Experimental study of the relationship between the vibro-acoustic parameters of the grinding process and the macro-roughness of the treated surface

OBRABOTKAMETALLOV Vol. 23 No. 3 2021 technology Further research was focused on studying the macro-roughness of the ground samples. The criterion for evaluating the macro-roughness was the value of the deviation from the cylindricality of the sample surface. GOST 24642–81 defines cylindricality as the maximum distance from the real surface points to the adjacent cylinder within the rated area, i.e., the cylindricality value of the measured samples is the difference between the maximum and minimum radii in two sections. Since cylindricality includes such parameters as roundness, straightness, and parallelism [GOST R ISO 230–1–2010] and can also indirectly characterize the size accuracy, this criterion should be recognized as the most convenient for a comprehensive assessment of macro-roughnesses. The analysis of data on the coordinates of the points of the real surface of the ground samples collected with the help of CMM made it possible to form polar lobe diagrams (round charts) (Fig. 5), which give a visual and graphical representation of the macro-dimensions of the processed workpieces. The analysis of the round charts showed that the irregularities of the sample surfaces, expressed in terms of the differences in the coordinates of neighboring points, have a non-constant character. It is known that the real profile of a surface obtained by mechanical processing is formed from the following components: deviations in shape, waviness, and roughness [19-21], each of which has a unique nature and characteristic parameters. The round charts have all the profile formation components. Using the round chart data, we determined the values of the samples’ cylindricality by the simplest calculations (Table 2). Based on the assumption that the shape of the samples was initially an ideal geometric cylinder, it was possible to construct a chart of the dependency of the cylindricality on the infeed rate over time (Fig. 6). We can see from the chart that the deviations from cylindricality increase over time and also have a direct dependence on the infeed rate. There is also a slight deviation from the similar nature of the deviations from cylindricality increasing over time at a feed rate of 0.8 mm/min. Such a deviation may be caused by the presence of the self-sharpening mode of the grinding wheel, which later passes into the stage of predominant blunting. For better visualization, Figure 7 shows the distribution curves of the radius values measured by the coordinate measuring machine – the so-called size distribution polygons for the workpieces ground at a feed rate of 0.8 mm/min. The sharpest peaks of the distribution curves are characteristic of the steady-state grinding mode (the second minute). In this mode, the spread of values is minimal and the deviations from the ideal geometric shape are small. The initial running-in stage is characterized by a flatter (hill-like) shape of the curve, which indicates a wider spread of values and less constancy of the size. After five minutes of processing, the distribution of the values forms two distinct hills, showing that the sample values drive towards two dominant values, which once again confirms the presence of a prominent deviation from cylindricality. According to [19], if a periodic component is not found when studying the profile of the workpiece, the empirical distribution law should be close to normal since there is no reason to believe that any technological factor has a dominant effect on the surface profile, changing its distribution from normal. This is exactly what we observe in Figure 7. The distributions of the values after the first and the second minute of grinding are Gaussian. The curve drawn according to the data collected after five minutes of grinding has two vertices. This may indicate deviations of the actual distribution from normal due to the presence of a systematic component in the profile; however, when we consider the round chart, it is obvious that the reason for this distribution is the presence of the cone shape of the ground surface. According to [19], the double-peaked shape is more characteristic during processing with a powerful systematic basis: turning, milling, rolling, etc. This phenomenon is characteristic of external circular grinding operations to a much lesser extent. Using the method of correlation and regression analysis again, we formed a mathematical model of the dependence of the deviation from cylindricality (Δ, mm) on the infeed rate (S R , mm/min) and the grinding wheel running time (t, min): ∆ = + ⋅ + ⋅ 0.006 0.012 0.046 . R S t (3)