The electromechanical drive system of electric vehicles (EVs) presents distinct challenges compared to conventional powertrains. The absence of torque converters and clutchers results in an underdamped system directly subjected to stochastic torque and speed fluctuations from both the motor and wheels. This leads to highly time-variant and random loading on the transmission components, profoundly influencing the dynamic meshing process of gears. Such conditions can initiate and propagate surface failures like pitting, ultimately compromising the reducer’s reliability. Micro-pitting, a surface fatigue phenomenon characterized by microscopic pits typically 10–500 μm in diameter and less than 10 μm deep, is often the precursor to macro-pitting failure. Therefore, a detailed investigation into the influence of micro-pitting on the lubrication performance of helical gears within EV reducers is critically important.
Accurate determination of the load distribution along the contact line is the fundamental prerequisite for analyzing the lubrication of helical gears. Common methodologies include the percentage of contact lines method, the minimum elastic potential energy (MEPE) method, and the finite element method (FEM). While the percentage method is simplistic and less accurate, and FEM is computationally intensive, the MEPE method offers a favorable balance between accuracy and efficiency.
The tooth of a helical gear can be discretized into a series of independent spur gear slices along its axis. For a single spur gear slice, the total elastic potential energy \( U \) under load \( F \) consists of bending energy \( U_b \), compressive energy \( U_c \), and shear energy \( U_s \), derived from Timoshenko beam theory:
$$U = U_b + U_c + U_s = \frac{F^2}{B} U_e(\xi_c)$$
where \( B \) is the face width, and \( U_e(\xi_c) \) is a normalized function dependent on the gear geometry and the contact point parameter \( \xi_c \), which relates the radius of curvature at the contact point to the base pitch. For a meshing position with \( n \) simultaneous tooth pairs, the total potential energy \( U_T \) and the total load \( F_w \) are:
$$U_T = \sum_{i=1}^{n} U_i = \frac{1}{B} \sum_{i=1}^{n} F_i^2 U_{e,i}(\xi), \quad \sum_{i=1}^{n} F_i = F_w$$
According to the principle of minimum potential energy, the load \( F_i \) carried by the \( i \)-th tooth pair that minimizes \( U_T \) is given by:
$$F_i = \frac{U_{e,i}^{-1}(\xi)}{\sum_{j=1}^{n} U_{e,j}^{-1}(\xi)} F_w$$
For a helical gear, the load per unit length \( \gamma_k \) for the \( k \)-th slice can be expressed as:
$$\gamma_k(\xi) = \frac{\varepsilon_\beta \cos \beta_b}{B} \frac{ U_e^{-1}(\xi_k) }{ \int_{l_c} U_e^{-1}(\xi) d\xi }$$
where \( \varepsilon_\beta \) is the axial contact ratio, \( \beta_b \) is the base helix angle, and \( l_c \) is the total length of all contact lines. The contact geometry of meshing helical gears can be modeled as the contact between two equivalent conical bodies. The entrainment velocity \( u \), sliding velocity \( v \), and equivalent radius of curvature \( R_{12} \) at any point on the contact line are calculated from the respective radii \( R_1, R_2 \) and angular velocities \( \omega_1, \omega_2 \):
$$u = \frac{R_1 \omega_1 + R_2 \omega_2}{2}, \quad v = R_1 \omega_1 – R_2 \omega_2, \quad \frac{1}{R_{12}} = \frac{1}{R_1} + \frac{1}{R_2}$$
The governing equations for the transient thermal elastohydrodynamic lubrication (TEHL) model are established as follows. The Reynolds equation incorporating the squeeze-film effect is:
$$\frac{\partial}{\partial x}\left(\frac{\rho h^3}{12\eta}\frac{\partial p}{\partial x}\right) + \frac{\partial}{\partial y}\left(\frac{\rho h^3}{12\eta}\frac{\partial p}{\partial y}\right) = u\frac{\partial(\rho h)}{\partial x} + \frac{\partial(\rho h)}{\partial t}$$
The film thickness equation, accounting for elastic deformation \( v \) and micro-pitting topography \( h_p \), is:
$$h(x,y,t) = h_0(t) + \Delta h(t) + v(x,y,t) + h_p(x,y,t)$$
The elastic deformation is given by the Boussinesq integral:
$$v(x,y,t) = \frac{2}{\pi E’} \iint_\Omega \frac{p(x’,y’,t)}{\sqrt{(x-x’)^2 + (y-y’)^2}} dx’ dy’$$
A simplified spherical dimple model is used to represent a micro-pit. For a pit with maximum depth \( h_{\text{max}} \) and radius \( r_p \), centered at \( (a’, b’) \), the depth is:
$$h_p = \begin{cases}
0 & \text{if } (x-a’)^2 + (y-b’)^2 > r_p^2 \\
h_{\text{max}} – r + \sqrt{r^2 – (x-a’)^2 – (y-b’)^2} & \text{if } (x-a’)^2 + (y-b’)^2 \le r_p^2
\end{cases}$$
The lubricant’s rheological behavior is described by the viscosity-pressure-temperature (Roelands) equation and a density-pressure-temperature equation:
$$\eta = \eta_0 \exp\left( (\ln(\eta_0) + 9.67) \left[ -1 + (1+5.1\times10^{-9}p)^{z_0} \left( \frac{T-138}{T_0-138} \right)^{-s_0} \right] \right)$$
$$\rho = \rho_0 \left[ 1 + \frac{0.6 \times 10^{-9}p}{1+1.7 \times 10^{-9}p} – 0.00065 (T – T_0) \right]$$
The force balance equation and the simplified energy equation across the film are:
$$F_i(t) = \iint p(x,y,t) \, dx \, dy, \quad k_f \frac{\partial^2 T}{\partial z^2} = -\eta \left( \frac{\partial u}{\partial z} \right)^2$$
The flash temperature on the two gear surfaces, \( T_a \) and \( T_b \), is calculated using the moving point heat source method, incorporating a heat partition coefficient \( f(x,y,t) \). The mid-film temperature \( T_m \) is approximated based on the surface temperatures and the frictional heat flux \( q \).
The model is solved using the multigrid method with a full multigrid/multilevel multi-integration (FMG/MLMI) technique. The highest grid level contains 65 × 513 nodes. Convergence criteria are set to \( 10^{-3} \) for pressure and load, and \( 10^{-4} \) for temperature. The results are validated against the classical Dowson-Higginson line contact minimum film thickness formula:
$$H_{\text{min-cal}} = 2.65 \frac{G^{*0.54} U^{*0.7}}{W^{*0.13}}$$
The analysis focuses on a pair of helical gears from the input stage of a two-stage EV reducer. Key parameters are listed below.
| Parameter | Symbol | Value |
|---|---|---|
| Number of teeth (pinion/gear) | \( z_1 / z_2 \) | 26 / 75 |
| Face width | \( B \) | 31 mm |
| Normal module | \( m_n \) | 1.6 mm |
| Normal pressure angle | \( \alpha_n \) | 20° |
| Helix angle | \( \beta \) | 30.5° |
| Center distance | \( a \) | 93 mm |
| Operating & Lubricant Parameter | Symbol | Value |
|---|---|---|
| Input speed | \( n_1 \) | 4000 rpm |
| Input torque | \( T_{in} \) | 95 N·m |
| Ambient viscosity | \( \eta_0 \) | 0.04 Pa·s |
| Ambient density | \( \rho_0 \) | 870 kg/m³ |
| Ambient/Bulk temperature | \( T_0 \) | 303 K |
| Elastic modulus (both gears) | \( E \) | 206 GPa |
A comparative analysis of the load distribution calculated by the percentage of contact lines method and the MEPE method reveals a significant dependency on the total contact ratio. For a high-contact-ratio gear pair (\( \varepsilon_{\gamma} \approx 4.5 \)), both methods yield very similar, smoothly transitioning load-sharing patterns. However, for a low-contact-ratio pair (\( \varepsilon_{\gamma} \approx 1.7 \)), the results diverge considerably. This indicates that for low-contact-ratio helical gears, the MEPE method is recommended for higher accuracy, whereas for high-contact-ratio gears, either method is acceptable due to the averaging effect of multiple contacting teeth.
The influence of thermal effects on the lubrication of helical gears is profound. A comparison between isothermal and thermal solutions at a specific meshing instant (32% of the path of contact) shows that thermal effects lead to a higher pressure peak, which shifts slightly towards the inlet zone, and a significantly reduced minimum film thickness. This degradation is attributed to the temperature-induced reduction in lubricant viscosity, which diminishes its load-carrying capacity. At the pitch point, where sliding is zero, the thermal and isothermal results converge as frictional heating is negligible. However, away from the pitch point, significant temperature rises occur in the mid-film and on the gear surfaces, directly correlating with the slide-to-roll ratio.
The transient effect, accounting for the time-dependent term \( \partial(\rho h)/\partial t \) (squeeze film action), is also critical. Comparing transient and steady-state (which neglects this term) solutions at the mid-point of engagement shows that the transient pressure and film thickness are both lower than their steady-state counterparts. This is because the steady-state solution, lacking the damping effect of squeeze motion, overestimates the film pressure and consequently the elastic deformation, leading to a thicker calculated film.
The presence of surface micro-pitting introduces localized disturbances in the lubricant film. The effects of pit diameter and depth are systematically analyzed and summarized below.
| Micro-pitting Parameter Variation | Effect on Pressure Profile | Effect on Film Thickness Profile |
|---|---|---|
| Increasing Diameter (constant depth) | Reduces maximum pressure in the Hertzian zone; smaller pits located in converging gaps can generate higher local pressure spikes. | Larger pits create deeper, less deformed cavities; smaller pits show more pronounced edge deformation due to higher local pressure. |
| Increasing Depth (constant diameter) | Decreases pressure in the inlet region of the pit; sharply increases pressure at the exit region. Very deep pits can generate pressures exceeding the smooth-surface maximum. | Creates a steeper film thickness gradient. Increased pressure at the exit leads to larger local elastic deformation, thickening the film at the pit’s trailing edge. |
While a single micro-pit does not drastically alter the global pressure or film thickness distribution, it induces severe localized pressure spikes at its edges due to the abrupt change in geometry. This phenomenon is critical, as these stress concentrations can accelerate fatigue crack propagation, leading to the growth and eventual coalescence of micro-pits into macroscopic spalling failures. The analysis for multiple pits confirms that the primary risk lies in these localized edge effects rather than a global loss of load capacity.
In conclusion, this analysis of EV reducer helical gears highlights several key findings. For accurate lubrication analysis, the minimum elastic potential energy method is preferable for calculating load distribution in low-contact-ratio gears, while both methods are adequate for high-contact-ratio designs. Thermal effects significantly deteriorate the lubrication state by reducing film thickness and increasing pressure. Transient squeeze-film effects lower both pressure and film thickness compared to steady-state assumptions, providing a more conservative prediction. Micro-pitting, modeled as spherical dimples, primarily affects lubrication through localized disturbances. Larger pit diameters reduce the effective load-carrying capacity, while greater depths cause significant pressure redistribution with high spikes at the exit edge. Crucially, the non-smooth edges of micro-pits act as stress concentrators, generating localized pressure surges that can initiate the propagation of surface fatigue, thereby serving as a nucleus for the development of macro-pitting damage in helical gears.