How to decarbonize steel
Today I will make a short excursion into materials science, consider one of its practical aspects. We will talk about such a phenomenon as decarburization of steel, the effects associated with it, methods of combating it, and for dessert I will describe a method of selling files at the price of forged laminate.
Disclaimer. Everything stated below is based on personal experience and training received at the university. I am aware that in essence I am describing Soviet technologies and that more advanced and diverse equipment exists now. However, what I saw is what I saw, unfortunately, nothing else was brought in. Again, as personal experience shows, there are not many places where the level of technological training is much higher than the Soviet one, and there are many where it barely reaches the Soviet level.
So, decarburization. In factory practice, it haunts heat treatment operators, controllers and technologists at almost every step and brings a lot of trouble. It affects large parts that have to be moved by a crane beam, and small elastic elements, and all sorts of shafts, gears, etc. It reduces the service life of the part and distorts the results of heat treatment control and often serves as a cause for holy wars between controllers and technologists. It is also, in a way, the engine of progress, forcing people to go crazy over the invention of furnaces with different atmospheres, improve the methods of applying galvanic coatings and come up with other different methods of combating it.
What is decarburization (decarburization in factory and craft slang)? When heated for hardening (this is about 800-900 degrees Celsius depending on the steel) in a normal air environment, active interaction of atmospheric oxygen with the surface of the part begins. Scale is formed, consisting mainly of Fe3O4, and along with the oxidation of iron, carbon begins to burn out. At such temperatures, carbon in steel is very mobile and, as it burns out on the surface, carbon from deeper layers tends to occupy the vacant place. And this continues until the part cools down.
The thickness of the decarburized layer depends on the holding time and can be several hundredths of a millimeter with a minimum holding time of 15-20 minutes up to 1-2 whole millimeters, if the holding time lasted a day or more, however, this will most likely be a defect since the dimensions of the part will float away from all conceivable tolerances purely due to the burning of the iron itself. Also, the rate of decarburization strongly depends on how much carbon is in the steel. In high-carbon steels, decarburization occurs much faster than in low-carbon steels. On average, for the most common structural steels such as 40X or 30XGSA on parts of average thickness of 20-30 mm, decarburization fluctuates somewhere in the range of 0.1-0.2 mm.
What does decarburization affect? First of all, it affects the results of heat treatment control. When you measure the hardness after heat treatment, you do not see the hardness that you actually got, but much less, because the surface of the part in a thin layer is softer than the base metal. Sometimes you can select a section on the part and clean it (just manually with a grinder with a petal disk), and then measure it again, but such treatment is not allowed with all parts. If you can’t do this and make sure that there is a decarburized layer, then the question arises – why did you not get the hardness? And there can be many reasons besides decarburization: the steel composition does not match the grade or the composition is at the lower limit of the grade composition, a pyrometer error and incorrect furnace temperature readings, a technologist error – the wrong modes, a heat treatment error and all sorts of combinations of these factors. From this point on, the technologist’s work turns into a real quest with a detective twist, and there is even an analogue of a laboratory with forensic experts – its own factory laboratory with microscopes, hardness testers, spark-gap devices for determining chemical composition, and everything else that a particular enterprise has enough money for.
The decarburized layer is harmful because it reduces the service life of the part. In general, the surface condition for cyclically loaded parts is very important – the fewer defects, the lower the probability of microcrack initiation and the longer the part can serve. Energetically, it is more difficult for a crack to initiate than to develop. If the surface is soft, then microcracks on it initiate quickly and then develop into the base metal, reducing the service life of the part. This applies to all kinds of shafts (rotation under static radial load is a classic example of cyclic loading), axles, gears, brackets and housings that operate under vibration conditions and other similar parts. Moreover, the thickness of the walls/part itself plays practically no role here, since fatigue cracks grow under the action of loads that are much smaller than destructive or plastically deforming. No matter what safety margin you include, failure will still occur. And decarburization will significantly bring this moment closer.
In the case of thin elastic elements, in addition to reducing hardness and service life, decarburization also significantly affects elasticity. If your spring has a thickness of 0.5-0.8 mm, of which 0.1 mm on each side is not elastic, but plastic and soft, then this is a bad spring. Moreover, even a minimum exposure of 10-15 minutes does not help here (less is not possible – you open the oven, which inevitably cools it down, plus heating the tooling, which is probably more massive than the parts), since elastic elements have a higher carbon content, not 0.3-0.4% as in structural ones, but 0.6-0.7% for spring ones, respectively, decarburization occurs faster.
Decarburization is combated by various methods. The most reliable is vacuum treatment and treatment in an environment with carburizing potential. There is also an option of treatment in an inert environment, as well as protective galvanic coating with copper, but as practice shows, their effectiveness varies from case to case. Experienced heat-treaters sometimes add coal to the furnace together with the charge (with a shovel, like stokers in the century before last), but this is more for self-reassurance.
Let's consider the methods of struggle in order.
Vacuum treatment
The cleanest, most trouble-free in terms of post-processing of the surface and geometry, and the most expensive method. Vacuum furnaces are a very expensive pleasure, requiring, in addition to the furnace itself, a lot of auxiliary equipment (vacuum pumps, its own cooling system, etc.). The furnaces themselves are quite delicate, do not like dirt and dust, which are usually in abundance in the heat treatment room. And vacuum furnaces should ideally be serviced by specially trained qualified people, which is usually difficult. In addition to the high cost of the process, vacuum processing is not suitable for all steels, and here's why.
When hardening steel, a fairly intensive cooling is required, which is usually provided by liquid media such as water, oil, soap solution, etc. These liquids will boil in a vacuum and ruin the entire vacuum (and the furnace at the same time, polluting it with themselves). Heat treatment specialists have a story about how to screw up a neighbor with vacuum equipment – throw a piece of sulfur into the feeder. The whole charm is that you will have to wash the furnace from the evaporated and settled sulfur everywhere for a long time and meticulously, and probably together with vacuum pumps. Smells like terrorism, right? Therefore, an ultrasonic bath is also required with a vacuum furnace to clean parts from coolant, grease and any other shop dirt. The process of vacuuming and de-vacuuming the furnace is long enough for the heated part to cool down and when removed, there can be no talk of any hardening. By introducing air into a heated furnace, we will reduce all the advantages of vacuuming to zero – scale and all the other delights will immediately appear (along with a whole bunch of possible breakdowns of our expensive furnace, which is not designed for such manipulations). This means that we can cool the part only inside the furnace itself. And cooling in a vacuum is only possible by radiation and has its own speed limitations. In practice, parts made of high-alloy stainless and heat-resistant steels with strict requirements for geometry and dimensional tolerances are usually processed in a vacuum furnace. Such steels are capable of hardening in air and vacuum is just what they need.
Shielding gases with carburizing potential
For these purposes, various gases and their mixtures are used, including endogas. Endogas is a product of incomplete oxidation of methane by air, consisting of nitrogen, hydrogen and carbon monoxide. The first two protect the surface of the part well from oxidation, the latter is directly responsible for carburization. Furnaces with endogas are also quite expensive and complex, equipped with an endogas generator with the ability to adjust the composition of the endogas, but still simpler and cheaper than vacuum ones, and are less demanding on cleanliness in the workshop. These furnaces are usually closely monitored, because a violation of the endogas generator operating mode can lead to an explosion (relevant requirements are sometimes prescribed even in the technical processes for heat treatment). An additional advantage of this furnace is that it is possible not only to protect the surface of the part, but also to carburize.
The process of heat treatment in endogas looks beautiful – the furnace lid has a hole for relieving excess pressure of endogas, from which a reddish-blue flame rises. When you open the furnace lid, it slightly flares with the same reddish-blue flame – this is endogas coming out of the furnace retort. These furnaces are used to process massive parts made of structural steels, the holding of which lasts for hours and in conventional furnaces will give a thick decarburized layer.
Use of inert gases
In theory, this method should be simpler, safer, cheaper and therefore better than processing with endogas. But in practice, things are a little different. Special furnaces with a protective environment, like furnaces with enogas, are expensive and very complex, often requiring a lot of auxiliary equipment (for example, a nitrogen station that receives nitrogen from the air and then supplies it to the furnace. The nitrogen station itself, in turn, requires a chiller for cooling). Therefore, with the tacit consent of the manufacturer, furnaces that were not initially designed for this are often equipped with argon, for example, chamber furnaces with a side door. The tightness of such a furnace is extremely low, especially where the door adjoins the walls, hearth and arch. Accordingly, to achieve any intelligible effect, you need to blow a lot of argon into such a furnace, which is expensive and generally not very healthy. Argon is a heavy gas and accumulates in the niches and pits of the workshop, of which there are plenty in the thermal room (the furnaces are large, recessed into the floor, all sorts of service recesses, etc.) and there is a risk of an accident with some mechanic suffocating in a pit where argon has accumulated.
Because of the low tightness, throwing coal together with the feed does not work well. In theory, coal burns with the oxygen deficiency, releasing carbon monoxide, which should protect the part. But in practice, it simply burns instead. It can be even worse – carbon dioxide, which is produced during normal coal combustion, oxidizes steel even more than air.
Galvanic coatings
Here, everything depends heavily on what kind of galvanic coating the company has. If the galvanic coating is weak, it usually doesn't work. Copper plating is usually used for protection – coating with a thin layer of copper. In practice, copper can peel off, oxidize itself, or is simply too porous and allows oxygen to pass to the steel surface. Another option is nickel plating. Nickel coating holds up and protects the part much better.
Changing the allowance during final processing
This is not strictly speaking a thermal method, but it is also operational. If, when developing the manufacturing process for a part, you include an allowance for finishing treatment that is greater than the decarburized layer obtained in the heat treatment room, then in the end the decarburization will be mechanically removed and will not affect the part. But this will definitely add work for the fitters in the heat treatment room to clean the part before inspection. After all, before the part is released from the heat treatment room, the inspectors must make sure that it is of the required hardness.
Well, that's it – we can move on to dessert.
It is interesting that steel with low and high carbon content is etched differently. What is etching? It is the treatment of ground (ideally polished) steel surface with acids. It is used to study the microstructure under a microscope. The difference is that the more carbon, the darker the steel after etching (the difference in phase composition and etchability of various phases, but that's another story). If you have an etched section of a sample with decarburization, then you don't even need a microscope to determine the latter – it is easy to see by the light border near the surface layer (the section is usually made on the section of the part, not on its surface). In order to see this, it is not necessary to polish the section, it will be visible on the ground surface. How can this be used? Take a workpiece of high-carbon steel (for example, U10, i.e. an ordinary Soviet rusty file will do just fine) and throw it into an oven with a temperature of 850 degrees for a day. During this time, our file will be strongly decarbonized to a depth of about 1-2 mm, which is what we need. After a day, we take it out, let it cool down and then – to the metalworker. When turning the blade, we get a cut of the outer decarbonized and inner high-carbon layers. Further etching will give a pattern very similar to what can be obtained by forging a package of low-carbon facings and a high-carbon filling – it is understandable, in fact, the same structure of the blade, only obtained in a different way (if we do not take into account the grain size, cementite network and crystallographic orientation of the fibers in steel, which is obtained during forging. But who can determine this by eye, heh-heh?) The difference lies mainly in the clarity of the boundaries between the layers – in the case of decarbonization, it is more blurred and an experienced user / craftsman can understand that something is not right. You can also be given away by the absence of underfusion, which is not such a rare phenomenon in forged laminate. You will have to cosplay as a very experienced blacksmith, confuse with all sorts of terms about multi-layer forging (in a three-layer laminate, yeah), etc. Then everything depends on how well you did the metalworking of the product in general and how well you gab. As far as I know (I haven't been following the market for a long time), forged laminate is 2-3-10-20 times more expensive than regular blacksmithing (it depends a lot on the name of the master, the quality of the metalworking and the impudence of the seller). You can simply get a different pattern on the blade in such a simple way, especially if you play around with some coatings (and there are some) or something else, without trying to look like a blacksmith. Unfortunately, then you will have to correct the structure of the workpiece obtained by such a long aging, and this is a rather expensive process. Or not correct it, but then your product will only be good for storage on a shelf.
On this cheerful note, your humble servant takes his leave and thanks you for your attention!
Author: Andrey Nozhovsky
