Schematic diagram of the Vickers indentation array. The indentation sequence is in numerical order. 

Context in source publication

Context 1

... of 3x3 Vickers indentations of varying loads and spacing were performed on surfaces of each material type to simulate damage produced by multiple impacts of sharp particles such as SiC grits. Each series of indentations was performed using a microhardness tester (MHT1; Matsuzawa Seiki Co. Ltd. Japan) with a diamond pyramid indenter (Vickers type). Tests were performed on polished surfaces (final polishing: 1 μ m) of each polycrystalline alumina material. The indentation load and indentation spacing were chosen to cover the regime in which cracks become linked to each other. The load range was 0.5 N to 3 N and the indentation spacings used were 40, 30, 20, 10 and 5 μ m (in x and y directions). Indentations were performed in the order shown in figure 1. The damage produced by a given load-spacing combination was examined using SEM. The amount of material removed for the 9-indentation grid, for the load/spacing combinations 3 N/10 μ m to 1 N/5 μ m was measured for each material using an UBM non- contacting microfocus measurement system (UBM non-contacting Microfocus Measurement System, APT & T Advanced Products & Technologies Limited, Botley, Oxford 0X2 0JJ). At the loads used, one indentation alone does not cause material loss by lateral fracture, but as subsequent indentations are produced nearby, damage may occur if they are closely enough spaced. The interlinking of microfractures present due to previous indentations with cracks from the new indentation leads to material removal from the surface. For a given load there is a critical indentation spacing below which such material removal occurs. Figure 2 shows SEM micrographs of four square grids of 9 indentations of 3 N load at 40, 30, 20 and 10 μ m spacings performed on each material type. For the 20 μ m indentation spacing, severe damage is observed on the specimen with G=14.1 μ m whereas less severe damage is observed on the specimen with G=3.8 μ m. No damage is seen on the specimen with G = 1.2 μ m. For the specimen with G = 1.2 μ m severe damage is found for loads of 3 N only when the indentation spacing is ≤ 10 μ m . Table I shows the critical indentation spacing needed to produce severe material loss. For any indentation spacing greater than the critical value at a given load, while local microfracture may occur for individual indentations, no crack interlinking occurs, thus no significant amount of material is removed from the surface. For any spacing lower than the critical value for that load, material removal always occurs. Where the damage level changed from zero to full interlinking between two arrays of different spacing, (e.g. Figure 2(c)), then the critical spacing was taken as the mean spacing of the two arrays. Based on these observations a “wear map” for polycrystalline alumina was constructed as shown in figure 3. For a given load the critical indentation spacing for crack linking giving substantial material loss is strongly grain size dependent; it is about two times greater for specimens of the coarsest grain size than for specimens of the finest grain size. The wet erosion experiments of Franco and Roberts [11-13] used the same alumina materials as studied here. SiC particles of ~650 mm diameter travelling at ~2.4 ms -1 were used; comparison of the sizes of isolated impact craters with the sizes of indentations indicates that each impact event at 90° incidence roughly corresponds to a 2N load indentation [11]. Table II compares the “wear rate” for loads of 2 N and the wet erosion rate per particle [11]. Both the wet erosive and simulated wear rates per particle increase by about one order of magnitude with grain size over the range of grain size studied, with the indentation arrays giving loss rates per contact ~2.5 0.5 times those from wet erosive wear. The difference in volume removed per indentation and per erodent particle could be attributed to several factors: (a) During impacts either the sharp or the round area of the SiC grits can strike the surface. Blunt (Hertzian) indentation requires higher loads to produce damage. This effect can be seen in the SEM micrographs shown by Franco et al. [11] for the early stage of erosion. (b) The grid spacings chosen for comparison with wet erosion are those just close enough for the cracks to coalesce. In wet erosion, impacts that produce damage will be on average closer and the volume removed per impact will thus be less. (c) The indentation tests were carried out in air, while the erosion tests were carried out submerged in water. The surface mechanical properties of alumina change with the moisture content of the test environment [16]; however, at the high loads used here, this is unlikely to be a significant effect. Nonetheless, the “wear rates” from the two methods scale with each other well, implying that for low-speed erosive wear, the mechanism of material loss is by damage linkage between successive close impacts, and that the indentation grid method can be used to predict relative erosive wear rates of ceramic materials. Arrays of closely spaced Vickers indentation can be used to produce “wear maps”, which provide a guideline for predicting the low-impact-velocity erosive wear resistance of brittle ...