In this column, we will discuss the different methods of carburizing, with a focus on resultant metallurgy.
Carburizing is the addition of carbon to the surface of low-carbon steels. It is generally accomplished at temperatures between 850–1,000°C. Once quenched, the high-carbon surface layer yields a high hardness martensitic case with excellent wear and fatigue resistance. This carburized case surrounds the tough, low-carbon steel core. The case hardness is primarily a function of the carbon content. There is little advantage in increasing the carbon content beyond 0.65 percent to increase hardness, however. Higher carbon content can increase microstructural properties such as wear, sliding contact fatigue, and rolling contact fatigue. A carbon content that is too high can result in excessive carbide networks or massive carbides.
The most common method is gas carburizing. In this method, the parts are surrounded by a carbon-rich atmosphere which supplies the carbon to the part. In most cases, this is an endothermic atmosphere which contains approximately 40 percent hydrogen, 20 percent carbon monoxide, 0.5 percent carbon dioxide, and the balance nitrogen. Carbon is supplied to the part surface by the reaction:
The carbon potential of the atmosphere is related to the relative partial pressures of CO and CO2 in the atmosphere:
where K is the equilibrium constant.
Carburizing is performed industrially by heating parts up to the carburizing temperature, at the desired carbon potential (typically 0.8 percent C). There are three basic methods of carburizing (Figure 1). The part is then quenched and tempered. In some cases, for very deep case depths, the part is carburized at high temperature (927°C and above) to shorten process times. The part is then cooled slowly to ambient temperatures, then reheated to an intermediate temperature (815-871°C) to allow the carbon to diffuse to obtain the desired carbon profile. After soaking, the part is then quenched and tempered. This approach has the advantages of a final austenite grain size that is smaller, improving ductility. The low reheat temperature places an upper limit on the amount of carbon dissolved in the austenite and reduces the amount of retained austenite in the case. The primary disadvantage of this method is the potential for increased distortion. Often, parts processed by this method are quenched in a press to reduce distortion.
For lighter case depths found in automotive (<0.6 mm), an alternative process was developed called the boost-diffuse technique. In this method, the part is introduced to the furnace at the carburizing temperature, with the carbon potential set near the saturation point of carbon in austenite for a short period of time. This brings the surface carbon well above the desired surface carbon of 0.8 percent. This is called the boost phase. The carbon potential is then dropped to the desired surface carbon potential of 0.8 percent. The carbon is then allowed to diffuse to the desired case depth. This method can reduce the carburization time by 25 percent. After carburizing, the part is then quenched and tempered. An additional benefit is that the residual compressive stress depth is significantly improved. This increases overall strength and fatigue life.
However, one common perception is “more is better.” Often, heat treaters will use carbon potentials greater than the saturation point of carbon in austenite in the mistaken perception that the carbon will be driven into the part faster. However, the steel cannot absorb carbon into the steel above the saturation point. The excess carbon then generates soot. This soot is carried into the oil and contaminates the furnace. At this point, the furnace must be burned out to eliminate the soot.
A comparison of straight carburizing and a boost-diffuse carburize cycle is shown in Figure 2. The desire was to achieve a case with a minimum depth of 0.6mm. For straight carburizing, the part is heated to 927°C and held at temperature for four hours, while being exposed to a surface carbon content of 0.80 percent C. For the boost diffuse cycle, the part is held at 954°C for 1.5 hours and then cooled to 843°C and held there for two hours. Comparing the resultant case profile, the boost-diffuse profile is deeper, and has a better profile for supporting tooth loads. The boost-diffuse also has a shorter furnace time at 3.5 hours, compared to the straight carburizing time of four hours.
Comparing a deeper case of 1.0 mm (Figure 3), the boost-diffuse has a better pattern and nearly identical depth. The cycle times for the straight carburization was 12 hours at 927°C, at a surface carbon potential of 0.8 percent C. The boost-diffuse cycle was held at 954°C for four hours, and then cooled to 843°C and held for four hours with a 0.8 percent C surface carbon potential. In this example that was not optimized in any manner, the boost-diffuse total cycle was eight hours, while the single carbon potential time was 12 hours to achieve the same case depth. This is, again, a significant time savings.
While the higher temperatures used in boost-diffuse carburizing cycles result in much shorter carburizing times, it is not without drawbacks. The higher temperatures involved cause more wear and tear on the furnace and grids, requiring more maintenance. Sooting, as mentioned previously, can also be a problem if trying to jam more carbon into the part. Excessive soot can shorten the life of the quench oil, and result in dirty or stained parts. Excess soot can also result in “tiger striping” which mimics the striping that results from oxidized oil. Much better carbon control is required to minimize soot formation.
In this article, we demonstrated some differences between single carbon potential carburizing and boost-diffuse carburizing. Significant time savings of 25 percent can be achieved with boost-diffuse carburizing, with an improved carbon profile. This improved carbon profile also achieves a better residual stress profile, contributing to improved fatigue life.
Should you have any questions about this article, or suggestions about further articles, please contact the editor or myself.