In the development of modern high-performance transmission systems, particularly within sectors such as aerospace, the operational demands on cylindrical gears have escalated dramatically. These components are now routinely subjected to extreme conditions characterized by high rotational speeds, elevated ambient temperatures, and significant mechanical loads. Under such harsh service environments, one of the predominant failure modes that emerges is scuffing—a sudden, severe form of adhesive wear that compromises gear life and system reliability. Scuffing occurs when the lubricating film between meshing tooth flanks collapses, leading to localized welding and subsequent tearing of surface material. This phenomenon is intrinsically linked to the thermal conditions at the contact interface, making the accurate prediction of tooth surface temperature a cornerstone of anti-scuffing design.
Internationally, two principal theoretical methods have been established to assess the scuffing load capacity of cylindrical gears: the flash temperature method and the integral temperature method. These form the basis of standards such as the ISO/TR 13989 series. In many regions, including China, national standards have been developed based on these ISO technical reports. However, the existence of multiple, sometimes slightly diverging, national standards can lead to confusion and differing results in engineering practice. For instance, within the Chinese context, engineers may reference GB/Z 6413.1 (flash temperature method), GB/Z 6413.2 (integral temperature method), and the older, specialized aviation standard HB/Z 84.4. A clear understanding of the theoretical underpinnings, formulaic differences, and practical implications of these standards is therefore essential for reliable gear design. This article delves into a comprehensive comparison of these methods, focusing on their application to involute cylindrical gears, to elucidate their similarities, differences, and respective domains of appropriate use.
Theoretical Foundations: Flash Temperature vs. Integral Temperature
The core concept uniting all scuffing evaluation methods is the critical role of surface temperature. The theory postulates that scuffing initiates when the temperature at the gear contact surpasses a critical value specific to the material and lubricant combination. The total contact temperature, $ \Theta_{B} $, is considered to be the sum of two components: the bulk temperature, $ \Theta_{M} $, and a flash temperature rise, $ \Theta_{fl} $.
The flash temperature method, rooted in the pioneering work of Blok, focuses on the instantaneous temperature spike that occurs at the point of contact. As two gear teeth slide against each other under load, frictional heat is generated within a very small contact area and over a very short time. This causes a local, transient temperature increase—the flash temperature—superimposed on the general bulk temperature of the tooth. The method involves calculating this flash temperature at numerous points along the path of contact. The maximum calculated flash temperature, $ \Theta_{flmax} $, is then added to the bulk temperature to determine the maximum surface temperature, $ \Theta_{Bmax} $. Scuffing risk is evaluated by comparing $ \Theta_{Bmax} $ to an allowable scuffing temperature, $ \Theta_{S} $.
$$ \Theta_{Bmax} = \Theta_{M} + \Theta_{flmax} $$
The integral temperature method, introduced later by researchers like Winter, adopts a more averaged approach. It argues that scuffing is influenced by the thermal history of the entire engagement cycle rather than just an instantaneous peak. This method calculates a mean value of the flash temperature along the path of contact, denoted as the average flash temperature rise $ \Theta_{flaint} $. A weighted version of this average rise is then added to the bulk temperature to obtain the integral temperature, $ \Theta_{int} $. The weighting factor accounts for the greater influence of the average temperature level on the scuffing process.
$$ \Theta_{int} = \Theta_{M} + C_{2} \cdot \Theta_{flaint} $$
Here, $ C_{2} $ is an empirical weighting coefficient, typically taken as 1.5 for spur and helical cylindrical gears.
The fundamental distinction lies in their sensitivity. The flash temperature method is inherently sensitive to local temperature peaks, which might be caused by specific geometric features or localized high load. The integral temperature method, by employing an averaging process, is less sensitive to such peaks and is considered more suitable for cases where no severe localized heating is expected. This theoretical difference dictates their scopes of application, as reflected in the standards.
Formulaic and Coefficient Comparison Across Standards
The Chinese standards GB/Z 6413.1-2003 and GB/Z 6413.2-2003 are technically identical to ISO/TR 13989-1 and -2, respectively. The aviation standard HB/Z 84.4-1984 is also based on the integral temperature concept but introduces simplifications and specific coefficient values tailored for aviation-grade hardened steel cylindrical gears. A detailed side-by-side comparison of their governing equations and the correction factors they employ reveals significant insights.
Bulk Temperature Calculation
The bulk temperature $ \Theta_{M} $ represents the temperature of the tooth body just before it enters the contact zone, determined by the overall thermal equilibrium of the gearbox.
- GB/Z 6413.1 (Flash Temp): $$ \Theta_{M} = \Theta_{oil} + 0.47 \cdot X_{s} \cdot X_{mp} \cdot \frac{\int_{A}^{E} \Theta_{fl} \, d\Gamma_{y}}{\Gamma_{E} – \Gamma_{A}} $$ This formulation integrates the flash temperature along the path of contact and modifies it with a lubricant supply factor $ X_{s} $ and a contact ratio factor $ X_{mp} $.
- GB/Z 6413.2 (Integral Temp): $$ \Theta_{M} = \Theta_{oil} + C_{1} \cdot \Theta_{flaint} $$ Here, $ C_{1} $ is a constant (typically 0.7) accounting for heat exchange conditions, and $ \Theta_{flaint} $ is the pre-calculated average flash temperature rise.
- HB/Z 84.4 (Aviation Integral Temp): $$ \Theta_{M} = (\Theta_{oil} + C_{1} \cdot \Theta_{flaint}) \cdot X_{s} $$ This mirrors the GB/Z 6413.2 form but reintroduces the lubricant supply factor $ X_{s} $ as a multiplier.
Flash Temperature / Average Flash Temperature Rise
This term quantifies the temperature increase due to friction at the contact.
- GB/Z 6413.1: Calculates instantaneous flash temperature: $$ \Theta_{fl} = \mu_{m} \cdot X_{M} \cdot X_{J} \cdot X_{G} \cdot \frac{(X_{\Gamma} \cdot \omega_{Bt})^{0.75} \cdot v_{t}^{0.5}}{a^{0.25}} $$ where $ \mu_{m} $ is the mean coefficient of friction, and $ X_{M}, X_{J}, X_{G}, X_{\Gamma} $ are coefficients for heat partition, mesh-in, geometry, and load sharing, respectively.
- GB/Z 6413.2: Calculates the average rise directly: $$ \Theta_{flaint} = X_{\epsilon} \cdot \mu_{m} \cdot X_{M} \cdot X_{BE} \cdot X_{\alpha\beta} \cdot \frac{(K_{B\gamma} \cdot \omega_{Bt})^{0.75} \cdot v_{t}^{0.5}}{|a|^{0.25}} \cdot \frac{X_{E}}{X_{Q} \cdot X_{Ca}} $$ This incorporates factors for contact ratio ($ X_{\epsilon} $), tip geometry ($ X_{BE} $), pressure angle ($ X_{\alpha\beta} $), load distribution along lines of contact ($ K_{B\gamma} $), running-in ($ X_{E} $), mesh-in ($ X_{Q} $), and tip relief ($ X_{Ca} $).
- HB/Z 84.4: Uses a simplified average rise formula: $$ \Theta_{flaint} = \mu_{mc} \cdot X_{M} \cdot X_{BE} \cdot \frac{W_{t}^{0.75} \cdot \tilde{v}^{0.5}}{\tilde{a}^{0.25}} \cdot \frac{X_{\epsilon}}{X_{Q} \cdot X_{Ca}} $$ Notably, it uses a prescribed, constant mean friction coefficient $ \mu_{mc} $ (often 0.114) and omits several factors like $ X_{\alpha\beta} $, $ K_{B\gamma} $, and $ X_{E} $.
Critical (Scuffing) Temperature Calculation
The allowable temperature limit is derived from standardized gear tests (like the FZG test) and adjusted for the specific application.
- GB/Z 6413.1: $$ \Theta_{S} = 80 + (0.85 + 1.4 \cdot X_{W}) \cdot X_{L} \cdot (S_{FZG})^2 $$ Where $ X_{W} $ is the welding factor for the gear material, $ X_{L} $ is the lubricant factor, and $ S_{FZG} $ is the test load stage.
- GB/Z 6413.2: $$ \Theta_{intS} = \Theta_{MT} + X_{WrelT} \cdot C_{2} \cdot \Theta_{flaintT} $$ This uses the test gear’s bulk temperature $ \Theta_{MT} $ and average flash rise $ \Theta_{flaintT} $, modified by a *relative* welding factor $ X_{WrelT} $ which compares the actual gear material to the test gear material.
- HB/Z 84.4: $$ \Theta_{sint} = \Theta_{MT} + C_{2} \cdot X_{W} \cdot \Theta_{flaintT} $$ This resembles the GB/Z 6413.1 approach but uses the integral temperature framework’s test data.
The selection and calculation of coefficients are where the standards diverge most practically. The table below provides a comparative overview of the key influencing factors considered by each standard.
| Factor Category | GB/Z 6413.1 (Flash) | GB/Z 6413.2 (Integral) | HB/Z 84.4 (Aviation) |
|---|---|---|---|
| Load Application | Application Factor $ K_A $, Dynamic Factor $ K_V $ | Application Factor $ K_A $, Dynamic Factor $ K_V $ | Application Factor $ K_A $ |
| Load Distribution | Face Load Factor $ K_{H\beta} $ ($K_{B\beta}$), Transverse Load Factor $ K_{H\alpha} $ ($K_{B\alpha}$) | Face Load Factor $ K_{H\beta} $, Transverse Load Factor $ K_{H\alpha} $, Helix Load Factor $ K_{B\gamma} $ | Face Load Factor $ K_{B\beta} $, Transverse Load Factor $ K_{B\alpha} $, Helix Load Factor $ K_{B\gamma} $ |
| Geometry & Mesh | Contact Ratio Factor $ X_{mp} $, Geometry Factor $ X_G $ | Contact Ratio Factor $ X_{mp} $, Pressure Angle Factor $ X_{\alpha\beta} $, Tip Geometry Factor $ X_{BE} $, Contact Ratio Factor $ X_{\epsilon} $ | Tip Geometry Factor $ X_{BE} $, Contact Ratio Factor $ X_{\epsilon} $ |
| Surface & Running-in | Roughness Factor $ X_R $ | Roughness Factor $ X_R $, Running-in Factor $ X_E $ | – |
| Material | Welding Factor $ X_W $ | Relative Welding Factor $ X_{WrelT} $ | Welding Factor $ X_W $ |
| Lubrication | Thermal Flash Factor $ X_M $, Lubricant Factor $ X_L $, Supply Factor $ X_s $ | Thermal Flash Factor $ X_M $, Lubricant Factor $ X_L $, Supply Factor $ X_s $ | Thermal Flash Factor $ X_M $, Supply Factor $ X_s $ |
Key observations from this comparison are:
- HB/Z 84.4 Simplifications: The aviation standard uses a fixed, relatively high friction coefficient $ \mu_{mc} $ and omits factors for lubricant type ($ X_L $), running-in ($ X_E $), and pressure angle ($ X_{\alpha\beta} $). It also does not use the relative welding factor, applying the material factor directly. These simplifications, while streamlining calculation for a specific domain, can significantly impact results.
- GB/Z 6413.2 Comprehensiveness: The integral temperature standard is the most comprehensive, incorporating factors for helix load distribution and a more nuanced material factor via $ X_{WrelT} $.
- Method-Specific Factors: The flash temperature method (GB/Z 6413.1) uses a path integration approach and specific geometry factor $ X_G $, while the integral methods rely on pre-derived average coefficients like $ X_{BE} $.
Case Study Analysis Using FZG A-Type Cylindrical Gears
To quantify the differences between these standards, a systematic case study was performed using the well-known FZG A-type test gear geometry. This cylindrical gear pair has a center distance of 91.5 mm, a module of 4.5 mm, and a face width of 20 mm. The pinion has 16 teeth and the gear 24 teeth. The material was assumed to be case-hardened steel (18CrNiMo7-6), and calculations were performed for a standard mineral oil. The safety factor calculation was omitted to focus purely on the comparative output of the calculated surface or integral temperature from each standard.
Effect of Input Torque / Load Stage
The input torque to the pinion was varied corresponding to FZG load stages 3 through 12. The results for the calculated contact temperature are summarized below.
| Pinion Torque (Nm) | GB/Z 6413.1 $ \Theta_{Bmax} $ (°C) | GB/Z 6413.2 $ \Theta_{int} $ (°C) | HB/Z 84.4 $ \Theta_{sint} $ (°C) |
|---|---|---|---|
| 35.3 | 81.4 | 76.8 | 106.6 |
| 135.3 | 119.8 | 116.6 | 179.7 |
| 239.3 | 158.7 | 157.1 | 256.2 |
| 372.5 | 207.4 | 207.7 | 354.4 |
| 534.5 | 267.2 | 268.7 | 474.5 |
Analysis: Across the entire load range, HB/Z 84.4 predicts consistently and significantly higher temperatures—approximately 30-50% higher than GB/Z 6413.2 at the highest load. This is primarily attributable to its use of the fixed, high friction coefficient $ \mu_{mc} = 0.114 $, whereas GB/Z 6413.2 calculates a lower, viscosity-dependent coefficient (approximately 0.058 for this case). In contrast, the results from GB/Z 6413.1 and GB/Z 6413.2 are in very close agreement (within 5%), demonstrating that for this standard geometry and condition, both the flash and integral methods yield similar predictions of the critical thermal state.
Effect of Oil Sump Temperature
The bulk oil temperature was varied from 40°C to 120°C under a constant high load (FZG load stage 12).
| Oil Temperature (°C) | GB/Z 6413.1 $ \Theta_{Bmax} $ (°C) | GB/Z 6413.2 $ \Theta_{int} $ (°C) | HB/Z 84.4 $ \Theta_{sint} $ (°C) |
|---|---|---|---|
| 40 | 240.3 | 242.7 | 390.5 |
| 60 | 267.2 | 269.7 | 476.8 |
| 80 | 292.8 | 295.4 | 557.6 |
| 100 | 317.4 | 320.1 | 633.0 |
| 120 | 341.4 | 344.1 | 703.4 |
Analysis: The trend is consistent. HB/Z 84.4 results are markedly higher. The agreement between the two GB/Z standards remains excellent across the oil temperature range. The differential between HB/Z 84.4 and GB/Z 6413.2 widens with increasing oil temperature, exceeding 50% at 120°C.
Effect of Gear Module
This is a critical investigation as it changes the fundamental geometry of the cylindrical gear. The module was varied from 2 mm to 7 mm, with proportional adjustments to gear geometry to maintain a similar center distance and ratio, under a high load condition.
| Module (mm) | GB/Z 6413.1 $ \Theta_{Bmax} $ (°C) | GB/Z 6413.2 $ \Theta_{int} $ (°C) | HB/Z 84.4 $ \Theta_{sint} $ (°C) |
|---|---|---|---|
| 2.0 | 557.7 | 685.1 | 1185.4 |
| 3.0 | 382.0 | 423.5 | 746.6 |
| 4.0 | 295.6 | 306.2 | 542.0 |
| 4.5 | 267.2 | 269.7 | 476.8 |
| 6.0 | 212.6 | 202.7 | 354.6 |
| 7.0 | 190.6 | 176.8 | 306.2 |
Analysis: This reveals the most significant finding regarding the two GB/Z standards. While HB/Z 84.4 remains the most conservative, the results for GB/Z 6413.1 and GB/Z 6413.2 diverge as the module decreases. For small modules (2-4 mm), GB/Z 6413.2 predicts higher temperatures than GB/Z 6413.1, with a difference of over 18% at a 2 mm module. This is likely due to the different ways the standards handle the influence of surface roughness and lubricant factors for contacts with smaller radii of curvature. For modules in the 4-7 mm range, typical for many power transmission cylindrical gears, their results converge and are in close agreement.
Effect of Rotational Speed
Pinion speed was varied from 1,000 to 10,000 rpm under high load.
| Speed (rpm) | GB/Z 6413.1 $ \Theta_{Bmax} $ (°C) | GB/Z 6413.2 $ \Theta_{int} $ (°C) | HB/Z 84.4 $ \Theta_{sint} $ (°C) |
|---|---|---|---|
| 1,000 | 243.9 | 246.1 | 440.2 |
| 3,000 | 323.4 | 326.6 | 557.1 |
| 5,000 | 376.1 | 379.9 | 623.2 |
| 8,000 | 439.5 | 444.1 | 691.9 |
| 10,000 | 476.9 | 482.0 | 727.5 |
Analysis: The established pattern holds: HB/Z 84.4 is conservative, and the two GB/Z standards show remarkable consistency across a very wide speed range. The temperature increase from 1,000 to 10,000 rpm is substantial (over 95% for GB/Z methods), highlighting the critical impact of sliding velocity on the thermal load of high-speed cylindrical gears.
Conclusions and Engineering Implications
This detailed comparative study of scuffing calculation standards for involute cylindrical gears leads to several definitive conclusions with direct implications for design and analysis:
- Conservatism of HB/Z 84.4: The specialized aviation standard HB/Z 84.4 consistently yields the highest predicted tooth temperatures, often 30-50% greater than those from GB/Z 6413.2 under identical conditions. This conservatism stems primarily from its use of a fixed, high mean friction coefficient ($ \mu_{mc} $) and the omission of several moderating factors such as the lubricant factor ($ X_L $) and running-in factor ($ X_E $). While this may provide a simple safety margin for its intended aerospace applications, it may also lead to over-designed components if applied generically.
- Convergence of Flash and Integral Methods (GB/Z 6413): For a wide range of operational parameters—including torque, oil temperature, and rotational speed—the calculated critical temperatures from GB/Z 6413.1 (flash temperature method) and GB/Z 6413.2 (integral temperature method) are in very close agreement (typically within 5%). This suggests that for standard cylindrical gear geometries and conditions where localized extreme flash heating is not a primary concern, either method can be used with confidence, and the choice may be based on convenience or specific standard compliance requirements.
- Critical Influence of Gear Module: A significant divergence between the two GB/Z standards occurs with small gear modules. For modules below approximately 4 mm, GB/Z 6413.2 (integral temperature) produces higher temperature estimates than GB/Z 6413.1. This indicates that the integral temperature method, with its specific set of averaging coefficients, may be more sensitive to the altered contact conditions in fine-pitch gears. For modules between 4 mm and 7 mm, their results are consistent. This finding is crucial for designers working with small-module, high-precision cylindrical gears, as the choice of standard could meaningfully affect the scuffing risk assessment.
- Parametric Sensitivity: The case studies quantitatively reaffirm the profound influence of operating conditions on the thermal state of cylindrical gears. Increasing the load (torque) from a mild to a severe stage can increase the calculated surface temperature by 250% or more. Doubling or tripling the rotational speed can nearly double the temperature. Even a rise in bulk oil temperature has a substantial additive effect. This underscores the necessity of precise operational data for accurate scuffing analysis.
In summary, the selection of a scuffing calculation standard for cylindrical gears should be a deliberate choice. GB/Z 6413.2 (integral temperature) offers a comprehensive and widely accepted approach suitable for most general power transmission applications. GB/Z 6413.1 (flash temperature) remains valuable, particularly for analyzing gears where local geometry might induce sharp thermal peaks. HB/Z 84.4, while simpler, provides a highly conservative result rooted in its historical aerospace context. Understanding these differences enables engineers to apply the most appropriate tool, ensuring both the reliability and the efficiency of their cylindrical gear designs.