Heating of loads in modern plasma nitriding furnaces is carried out by a combination of the radiative heat transfer from the wall or furnace resistance heaters and heat transfer from the glow discharge. We will look at the effect of the workpiece emittance on its temperature.


Radiative heat transfer under vacuum is well known and described in professional literature.[1] Transfer of the heat to the workpiece from the glow discharge has not been researched as thoroughly.[1-6]

    The analysis is complicated by the component of the glow discharge. It has been proven that the heat input from the glow discharge is a combination of the effect of ion and neutrals bombardment of the cathode, the latter being about 85%.[6] It can also be assumed that this input does not depend on emittance of the cathode. However, it was also shown that the two geometrically identical workpieces with different emittance and heated solely by the glow discharge have various temperatures, the high-emittance piece being cooler.[6] That is because the cathode loses heat, depending on its emittance. Thus, the total heat absorptance by the cathodic workpiece in cold-walled furnaces is (see Equation 1), where q+ is radiosity and q- is irradiation from plasma.[6]

    Obviously, if emittance of the one workpiece is higher than the other one, their temperatures are not identical. Therefore, temperature of the complex loads containing parts of various emittances in cold-walled plasma furnaces may not be uniform.

    The total heat absorptance by the cathodic workpiece in hot-walled furnaces is (see Equation 2), where qw is the heat radiated by the wall of the furnace and intercepted by the workpiece.

    Radiative heat absorptance of the object depends on its emissive power, and that is why the workpiece with the higher emittance will absorb more heat radiated from the wall of the furnace than the low-emittance workpiece. As a result, there are two different mechanisms of heating in plasma nitriding furnaces: one is by radiation from the wall and the other one is the bombardment by plasma particles having momentum toward the cathodic workpiece. They both are dependent on emittance of the workpiece but in the opposite way.

    Therefore, it can be expected that the heating component of the glow discharge and its dependence on emittance of the workpieces is minimized in the hot-walled plasma furnaces. The hot-walled plasma nitriding furnace becomes more similar to conventional vacuum furnaces. It needs to be seen to what degree the wall-heating component prevails in controlling temperature of the workpiece.

    The experiments carried out demonstrate that this is indeed the case, but it is also shown that the plasma-heating component must not be ignored if good temperature uniformity in the furnace is to be achieved.



Plasma nitriding experiments were carried out at 540°C (1004°F) in a mixture of 30% nitrogen and 70% hydrogen at a total pressure of 2.5 mbar for 12 hours. The arbitrary selected voltage was 540 V. In experiment number one, two geometrically identical test cylinders in the as-ground condition were heated in the plasma nitriding unit “100/180 Duo Rubig” manufactured by Rübig Engineering. Fig. 1 shows that each cylinder had a thermocouple inserted in it.

    The experiment demonstrated that perfect temperature uniformity between the two workpieces was achieved (Fig. 2). The temperature difference between the two cylinders was not larger than +/-1°C during the entire heating cycle. Therefore, it could be assumed that the conditions of the experiment for comparing the effects of the differences in emittance of the two cylinders were nearly perfect.

    Experiment number two demonstrated what happens when the emittance of the two workpieces is not the same. Temperature graphs of the two cylinders shown in Fig. 3 can be divided into four segments:

I – Radiative heating during ramping to final temperature (wall on only), pressure 0.9 mbar of nitrogen

II – Mix radiative and plasma heating during ramping to final temperature (wall and plasma on), pressure 1.0-1.25 mbar of 30% nitrogen-70% hydrogen

III – Mix radiative/plasma heating during the nitriding/soaking step at nominal temperature of 540°C (wall and plasma on), pressure 2.5 mbar of 30% nitrogen-70% hydrogen

IV – Cooling to room temperature with nitrogen circulated by the fan (wall and plasma off), pressure 800-900 mbar

    It is clear that radiative heating of the workpiece by the wall prevails over plasma heating during the two first segments of the heating experiment, and the cylinder with the higher emittance gains more heat. Its temperature is higher than the temperature of the cylinder with the lower emittance.

    In the third step of the experiment, the situation reverses. As shown in Fig. 3, the temperature of the workpiece with lower emittance starts to prevail over the temperature of the other piece shortly after the process stabilizes. It should be noted here that nitriding increases emittance of ferrous alloys. The low emittance increases comparatively quickly to the level of 0.5, and the higher emittance increases to about 0.6 at the same time.[6] Therefore, temperatures of both cylinders get closer together with the processing time once the difference in the emittance becomes smaller. Nevertheless, temperature difference between the two workpieces was about 10°C at the end of the 12-hour nitriding step. That difference was still much smaller than in the cold-walled plasma nitrider used previously for the identical cylinders, which generated a wider spread of temperatures up to 40°C.[6]

    Essentially, heating of the workpiece in plasma nitriding furnaces can be divided into two main regions: one with “weak plasma” regime and the other one with “intense plasma” regime (Fig. 4). In the weak plasma, the workpiece with the low emittance will have a lower temperature than the workpiece with high emittance. The situation reverses itself in the intense plasma regime. It has to be understood that both of the regimes are characteristic of specific metallurgical results, mainly for thickness of the compound layer.[7]



The experiments clearly showed that the emittance of the cathodic workpiece treated in the plasma nitriding furnace equipped with resistance heaters has an effect on its temperature. The higher-emittance workpiece achieves lower temperature than the low-emittance piece. Those temperature differences are of technical significance and must not be ignored when planning nitriding cycles.

    It should be noted here that the ∆T will depend on proportions between the wall heating and plasma heating. It is important how much power is delivered by plasma and how much by the wall. The plasma intensity also has metallurgical consequences. IH


For more information: Contact Edward Rolinski DrEng, Dr-habil (aka Dr. Glow), VP technology, Advanced Heat Treat Corp. 2825 MidPort Blvd. Waterloo, IA 50703; tel: 734-243-0063; fax: 734-243-4066; e-mail: doctorglow@ion-nitriding.com. Corporate contact inform-ation is: tel: 319-232-5221; fax: 319-232-4952; e-mail: www.ahtweb.com



1. M.F. Modest, Radiative Heat Transfer, Second Edition, Published by Academic Press, 2003, 822 pages

2. A. Marciniak and T. Karpinski, “Comparative Studies on Energy Consumption in Installations for Ion and Gas Nitriding,” Industrial Heating, 1980, 4, p. 42-44

3. A. Marciniak, “Heat and mass transport during nitriding in a glow discharge,” PhD Thesis, Warsaw Technical University, Warsaw, 1983

4. A. Marciniak, “Non-uniform Heating Effects During Treatment in a Glow Discharge,” Thin Solid Films, 156 (1988), p. 337-344

5. M.Q. Brewster, Thermal Radiative Transfer and Properties, John Wiley & Sons, Inc., 1998

6. E. Rolinski, J. Machcinski, T. Larrick and G. Sharp, “Effect of Cathode Workpiece Surface Emissivity on Temperature Uniformity in Plasma Nitriding.” Surface Engineering, 2004, Vol. 20, No 6, p. 426-429

7. J.G. Conybear and B. Edenhofer, (1988), “Progress in control of plasmanitriding and –carburizing for better layer consistency and reproducibility,” Proc. 6th International Conference on Heat Treatment of Materials, 28-30 Sept., p. 381-394