Influence of the abradant’s composition on the selected physical properties in the process of front grinding of surfaces with microcrystalline sintered corundum grinding wheels

This paper presents the results of experimental research on the influence of the composition of the abradants applied in the microcrystalline sintered corundum grinding wheels at the maximum temperature obtained in the grinding zone, maximum grinding power thrust component, grinding power feed component, active grinding power, and volumetric abradant wear. Tests were conducted in the conditions of front finishing grinding of flat surfaces of the objects made of corrosion-resistant steel, with austenitic structure. In our tests, we used a mini-grinder for flat surfaces, controlled by a CNC system, shank mounted grinding wheels made of mixed electrocorundum-cubitron abradant, a piezoelectric dynamometer, a thermal camera, a digital power analyzer, and a laser gauge. The test results were subjected to statistical and substantive analyses. It was found that, together with a proportional increase of the microcrystalline sintered corundum content in the grinding wheel’s material, the values of the indicated physical properties displayed a decreasing tendency that could be determined by the trend functions, in the form of second-degree polynomials, without interaction. It was further found that such changes in the abradant composition significantly influenced the changes of the values of selected physical properties, and the test results well-approximated by the test object functions, in the form of third-degree polynomials, without interaction. The test object functions were characterized by the occurrence of local extremes whose locations indicated that it was the most beneficial to operate grinding wheels with the abradant content belonging to the range of 7CB–9CB, in the assumed testing conditions.


Introduction
In 1910, Jeppson and Saunders developed a method of producing white noble electrocorundum by alumina melting in an electric-resistant arc furnace. They obtained a new abradant whose properties were considerably different from those of ordinary electrocorundum. White noble electrocorundum is primarily characterized by a higher chemical purity and a lower quantity of glassy phase in comparison to ordinary electrocorundum. Consequently, allotriomorphic grains of abradant grains have weaker bonds between each other, less glassy phase is concentrated on the grain edges, and the spaces between the grains are more open in each abradant grain [1]. Those properties allow for temperature reduction in the grinding zone and obtaining the so-called "cold grinding" effect. For that reason, white noble electrocorundum is mostly applied in the conditions of finishing machining. However, upon the material's development, it was found that the production process was unecological and required a lot of energy. In addition, the process made it impossible to produce grinding grains with required shape and dimensional proportions ( Fig. 1) [2,3].
Microcrystalline sintered corundum (that is how it was called in international literature) was produced for the first time by the 3M Company in 1981 under the commercial name of CubitronTM. Presently, microcrystalline sintered corundum is manufactured by five companies based on slightly differing technologies. For that reason, particular material variations display different properties than those of CubitronTM and they are marked individually [4,5].
The grains of microcrystalline sintered corundum (which is also sometimes called submicrocrystalline sintered corundum) are composed of strongly sintered particles, with the dimensions that are usually smaller than 1 μm (Fig. 3) [4,8,9].
In comparison to noble electrocorundum, the abradant grains of microcrystalline sintered corundum are characterized by larger ductility and durability as a result of self-sharpening capability on the microscale, with slightly lower microhardness at the same time. Owing to those properties, grinding wheels made of microcrystalline sintered corundum have a very wide range of industrial applications, also in the area of dry grinding or grinding in the MQL conditions [9][10][11][12].
Presently, microcrystalline sintered corundum grinding wheels are usually made of mixed abradant containing white noble electrocorundum and microcrystalline sintered corundum, in various volumetric proportions. The proportions are reflected in the abradant marking system. For example, abradant 5CB means that 50 % of the whole abradant volume is filled with microcrystalline sintered corundum (e.g., CubitronTM: for the needs of this study, we identify it by the CB abbreviation), while the remaining 50 % is filled with white noble electrocorundum 99A. Such grinding wheels are usually fabricated with ceramic binder or, recently, a glass-crystalline one (Fig. 4). Currently, the market offers only a range of grinding wheels made of microcrystalline sintered corundum composed of oddnumbered products, e.g., 1CB, 3CB, 5CB, 7CB, and 9CB. Grinding wheels with the abradant made of 100 % microcrystalline sintered corundum are produced only for special applications [4,9,10].
The catalogs of particular manufacturers mostly specify the general rule of the abradant's composition selection, which can be reduced to the principle that with higher hardness of the top surface of the workpiece (TSW), larger length of the machining path by a single grinding grain, and larger grinding depth, or higher relative grinding effectiveness to be obtained, we should increase the microcrystalline sintered corundum content in the grinding wheel's abradant volume [2,9]. That rule suggests linear or monotonic dependence between the abradant content and the physical parameters [10,   [13][14][15] selected for the purpose of this study. This suggestion is not, however, confirmed by the operating examples specified by manufacturers. Their examples clearly indicate that intermediate abradant composition applied in the grinding wheel production often prove to be optimum for the selected grinding conditions owing to the criteria that are customarily assumed for the evaluation of the grinding wheel's grindability or its physical parameters. The purpose of the present study was to find the answers to the questions: will the increase of the microcrystalline sintered corundum content in the abradant volume significantly influence the changes of the physical parameters considered in this study and are such changes linear in the process of front finishing grinding of the surfaces of materials made of stain-resistant steel, with austenitic structure?

Description of experimental tests
The design of experimental tests of the front finishing surface grinding process, using microcrystalline sintered corundum grinding wheels, with various abradant composition, is presented in Fig. 5.
Output parameter description:    Experimental tests were conducted in accordance with a determined complete static plan, PS/DK, treating the front finishing grinding process as five separate elementary objects. The same grinding effectiveness was applied in all the test objects. The test results for each test object took into account five values measured in five grinding cycles (five replications).
The following dependences were the test object functions: θ max =f(XX), F p =f(XX), F f =f(XX), P c =f(XX), and V m = f(XX), where XX was the changing abradant composition. First, the least square method was applied to determine the trend curves for each of the samples, taking into account the functions in the form of second-degree polynomial, without interaction. Next, approximation of test results was conducted by the computer equalization method, with the functions in the form of third-degree polynomials, without interaction, and the adequacy of those functions was checked at the significance level of α=0.05. In case of both trend curves and approximating functions, the values of determination coefficients R 2 were calculated. The maximum temperature in the grinding zone was measured with the use of FLIR SC620 thermal camera, with the accuracy of +/− (2 % v2), operating in a set with a microcomputer equipped with ThermaCAM Researcher Pro 2.10 software. However, the measurements of active grinding power were made with the use of an NDN DW6090 digital power analyser (Fig. 6). The grinding force component measurements were made with the use of a KISTLER 9257B piezoelectric dynamometer (Fig. 7). The dynamometer was also connected to a PC, which recorded digital signals and carried out proper data processing. The measurements of the axial wear of the grinding wheel were taken with the use of a KEYENCE laser gauge (an LJ-G030 sensor and an LJ-G5001(P) controller), followed by conversion in respect of volumetric wear (Fig. 8). Regardless of the measurements specified in the input test parameter design, each time microscopic observation of the active grinding wheel surface (AGWS) was carried out and the topography of the top machined surface was checked using profilographometer TALYSURF FORM 50 from Taylor Hobson (Fig. 9). Those additional observations were carried out owing to the need of substantive evaluation of the causes of identified phenomena.
A selected thermal image of the grinding zone area, with a temperature record graph, is presented in Figs. 10 and 11. A selected printout from the power and grinding moment

Test results and statistical analysis
The experimental test results relating to the influence of the abradant composition on the selected physical parameters in the process of front finishing grinding of flat surfaces of corrosion-resistant steel, using grinding wheels made of microcrystalline sintered corundum, with various abradant compositions, are presented in Table 1.
The results were subjected to statistical analysis. In the first stage, the analyst tried to establish whether the pro-portions of two variances (in the systems of "u," 3CB to 1CB, 5CB to 1CB, 7CB to 1CB, and 9CB to 1CB) for the same input value of z i would be higher than one could expect when the samples had been collected from the same population.
Zero hypotheses were set as to variance equality. To check those hypotheses, the expressions of S 2 (z i/u ) were used as estimations of the variance of variables z i/u (based on samples), with Fisher-Snedecor statistical distribution. The results of variance analysis by test F, on the example of variable z 1/u , is presented in Table 2.
The comparison of the calculated values with the table values of F does not allow to refute zero hypotheses, with variance equality at the significance level of α=0.05. An analogous situation occurs in the cases of the remaining variables z i . In the second stage of the statistical analysis of test results, the analyst checked whether there was a significant difference between the two average values z i (from both test result groups, in each "u" system), for the same input value z i . Zero hypotheses were set as to equality of respective average values in particular "u" systems. Those hypotheses were verified by the Student's t test, proper for checking the difference between the two average values. The results of the analysis with the test t of the differences between the average values, on the example of variable z 1/u , are presented in Table 3.
The comparison of calculated values of t with the table values of t allows us to state that there are significant differences between the respective average values in the particular systems "u," and the zero hypotheses are not true at the significance level of α=0.05. An analogous situation occurs in the case of the remaining variables z i . In the third stage of the statistical analysis of Fig. 12 Selected course of grinding force components; 5216 30×30× 6 5CB60L7V grinding wheel Fig. 11 Selected graph of the temperature record made during grinding with a 5216 30×30×6 5CB60L7V grinding wheel (blue line: θ max , red line: θ s -maximum temperature in the grinding wheel) experimental test results related to the influence of the abradant content on the selected physical parameters, trend curves were designed, in the form of seconddegree polynomials, for particular test objects, with the calculation of the determination coefficients R 2 , followed by test result approximations by the functions in the form of third-degree polynomials, without interaction, plus determination of their adequacy at the significance coefficient of α=0.05. Since for each test object F TOF ≤ F 0.05; 2; 6 = 5.14, (where TOF = Test Object Function), there are no grounds for refutation of the hypotheses that the determined TOF values are adequate (Figs. 13, 14,  15, 16, and 17). Standard deviations of the i-th physical values are included in Table 1.

Merits-related tests results analysis
Even a preliminary analysis of the obtained test object functions indicates the existence of local extremes. The increase of the values of the selected physical parameters during front finishing grinding of surfaces with microcrystalline sintered corundum grinding wheels of the 1CB-3CB range is noticeable, just as the parameter minimization when machining with the grinding wheels of the 7CB-9CB range.
The TSW analysis and AGWS microscopic observations indicated that the cause of the occurrence of the maximum when grinding with a 3CB grinding wheel consisted in the appearance of the silting (gluing) process on the active grinding wheel's surface by the microchips which settled in the inter-blade spaces (Fig. 18). After those microchips were cut off from the active grinding wheel's surface, they were stuck to the workpiece's surface (Fig. 19), thus increasing the friction forces in the grinding process, and consequently causing increase of the selected physical values. The clogging of the active grinding wheel's surface results from the dominating grinding grain wear mechanism in case of 3CB abradant composition (Figs. 20 and 21). The increased friction between the grinding grains and the workpiece causes increase of the machining force value and of the grinding wheel's wear, as well as the increase of machining temperature and power (Figs. 13, 14, 15, 16, and  17). However, in case of 7CB abradant composition, the decrease of the physical values under analysis was caused by the intensification of the contribution of the grinding grain self-sharpening mechanisms to the process of grinding wheel's active surface wear (Figs. 22 and 23). The change of the dominating grain wear mechanism component resulted not only from the increase of the proportion of the Cubitron grinding grains with the self-sharpening capability in the abradant composition, but probably also from the changing minimum micromachining depth of single grinding grains, when micromachining after the traces of preceding grains (Figs. 20 and 22; Tables 4 and 5).

Conclusions
Based on the statistically and substantively analyzed test results, it was found that, in the conditions of front finishing grinding of flat surfaces of the objects made of austenitic corrosion-resistant steel, using microcrystalline sintered corundum grinding wheels, the following regularities were found: & The composition of the abradants of the grinding wheels used in tests indicated their significant influence (at the significance coefficient of α=0.05) on the maximum temperature in the grinding zone, thrust power component, grinding feed power, active grinding power, and volumetric grinding wheel wear. & Synergic influence of the mixed 3CB abradant on the machined surface caused the occurrence of the wear phenomenon on the active grinding wheel's surface, in the form of gluing (silting), in the assumed machining conditions, which resulted from the specific properties of the ground material (in this case of low thermal conductivity of austenite, its resistance to microchip corrosion, and the adhesion capability). & In the assumed and similar grinding conditions, it is recommended to apply the grinding wheels with the 7CB-9CB abradant content.
Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.