This article examines the technology of cross-wedge rolling (CWR) in the production of railway axles, as well as new equipment for its realization. Some theoretical bases for the rolling of such axles have been considered and confirmed, resulting in axles of higher fatigue strength. The new method enables a significant increase in processing speed; improved rough axle accuracy; and decreases in scrapped pieces, metal consumed and labor content.


Railway axles such as that shown in Figure 1 are key components in rail transport. As such, very strict quality and reliability requirements are applied to them since their failure in active service can result in catastrophe. Such a failure in service is caused by metal fatigue, which takes place at stresses that are less than the material yield point.


Cross-wedge Rolling Video Still
See a video of the cross-wedge rolling process online at

Mechanics of Axle Failure

Fatigue failure has three stages:

  • The first stage is the initiation of a fatigue crack in the surface layer of an axle near an area of stress concentration. This area can be near microcracks, micropores, non-metallic inclusions It will likely occur close to the axle’s surface – near scratches, scale, metal oxides and other metal defects. The initiation of a crack in the surface layer is likely because this is the place where the highest tensile stresses are during axle loading in service.
  • The second stage is the growth of the crack under the influence of a cyclical load. This stage of failure can take a long time and is dependent on the number of loading cycles, as may be described or anticipated on a Wohler curve.
  • The actual failure occurs quickly, accompanied by the generation of a coarse-grained structure at the surface of the fracture.


About the Blanks

Aside from the traditional manufacturing requirements of high productivity and economical metal and energy consumption, failure mechanics dictate that the technology of manufacturing railway axles at the stage of a forging blank has an additional requirement of obtaining high fatigue strength.

Consequently, the manufacture of a railway axle using a cast blank is prohibitive in that it does not ensure the required quality. Traditionally, railway axles are manufactured using blanks forged with hydraulic presses, radial forging or cross-screw rolling. It is possible that a continuously cast blank may also be used. The forming of the blank refines the metal structure, and it increases its fatigue strength. The standard for a railway axle anticipates a U-setting ratio of U ³ 3; that is, the reduction of a steel ingot’s cross-sectional area by no less than a factor of three.

Forging these blanks with hydraulic presses, producing only four units per hour, is extremely ineffective, so we have focused on comparative production with a radial forging machine (GFM SX-40), an axle-rolling machine (Mill 250) and a cross-wedge rolling mill (WRL20360 TS-03). Of the first two, the cross-screw axle-rolling machine (Mill 250) is preferable because it ensures a productivity of 60 axles per hour, and it is 1.5-2.0 times more productive than using the radial forging machine (Table 1).

Specification of Rough Railway Axle Manufacturing Methods
Table 1. Specification of rough railway axle manufacturing methods


Cross-Wedge Rolling Mill

The authors have developed a method of cross-wedge rolling (CWR) railway axles, the specification of which is given in Table 1. This technology has yielded the highest productivity – up to 120 axles per hour (less allowances and tolerances) – and achieves minimal metal consumption that helps minimize labor input doing finishing work at the lathe.

The high accuracy of a cross-wedge-rolled piece[1] allows the exclusion of some additional operations of other technologies such as the cutting of end wastes and the correction of an axle, which takes place after plastic deformation. The CWR mill (Figure 2) has two movable plates, which, along with flat tools, ensure the relatively high accuracy and linearity of a forged axle. CWR mills equipped with flat tools ensure a decrease of up to 10 percent in the cost of axles in comparison with those with roller tools. The life expectancy of a CWR flat tool, including midlife repairs, is no less than 500 000 axles, and it decreases tool cost per axle to 1.7%.

The weight of a CWR mill is 260 tons. It is 1.7 times less than that of a radial forging machine (GFM SX-40). Thus, the price of the CWR mill is lower. Moreover, a mill that is vibration-free can be set on the shop floor without a specially-constructed foundation bed. Because the operation of the mill is fully automated, round-the-clock operation is possible. A railway axle is manufactured with each mill cycle.



Evaluation of railway axle stress-strain and microstructural data from the rolled metal structure yields an increase in fatigue strength of the axles manufactured using the CRW method.

Consequently, the technology and equipment for the CWR of railway axles:

  • Yields higher fatigue strength values in axles
  • Yields at least a twofold increase in equipment productivity
  • Improves metal economy and a decrease in labor costs per part
  • Ensures the environmental safety of axle manufacture
  • Offers the opportunity to decrease the diameter and weight of axles, which has the potential to improve rail transportation economy

Compared with other forging methods, CWR ensures the forming of favorable properties symmetrically from the center of the axle and ensures better coaxial alignment of axle steps because of their continuous contact with the tool during the entire rolling cycle.



  1. Shchukin V.Y., Kozhevnikova G.V.; “New Trends in Development of Cross-Wedge Rolling Technology,” FORGE (USA). Oct. 26, 2015
  2. Kozhevnikova G.V., Shchukin V.Y.; “Cross-Wedge Rolling Theory, Technology and Equipment,” Industrial Heating (USA). Nov. 2014


Contact Val Leyzeruk, Nesso Automation Solutions, Detroit, Mich.; tel: 248-254-3773; email: The authors are Dr. Grazhina V. Kozhevnikova, research scientist at Physical Technical Institute of NAS Belarus; Aliaksandr O. Rudovich, AMT Engineering Center, Belarus; and Dr. Valery Y. Shchukin, leading research scientist, Nauchno-Tekhnologichesky Park BNTU “Politeknik,” Belarus.