Part orientation during quenching

In this column, I will discuss the effect of gear orientation as it enters the quench. This is a critical part of controlling the distortion and residual stress of a heat-treated part. Distortion and residual stress are not just a material or quenchant selection problem. It is also a fixturing and orientation problem.

Introduction

The problem of distortion is universal across industry. A study by Thoben [1] indicated that the 1995 losses from heat treatment alone in the German machine, automotive, and transmission industry exceeded 850M € (1 billion USD). This does not include the rest of Europe, Asia, or the Americas. The problem is truly immense.

During quenching, there is a volumetric contraction as the work piece cools. As austenite transforms, there is a volume expansion. Martensite expands the most, as a function of the carbon content. Other phases can form, such as bainite or pearlite, which have much smaller volumetric expansion. Generally outer fibers transform first, then inner fibers. This is further complicated by the effect of martensitic transformation start temperatures. Depending on the relative Ms temperatures and the thickness of a part, it is often possible that the core will transform first due to a much higher Ms temperature. As the remaining part transforms, a stress reversal occurs.

When a hot part is quenched, there are three coupled components that drive distortion. First, there is non-uniform heat transfer rate. Secondly, is phase transformation that occurs non-uniformly. This is transformation from austenite to martensite, bainite, or ferrite and pearlite. Lastly, there are mechanical constraints and stiffness differences within the part.

Part Orientation

Part orientation affects each of the components of heat transfer, phase transformation, and mechanical properties. Orientation changes which surfaces contact the quenchant first, and how the vapor blanket forms and detaches from the part. Further, gear teeth and other surfaces can retain or trap the vapor phase, slowing the quench rate. Sagging of the soft hot part can also occur while the part is at the austenitizing temperature, yielding a distorted part before it ever contacts the quench. In a full load, part contact can locally slow down the local quench rate, giving rise to local soft spots. Parts or baskets near the part can also shield the parts, effectively slowing down the quench rate.

Even for nominally symmetric parts, experiments show that simply changing from vertical to horizontal racking significantly changes residual stresses and distortion patterns. One study [1] on cylinders reported that orientation (horizontal vs vertical) and martensite‑start temperature were the dominant factors in hoop and radial stress development during quenching, more influential than immersion rate under the tested conditions.

Complex gears have non-uniform stiffness. Lightweight gear designs with reduced web thickness and asymmetrical webs showed more tilting and rim distortion than symmetric, thicker designs. For instance, increasing asymmetry in the web position increased rim tilt, especially during early quenching when upper and lower regions of the rim shrank asymmetrically [2].

With different heat-transfer rates on different portions of the part from orientation effects, different microstructure can result, or the timing of transformation can change, contributing to distortion and residual stresses [3] [4] [5].

Consider a carburized steel gear with a thin web and a relatively heavy rim. The case is carburized to a typical depth of 0.7 mm effective depth, then quenched in oil, in two different orientations — one vertical, and the other horizontal.

For the horizontal gear, the top side of the gear cools faster because the vapor phase is less stable, with rapid dissipation of the vapor phase. The bottom surface tends to capture and retain the film, and once nucleate boiling begins, the bubbles are trapped by the web, effectively slowing down heat transfer by insulating the part. The outer rim experiences roughly equal heat transfer around the periphery of the part. The net effect is that the part tends to dish (concave downward), and the rim may tilt, increasing part runout. However, the part tends to have good roundness [6] [2] [5]. This non-uniformity can be seen by examining the fluid flow around the part (Figure 1). One interesting detail in the figure is the flow through the center hole in the hub. Higher velocity is evident in the bore, with the highest velocity (highest heat transfer) toward the bottom of the bore. This would suggest that shrinkage of the bore may occur due to an earlier martensitic transformation from cooling faster.

In the second case, the gear in Figure 1 is racked vertically (Figure 2), with the gear supported through the central hub. In this case, the faces experience similar heat-transfer conditions on opposite faces. Depending on the stiffness of the gear, the distortion shifts from dishing, to ovalization of the gear, and out-of-roundness [5] [6] [2].

In this case, the velocity is greater along the bottom of the gear, with potential trapping of vapor from quenching along the upper rim. The bore has virtually no flow within it, with the upper surface potentially trapping the vapor film. Because of the asymmetry of the velocity (and heat transfer) between the upper and lower surfaces, the gear may become out-of-round. Racking gears vertically is important when the geometry of the gear is stiff enough that hanging will not ovalize the bore excessively, and flatness/runout is critical [7].

Generally, if you want the gear to be flat, it should be hung vertically, but if you want the gear to be round, the gear should be mounted horizontally. Because large gears (and their processing) vary a lot, orientation is usually validated empirically (trial runs, distortion mapping) rather than taken from a formula [8].

Conclusions

Part orientation during quenching affects distortion by changing heat transfer and quenchant flow around the part. This in turn shapes the temperature and transformation gradients. Even with identical material and quenchant, changing the orientation can alter the overall distortion.

Should you have any comments regarding this article, or suggestions for additional articles, please contact the editor or myself. 

References

1. D. S. MacKenzie and D. Lambert, “Effect of Quenching Variables on Distortion and Residual Stresses,” in Proceedings of the 22nd Heat Treating Society Conference and the 2nd International Surface Engineering Congress, Indianapolis, IN, 2003.
2. J. Kagathara, T. Lübben and M. Steinbacher, “Effect of Design Modification on the Distortion Behavior of a Complex Counter Gear,” Materials, vol. 14, pp. 1-18, 2021.
3. D. S. MacKenzie and B. L. Ferguson, “Effect of Alloy on the Distortion of Oil Quenched Automotive Pinion Gears,” in International Conference on Thermal Process Modeling and Computer Simulation, Shanghai, China, 2010.
4. D. S. MacKenzie, Z. Li and B. L. Ferguson, “Effect of Quenchant Flow on the Distortion of Carburized Automotive Pinion Gears,” HTM, vol. 63, no. 1, pp. 15-21, 2008.
5. A. L. Banka, B. L. Ferguson and D. S. MacKenzie, “Evaluation of Flow Fields and Orientation Effects Around Ring Geometries During Quenching,” J. Mat. Eng. Performance, vol. 22, no. 7, pp. 1816-1825, 2013.
6. D. H. Herring and G. D. Lindell, “Reducing Distortion in Heat-Treated Gears,” Thermal Processing, no. June, pp. 27-35, 2004.
7. N. Bugliarello, C. Zimmerman, S. Richardson, R. Perkins, D. McCurdy, D. Giessel and B. George, “Heat Treat Processes for Gears,” Gear Solutions, no. July, pp. 38-51, 2010.
8. B. L. Ferguson and Z. Li, “Typical Heat Treatment Defects of Gears and Solutions Using FEA Modeling,” Thermal Processing, no. April, pp. 34-41, 2013.
9. D. S. MacKenzie, A. Kumar, H. Metwally, S. Painganker, Z. Li and B. L. Ferguson, “Prediction of Distortion of Automotive Pinion Gears during Quenching Using CFD and FEA,” J. ASTM Intl., vol. 6, no. 1, pp. 1-10, 2009.