Redefining Quenching Technology
By Aymeric Goldsteinas and Jake Hamid
As metallurgical technology seeks to ever improve and provide heat-treating solutions, dry gas quenching stands out as the best match for industry needs. Gas quenching has several noteworthy benefits compared to other quench systems. It is the only dry quenching that exists and, therefore, eliminates all environmental or safety problems connected with liquid quenching. It is not only the safest and most environmentally friendly system, but it is also the system that produces the least distortion.
In this paper, authors examine why defining gas quenching in bar pressure no longer applies and why a new set of definitions are needed, as well as how these new definitions will allow for a better understanding of which steel and cross-section can be hardened in gas. Additionally, they will discuss results from studies regarding quenching chamber design, material cross-section and quenching in liquid or gas.
The production of a hardened core or hardened surface for metals requires a heat treatment process consisting of heating to a sufficiently high temperature and then a rapid cooling to room temperature. When it comes to quenching, uniformity is of the utmost importance. Therefore, inadequacies of any system that may be detrimental to the process and results need to be addressed. As metallurgical technology seeks to provide solutions for such deficiencies, certain quench techniques, namely dry gas quenching, seem to more efficiently match the needs of the industry.
For ferrous alloy metal, the objective is to increase the temperature above the upper critical temperature (Ac3). This is dependent on the composition of the alloy, therefore the metal is completely in the austenite phase. The alloy is then rapidly cooled by a quenching fluid or gas so that it can be converted into the harder martensite phase. A sufficient, fast cooling rate is needed to minimize the formation of the bainite and pearlite phases, which are softer than the martensite phase and will negatively impact the physical properties of the steel.
The key to accomplishing this process is the uniform removal of heat from the surface of the metal part. A continuous cooling curve, shown in Fig. 1, depicts the cooling rate of a ferrous alloy.
Figure 1: Continuous cooling curve for a ferrous alloy.
Quenching oils of different qualities exist with the quenching severity depending on their physical properties, the most important one being the viscosity. Oil, just as water, exhibits a pronounced vapor phase followed by a nucleate boiling phase with a very rapid heat transfer in the temperature range between 572 °F and 1,112 °F (300 °C and 600 °C). The three stages for the cooling of an oil quench are illustrated below in Fig. 2.
Figure 2: The three cooling stages of an oil quench.
These stages of cooling may not occur at all points on a part at the same time. During the oil quench nucleate boiling phase, extremely high, instantaneous heat transfer coefficients can be achieved. This is a distinct advantage in the temperature range where pearlitic transformation occurs (this advantage is not shared by gas quenching). With the breakdown of the vapor phase at the onset of boiling, however, the so-called Leidenfrost effect occurs. The result is a totally non-uniform heat transfer rate on various surfaces of different parts, and it is dependent on a number of variables and factors. This uneven transitory step creates huge temperature differentials and is the major factor in distortion when quenching in these media.
Gas quenching is a single-phase quenching of a pure, convective type. Gas type, gas pressure and gas velocity are the main factors. The gas quench cells are equipped with powerful fans and are capable of injecting gases typically up to 20-bar positive pressure in conjunction with heat exchangers using chilled water to quickly remove heat from the quenching gases. The most common quenching media is high-pressure nitrogen gas.
The non-uniform cooling of parts associated with liquid quench, which has a vapor phase, can be eliminated and replaced with gas quench which has a more uniform cooling rate, allowing for more uniformity and pure convection. One major benefit of this is that part distortion can be greatly reduced. High-pressure gas quench can sometimes eliminate the need for post-heat treatment straightening or clamp-tempering operations, reduce grind stock allowances and hard machining or replace more costly processes, such as press quenching.
Quench Rate Comparison between Oil Quench and Gas Quench
It is frequently stated that the intensity of gas quenching can be adjusted to a similar cooling rate as liquid quenching. As can be seen from Table 1, the mean heat transfer coefficient over the whole cooling from 212 °F to 1,652 °F (100 °C to 900 °C) of quenching a full load in a high-volume gas flow can be correlated to quenching in liquids. Thus, in looking at the overall cooling rate from start to finish, quenching in nitrogen gas at 10 bar (gas velocity at 10 meters per second) compares to quenching a full load in a molten salt bath; whereas, the average cooling rate of an agitated, high-grade fast oil can be equalized by quenching in hydrogen gas at a pressure of 40 bar.
|500||Salt||Nitrogen at 10 bar|
|1,000||Oil (non-agitated)||Helium at 20 bar|
|2,000||Oil (agitated)||Hydrogen at 40 bar|
|3,000||Water||Hydrogen at 100 bar|
Table 1: Comparison of heat transfer coefficient.
To conclude from this comparison of average cooling rates that quenching a load in 40-bar hydrogen gas would lead to the same metallurgical result with respect to hardness, case depth and metallurgical structure of the steel is a wrong assessment. The temperature and time dependence of the cooling rate is totally different in a liquid, with a pronounced nucleate boiling phase and pure convection cooling. Fig. 3 demonstrates this effect with the comparison of the temperature-dependent cooling rate of gas to oil quench.
Figure 3: Cooling rate comparison of high-pressure gas quench and cold oil quench.
It becomes apparent that even with the highest cooling rate possible (close to 122 °F [50 °C] per second) in high-pressure gas quenching, the peak of the gas does not even come close to the peak of the oil cooling rate in the nucleate boiling phase, where maximum values of 212 °F (100 °C) or even 302 °F (150 °C) per second are possible. As these high cooling rates during nucleate boiling take place in the important phase of steel quenching (demonstrated on CCT diagrams where the ferrite and perlite nose are located), the quenching of low-hardenability steel in oil will lead to a pure martensitic structure, whereas the quenching in high pressure will lead to a hardened structure also having also perlite and ferrite. This despite the fact that the average cooling rate of both quench systems are equal.
Thus, a large uncertainty exists in gas quenching for the prediction of hardness and structure of quenched steel components.
A procedure developed by Ipsen to predict the hardness and structure after gas quenching makes use of the necessary cooling rate in the temperature region of the perlite and ferrite formation (i.e., between 1,472 °F and 932 °F [800 °C and 500 °C]). If, during a given point in the quenching process, the necessary cooling rate to avoid perlite and ferrite formation is reached or exceeded, then one can be sure about the results to be achieved.
Defining these cooling rates from given CCT diagrams leads to the next question: what is the necessary heat transfer coefficient for given diameters of the work pieces in order to reach the specified cooling rate in the core of those pieces?
By solving the heat conduction equation for the given problem or respective approximation formulas, the necessary heat transfer coefficient can be estimated. The results of such an estimation is shown in Table 2.
|Steel Grade||Core Hardness||Cooling Rate (°C/s)||λ||Heat Transfer Coefficient α|
|Ø 20 mm||Ø 30 mm||Ø 40 mm|
|55NiCrMoV6||> 57 HRC||38.3 °F (3.5 °C)||0.85||190||275||350|
|X210Cr12||> 64 HRC||40.64 °F (4.8 °C)||0.63||260||370||480|
|90MnV8||> 64 HRC||42.08 °F (5.6 °C)||0.54||300||430||550|
|42CrMo4||> 54 HRC||140 °F (60 °C)||0.05||> 2,000||> 2,000||> 2,000|
|42CrMo4 mod||> 54 HRC||38.84 °F (3.8 °C)||0.80||350||460||600|
|16MnCr5||> 300 HV||59 °F (15 °C)||0.20||800||1,150||1,500|
|20MoCr4||> 300 HV||73.4 °F (23 °C)||0.13||1,250||1,800||> 2,000|
|15CrNi6||> 300 HV||39.2 °F (4 °C)||0.75||200||320||400|
Table 2. Correlation of material requirement, cooling rate and α-value.
A more useful method is, therefore, the empirical measurement of the heat transfer coefficients in each gas quench system. This can be done very easily with the Ipsen flux sensor.
Gas quenching has tremendous benefits compared to other quench systems. It is the only dry quenching that exists and, therefore, eliminates all environmental or safety problems connected with liquid quenching. It is not only the safest and most environmentally friendly system, but also the system producing the least distortion. This comes about because of the single-phase, pure convective type of cooling and the high flexibility of adjusting the cooling parameters to the needs of each special case.
In many cases, the first attempt on gas quenching a component does not lead to reduced distortion because the load and gas flow considerations are not yet optimized. Yet using the existing knowledge about laminar and turbulent gas flow, in addition to a gas-quenching-adapted and -adjusted load configuration, will almost always lead to much lower distortion results as compared to oil quenching, and even salt bath quenching.