Dimensional changes during tempering

During tempering, much of the distortion due to elastic stresses from thermal and transformational stresses during quenching is removed due to stress relief.

0
612

In this column, I will discuss the dimensional changes that occur during tempering, with a short discussion of the underlying physical metallurgical changes.

Introduction

After quenching, the martensite present is hard and brittle. To reduce hardness, and to improve toughness, the part is tempered. This involves reheating the part to some temperature below the ferrite to austenite transition temperature and holding for a short period of time. This time varies, but a standard rule of thumb is one to two hours per inch of thickness.

There are three stages of tempering that a part will experience when continuously heated to the tempering temperature [1]:

(1) 80°-160°C: Precipitation of carbon rich ε-carbide. The carbon in the martensite is reduced to approximately 0.3 percent.

(2) 230°-280°C: Decomposition of retained austenite to bainite.

(3a) 160°-400°C: Formation and growth of cementite (Fe3C) at the expense of ε-carbide.

(3b) 400°-700°C: Continued growth and spheroidization of cementite.

For high chromium steels, the range of decomposition of retained austenite is moved to higher temperatures. The product of decomposition (martensite or bainite) depends on the tempering time and temperature. Bainite transforms isothermally, whereas the martensite will form upon cooling from the tempering temperature.

Secondary hardening, resulting from the precipitation of finely dispersed complex carbides, and the conversion of retained austenite to martensite, occurs at about 500°C. This occurs in high-alloy steels, most commonly containing molybdenum, tungsten, titanium, chromium, and vanadium.

Dimensional changes during tempering

In a hardened part prior to tempering, the shape of the part or distortion is dependent on transformational stresses and thermal stresses that have occurred during hardening. During the tempering process, martensite transforms to ferrite and cementite, creating a continuous change in volume (Figure 1). This would tend to result in the part assuming the original shape.

Figure 1: Schematic representation of the effect of microstructural changes to the volume change of a part being tempered [1].

An as-quenched part may not be 100 percent martensite, but a mixture of bainite and pearlite, depending on the quenching conditions. From the thermal stresses or transformational stresses occurring during quenching, there may be some plastic deformation. This would prevent the shape reverting to its original, prior to heat-treatment shape. The volume change associated with a complete transformation of retained austenite to various microstructures is shown in Figure 2 [2].

Figure 2: Volume change associated with transformation of austenite to different microstructures [2].

In the region of 230°-280°C, decomposition of retained austenite to bainite occurs. This results in an increase in volume. The increase in volume is dependent on the percentage of retained austenite present and the carbon content.

Depending on the amount of retained austenite present, and the amount of martensite transformed to tempered martensite, the change in volume could be negative or positive.

In high-alloy tool steels, when tempered at 500°-600°C, fine secondary carbides are precipitated. These complex carbides containing the alloying elements such as vanadium, molybdenum, and others, reduce the alloy content in the retained austenite. This also reduces the carbon content in the retained austenite. The reduction in alloying content, and the reduction of carbon in the retained austenite, increases the martensite start temperature in the remaining retained austenite. After tempering is complete, then the retained austenite present will transform to martensite, with an accompanying increase in volume.

Change in stress state

After quenching, there are considerable residual stresses. These stresses are caused by thermal stresses from the outside cooling faster than the inside, resulting in both elastic and plastic stresses. There are also transformational stresses, from the transformation of austenite to martensite (and other microstructures if transformation is not completely to martensite). These stresses are both elastic and plastic. The combination of these stresses contributes to the change in shape during quenching.

In a study by Brown and Cohen [3], a navy split ring, fabricated from SAE 52100, was heat treated and quenched. The as-quenched ring was stressed to 415 N/mm2 and tempered for one hour at different temperatures. After tempering, the rings were measured and the amount of stress reduction was measured (Figure 3). Even after tempering at low temperatures of 250°C, 85 percent of the stress was relieved. This would indicate that most of the distortion due to elastic stresses was removed. Any plastic stresses present would also be removed, but the distortion of the part remained.

Figure 3: Relaxation during tempering of as-quenched rings of SAE 52100 [3] [1].

Conclusion

During tempering, much of the distortion due to elastic stresses from thermal and transformational stresses during quenching is removed due to stress relief. Further, the decomposition of martensite to tempered martensite further reduces the distortion of the part. Some growth can occur in high-alloy steels or tool steels containing high alloy levels of chromium, vanadium, molybdenum, and tungsten due to the precipitation of fine complex precipitates during secondary hardening. Retained austenite decomposition to bainite can result in growth of the part during intermediate tempering temperatures.

Should there be any questions regarding this article, or suggestions for further columns, please contact the writer or editor. 

References

  1. K. E. Thelning, Steel and its Heat Treatment, 2nd ed., London: Butterworths & Co., 1984.
  2. B. S. Lement, Distortion in Tool Steel, Metals Park, OH: ASM, 1959.
  3. R. L. Brown and M. Cohen, “Stress Relaxation of Hardened Steel,” Metal Progress, no. February, pp. 66-71, 1962.