Wear phenomena of grinding wheels with sol–gel alumina abrasive grains and glass–ceramic vitrified bond during internal cylindrical traverse grinding of 100Cr6 steel

This article presents the results from a traverse internal cylindrical grinding process conducted using grinding wheels with sol–gel alumina abrasive grains and varying volumes of glass–ceramic vitrified bond. The aim was to determine how changing the bond volume within the range of Vb = 11.5–14.5 % influenced the wear phenomenon of the microcrystalline sintered corundum abrasive grains and the ceramic bond bridges with glass-crystalline structure. The paper presents grinding wheels with conic chamfer and the most important wear phenomena of the grinding wheel components occurring in the area of contact between the tool and the workpiece material. The work includes the methodology of the experimental tests on the traverse internal cylindrical grinding of bearing rings made from steel 100Cr6 (62 ± 2 HRC). A wide range of test results were analyzed, including machined surface roughness parameters (Ra, Rz, Sm, Δa), grinding power, grinding wheel volumetric wear Vs, material removal Vw, G-ratio, root-mean–square roundness deviation from mean circle (Δ, rms), grinding wheel edge wear, grinding wheel microtopography parameters (Sa, St, Sdr, Sds), and SEM images of the wear marks on the grinding wheel active surface. The dominant wear phenomena of the grinding wheels with conic chamfer and the influence of the bond volume share on its intensity were experimentally determined. The results show that the decrease in bond volume from Vb = 14.5 to 11.5 % increases the grinding wheel life by twofold. This was due to a decrease in abrasive wear and plastic flow phenomena and to a greater fracture wear (mostly fatigue and thermal-fatigue). These events led to a periodic shedding of the oxide layer and the plastically deformed grain layer and also revealed the crystals’ sharp edges, located below the plastically deformed surface layer.


Introduction
Modern grinding processes provide a high quality of machined surface with constantly improving material removal rates. The quality is obtained by, among other procedures, applying new abrasive materials, developing new kinematic variations, and modifying grinding process parameters or by introducing modifications in the grinding wheel structure. These elements are often interrelated, with, for example, the effective realization of a single working pass of the grinding wheel being connected to the proper modification of the grinding wheel's macrogeometry, i.e., the shaping of the conic chamfer.
Grinding wheels with conic chamfer were developed for the needs of traverse grinding processes, in which the whole machining allowance is removed in a single working pass of the grinding wheel. The most common traverse processes include external cylindrical continuous path-controlled grinding (CPCG), where grinding wheels from superhard abrasive grains are used [1][2][3][4][5][6]. Methods of traverse grinding with grinding wheels made from Al 2 O 3 grains are also being developed [7][8][9][10][11].
In relation to internal cylindrical grinding processes, the subject literature offers only descriptions of processes called peel grinding or internal deep traverse grinding [12], as well as an example of traverse cylindrical grinding with the Quickpoint method [13]. This method consists of reducing the area of contact between the grinding wheel and the workpiece to the aforementioned point-hence the name of the method-by deflecting the grinding wheel axis in relation to the workpiece axis by an angle of α k =0.5° [14,15].
Grinding wheels from cubic boron nitride (CBN) with the total height in the axial direction T<6 mm, characterized by a conic rough grinding zone, are used in the peel grinding and traverse grinding processes. Such grinding wheels make it possible to remove the total machining allowance of a e tot =0.15 mm in a single pass. The grinding power is also decreased and therefore the tendency to shape the conic opening is limited. Moreover, flexibility is improved compared to conventional cylindrical grinding, while the high process reliability and high quality finish workpiece are maintained [12].
The conic chamfer angle χ is crucial to the rough grinding process course, and it is dependent on a number of parameters such as the grinding wheel coasting, volume of the removed machining allowance, grinding wheel height, as well as the requirements concerning surface quality. The latter determines the finish grinding and sparking-out zones' width. As a result of grinding wheel wear, the rough grinding zone is moved into the finish grinding and sparking-out zone, which causes its shortening.
Application of increased grinding speed v s =60-80 m/s and a properly shaped grinding wheel active surface (GWAS) macrogeometry, divided into the conic zone with angle χ=5°and the cylindrical finish grinding zone (2-4 mm long), enables the achieving of the material removal rate of as much as Q w =24-37.7 mm 3 /s. The sparking-out zone, which is then long enough, guarantees the attainment of a machined surface roughness on the level of Rz<1 μm [12].
The described grinding wheel geometry makes it possible to increase the material removal rate through increasing the axial table feed speed v fa . Even though this leads to an increased rough grinding zone load, it is possible to obtain a good quality surface thanks to multiple grinding of the workpiece by the cylindrical zone of the finish grinding. As a result, with machining efficiency similar to hard turning, considerably better surface quality is obtained [12].
As up to now mostly expensive grinding wheels from cubic boron nitride, which required specialized grinding machines with considerable stiffness and large grinding speed (v s >60 m/s), have been used in traverse grinding processes, the need to develop new grinding wheel constructions from far cheaper grains that could also be used in traditional grinding machines has arisen.
2 Grinding wheels with Al 2 O 3 abrasive grains for internal cylindrical traverse grinding In the case of the application of cheaper abrasive grains on the basis of Al 2 O 3 , it was necessary to considerably increase the total grinding wheel height (T=20 mm) due to the lower cutting ability of such abrasives as compared to CBN grains [16][17][18][19]. Figure 1 presents a construction scheme (Fig. 1b) and the axial microgeometry profile with conic chamfer (Fig. 1c), designed for traverse internal cylindrical grinding. What is also included are microscopic images of the grinding wheel active surface and single microcrystalline sintered alumina abrasive grains of the SG™ type produced by Norton company (Fig. 1a), which the described grinding wheels were made from.

Abrasive grains
Sintered microcrystalline alumina is characterized by similar hardness and greater ductility than fused alumina oxide grains, which translates into greater resistance to fragile cracking. Compared to white fused alumina, sintered microcrystalline alumina grains are 10 % harder. Their shape is similar to the typical one but is characterized by greater uniformity of shape and size. The grains are composed of a large number of minor crystallites, connected through a thin crystalline layer ( Fig. 2) [20].
As the grains of conventional white fused alumina are characterized by a polycrystalline structure with preferential crystallographic planes, they crack parallel to these planes (the crystallographic layout) under the influence of strain. Such a process results in the loss of considerable grain mass, which is additionally enhanced by defects in the grain that cause microchipping. The dressing process also leads to dulling of the grinding wheel active surface through flattening of the grain corners. The reason for this is the fact that white fused alumina grains are usually single crystallites [16][17][18][19]21].
A sintered microcrystalline alumina grain has a submicrocrystalline structure with particles sized 0.1-5 μm (however, usually <1 μm). For reference, the size of sintered alumina crystallites is similar to the nominal grain size or slightly smaller. Such a structure results in a greater value of the critical coefficient of grain strain intensity K Ic . Proper selection of grinding and dressing parameters makes it possible for the phenomenon of abrasive grain self-sharpening to occur, which is caused by microparticles breaking out from it, which, in turn, lead to the creation of sharp corners and cutting edges. As a result, it is possible to limit the grinding wheel wear through the creation of a greater number of active apexes on its active surface [16][17][18][19]21].

Glass-crystalline bond
The described grinding wheels were made using glasscrystalline bond from the CMAS system (CaO-MgO-Al 2 O 3 -SiO 2 ) [22][23][24]. It is characterized by good wetting, while the presence of the basic crystalline phase of the pyroxene type improves its resistance to abrasive wear and microhardness of the material as compared to glass bond [25]. Figure 3 presents the phase diagram of the applied bond.
The analysis was made using the X-ray diffractometer Siemens D5000, which utilizes CuK radiation (counting time 5 s, sample rotation 1 rpm, measurement range 10-60 s). The bond diffractometer revealed the presence of two crystalline phases from the pyroxene group, augite and diopside, in the sample. Table 1 presents the most important structural and physicalchemical properties of the applied glass-crystalline bond as compared to the glass bond (amorphous) and microcrystalline sintered alumina grains.

Wear phenomena of the GWAS
The most important GWAS components wear phenomena include abrasive and facture wear [16,[27][28][29]. These Grinding wheel with conic chamfer for internal cylindrical traverse grinding: a SEM images of single abrasive grains and grinding wheel active surface, b construction scheme, and c microgeometrical axial profile of the grinding wheel active surface with an exposed conical chamfer angle χ processes take place in both the abrasive grains and the bond and therefore the following can be discerned: -Abrasive wear of the active abrasive grains vertexes, -Abrasive wear of the bond in the areas of contact with the machined material, -Fracture wear of the abrasive grains due to mechanical and thermal shock loads, -Fracture of bond bridges, -Fracture at the interface between abrasive grain and bond bridge, and -Smearing of the grain vertexes and intergranular free spaces with the machined material. Figure 4 presents examples of microscopic views illustrating the basic processes of grinding wheel component wear.
Dulling of the grain vertexes is caused by the abrasive wear and occurs as a result of a number of phenomena occurring on the microscale such as the following [16]: -Wear due to frictional interaction between workpiece and abrasive grain, -Plastic flow of the abrasive grain at high temperature and pressure, -Crumbling of the abrasive grain due to thermal diffusion and microscale mechanical impact, and -Chemical reaction between abrasive and workpiece material at elevated temperatures and in the presence of grinding fluids.
The most important phenomena causing fractures in the form of cracking or chipping of the abrasive grain particles, as well as chipping of the bond particles and fracturing of its bridges, include the following: -Effective wear caused by the mechanical shock load, -Fatigue wear caused by recurrent influence of the mechanical shock load, and -Thermal-fatigue wear in which the fatigue wear is additionally supported by the periodically changeable or intermittent thermal shock load.

Experimental tests
When grinding wheels made from conventional abrasive grains (on the basis of Al 2 O 3 or SiC) are used in industrial conditions in automated cylindrical grinding (e.g., in the bearing ring industry), the dressing procedure is usually performed after each working pass. Such a strategy is used in order to maintain full repeatability of the machining effects. As a result, most of the grinding wheel volume is lost not because Fig. 2 Microscopic view of the SG™ grains microstructure: a magnification ×1000 and b magnification ×10,000 Fig. 3 X-ray powdered crystal pattern of the glass-crystalline bond, thermally treated at 1000°C for 1 h, at which point the grinding wheels are produced with this bond-the intensity of the deflected X-ray beam for an angle of beam reflection within the range from 10 to 60° [26] of the wear phenomena but because of the cutting ability being renewed (conditioning, shaping, and sharpening). Application of grinding wheels with conic chamfer requires far more from the dressing procedures because of the necessity for precise shaping of the grinding wheel macrogeometry. For this reason, it is crucial to look for possibilities to prolong the life of those grinding wheels that usually work with two or three times greater machining allowance resulting from the traverse grinding process.
The purpose of the experimental tests was to determine the most advantageous volume of glass-crystalline bond from the perspective of extending the grinding wheel life in the process of traverse internal cylindrical grinding of steel 100Cr6. The experiment's results were used to determine the dominant Ua=20 kV 10 m

Methodology of experimental tests
The experimental tests were carried out on a station equipped with a universal grinding machine RUP-28P. The applied values of the key machining parameters (a e =0.20 mm and v fa =2.0 mm/s) correspond to grinding with material removal rate Q w ≈24 mm 3 /s. What was registered during the tests were a grinding power ΔP, root-mean-square roundness deviation from mean circle (Δ, rms), and grinding wheel outer diameter changes d s , which were indicative of the volume wear V s . The grinding wheel outer diameter was measured on the cylindrical part, while the perimeter profiles in the roundness error measurements were registered on the conic and the cylindrical areas of the grinding wheels. After the trials were finished, workpiece surface roughness parameters were determined, microtopography measurements were made, and the SEM observations of the examined grinding wheel active surfaces were performed. Exceeding the limit value of grinding power gain ΔP l =1000 W, which was determined on the basis of previous research on traverse internal cylindrical grinding with such grinding wheels [30,31], was adopted as the criterion of cutting ability loss and the end life of the examined grinding wheels. It was agreed that over that value, the risk of defects appearing on the workpiece surface in the form of grinding burns increased dramatically. Table 2 includes a presentation of the conditions of the performed experimental tests.

Grinding wheels
Three variations (C1, C2, and C3) of the type 1 grinding wheel sized 35×20×10 mm were used in the tests, which were made from microcrystalline sintered alumina grains SG™ number 46 and glass-crystalline bond. The three variations differed only in the bond volumes which were V b =14.5 % (grinding wheel C1), V b =13.0 % (grinding wheel C2), and V b =11.5 % (grinding wheel C3), which corresponded to three classes of grinding wheel hardness M, L, and K, respectively. Conic chamfer with the angle χ=0.72°and width b=16 mm, selected according to the machining allowance (a e =0.20 mm), was shaped on the grinding wheel active surfaces.
In order to make quick shaping of the conic chamfer with specific geometrical parameters on the grinding wheel active surface possible, it was necessary to develop a special automated dressing device [32]. The most important element of the device, as presented in Fig. 5, is the sliding table on ball slideways to which the blade mounting of the single-grain diamond dresser and the micrometric screw, which was used for precise setting-up of the conic chamfer angle, were fixed.
The theoretically and experimentally determined precision of conic chamfer shaping in the range χ=0-1.5°is approximately ±3 % [33]. Table 3 presents the characteristics of the three grinding wheel variations used in the experimental tests.

The applied measuring systems
The grinding results were analyzed, and the grinding wheel active surface wear phenomena were identified on the basis of the measurements and observations carried out using a variety of modern measurement systems. The machined surface roughness parameters were determined on the basis of the profiles registered by stylus profilometer Hommel-Tester T8000 by Hommelwerke GmbH (Germany). The grinding wheel active surface microtopographical measurements were made using the multi-head measuring system Talysurf CLI 2000 by Taylor-Hobson Ltd. (UK), which performs the measurement using the laser triangulation method. The GWAS microscopic images were registered on the scanning electron microscope JSM-5500LV by JEOL Ltd. (Japan). The correct identification of objects observed on the SEM images was confirmed with an energy-dispersive X-ray spectroscopy (EDS) module INCAPentaFET-x3 from Oxford Instruments (UK). Values of the grinding wheel roundness errors were determined with the roundness measuring instrument Rondcom 44 by Carl Zeiss AG (Germany). Moreover, in order to estimate the edge wear, the microgeometry of the axial profile of the grinding wheels with conic chamfer was measured on bench-type multisensor coordinate measuring machine VideoCheck-IP 250, produced by Werth Messtechnik GmbH (Germany).  The test results obtained show that the C3 grinding wheel, characterized by the lowest bond volume V b =11.5 %, had the longest life. It was used for grinding eight rings until the grinding power limit value ΔP l =1000 W was exceeded. The remaining grinding wheels, with greater bond volume, were used to process only between 4 and 5 rings (Fig. 6).
The charts illustrating the changing values of the selected parameters describing machined surface roughness, included in Fig. 6a-d, indicate that satisfactory surface roughness was obtained for all grinding wheels. What is also crucial are the minor dispersions in the values of the determined parameters, which indicate the repeatability of the obtained machining effects. The tendency to rise along with the grinding wheel operating time prolongation is visible; however, the tendency is relatively minor. The greatest influence of the changes caused by the grinding wheel component wear on the grinding course can be observed in the chart covering grinding power gain ΔP (Fig. 6e). This parameter reacts with great precision to any increase of friction in the machining process, which is indicative of dulling of the abrasive grain vertexes on the GWAS.  Fig. 7 show that as the grinding wheel operating time increases, the mean values of the machined surface roughness parameters grow ( Fig. 7a-d). The ranges of these changes were, respectively, ≈37 % for parameters Ra and Rz, ≈37 % for parameter Rz, ≈22 % for parameter Sm, and ≈7 % for parameter Δa. The mean value of grinding power gain was insignificant and (within the range of 832-935 W) without any visible tendency (Fig. 7e). Figure 8 presents a comparison of the total material removal volume V w (Fig. 8a), volumetric grinding wheel wear V s (Fig. 8b), and the G-ratio G=V w /V s (Fig. 8c) obtained with the applied grinding wheels. Analysis of the charts presented in Fig. 8 shows that for all grinding wheels, the G-ratio barely changed and ranged from 118.9 (grinding wheel C1) to 93.6 (grinding wheel C3), while a decreasing tendency in the function of decreasing the volumetric share of the bond is visible. Figure 9 contains charts presenting changes in values of parameters describing root-mean-square roundness deviation from the mean circle (Δ, rms) measured on the conic and cylindrical areas of the examined grinding wheels with the registered peripheral profiles. The initial values (Fig. 9a), corresponding to V w =0 mm 3 (measured after the dressing procedure), were compared with the values determined on  (Fig. 9b).
In the case of C1 and C2 grinding wheels, a slight decrease in the value of parameters describing root-mean-square Fig. 9 The changing values of the maximum roundness deviation Δ and root-mean-square deviation from the mean circle rms and circumference profile of the tested grinding wheels, measured using roundness measuring instrument Rondcom 44 by Carl Zeiss AG: a after dressing and b at the end of life roundness deviation from the mean circle after grinding is visible (Fig. 9b), as compared to the values determined after the dressing procedure (Fig. 9a). The case is different when it comes to the C3 grinding wheel which remained operational the longest. The measured Δ values after work are considerably greater than after dressing for the C3 grinding wheel, especially in the most loaded rough grinding conic zone (Δ=22 μm after dressing and Δ=31 μm after termination of life).
A comparison of the registered axial profiles of the examined tools after work (Fig. 10) shows that as the bond volume decreases, the grinding wheel edge wear increases. The greatest change in the shape of the axial outline, shaped in the dressing procedure, relates mostly to the conic area, which is most loaded because of the rough grinding procedure. The outline of the C3 grinding wheel (Fig. 10c) becomes considerably more visible the more smoothed out the active abrasive grain vertexes are at the beginning of the conic zone, as compared to grinding wheels C1 and C2 (Fig. 10a-b).
The comparison of macrotopographies of the active surfaces of the examined grinding wheels before and after the machining (Fig. 11) shows that in the case of grinding wheels C1 and C2 after work wear smears/loads the intergranular free spaces with machined material chips ( Fig. 11a-b). The surface of these grinding wheels is dark, which means that ground products remain in the intergranular free spaces. Figure 11c, which presents the C3 grinding wheel surface, does not have Fig. 10 Axial profiles of the tested grinding wheel active surfaces at the end of life: a grinding wheel C1, b grinding wheel C2, and c grinding wheel C3 (measured using bench-type multisensor coordinate measuring machine VideoCheck-IP 250 by Werth Messtechnik GmbH) Fig. 11 Macroscopic images of the active surface of the tested grinding wheels after dressing and at the end of life: a grinding wheel C1, b grinding wheel C2, and c grinding wheel C3 visible smears/loads, while lighter areas are visible, which result from new cutting vertexes being revealed as a result of microchipping of the active abrasive grains and glasscrystalline bond bridges. Figure 12 presents the registered GWAS microtopographies after grinding, alongside the values of the selected roughness parameters, as well as analyses of islands with the threshold value of 0.2 mm from the highest topographical point.
The comparison of the Sa and St parameter values shows that all the grinding wheels were characterized by similar heights of active surface unevenness after work, while the greatest Sa value was determined for the C3 grinding wheel (Fig. 12c). Greater differences were noted for the values of developed interfacial area ratio Sdr and density of summits of the grinding wheel active surface Sds. What is clearly visible in the case of both parameters is the growing Fig. 12 Collection of selected analyses of microtopographies of the tested grinding wheels at the end of life: a grinding wheel C1, b grinding wheel C2, and c grinding wheel C3 (measured using multi-head measuring system Talysurf CLI 2000 by Taylor-Hobson Ltd.) tendency for a progressive decrease of the bond volume in the grinding wheel.
Moreover, the island analysis was carried out in order to determine a group of parameters describing the geometric features of the active abrasive grain vertexes and the intergranular spaces. It shows that as the bond volume decreases, the active surface of the examined grinding wheels, after work, is characterized by a greater number and volume of the abrasive grain vertexes.
In order to perform the visual evaluation of the condition of the examined grinding wheels' active surface, SEM observations were carried out, whose selected results are compared in Fig. 13.
The SEM images show the majority of abrasive wear of the active abrasive grain vertexes. Moreover, in the areas of the GWAS contact with the machined surface, microsmearings with chips of the machined material and grinding wheel wear products, such as chipped bond and grain fragments, were created. The chips located in the grinding zone were embrocated into the microfractures located on the smoothedout grain surface. However, no extensive intergranular free space smearings were observed. What was noticed on the surface of all examined grinding wheels were chips of the machined material characterized by a continuous shape, resulting from the characteristics of the machined steel 100Cr6, as well as significant size diversification. Also visible are chips the length of 150 μm shaped during rough grinding, as well as microchips the size of a few micrometers, created during the last stage of the machining process, when sparking out of the surface takes place. There were also spherical chips on the surfaces of the C1 and C2 grinding wheels (Fig. 13c, d, g), and melted machined material congealed on the smoothed-out abrasive grain vertex was observed (Fig. 13h).

Discussion of the test results
The test result comparison presented in Section 4.4 allowed for a multicriterion evaluation of the grinding wheel surface element wear phenomena in the process of traverse internal cylindrical grinding in steel 100Cr6 with increased material removal rate. Among the three grinding wheels selected for the tests, the longest life was registered in the C3 grinding wheel (Fig. 6), characterized by the lowest volume of glasscrystalline bond V b =11.5 %. It can be therefore assumed that minimizing the bond volume in the grinding wheel contributes to easier chipping of the dulled abrasive grain microvertexes that result from reductions in the abrasive grain binding strength. As a result, the C3 grinding wheel maintained its cutting ability for approximately twice as long than in the case of grinding wheels with greater bond volume (C1 and C2). The expected workpiece top layer properties were maintained, along with its geometry and low roughness (Fig. 7a-d), without a noticeable influence on the grinding power (Fig. 7e) and with only a slight decrease of the obtained G-ratio (Fig. 8c).
In the case of the greater bond volume (grinding wheels C1 and C2), the grains were held in place by the bond bridges for a longer period of time, which contributed to the smoothing out of the active abrasive grain vertexes and simultaneous decrease in the intergranular space depth. As a result, limitation of the root-mean-square roundness deviation from the mean circle was observed, both in the conic and the cylindrical zones of these grinding wheels. The C3 grinding wheel worked for the longest duration, and it also registered the greatest volume wear (V s =40.61 mm 3 , Fig. 8b), as a result of which the circumferential outline geometry was most influenced, as compared to the other examined grinding wheels, by the wear phenomena. The shaped GWAS microgeometry was characterized by a greater root-mean-square roundness deviation from the mean circle values and visibly greater free intergranular spaces (Fig. 9b) than after dressing (Fig. 9a). This is indicative of a progressive bond bridge microchipping process caused by the wear of the active abrasive grain vertexes. This also contributed to the greatest change of the C3 grinding wheel axial outline shape in relation to the initial conic chamfer outline shaped in the dressing procedure (Fig. 10c).
The microscopic view of the C3 grinding wheel after grinding was characterized by the occurrence of brighter areas, which resulted from new cutting vertexes being revealed (Fig. 11c), which were visible on the surfaces of C1 and C2 grinding wheels ( Fig. 11a-b). Moreover, the parameters determined, on basis of the grinding wheel surface microtopographies, showed that the C3 grinding wheel was characterized by the greatest number of vertexes (Sds=56.3 pks/mm 2 ), the greatest developed interfacial area ratio (Sdr=283 %), the greatest number (194) and volume of islands (0.00123 mm 3 ), as well as the greatest surface share of the islands (20.7 %) from all of the examined tools (Fig. 12). This means that for the examined grinding process, with the applied parameters, the active abrasive grain vertexes and bond bridge load made it possible to partially renew the C3 grinding wheel cutting ability during a relatively long work period.
The long working life of abrasive tools with microcrystalline sintered corundum grains is mostly dependent on the conditions in which the grains work and on the dominant wear processes, including the complex system of phenomena combining the effects of chipping, plastic deformations, and chemical reactions [34,35]. The microfractures existing in the abradant develop into fracture systems when the strains grow to the critical value, which further leads to chipping of grain microfragments [36]. The grains of microcrystalline aluminum oxide can also be worn out as a result of plastic deformations, which result from slippage and twinning of single allotriomorphic grains [37]. It was also proved that with the limited number of slip systems, the intensity of grain wear is highly dependent on the crystals' orientation [38]. The process of plastic deformations of the polycrystalline alumina is closely related to the size of crystals and level of strain [39,40]. As the crystals and their load get smaller and smaller, the grain deformation process changes from the dominant slippage into diffuse creep and slippage along the crystal edges. However, only recently have the processes of microcrystalline sintered corundum grain wear become recognized. Existing descriptions have mainly focused upon grain crystal microchipping and their self-sharpening through the revealing of numerous sharpened corners. Most works omitted, however, the relationship between the chipping, plastic deformations, and diffusion that take place on the active abrasive grain vertexes.
Works [34,35] were the first to present the processes of grinding and Al 2 O 3 grain wear in the microcrystalline structure in a detailed and comprehensive way. They showed that the excellent cutting properties of these grains are closely connected with the tribological properties of the smoothedout cutting vertex surfaces in the area of contact with the machined material. Under the influence of strain and temperature, the vertexes of the microcrystalline sintered corundum abrasive grains are subject to plastic deformation, thus creating flat surfaces of contact with the machined material. The contact of the abrasive grain vertex with the machined material causes dispersion of a large amount of energy as well as a local temperature rise, which causes melting of its surface layer. The alloy is composed of a lowest melting temperature phase, which in the case of 100Cr6 steel results in the creation of iron oxide FeO, in the liquid phase, with high cohesiveness. This limits both the wear and the friction coefficient in the contact area dramatically.
The catastrophic chipping of the largest grain fragments is slowed down on the microcrystalline borders, thanks to which the fractures remain close to the surface and do not propagate into the grain as in the case of the polycrystalline Al 2 O 3 grains. The excellent cutting ability of the microcrystalline grains can also be explained by the stable conditions of slippage on flat grain surfaces, created behind the cutting edges.
This research presents results which confirm such a model of microcrystalline sintered corundum grains and indicate that the dominant phenomena of wear of the active microcrystalline abrasive grain vertexes end with the creation of flat surfaces as a result of abrasive wear and plastic flow. This phenomenon was common on the active surfaces of all the examined grinding wheels and did not have a negative influence on the obtained material removal rate or the quality of the machined surface. The presence of spherical chips on the surfaces of C1 and C2 grinding wheels (Fig. 13c, d, g), as well as signs of the melted machined material (Fig. 13h), is indicative of very high temperatures in the microareas of contact between the abrasive grains and the workpiece. It needs to be assumed that the chip temperature temporarily reached the ground steel melting temperature (1425-1540°C [41]), causing formation of spherical chips that were empty inside and whose shape results from the surface strain of the liquidized steel. This confirms the possibility of plasticizing the active SG grain vertexes, whose crystals are physically stable up to the temperature of 1500-1700°C [42]. It was also observed that the phenomenon of shedding the oxide layer and the plastically deformed grain layer occurs periodically on the smoothed-out surfaces of active cutting vertexes. This process is caused by fracture wear (mostly fatigue and thermal-fatigue) as a result of which the sharp crystal edges, located below the plastically deformed surface layer, are revealed (Fig. 14).
Bridges of the glass-crystalline bond, which were characterized by an approximately 30 % volume of the crystalline phase (composed of diopside and augite crystals), were worn out mostly as a result of the abrasive wear caused by their interactions with the machined surface. Analysis of the experiment's results show that for grinding wheels C1 and C2, the fracture wear share was low and was due mainly to the bond's resistance to fracture spreading, caused by the inclusion of the crystalline phase that has a significant influence on the bond's mechanical properties (K IC =1.15 MPa⋅m 1/2 ). By limiting the bond volume to 11.5 % of the grinding wheel (grinding wheel C3), a similar effect to the intensity of wear phenomena occurrence, in the abrasive grains and in the bond, was obtained, which facilitated the periodical renewal of the grinding wheel active surface cutting ability, which took place in microareas of the abrasive grains and the bond.

Conclusions
Proper selection of the proportions of particular grinding wheel components (abrasive grains, bond, and intergranular free spaces) affects the tool structure and hardness. The presented tests were aimed at determining the extent to which changing the bond volume within the range of V b =11.5-14.5 % (which corresponds to a change of the grinding wheel hardness from class K to M) influenced the wear phenomenon of the microcrystalline sintered corundum abrasive grains and ceramic bond bridges with glass-crystalline structure.
The tests were carried out in the process of traverse internal cylindrical grinding in steel 100Cr6, characterized by a relatively high material removal rate (Q w ≈24 mm 3 /s) and difficult realization conditions, resulting from the long area of contact between the grinding wheel and the machined material.
Comprehensive analysis of the test results obtained allowed for formation of the following conclusions: & The longest grinding wheel life was reached in the process of grinding with the grinding wheel possessing the lowest bind volume (C3), which resulted from the change in the share of the dominant wear phenomena taking place on the GWAS; & In the case of grinding wheels C1 (V b =13.0 %) and C2 (V b =14.5 %), the dominant phenomena were the abrasive wear and plastic flow of the active abrasive grain vertexes, accompanied by the creation of microsmearings of the grinding products in the area of contact between the grinding wheel and the machined material. The excessive force of grain binding by the bond bridges did not make it possible to renew the grinding wheel cutting ability and thus enable abrasive grain vertexes self-sharpening. As a result, the significant share of friction in the material removal process, alongside the simultaneous hindering of the provision of the coolant, caused increased temperatures in the microareas of the active abrasive grain vertexes contact, as proven by the presence of spherical chips on the GWAS; & Analysis of the wear marks on the C3 grinding wheel active surface (V b =11.5 %), combined with the other test results, made it possible to conclude that in this case, the process for cyclical revealing of new, sharp cutting edges on the surface of microcrystalline sintered corundum grains took place. The weakening of the grain binding force, resulting from limitation of the bond volume in the grinding wheel, allowed for a twofold prolongation of the grinding wheel life. This was caused by limiting the share of the abrasive wear and plastic flow phenomena, as well as a greater occurrence of the fracture wear, mostly fatigue and thermal-fatigue. In turn, this led to periodic shedding of the oxide layer and the plastically deformed grain layer, as well as to the revealing of the crystals' sharp edges, located below the plastically deformed surface layer.