frac{d(\bar{\rho} H)}{dX} + \frac{\partial (\bar{\rho} H)}{\partial T}$$
where \(\epsilon = \frac{\bar{\rho} H^3}{\bar{\eta} \lambda}\) and \(\lambda = \frac{12 \eta_0 U R^2}{p_h b^2}\).
2. The Film Thickness Equation (including roughness):
The surfaces of the worm and roller are not perfectly smooth. A transverse roughness pattern is assumed for both surfaces, modeled by a simple cosine function:
$$s_w(x,t) = A_w \cos\left[\frac{2\pi}{l_w}(x – v_w t)\right]$$
$$s_g(x,t) = A_g \cos\left[\frac{2\pi}{l_g}(x – v_g t)\right]$$
where \(A_w, A_g\) are roughness amplitudes and \(l_w, l_g\) are wavelengths. The combined dimensionless roughness function is \(S(X,T)\). The film thickness equation, which includes elastic deformation and this roughness, is:
$$H(X) = H_0 + \frac{X^2}{2} – \frac{1}{2\pi} \int_{X_{in}}^{X_{out}} \ln|X – X’| P(X’) dX’ – S(X)$$
3. The Viscosity-Pressure Equation (Roelands):
$$\bar{\eta} = \exp\left\{ (\ln(\eta_0) + 9.67) \left[ \left(1 + 5.1 \times 10^{-9} p_h P\right)^{z} – 1 \right] \right\}, \quad z = \frac{\alpha}{5.1 \times 10^{-9}(\ln(\eta_0) + 9.67)}$$
4. The Density-Pressure Equation (Dowson-Higginson):
$$\bar{\rho} = \frac{1 + \frac{0.6 \times 10^{-9}}{p_h} P}{1 + \frac{1.7 \times 10^{-9}}{p_h} P}$$
5. The Load Balance Equation:
$$\int_{X_{in}}^{X_{out}} P(X) dX = \frac{\pi}{2}$$
To solve this highly nonlinear integro-differential system, the multi-grid method is employed due to its computational efficiency for EHL problems. The pressure is solved using a full approximation scheme (FAS) on a multi-grid hierarchy, while the film thickness integral is computed using the multi-level multi-integration (MLMI) technique. The discrete Reynolds equation on a grid with spacing \(\Delta X\) is:
$$\frac{\epsilon_{i-1/2} P_{i-1}^{k+1} – (\epsilon_{i-1/2}+\epsilon_{i+1/2}) P_i^{k+1} + \epsilon_{i+1/2} P_{i+1}^{k+1}}{\Delta X^2} = \frac{\bar{\rho}_i^* H_i^{k+1} – \bar{\rho}_{i-1}^* H_{i-1}^{k+1}}{\Delta X} + \frac{\bar{\rho}_i^* H_i^{k+1} – \bar{\rho}_i^* H_i^{k}}{\Delta T}$$
The film thickness equation is discretized as:
$$H_i = H_0 + \frac{X_i^2}{2} + \frac{1}{\pi} \sum_{j=1}^{n} K_{ij} P_j – S(X_i)$$
where \(K_{ij}\) are the discrete influence coefficients. The load balance is enforced as:
$$\Delta X \sum_{j=1}^{n-1} \frac{P_j + P_{j+1}}{2} = \frac{\pi}{2}$$
The solution process iteratively adjusts the rigid film thickness \(H_0\) until the pressure distribution satisfies both the Reynolds and load balance equations within a specified tolerance. A W-cycle is typically used across 5-6 grid levels, with Gauss-Seidel relaxation on each level. The initial guess for a given meshing instant often uses the converged solution from the previous instant to speed up convergence.
The analysis is performed for a specific set of design and operating parameters, as summarized in the following table:
| Parameter | Symbol | Value |
|---|---|---|
| Number of Worm Threads | \(Z_1\) | 1 |
| Number of Wheel Teeth | \(Z_2\) | 25 |
| Center Distance | \(A\) | 125 mm |
| Roller Radius | \(R_k\) | 6.5 mm |
| Roller Offset | \(c_2\) | 7 mm |
| Throat Form Factor | \(k_1\) | 0.4 |
| Roller Inclination Angle | \(\gamma\) | 6° |
| Input Power | \(P_{in}\) | 5 kW |
| Input Speed | \(n_1\) | 1450 rpm |
| Ambient Viscosity | \(\eta_0\) | 0.028 Pa·s |
| Ambient Density | \(\rho_0\) | 870 kg/m³ |
| Pressure-Viscosity Coefficient | \(\alpha\) | 2.2 × 10⁻⁸ m²/N |
| Effective Elastic Modulus | \(E’\) | \(2.3 \times 10^{11}\) Pa |
| Roughness Amplitude (Worm & Roller) | \(A_w, A_g\) | 0.06 µm |
| Roughness Wavelength (Worm & Roller) | \(l_w, l_g\) | 12 µm |
The engagement cycle of one conjugate screw gear pair is discretized into 100 instants. Results for both smooth and rough surface assumptions are compared at key instants corresponding to single, double, triple, and quadruple-tooth contact phases. The pressure and film thickness profiles reveal significant effects of surface roughness.
For the smooth surface case, the pressure profile exhibits the classic EHL shape: a sharp pressure spike near the outlet followed by a rapid drop. The central film thickness is relatively flat. As the tooth pair moves through the mesh, the maximum Hertzian pressure \(p_{max}\) and the minimum film thickness \(h_{min}\) vary. Typically, \(p_{max}\) peaks and \(h_{min}\) reaches its lowest value when the contact is near the worm’s throat region, where the load per unit length is highest and the entrainment velocity is often at a minimum for this screw gear geometry.
Introducing surface roughness dramatically alters the pressure distribution within the high-pressure contact zone. Each microscopic asperity acts as a local obstruction to the flow, generating a corresponding local pressure peak. The pressure profile is no longer smooth but exhibits high-frequency oscillations superimposed on the overall Hertzian-EHL pressure curve. The amplitude of these oscillations is most pronounced during the middle stages of engagement (e.g., triple-tooth contact), where the nominal pressure is high. The global maximum pressure \(P_{max}\) for the rough case is consistently higher than that for the smooth case. This is a critical finding, as elevated pressure peaks can accelerate surface fatigue (pitting).
The film thickness is also affected, though its response is less instantaneous due to the integral nature of the deformation. The roughness causes the film to undulate, with local reductions in film thickness occurring under the asperities. Consequently, the minimum film thickness \(H_{min}\) in the rough contact is lower than its smooth-surface counterpart. This reduction effectively decreases the specific film thickness (\(\lambda = h_{min} / \sigma\), where \(\sigma\) is the composite roughness), pushing the lubrication condition towards the mixed or boundary regime, which is undesirable for wear and efficiency in a precision screw gear.
The influence of key design parameters of this screw gear on the rough EHL performance is investigated by varying one parameter at a time while holding others constant. The analysis is focused on the most critical meshing instant (often triple-tooth contact). The effects are summarized below:
| Design Parameter | Effect on Rough EHL Pressure | Effect on Rough EHL Film Thickness | Physical Interpretation |
|---|---|---|---|
| Roller Radius \(R_k\) (Increase) | Increased pressure fluctuations; higher max pressure peak. | Decreased minimum film thickness. | Larger roller reduces effective contact curvature, increasing Hertzian contact width and pressure for the same load. This wider, high-pressure zone interacts more severely with roughness, destabilizing the film. |
| Throat Form Factor \(k_1\) (Decrease) | Sharply increased pressure fluctuations and peak. | Significant decrease in minimum film thickness. | A smaller \(k_1\) leads to a smaller worm throat diameter, increasing local curvature and Hertzian pressure. This severely hampers the formation of a protective hydrodynamic film under rough conditions. |
| Roller Offset \(c_2\) (Increase) | Increased pressure fluctuations; higher max pressure. | Decreased minimum film thickness. | Increased offset alters the meshing geometry, typically reducing the instantaneous entrainment velocity and/or increasing sliding, both detrimental to film formation and stability against roughness. |
| Inclination Angle \(\gamma\) (Increase) | Moderate increase in pressure fluctuations and peak. | Moderate decrease in film thickness. | A larger inclination angle changes the contact line orientation and kinematics. It generally increases the sliding component of motion relative to rolling, which is less favorable for building a thick EHL film and makes the film more susceptible to roughness disturbances. |
Among these, the throat form factor \(k_1\) exhibits the most pronounced influence on the EHL performance of this screw gear. The roller radius and offset have a secondary but significant effect, while the inclination angle, though influential, has a relatively smaller impact within a practical design range.
In conclusion, the elastohydrodynamic lubrication analysis of the inclined double-roller enveloping hourglass worm drive, accounting for realistic surface roughness, provides critical insights for the design and performance prediction of this advanced screw gear. The numerical modeling, based on a line-contact EHL formulation solved with the multi-grid method, successfully captures the complex interaction between the micro-geometry of roughness and the macro-scale lubrication physics. The key findings are:
- Roughness is Detrimental: Surface roughness cannot be neglected in the EHL analysis of precision worm gears. It induces high-frequency pressure oscillations, elevates the maximum contact pressure peak, and reduces the minimum lubricant film thickness compared to smooth-surface predictions. This pushes the lubrication condition towards a less favorable regime, increasing the risk of wear and surface-initiated fatigue.
- Critical Meshing Zone: The most severe lubrication condition, characterized by the thinnest film and highest pressure peaks under roughness, typically occurs during the middle phase of engagement (e.g., triple-tooth contact), when the load share and contact pressure are near their maximum.
- Design Parameter Sensitivity: The lubrication performance is highly sensitive to core screw gear design parameters. To promote robust EHL film formation and mitigate the adverse effects of roughness:
- The throat form factor \(k_1\) should not be chosen too small.
- The roller radius \(R_k\) and the roller offset \(c_2\) should not be excessively large.
- The roller inclination angle \(\gamma\) should be optimized, as larger angles tend to degrade lubrication.
A holistic design approach should balance these parameters through EHL-based optimization to ensure the screw gear operates with a sufficient and stable lubricant film.
This analysis underscores the importance of integrated tribological design in high-performance screw gear systems. Future work could extend this model to include thermal effects, non-Newtonian lubricant behavior, and a more detailed analysis of the transient loading and kinematics throughout the entire mesh cycle of the screw gear.