We cannot imagine modern life without rolls, which are a key component for the production of rolled steel products. Rolled steel products represent essential components in every major industrial sector, and can be found at every step. From the first devices for rolling sheets and rods designed by Leonardo Da Vinci, the process of rolling has been under constant development and is today a complex and perfected process. Due to constant strong tendencies towards higher product quality and higher productivity, improvements in the field of rolls are especially important. In the past hundred years, there has been a rapid development of various types of roll grades for hot rolling, along with the development of different roll manufacturing technologies. In recent years, centrifugally cast composite rolls with a working layer of high-speed steel have established themselves as standard for rougher rolling stands and initial stands for finishing rolling. These rolls contain a high proportion of very hard carbides in the working layer and show excellent wear resistant properties at elevated temperatures on the rolling mill. Further possibilities for improvements lie, among other things, in the refinement of microstructural components of the working layers of the rolls.
Refinement of the microstructure is the only known process that has a positive effect on both strength and toughness, and refinement also has a positive effect on a number of other properties. In practice, microstructure refinement is usually achieved by increasing the cooling rate of the melt, mechanical deformation, or by inoculation. In the roll manufacturing, the inoculation process is the most suitable process of grain refinement due to technological limitations. In recent years, inoculants based on rare earth elements have begun to be used for inoculating different steels. These form fine oxides and sulfides in the melt. These inclusions of rare earth elements have very good crystallographic matching with austenite and act as nucleation sites.
In this work, we conducted a test of a cerium-based inoculant. At the industrial level, two work rolls for hot rolling of sheet metal were produced with a standard production process of centrifugal casting of the working layer and static casting of the roll core. The melt of the working layer of roll A was not inoculated, while the melt of the working layer of roll B was inoculated. For both analyzed alloys, we carried out solidification simulations using the Thermo-Calc software with models of equilibrium solidification and solidification according to the Scheil-Gulliver model. From the working layers of the rolls, samples were cut before heat treatment. We examined these samples using light and scanning electron microscopy (SEM). When examining the samples with a light microscope, we first etched them with Nital and examined them. We then re-ground the samples, polished them, and etched them with a color etchant called Groesbeck's reagent. From the working layers of both rolls, after completing heat treatment, we made specimens for tensile tests, toughness tests, wear tests, and bending tests.
Results of the Thermo-Calc simulations show, that solidification by both simulation methods begins with austenite, followed by the eutectic reaction of austenite and MC carbide. In the simulation according to the equilibrium model, this reaction proceeds to the end of the solidification range, while in solidification according to the Scheil-Gulliver model, it is followed by eutectic reactions of M2C and M7C3 carbides. Upon examining the samples with SEM, we noticed two types of eutectically solidified carbides in the microstructure. One type of observed carbides was rich in vanadium, the other in molybdenum. In the analysis of the samples with SEM, we also noticed larger differences in size and shape of the vanadium-rich carbides. In roll B, the vanadium eutectic carbides were larger, more spherical in shape and more homogeneously distributed. In light microscopy, we mainly noticed previously observed differences in shape, size, and distribution of vanadium carbides. It was also noted that the proportion of M2C eutectic carbides in roll B was smaller, and the carbide network of this type of carbides was less connected. In the analyzed sample of roll B, a lower proportion of pearlite was also noticed, compared to the sample of roll A. In the analysis of samples etched with Groesbeck's reagent, we detected the occurrence of three types of eutectic carbides: MC, M2C, and M7C3. In the analyzed samples, we also compared the total proportion of MC and M2C carbides, confirming the previously noted detail that inoculation of the melt affected a reduction in the total proportion of these carbides by 2,26 area percent. Comparing the results of mechanical tests, we found the most significant differences in the measured values of wear resistance and the modulus of elasticity. In both of these cases, roll B showed more favorable properties: 9,1 % higher in the value of the elastic modulus and 18,6 % less in the amount of worn material on the samples. The average value of tensile strength was 1,1 % lower in the inoculated roll, the average value of impact toughness was 4,7 % lower in the inoculated roll, and the average value of bending strength was also lower in the inoculated roll by 3,6 %.
Based on the results, it was concluded that the inoculation of the high-speed steel melt for the working layers of the rolls did not have notable grain refinement effect. Inoculation had the greatest effect on the size, shape, and distribution of vanadium carbides. In the inoculated sample, these were larger, more evenly distributed, and more spherical in shape. It was also observed that there was a decrease in the overall total proportion of M2C and M7C3 carbides, which consequently also contributed to a reduction in the carbide network. In the sample of the inoculated roll, we noticed a smaller overall proportion of pearlite in all analyzed areas. Comparing the results of measurements of mechanical properties, we did not detect major differences except in the amount of worn material during the wear test and the modulus of elasticity of the specimens. It was found that inoculation positively affected the average of these two properties, while it had a negative effect on the other measured properties averages. In percentage terms, the positive impact on mechanical properties was greater than the negative.