In the manufacturing of automotive components, heat treatment defects are a critical concern, particularly for gears where dimensional stability is paramount. Among these, flatness deformation in rear axle driven bevel gears after carburizing and quenching poses significant challenges, leading to issues such as premature wear, tooth breakage, increased noise, and reduced service life. As an engineer involved in this field, I have conducted extensive research to address these heat treatment defects by investigating the root causes and optimizing process parameters. This article presents a first-person account of our experimental study aimed at controlling flatness deformation through three key factors: the isothermal normalizing condition of gear blanks, the selection of quenching temperature, and the loading method during heat treatment. Our goal is to provide a comprehensive analysis using tables and formulas to summarize findings, ensuring that the insights can be applied to mitigate similar heat treatment defects in industrial settings.
Heat treatment defects, especially distortion, are inevitable in gear production due to the complex interplay of thermal and transformation stresses. For rear axle driven bevel gears, flatness deformation directly impacts gear accuracy and performance. The primary mechanisms behind these defects include non-uniform microstructures, thermal gradients during heating and cooling, and creep under high-temperature loading. To tackle this, we employed a systematic approach involving orthogonal experiments to analyze the effects of various factors. The material used was 20CrMnTi steel, a common choice for automotive gears due to its good hardenability and carburizing properties. The technical requirements included a case depth of 1.5–1.9 mm, surface hardness of 58–63 HRC, core hardness of 320–450 HV30, and flatness limits of ≤0.06 mm for the outer plane and ≤0.10 mm for the inner plane. These specifications are crucial for minimizing heat treatment defects and ensuring gear reliability.
The equipment utilized in our study included an isothermal normalizing line for preprocessing gear blanks and a double-row continuous carburizing line for the final heat treatment. The carburizing line consisted of heating, soaking, carburizing, diffusion, and quenching zones, with automated control over carbon potential, temperature, and cycle programs. To design the experiments, we adopted a full factorial orthogonal design with three factors at two levels each, resulting in eight experimental groups. Each group comprised 120 gear samples to ensure statistical significance, and we measured flatness using precision instruments while examining microstructures and hardness to correlate with heat treatment defects. This rigorous methodology allowed us to isolate the impact of each factor on deformation.
One of the primary contributors to heat treatment defects is the initial microstructure of the gear blank. Isothermal normalizing enhances the uniformity and stability of the microstructure, which in turn reduces irregular deformation during subsequent carburizing and quenching. The process involved heating the blanks to 930°C for 4 hours, followed by rapid cooling for 7 minutes, and then isothermal holding at 650°C for 4 hours before air cooling. We compared gear blanks that underwent isothermal normalizing with those that did not, assessing their microstructures and hardness. The results showed a significant improvement: normalized blanks exhibited a Grade 2 pearlite-ferrite structure with a hardness of 177 HB, whereas non-normalized blanks had a Grade 5 structure with 163 HB. This uniformity is vital for minimizing heat treatment defects, as it ensures consistent transformation behavior during quenching.
To quantify the effect on flatness deformation, we processed 240 normalized and 240 non-normalized gears using the same carburizing and quenching parameters, with variations in loading methods. The carburizing was performed at 930°C with a carbon potential profile, followed by direct quenching in fast oil. The flatness measurements revealed a stark contrast in defect rates, as summarized in Table 1. The data clearly indicates that isothermal normalizing substantially reduces heat treatment defects, with normalized gears achieving up to 98.3% compliance in flatness specifications. This underscores the importance of preprocessing in controlling deformation.
| Loading Method | Condition | Compliance (Outer ≤0.06 mm, Inner ≤0.10 mm) (%) | Marginal Deformation (Outer 0.07–0.10 mm, Inner 0.11–0.15 mm) (%) | Excessive Deformation (Outer >0.10 mm, Inner >0.15 mm) (%) |
|---|---|---|---|---|
| Vertical (Hanging) | Normalized | 74.2 | 18.3 | 7.5 |
| Vertical (Hanging) | Non-Normalized | 65.0 | 20.8 | 14.2 |
| Horizontal (Flat) | Normalized | 98.3 | 1.7 | 0.0 |
| Horizontal (Flat) | Non-Normalized | 53.3 | 26.7 | 20.0 |
The second factor we investigated was the quenching temperature, which directly influences thermal and transformation stresses, key drivers of heat treatment defects. Higher quenching temperatures can lead to coarser martensite, increased stress, and greater distortion. We tested two temperatures, 850°C and 870°C, using the same carburizing cycle and loading method for consistency. The carburizing process involved zones at 900°C, 910°C, 930°C, and 900°C with controlled carbon potentials, followed by quenching at the specified temperature. Flatness compliance rates are presented in Table 2, showing that 850°C quenching resulted in a higher compliance rate of 98.3% compared to 89.2% at 870°C. This reduction in heat treatment defects at lower temperatures aligns with theoretical expectations, as the stress magnitude can be approximated by the formula for thermal stress: $$\sigma_{thermal} = E \alpha \Delta T$$ where \(E\) is the Young’s modulus, \(\alpha\) is the coefficient of thermal expansion, and \(\Delta T\) is the temperature change during quenching. A lower \(\Delta T\) at 850°C reduces stress, thereby minimizing deformation.
| Quenching Temperature (°C) | Compliance (Outer ≤0.06 mm, Inner ≤0.10 mm) (%) | Marginal Deformation (Outer 0.07–0.10 mm, Inner 0.11–0.15 mm) (%) | Excessive Deformation (Outer >0.10 mm, Inner >0.15 mm) (%) |
|---|---|---|---|
| 850 | 98.3 | 1.7 | 0.0 |
| 870 | 89.2 | 10.0 | 0.8 |
Microstructural analysis supported these findings, with both temperatures yielding acceptable martensite and retained austenite (M+A’) levels of Grade 2–3 and core structures of lath martensite with minimal ferrite. However, the slight increase in deformation at 870°C highlights the sensitivity of heat treatment defects to quenching parameters. Additionally, the transformation strain during martensitic formation can be modeled as: $$\epsilon_{trans} = \beta \Delta V$$ where \(\beta\) is a material constant and \(\Delta V\) is the volume change. Lower temperatures mitigate this effect, contributing to better flatness control.
The third factor, loading method, addresses creep-induced deformation during prolonged high-temperature exposure. Gears can deform under their own weight during carburizing, and the method of placement affects how they enter the quenchant, influencing cooling uniformity. We compared horizontal (flat) loading and vertical (hanging) loading, as illustrated in Figure 1 (refer to the image link for visualization). The results, summarized in Table 3, demonstrate that horizontal loading significantly reduces heat treatment defects, with a compliance rate of 98.3% versus 74.2% for vertical loading at 850°C quenching. This is because horizontal loading minimizes gravitational creep and ensures symmetric cooling, whereas vertical loading can lead to uneven stress distribution. The creep strain rate at high temperature can be described by the Norton-Bailey law: $$\dot{\epsilon}_{creep} = A \sigma^n e^{-Q/RT}$$ where \(A\) and \(n\) are constants, \(\sigma\) is stress, \(Q\) is activation energy, \(R\) is the gas constant, and \(T\) is temperature. Horizontal loading reduces \(\sigma\), thereby lowering creep-induced deformation.
| Loading Method | Compliance (Outer ≤0.06 mm, Inner ≤0.10 mm) (%) | Marginal Deformation (Outer 0.07–0.10 mm, Inner 0.11–0.15 mm) (%) | Excessive Deformation (Outer >0.10 mm, Inner >0.15 mm) (%) |
|---|---|---|---|
| Horizontal (Flat) | 98.3 | 1.7 | 0.0 |
| Vertical (Hanging) | 74.2 | 18.3 | 7.5 |
To integrate these factors, we conducted a full factorial orthogonal experiment, with the three factors—normalizing condition, quenching temperature, and loading method—each at two levels. The comprehensive data from all eight experimental groups are presented in Table 4. This table allows for an analysis of interactions between factors, revealing that the combination of isothermal normalizing, 850°C quenching, and horizontal loading yields the highest compliance rate, effectively minimizing heat treatment defects. The overall deformation can be expressed as a function of these factors: $$D = f(N, T, L)$$ where \(D\) is deformation magnitude, \(N\) is normalizing condition, \(T\) is quenching temperature, and \(L\) is loading method. Our empirical data suggest a synergistic effect, with normalizing providing a stable baseline, lower temperature reducing stresses, and horizontal loading mitigating creep.
| Loading Method | Quenching Temperature (°C) | Isothermal Normalized | Non-Normalized | ||||
|---|---|---|---|---|---|---|---|
| Compliance | Marginal | Excessive | Compliance | Marginal | Excessive | ||
| Vertical (Hanging) | 850 | 74.2 | 18.3 | 7.5 | 65.0 | 20.8 | 14.2 |
| 870 | 80.0 | 11.7 | 8.3 | 45.8 | 53.3 | 0.8 | |
| Horizontal (Flat) | 850 | 98.3 | 1.7 | 0.0 | 53.3 | 26.7 | 20.0 |
| 870 | 89.2 | 10.0 | 0.8 | 23.3 | 32.5 | 44.2 | |
The mechanisms behind these heat treatment defects are further elucidated by considering the combined stresses during quenching. The total stress \(\sigma_{total}\) can be approximated as the sum of thermal stress and transformation stress: $$\sigma_{total} = E \alpha (T_q – T_m) + K \Delta \gamma$$ where \(T_q\) is the quenching temperature, \(T_m\) is the martensite start temperature, \(K\) is a constant, and \(\Delta \gamma\) is the volume change due to phase transformation. By optimizing the factors, we reduce both components, leading to better flatness control. For instance, isothermal normalizing homogenizes the microstructure, reducing localized \(\Delta \gamma\), while lower \(T_q\) decreases \(E \alpha (T_q – T_m)\).
In production validation, we applied the optimized parameters—isothermal normalizing with microstructure controlled to Grade ≤2, 850°C quenching, and horizontal loading—to a batch of 2,200 gears. Random sampling of 200 blanks confirmed consistent hardness (168–179 HB) and microstructure (Grade 2). After carburizing and quenching, the flatness compliance rate reached 95.1%, with all gears meeting the technical requirements for case depth, hardness, and microstructure. This demonstrates the practical effectiveness of our approach in mitigating heat treatment defects on an industrial scale.
Throughout this study, we have emphasized the importance of addressing heat treatment defects through a multifaceted strategy. The recurrence of the term “heat treatment defects” in our analysis underscores their pervasive impact on gear quality. By leveraging experimental data and theoretical models, we have shown that controlling flatness deformation is achievable through precise management of preprocessing, quenching parameters, and loading configurations. Future work could explore advanced simulation techniques to predict deformation, but our findings provide a robust foundation for minimizing heat treatment defects in rear axle driven bevel gears and similar components.