Spheroidizing Annealing of ShKh-15 Steel

N. D. Tyuteva

Annealing of ShKh-15 steel has the purpose of producing a material that has good machinability when chips are removed on metal-cutting machine tools. As a result of annealing, a steel should be obtained with a structure favorable for the subsequent heat treatment of the finished bearing ring.

These requirements are satisfied by the structure of fine-grained homogeneous pearlite and spheroidized carbides. But the structure must not be too dispersed, and this circumstance determines the hardness of the metal within the limits of 183–207 Brinell.

There have been quite a number of investigations on the heat treatment of steel of the ShKh-15 type, but among them the works of Rauzin [1], Galikeev and Zabolotskii, Zabolotskii and Usvatova, and some others [2] stand out.

In our work we dealt with specimens of rings of ball and roller bearings.

The usual thermal regime in forging rings with diameters from 55 to 80 mm is as follows: maximum heating for forging 1050°, forging finishing temperature 900°. The forging temperature fluctuates within limits of 150°. Sometimes, despite a constant forging regime, rings are rejected on the basis of final hardness before mechanical machining. This is explained by the different character of cooling of the rings after forging. Cooling after forging depends on many causes. Sometimes hot rings are scattered over the shop floor; sometimes they are stacked in piles; their cooling rate is different in winter and in summer. The initial structure of specimens cut from different rings differs in the size of the plates of pearlitic cementite and in the character of the carbide network.

In annealing in furnaces, one must take into account the distribution of temperatures over the cross section of the charge. The final structure of the rings is affected by their structure after forging, their arrangement in the annealing furnaces, and the heating and cooling regime.

According to data from some investigators, the best machinability is obtained after annealing at 780°. Rings annealed at other temperatures are machined poorly, and it is impossible to correct poor machinability on machine tools by changing the cutting regime.

Rauzin [1] also speaks of the influence of the initial structure before annealing on the mechanical properties of ShKh-15 steel. In his opinion, only an optimal concentration of solid solution, caused by the presence in the structure of homogeneous fine pearlite, will give the best results in all respects. He gives a table in which the hardness and microstructure of the steel are related (Table 1).

According to Galikeev and Zabolotskii [2], rejection of rings on the basis of structure can be eliminated only by additional heat treatment, while improvement of the structure of annealed forgings will make it possible to reduce the allowance for turning.

Table 1

Rausin [1] emphasizes the influence of the initial structure on the final result of annealing.

The initial structure is affected both by the temperatures at the beginning and end of forging and by the size of the forgings. The larger the forging, the coarser the structure, and during annealing, in the temperature range 770—800° with holding for 1.5 and 3 hours and cooling rates of 10 and 20° per hour, only the initial structure of fine-grained pearlite, formed during accelerated cooling, is spheroidized.

When a coarse-grained initial structure with a thick carbide network, obtained under the same annealing conditions during slow cooling in stacks of small rings or in the case of large-sized products, is annealed, a structure is obtained with retained regions of lamellar pearlite and a carbide network. The coarser the initial structure was, the more regions of lamellar pearlite and network there will be. As the annealing temperature is raised and its duration increased, the amount of lamellar pearlite decreases and the carbide network coagulates. After annealing at 830° with slow cooling, instead of a carbide network rounded carbides and a field of granular cementite are obtained. If fine lamellar pearlite is annealed at 830°, the structure will be worse than when annealed in the range 770—800°. It will be more nonuniform. The nonuniformity is explained by processes occurring during the annealing itself, namely: dissolution of carbides, diffusion, and crystallization. On heating, the finest carbide grains dissolve first. The larger ones, without dissolving completely, become centers for subsequent crystallization and coarsen during cooling. As a result, greater nonuniformity of the granular structure is obtained than when annealing from lower temperatures.

The conclusions reached by Rausin amount to the following:

1. The coarser the initial structure, the higher the annealing temperature must be.
2. When annealing fine structures, a lower temperature is necessary.
3. Fine initial structures can be obtained by accelerated cooling of forgings.
4. Stacking products after forging in piles cannot be recommended.
5. The structure of forgings is determined not by the heating temperature during forging, but by the subsequent cooling rate, and slow cooling is more harmful than a high heating temperature.

One cannot agree with the author’s conclusion that “the structure of forgings is determined not by the heating temperature during forging-stamping, but by the cooling rate after it.” The initial structure is undoubtedly affected both by the heating temperature during forging and by the rate of subsequent cooling.

There is no doubt that, upon overheating, the austenite grain grows, and secondary carbide is precipitated precisely along the boundaries of the austenite grains. Only in the case where there is certainty that the austenite grain has not changed at all can one agree with the author. And this will be only in the case of forging rings from one bar, or from several bars of a quite definite heat that has been tested for its sensitivity to overheating.

If, after die forging, a network cementite is obtained which does not coagulate during annealing, then normalization of the forgings may be carried out.

Before normalization the steel is heated to such a temperature that all carbides pass into solid solution. With rapid cooling the carbides may precipitate in the form of a broken network. During annealing such carbides readily become granular.

The question of introducing normalization as an additional heat-treatment operation evokes the most varied attitudes on the part of heat treaters.

At some plants normalization is carried out after annealing, before hardening, and occupies the following place in the technological process: die forging, annealing, sawing, normalization.

Normalization is recommended to be carried out at a temperature of 880–900°; sometimes a higher temperature is used—925–950°. The holding time at the indicated temperatures is about one hour.

The Timken firm gives the following heat-treatment schedule: normalization after forging at 900–950°, annealing at 780°. The hardness after annealing is about 200 Brinell. If a hardness after hardening above \(R_c = 66\) is required, then additional normalization at 950–975° before hardening is recommended.

In Rausia’s opinion, normalization generally has an unfavorable effect on the properties of ball-bearing steel, making it prone to quench cracks. On the other hand, normalized rings have, after hardening, a higher and more uniform hardness, although the very same result is obtained with rings having fine-grained pearlite. It has been observed that normalized rings warp more severely.

On the basis of literature data one can draw certain more or less definite conclusions. The best structure of forgings is fine lamellar pearlite without a carbide network. This occurs if die forging is finished at 850° and the rings cool sufficiently rapidly after die forging. After normal annealing (at 770–790°), in this case a homogeneous fine-grained pearlite is obtained. Of course, in this case normalization is not needed.

If, after forging, such an optimal structure has not been obtained, normalization is necessary; when it is properly carried out, the subsequent annealing gives a quite favorable structure.

Normalized steel after hardening is harder, although it should be noted that hardness still does not determine the quality of the bearing.

A coarse structure after normalization is improved only slightly. To improve a coarse structure with a large carbide network, normalization from 950° is necessary. If cooling during normalization was slow, then normalization will be harmful—a carbide network forms again.

Apparently, it may be regarded as firmly established that, as a result of heating to 900–920°, globular carbides dissolve only in normally annealed steel with granular pearlite. The Brinell hardness of such steel is 187–196.

Coarse globular carbides remain after normalization at the indicated temperature.

Normalization with slow cooling gives a carbide network. This is not so noticeable when heated to 900°. When the temperature is raised to 925°, the carbide network becomes clearly pronounced.

As a rule, overheating defects in the forging of steel are weakened after normalization, but are not completely eliminated.

Author’s Experiments

For metallographic investigation, rings were selected at the plant during the forging of rings for roller bearings 58 mm in diameter.

From a bar with a single heating on a horizontal forging machine, four rings are forged. The forging temperature of the first and fourth rings proved to be different. The difference reached 50 degrees and more. Therefore, polished sections were prepared from all four rings.

After forging, the rings are cooled in the shop in stacks or one by one on the shop floor. We investigated both types. No sharp difference was found in the structures of the first and second ring cooled in a stack, or of the first and last ring cooled one by one on the floor. But in structure the first ring cooled in a stack differs very strongly from the last ring cooled on the shop floor. The difference in the structures consists in the thickness of the carbide network. In rings slowly cooled in stacks from a higher temperature, the thickness of the carbide network is, naturally, greater.

It should be noted that the size of the austenite grain was found to be the same in all these cases.

When etching the polished section with a solution of nitric acid in ethyl alcohol, it is difficult to determine the character of the network of structurally free carbides, especially when it is thin and broken. Therefore, for a clearer revelation of the network, etching with a hot solution of sodium picrate may be recommended.

For experimental anneals in the laboratory and for studying the structure, rings were selected at the plant at different times over the course of a month, in order to cover as wide a range as possible of variation in the chemical composition of ShKh-15 steel.

The ring forged first from the heated bar, that is, the hottest ring, forged at the higher temperature, had the structure of coarse-grained pearlite with a well-pronounced network of secondary carbide. In the ring, especially the cold one, there was lamellar pearlite.

Rings enter annealing after forging. The temperature at the beginning and end of forging at the plant was not determined regularly. After forging, the rings were cooled chiefly in stacks and only as an exception individually. Annealing of the rings is carried out according to the following schedule. The heating rate is set by the furnace design. Since the furnace is charged at a low temperature and heating proceeds slowly, it is considered that by the time the temperature of normal annealing—780–800°—is reached, the entire charge has been heated through. At a temperature of 780–800° the rings are held for 40–50 minutes.

The holding time depends, of course, on the number of rings being annealed and on the heating rate. The faster the heating, the more time the rings must be held at the annealing temperature.

The cooling rate is 15° per hour down to a temperature of 550°. The subsequent cooling rate is not taken into account.

The result of annealing is strongly affected by the character of the initial structure of the steel. We set ourselves the goal of providing such an annealing schedule that would ensure the production, in the structure, of fine-grained carbides well ob—

machinability and suitable hardness (183–207 Brinell) even in the case where there is an unfavorable initial structure. Such an unfavorable structure may be considered to be coarse lamellar pearlite and a massive carbide network with a large pearlite grain size.

The stamping process does not always proceed smoothly. Sometimes (especially in flame furnaces) overheating is observed and, as a consequence, a coarse austenite grain and a coarse carbide network. We heated several rings to a temperature of 1050° and cooled them slowly at a rate of 300° per hour to a temperature of 500°, and then at a rate of 200° per hour to 100°.

Fig. 1

Fig. 2

Figure 1 shows the structure of a ring annealed in this way. The specimen was etched with sodium picrate and photographed at a magnification of 500 times. In the photograph the almost continuous structure of the carbide network is very clearly visible. The principal background—pearlite—was almost not revealed by etching with picrate.

In Fig. 2 the same specimen is shown, but etched with nitric acid; the magnification is the same. The cementite network is almost not visible. But the structure of the principal background can be examined; it is coarse lamellar pearlite.

Fig. 3

Figure 3 presents the structure of steel annealed according to the plant instructions: holding for 30 minutes at 780° and cooling at a rate of 15° per hour. The specimen had previously been heated to a temperature of 1050° and cooled at a rate of 300° per hour to 500° and at a rate of about 200° per hour from 500° to laboratory temperature. The carbide network began to break up only in places. If the process of spheroidization of the carbide network had proceeded to completion, the structure would still have been unsatisfactory: the carbide grains formed in place of the network are too large. For final spheroidization of the network, holding at 780° for several tens of hours is necessary.

The structure of a specimen from a ring heated to 1100° and cooled in air is shown in Fig. 4, etched with nitric acid. As can be seen, at this temperature the austenite grain becomes very greatly coarsened. The principal structure is martensite and troostite; areas of austenite are noticeable.

Figures 5, 6, and 7 show structures (etched with nitric acid) of specimens cut from rings annealed according to the plant instructions, but with holding for five hours at 780°. As can be seen, the structure is completely spheroidized. There is no noticeable difference between secondary carbides and the carbides of granular pearlite. Such a homogeneous structure

was obtained after annealing oil-hardened rings that had had a very coarse structure after forging.

As has already been said, the literature contains the most contradictory data on the benefit of normalizing. Therefore, rod steel 12 mm in diameter from two heats was heated to a temperature of 1050°, under conditions precluding decarburization, held for two hours, and cooled so slowly that a coarse carbide network precipitated.

Fig. 4

Fig. 5

The overheated specimens were normalized from a temperature of 950°. During normalizing the specimens were underhardened, as indicated by the increased hardness and the troosto-martensitic structure.

The normalized specimens were annealed according to the normal schedule: holding at 770–780° for one hour and cooling to 500° at a rate of 15° per hour. In all cases, after normalizing and annealing, the carbide network did not completely disappear.

Fig. 6

Fig. 7

Therefore, in order to correct the structure, we decided to apply oil hardening from 950° followed by high tempering. On heating above the \(SE\) line, all carbides began to pass into solid solution. During cooling, a certain number of carbide crystallization centers appears, depending on the cooling conditions. The carbide crystals that form grow at a rate whose magnitude is also determined by the cooling conditions. These two circumstances determine the structure of the steel obtained as a result of cooling under one or another set of conditions.

During air cooling it is practically impossible to change the number of crystallization centers and the rate of crystal growth, since we are constrained by the cooling conditions. But in the case of a hardened structure it is possible, within wider limits, to vary the conditions of the transformation of martensite into troostite and sorbite.

At different tempering temperatures and with different holding times during tempering, carbide crystals of different dispersity can be obtained. During annealing after tempering, the carbide crystals of such a structure are partly retained and provide the basis for a granular structure.

We settled on a tempering temperature of 500° with a prolonged holding time. At 500° sorbitic carbides are precipitated; prolonged holding leads to the formation of a clearly expressed fine-grained carbide structure, which is also retained during subsequent annealing.

Very good results with ShKh-15 steels of various heats were obtained by heat treatment according to this schedule. Rings with a coarse carbide network are quenched in oil from 950°. At this temperature a sufficiently long holding time is necessary so that the carbide network has time to dissolve in the austenite. This is followed by holding for two hours at 500–550° and then by ordinary annealing at 770–780°.

Prolonged annealing at a comparatively low temperature (not exceeding 770°) is also of interest. Good machinability on machine tools, suitable hardness (183–207 Brinell), and a very good structure were obtained after annealing at 770° for 8 hours, followed by cooling at 15° per hour.

Conclusion

Since the initial structure has a very strong influence on the result of spheroidizing annealing of forgings made of ShKh-15 steel, careful control of temperatures during and after forging is mandatory. This is the only, albeit indirect, method of structural control that is easy to establish in the forging shop.

Cooling of parts should be carried out not in stacks, but by laying them out on the shop floor; only after the forging has blackened may they be gathered into piles.

Selective control of the structure of stamped rings is necessary. If a coarse carbide network is obtained in the forgings, it can be broken up only by prolonged annealing. As a result of such prolonged annealing, a granular structure can be obtained, but in place of the coarse network large carbide grains will remain. As a result of the coarsening of the grains, the hardness will decrease sharply. A consequence of excessive reduction in hardness, along with a coarse-grained structure, may be poor machinability by chip removal.

If a coarse carbide network is obtained in the forgings, we recommend quenching in oil followed by annealing. If after stamping a normal structure has been obtained, then the usual factory annealing may be applied. Annealing with preliminary quenching gives the most favorable structures.

LITERATURE

1. Rauzin Ya. The influence of the initial structure on the results of annealing of chromium-carbon steel, Podshipnik, 1940, No. 5.

Rauzin Ya. The influence of the initial structure on the mechanical properties of hardened steel, Stal’, 1941, No. 4.

Rauzin Ya. The influence of cooling rate on the structure of forgings made of ShKh-15 steel, Podshipnik, 1939, No. 5.

Rauzin Ya. The influence of normalization on the properties of bearing steel, Podshipnik, 1939, Nos. 10–11.

Rauzin Ya., Maslennikov M. I., et al. The influence of normalization on the structure of bearing steel, Podshipnik, 1940, No. 2.

Rauzin Ya., Kondrat’ev M. The influence of cooling rate on the structure of forgings made of ShKh-15 steel, Podshipnik, 1939, No. 5.

1. Galikeev A. Kh., Zabolotskii I. Ya. Machinability of ShKh-15 bearing steel, Podshipnik, 1940, No. 9.

2. Zabolotskii I. Ya., Usvatov S. L. Annealing of rings with accelerated cooling in one well, Podshipnik, 1940, No. 10.