Carbon restoration in vacuum furnaces with LPC

As a philosopher would say, everything good is bad for something… Carburizing under reduced pressure, or vacuum carburizing or low presssure carburizing has the advantage of being very fast. On the one hand, because unlike carburizing in the ENDO atmosphere, we work with a clean environment where carbon transfer from the atmosphere has a minimal amount of barriers, and on the other hand, because thanks to the vacuum, we can go to higher process temperatures, which will speed up the carburizing process even more. But if we only need to restore the carbon on the surface, i.e. to repair the decarburized layer created during casting or forging, then this speed is detrimental, because we only move in small layers with CHD from 50 to 100 µm with times only in the order of minutes (Fig. 1).

Fig. 1 – Dependence of CHD on time and temperature for short diffusion depths

LPC is based on the fact that we inject a certain amount of C₂H₂ into the vacuum chamber in such a way as to achieve saturation of the environment for the formation of Fe₃C iron carbides. In the next phase, only in the presence of nitrogen, the layer is diffused and the carbon content decreases. These both segments are repeated several times, while the carburizing time, the saturating period, is in minutes and changes only minimally, the diffusion period gradually increases with each subsequent segment. The final period of carburizing is then set so that we get the carbon content to the eutectoid amount, roughly 0.75 wt% C (Fig. 2). The total time of the carburizing process is then expressed by the following formula:

Time LPC = ∑(t _{saturation period_C2H2}) +  k  *  ∑(t _{diffusion period_N2})

where k is approximately 5.

Fig. 2 – Schematic illustration of the vacuum carburizing process from the point of view of saturation and diffusion periods [ECM]

The carburizing process is shown schematically in the next picture

Fig. 3 – Schematic illustration of the vacuum carburizing process on Fe-Fe3C diagram

 For carbon restoration, the process should proceed similarly, however, due to the speed of carbon diffusion, we usually only have one attempt to achieve the optimal result. One saturation period and one diffusion period. During the diffusion period, we do not reach a carbon content corresponding to the eutectoid amount, but corresponding to the chemical composition of the steel, e.g. 0.3 to 0.4 wt% C.

Fig. 4 – Schematic illustration of the carbon restoration on Fe-Fe₃C diagram  

 What does this imply? Unlike carburizing in multi-purpose furnaces, where we work with an ENDO atmosphere, and where we simply set the Cp potential to a value corresponding to the carbon content in the steel, with LPC we have to test the process. It depends on what happens next. Usual processes are

• carbon restoration and normalization
• carbon restoration and quenching into oil
• carbon restoration, normalizing and quenching into oil

Fig. 5 – A graphical illustration of the effect of the atmosphere on decarburizing or carburizing [1]

In the first case, nothing fundamental will happen, we will have a ferritic-pearlitic structure with a greater or lesser amount of pearlite. In the second and third case, the hardness of the surface after quenching will change depending on the amount of carbon. If we overcarburize the parts, after quenching  they will have a higher hardness on the surface and machinability will deteriorate, if we do not reach the eutectoid point,  then we will have a problem with achieving the hardness corresponding to the selected type of steel

Fig. 6 – Effect of carbon content on the hardness of martensite after quenching [2]

What to say in conclusion? It is feasible, but how to manage it? For carburizing, the following equation applies:

For decarburizing:

Where

• c in wt. % is the required concentration of carbon after the diffusion step, in the case of carbon restoration it is the amount of carbon in the steel, e.g. 0.4 wt.%
• c₁ in wt. % is the initial concentration before the start of diffusion, i.e. roughly a concentration of 1.3% wt.% for Fe₃C
• c₀ in wt. % is the concentration of carbon in the steel we are carburizing, and in this case c = c₀
• x is CHD in m
• D is the diffusion flux of carbon for a given temperature into austenite in ms^-1, see Fig. 7
• t is the diffusion segment time in seconds

The transfer of carbon into steel expressed by the coefficient D is temperature dependent and increases with increasing temperature

Fig. 7 – Rate of carbon diffusion into austenite as a function of temperature [3]

If we set the endpoint of the diffusion segment to, for example 0.1 mm, then the diffusion segments needed to reduce the amount of carbon, for example, from 1.3 wt.% to 0.4 wt.% will have times according to the diagram in Fig. 8. Depending on the temperature from 8 to 15 minutes.

Fig. 8 – The diffusion segment time required to reduce the carbon content from 1.3 wt. % to 0.4 wt. % depending on the temperature [3]

The result of our efforts is that it is feasible, although with greater complications than with classic gas carburizing. However, since LPC equipment is specifically designed for true carburizing, the benefits of the LPC process will more than compensate for the carbon restoration problems

[1] Pavel Stolař, Chemicko-tepelné zpracování, 2004

[2] Daniel Herring,  https://www.industrialheating.com/articles/90008-martensite, 2011

[3] http://www.fsiforum.cz/upload/soubory/databaze-predmetu/BUM/bum2008/bum-2008-zaklady-difuze.pdf

Jiří Stanislav

December 17,  2023