Hard materials free of WC and Co for tribological applications: WB4 ultrahard ceramics and TiC cermets.

Abstract

This thesis addresses the problem of finding wear-resistant materials that could substitute WC-Co alloys in specific tribological applications. The main challenge lies in the versatility of these alloys to cover a broad range of physicochemical properties by merely varying the WC/Co ratio and the grain size of the ceramic phase. This is not possible when using other combinations of ceramic and metal since, in general, the interfacial strength is lower than that observed between tungsten carbide and cobalt. The need to find alternatives to the WC-Co system arises from the strategic value of these raw materials, currently controlled by the People's Republic of China. Furthermore, batteries developed for electric vehicle motors contain a significant amount of cobalt, exacerbating the supply risk of this metal in the European Union. This thesis proposes the development of two very different materials that could replace WC-Co alloys in two sectors of interest: materials used in shot blasting or water jet cutting and guiding systems in hot rolling equipment. For the former, tungsten tetraboride, a new metallic boride with hardness exceeding 40 GPa, has been considered, and for the latter, the development of cermets based on titanium carbide and iron is suggested. The first set of results in this thesis focuses on the development of ultrahard materials based on tungsten tetraboride that could substitute WC binderless carbides in applications requiring high erosive wear resistance (i.e., nozzles for water cutting or shot peening). This objective has been addressed applying hot isostatic pressing (HIP) technique to both as-received WB4-B and WB4-B-Ta powder mixtures. Overstoichiometric B/W ratios and Ta additions were selected for their effect on stabilizing the ultrahard WB4 phase, as described in the literature. This is a difficult task, since the stability of WB4 requires a boron activity much higher than that corresponding to a B/W ratio of 4. Porosity removal was more efficient in the alloy containing metallic tantalum, achieving near full density at temperatures 300 ºC lower than those reported so far for these materials. The WB4 phase is better stabilized by HIPing at 1350 ºC than at 1100 ºC. This is due to the formation of TaB2, which at 1100 ºC, likely occurs by direct reaction between metallic Ta and the surrounding WB4 particles. At 1350 ºC, diffusion is enhanced and the reaction between free B and Ta particles becomes more probable. The nanohardness of WB4 HIPed specimens reaches 43 GPa, that is, 43 % higher than the highest reported for binderless WC. Indentation toughness is similar to that reported for WC-1 wt. % Mo2C (5.6 and 6.6 MPa·m1/2 respectively). Contrary to that reported by other authors, it has been confirmed that metallic Ta additions enhance the decomposition of tungsten tetraboride into tungsten diboride and Ta-rich borides. According to XRD data and SEM analyses, this phenomenon is more pronounced when HIPing is made at lower temperatures, which is against thermodynamic calculations. This is probably related to kinetic effects, since direct reaction between free B regions and metallic Ta powders requires higher mobility than that observed after HIPing at 1100 ºC. The latest results were obtained by investigating the possibility of producing cermets based on WB4-B-TaB2 powders with Ni additions. It was confirmed that densification of WB4-B-TaB2 is notably activated by Ni additions, reducing by 250 ºC the temperature needed for porosity removal. This is likely due to the formation of a liquid phase above 1007 ºC, compatible with the formation of Ni4B3 and NiB borides on heating. These borides are mainly formed by direct reaction between free B and Ni powders, although some boron could also be available from the partial decomposition of WB4 into W2B5. As no metallic nickel remains after HIPing, the toughness of these composites is very low. On the other hand, strength and toughness of WB4-B-TaB2-Ni alloys are notably improved by TiAl3 and Zr additions. Although Ni containing borides are still present in these alloys, there are also Ni-Al rich phases free of boron which provide a significant toughening effect, as confirmed by indentation cracking. These B-free regions remain unbroken joining the crack lips, although it progresses forward through boride grains. Highest fracture strength values correspond to the combined addition of TiAl3 and Zr powders to WB4-B-TaB2-Ni mixtures (≈ 1 GPa). However, in these materials WB4 grains are fully decomposed into a combination of mixed borides. The challenge associated with manufacturing components from TiC-Fe alloys for steel wire guiding in hot rolling equipment lies in their difficulty to be sintered. In this thesis, fully dense TiC-Fe-Cr-Mo based cermets have been sintered from two different alloys: TiC-Fe-Cr3C2-Mo and TiC-Fe-Cr3C2-Mo2C. It has been investigated the effect of metallic molybdenum or molybdenum carbide additions as wetting activators during liquid phase sintering of TiC-Fe powder mixtures. The addition of molybdenum in these systems is carried out to increase the wetting properties by the formation of the so-called core-rim structures, typical of TiC based cermets. These structures are formed by Ti-rich cores (α’ phase) surrounded by shells (α’’ phase) comprised of complex cubic carbides. Chromium was added into the powder mixtures as carbide, since its oxidation resistance is higher than that of metallic Cr and it is brittle enough to distributed homogeneously with the other constituents of the alloy during the mixing and milling process. In addition, the effect of pressure in the furnace chamber on porosity removal and on the formation of surface compositional gradients has also been explored. Results show that densification of TiC-Fe-Cr-Mo cermets is strongly affected by the selection of starting powders and the vacuum condition used during the sintering cycle. Liquid phase sintering is enhanced by using a lower pressure during the heating ramp of the sintering cycle (i.e., 10-5 mbar vs. 10-2 mbar), confirming that carbothermal reduction of the most stable oxides present in these materials requires very low oxygen activity. The massive evaporation of the binder phase observed at 10-5 mbar is avoided by injecting Ar in the sintering chamber above 1300 ºC. However, cermets sintered in this condition present some residual porosity due to Ar entrapment. Cermets with Mo2C additions lead to higher densities than those based on metallic Mo. This is likely due to the finer particle size of the former which likely accelerate diffusion kinetics associated to the formation of “core-rim” structures. Carbon losses after sintering are also higher in compositions based on Mo2C additions, suggesting that these powders enhance carbothermal reduction of oxides present in the surface of starting powders. Significant migration of the binder phase towards the cermet surface is observed in TiC-Fe-Cr-Mo cermets when sintering is carried out at low vacuum levels (i. e., 10-2 mbar). Compositional gradients produced by this migration are likely related to oxidation occurring once the closed porosity state is reached. Best results are obtained with 10-5 mbar of pressure and injection of 1.2 bar of Ar at 1300 ºC (named HV-1 cycle) and with 10-2 mbar of pressure without Ar injection (named LV-2 cycle). These two cycles lead to materials with low porosity levels, low mass losses and homogeneous microstructures. These materials achieve significant hardening through air-quenching from 950 ºC, resulting in properties suitable for hot wear applications. Another key aspect of TiC-Fe-Cr-Mo based cermets is the carbon partitioning between the ceramic and the metallic phase during sintering, since the latter is susceptible to hardening by austenitizing and quenching. It has been found that relatively low differences in the carbon content of the FeCrMo binder phase induces significant microstructural changes both after sintering and after subsequent thermal treatment. Thus, the alloy with higher carbon content (that with metallic Mo) presents higher precipitation of Cr-rich M7C3 carbides at the (Ti1-x,Mox)yCz - metal interface and certain amount of retained austenite. In the alloy with lower carbon content (that with Mo2C), there is no retained austenite, and the precipitation of Cr-rich carbides is less abundant. By means of EDS-TEM analyses of the metallic matrix of TiC-Fe-Cr-Mo cermets after sintering, it was found to be alloyed with Cr and Mo. The contents of these elements are similar to those reported for hot work steels (i.e., AISI H11 or H13). However, these metallic matrices are very different as they do not contain Si, Mn or V. With the aim of studying the hardenability of these materials, thermal treatments have been carried out after sintering using a quenching dilatometer. Volumetric changes associated to phase transformations have been measured by austenitizing at 950 ºC for 20 min and subsequently cooling at different rates (from 100 ºC/s to 0.1 ºC/s). Carbon differences are observed to shift bainitic transformation, with its onset found at approximately 1 ºC/s. As expected, this transformation is shifted towards higher temperatures as the total carbon content of the alloy decreases. CCT diagrams built on the basis of dilatometric results confirm that bainitic transformation occurs at higher cooling rates than in standard hot work steels. Anyhow, the metallic matrices present in these TiC-Fe-Cr-Mo cermets are air-quenchable at approximately 1 ºC/s, inducing a hardness increase of 30 % with respect to that of as-sintered materials. Precipitation of M7C3 and M23C6 carbides is observed by means of SEM/TEM analyses and XRD measurements. These findings agree with thermodynamic calculations made by Thermocalc® software. Although it still unclear, preliminary results suggest that the precipitation of M23C6 carbides is related to the progression of the bainitic transformation.