Testing the Effectiveness of Hybrid Milling and Surface Burnishing in Improving the Wear Resistance of Machine Parts Made of Structural Steel

Abstract

Due to the need to form a surface layer with specific operating properties, recent years have seen an increased interest in surface strengthening treatment, which aims to create a surface layer that improves the durability of parts. With a view to the economics of the machining process, it is common to combine shaping milling, characterised by high volumetric efficiency, with finishing burnishing, during which significant forces are applied. In the literature, one of the important limitations of such technological operations is the value of residual stresses, excessive values of which can lead to the flaking and falling off of surface fragments. In the present study, the authors put forward the research hypothesis that, in addition to stresses, the geometry of the machining roughness is also important and may contribute to faster tribological wear than stresses. It has been shown that what is important in hybrid machining is not so much the height of the resulting irregularities and the effectiveness of their levelling by burnishing, but the geometry of the irregularities. After milling, surfaces with small, regular irregularities with smooth peaks and shallow valleys were found to be the best in tribological tests. Such roughness can be plastically levelled out during burnishing. On the basis of the experimental studies carried out, it was shown that a higher burnishing force does not always lead to higher wear resistance.

1. Introduction

Proper shaping of the surface (by strengthening it) can extend the service life of machine parts and contribute significantly to their durability while reducing the amount of lubricants used. As a result of the need to form a surface layer with specific performance properties (see, for example, [1,2,3,4,5]), there has been an increased interest in recent years in surface strengthening treatment, which aims to produce a surface layer that improves the strength of a machine part. At the same time, in [6,7], it was shown which parameters of the geometrical texture of the surface are important in determining the properties of the surface layer and how their values change during wear. It should also be noted that wear is related to the surface microstructure of the machine part [8].

Tribological wear is generally an indirect cause of machine failure [9]. It can have various backgrounds and causes. In the operation of mechanical equipment, tribological wear processes include abrasive wear, adhesive wear, wear by oxidation, fatigue wear, chemical and electrochemical (corrosion) wear, cavitation wear, thermal wear and fretting [6]. The scope of contemporary scientific and technical research is most often concerned with determining the relationships associated with abrasive and fatigue wear processes and their impact on the loss of product functionality [10,11,12].

It can be assumed that surface wear is a type of wear in which local cohesion loss and associated material loss are caused by fatigue as a result of cyclic contact stresses in the surface layers of the parts to be joined. As a result of superimposed deformations in the surface layer, fatigue-induced surface microcracks form, subsequently progressing to macrocracks. The final result is the detachment of pieces of material from the core. Once the limit number of load cycles is exceeded and the fatigue limit is reached by the individual microvolumes of material in the surface layer, mass loss occurs.

Thus, strengthening the surface layer of mating parts (by introducing a favourable stress state) should contribute significantly to the mechanical resistance of parts to fatigue wear. One of the more popular strengthening treatments is burnishing, which is combined with machining operations such as milling [13,14,15], turning [16,17,18,19] or other methods of volumetric and ultrasonic machining [20,21,22,23].

The burnishing operation during the realisation of hybrid technological processes on CNC machine tools leads to an effective reduction in the height of roughness (through high efficiency and low cost). The degree of change in roughness after finishing burnishing in relation to the initial roughness after a machining operation (turning or milling) can be determined from the following relationship:

$$K={\displaystyle \frac{Roughness of the machined surface }{Roughness of the burnished surface}}$$

(1)

where K is the technology effectiveness factor.

Depending on the way in which the burnishing is carried out and the type of material to be processed, calculated according to Equation (1), the technology effectiveness factor can take on a value of more than 5 [24,25] or even 10 [26,27].

The idea of effectiveness in levelling surface irregularities is confirmed by numerous published scientific results [28,29,30,31]. One of the more frequently emphasised limitations of burnishing is the so-called residual stresses [32,33,34]. Their maximum value should not be exceeded, as the strength of the material may be exceeded during the service life of the part (after the introduction of service stresses). As a result of exceeding the maximum strength of the material, there may be a local loss of material cohesion between the surface and the core in the form of microcracks (which often occur below the surface). The development of microcracks over an extended period of time results in the flaking and falling off of larger parts of the surface, contributing to accelerated wear of the parts.

Various studies have been carried out to analyse the burnishing process using experimental, numerical and analytical approaches. Rodríguez et al. [35] carried out an experimental and numerical approach to investigate the surface properties, hardness and residual stresses after the ball burnishing process. They found that the ball burnishing process significantly improves surface properties and residual stress values. El-Taweel and El-Axir [36] conducted an experimental study based on a response surface methodology to analyse the effects of burnishing speed, burnishing feed rate, number of passes and burnishing force on the surface roughness and hardness of turned parts. Analysis of variance showed that the burnishing force and feed rate had the greatest effect on surface roughness and hardness. Gharbi et al. [37] carried out an experimental study using the Taguchi experimental design to determine the effect of burnishing parameters on the surface properties of AISI 1010 flat surfaces using a multiball tool. The results obtained from the analysis of variance showed that the burnishing force has a significant effect on both hardness and surface roughness. Korzynski et al. [38] developed a hydrostatic pressure-loaded burnishing tool for machining chrome-coated surfaces. They found that coating operations induce tensile residual stresses on the specimens and reduce fatigue life by causing microcracks on the surface. However, after ball polishing, surface quality and fatigue life are improved by removing surface irregularities and reducing cracks. In addition, the residual tensile stress on the surface of the coated specimens is changed to compressive stress as a result of slip ball polishing. Avilés et al. [39] investigated the effect of a low plasticity ball-burning process on improving the fatigue strength of normalised AISI 1045 steel. Both untreated and blackened ball-burnished specimens were subjected to rotary bending fatigue tests up to 3.25 × 10⁶ cycles. It was shown that the fatigue strength of the ball-burnished specimens, compared to the untreated specimens, improves by approximately 21.25%. Travieso-Rodriguez et al. [40,41] also integrated longitudinal vibration with a burnishing tool to improve the surface characteristics of G10380 steel. They analysed the surface roughness, hardness and residual stresses of the burnished parts. The results showed that the application of vibration to the burnishing tool significantly improves the roughness of the burnished surface, but less robust results were observed in terms of specimen hardness and residual stresses. Revankar et al. [42,43] applied ball burnishing to the machining of titanium round bars. They investigated the surface roughness, hardness, wear resistance and residual stress of the specimens subjected to burnishing. They found that the use of ball burnishing significantly improves the respective properties compared to turned specimens.

It can be seen from the above observations that most of the research was carried out based on experimental studies. However, in order to better understand the mechanism of ball-crimping and extend its application, theoretical studies are very useful. In this case, few studies have been carried out to analyse the burnishing forces or the surface generation mechanism. Luo et al. [44] carried out an analytical study to predict the burnishing force. They used the elastic Hertz contact model to determine the coring depth threshold for elastic deformation. In turn, they obtained the values of the pressing depth for elastic-plastic and plastic deformations. The results obtained from their analytical model were in good agreement with the experimental results. Hiegemann et al. [45,46] developed a theoretical model to predict the surface roughness and pressure of the coated surface. They also used the Hertz contact theory with appropriate simplifications to include the effect of burnishing depth in the model. The results showed that the experimental results were consistent with the analytical approach. Korzynski [47] developed an analytical model based on micromechanical slip line theory to correlate the relationship between surface roughness and burnishing force. The model included the mechanical parameter of the material, the burnishing force, the number of surface roughness over a given length, the angle of surface roughness and the geometrical parameters of the burnishing process, such as the ball diameter. With the help of this model, the surface roughness of the specimen can be predicted after a one-time burnishing process.

Complementing the research presented above is our research described in this paper. It describes tribological tests carried out on surfaces that were characterised by a similar height of roughness. However, the surface irregularities were created under different technological conditions. The first case was the combination of fine (i.e., labour-intensive) milling with light surface burnishing, the result of which is shown in Figure 1a (as an example of an economically unsustainable operation). The second case was the combination of shaping milling with heavy surface burnishing, the result of which is shown in Figure 1b (as an example of an economically sustainable operation).

During the experiment that is the subject of this article, milling and burnishing operations of specimens were carried out on a single CNC machine tool (see Figure A1 in Appendix A), achieving a similar state of the geometrical texture of their surfaces under different process parameters. The aim of this experiment was to investigate how changes in the geometry of surface irregularities after milling and burnishing conditions contribute to changes in the tribological properties of product surfaces. The conclusions were based on an examination of the surface wear process.

In the literature related to the study of the state of the surface geometrical texture, one can find many works related to the modelling of the height of surface irregularities after the application of typical machining processes. The area of interest is usually models determining the values of indices such as Ra or Rz, which are most often used in technological documentation and technical drawings [48,49]. There is also a lot of work on modelling surface properties related to wettability [50,51,52], the ability of the surface to hold paint coatings [53,54,55] and ensuring adequate strength of adhesive joints [4,56,57].

The most commonly described models take into account the technological parameters of the cutting process for materials such as steel, cast iron or plastics. As mentioned in the Introduction, there is a lot of work on modelling the surface roughness after burnishing. In contrast, relatively little attention has been paid to the study of the functionality of surfaces obtained by hybrid manufacturing. This is generally due to the considerable difficulty of such studies, the degree of compilation of the process model itself and nevertheless the high demand for computing power. Therefore, this area is dominated by experimental work in which the influence of the technological parameters of the two treatments on the basic indices of roughness height, such as Ra (or Sa), Rz (or Sz) and the degree of isotropy, was most often determined [29,58].

2. Materials and Methods

The actual interaction of the friction surfaces of two solid bodies sliding against each other depends on the contact surfaces and the physical properties of the materials of the parts rubbing against each other. The surfaces of all solids, especially machine parts, have a certain state of roughness. The surface geometry of machine parts is shaped by the machining process. The machined surface has irregularities, resulting from errors in shape, position, waviness and roughness, which characterise the surface geometry obtained during machining. The specimens to be tested were prepared in such a way as to take into account all the effects that usually accompany production under industrial conditions.

2.1. Preparation of Specimens for Testing

Six specimens of C45 steel with a hardness of 189HV30 ± 6 and dimensions of 20 mm × 20 mm × 20 mm were used for the tests. A Mitsubishi SRFT10 finger milling cutter (Mitsubishi Materials Corp, Chiyoda-Ku, Tokyo, Japan) with a 5 mm radius VP15TF round insert was used for shaping milling. Roughing of the specimens and shaping milling were carried out on a MIKRON VCE 500 machining centre (Haas Automation, Inc., Oxnard, CA, USA) at a cutting speed of v_c = 80 m/min and a depth of cut of a_p = 0.3 mm (see Figure 2). After machining, the machine was rearmed for burnishing, which was carried out using a hydrostatic head powered by a hydraulic unit [59] with a 10 mm diameter spherical working tip made of Al₂O₃ ceramic material.

An emulsion from Statoil (Statoil, Stavanger, Norway), commercially named ToolWay, with a concentration of 12%, was used as the machining fluid and the hydraulic fluid to feed the hydrostatic head. The burnishing force was set according to the values adopted in the experimental plan shown in

Table 1. Technological parameters of milling and burnishing operations with the results of profile measurements, where the expanded uncertainty of parameter measurement was URa = 0.35 μm, URz = 0.94 μm and URmax = 0.59 μm.

Specimen	fm (mm)	Fb (N)	Ra 1 (μm)	Rz 1 (μm)	Rmax 1 (μm)
D10	Reference milling		5.56	21.6	22.2
D8	0.4	600	0.94	4.44	4.87
D6	0.4	200	0.88	4.00	4.46
G10	Reference milling		6.29	25.3	27.2
G8	0.7	600	2.94	10.6	11.1
G6	0.7	200	2.64	9.80	10.6

¹ Average value of the parameter.

.

In the experimental study, three specimens were prepared on which one measurement each was taken. A normal distribution was adopted for the obtained Rx values, for which the uncertainty expansion factor k = 2 (covering 95% of cases around the mean value). The value of the standard deviation S from the specimens (with 3 measurements) was determined from the spread of the received measurement values URx = 2S. Table 1 also summarises the results of profile measurements taken directly on the machine tool according to EN ISO 21920 [60,61,62].

The value of the burnishing force during machining was monitored using a piezoelectric force gauge and a Kistler 5019 multichannel charge amplifier (Kistler Instrumente AG, Winterthur, Switzerland) so that its variation did not exceed ±5% of the nominal value. During the burnishing process, in addition to the burnishing force F_b, the value of the transverse feed rate for milling f_m was changed in steps. The experimental plan was carried out at a constant burnishing speed of 8 m/min. When switching the machine between milling and burnishing operations, reference profile roughness measurements were taken over a length of l_t = 4.8 mm. The measurements were taken using a HOMMEL ETAMIC T1000 portable profilometer (HOMMEL ETAMIC, Villingen-Schwenningen, Germany) equipped with a TKU 300/600 inductive sensor. Three measurements were taken each time, while Table 1 shows the average values from the results obtained.

Measurements of the values of the profile roughness parameters were taken immediately after the finishing burnishing operation. Preliminary surface tests showed a roughness reduction of approximately 20% to as much as 50% of the roughness height values obtained after milling.

Surface hardness measurements were performed with a Vickers indenter (LV700AT, Leco, St. Joseph, MI, USA) at a load of 1 kg.

Precise measurements of the spatial state of the geometrical texture after milling and burnishing, together with an analysis of the tribological test traces, were developed in the next part of this study according to the experimental plan.

2.2. Shaping the Surface Geometrical Texture

To measure the spatial state of the surface geometrical texture, an Altisurf A520 multi-sensor measurement system (Altimet, Thonon-les-Bains, France) was used, which was equipped with a MPLS (Multi Point Line Sensor) confocal white light sensor with an operating range of up to 400 μm and a resolution in the optical axis of the instrument of 8 nm. During one measurement pass, 180 profiles were recorded on a selected section of the surface with a width of 1.6 mm.

A representative section of the surface bounded by dimensions of 4.0 mm × 4.0 mm was taken in this study. Digital processing of the recorded point cloud and determination of selected values of the surface geometrical texture indices were carried out using the MCubeMAP 8.1 software (Mitutoyo, Kanagawa, Japan).

The following set of indices was chosen to characterise the changes occurring on the surface after burnishing, and in subsequent tests of the resistance of burnished surfaces to tribological wear [63,64,65,66]:

• Surface roughness height indices such as Sa (arithmetic mean surface height), Sz (maximum surface height), Sp (maximum surface peak height) and Sv (maximum surface valley depth);
• Indices based on changes in the Abbott–Fireston curve such as Vmp (surface material peak volume), Vmc (surface material core volume), Vvc (surface void core volume) and Vvv (surface void valley volume).

2.3. Surface Wear Tests

Four independent surface wear tests were performed on each specimen (four independent wear tracks—see Figure 3a). A TRN S/N 18–324 Pin-on-Disk Tribometer (CSM Instruments, Peseux, Switzerland) was used to test the degree of surface wear in accordance with ASTM G99 [67,68]. Between successive series of tribological tests, the ball was changed and the surfaces of the specimens were cleaned in acetone using an ultrasonic cleaner. The tests were conducted at an air temperature of 22–24 °C and a relative humidity of 48–50%. A 6 mm diameter alumina ball loaded with a force of 1 N was used as a counter probe. A reciprocating motion with a stroke of 8 mm was used. The maximum linear velocity during testing was 30 mm/s (corresponding to a frequency of 3.75 Hz). The specimens were positioned during the test so that the direction of the wear tracks formed was parallel to the surface texture directionality left by the burnishing tool. The degree of surface wear was determined by the friction path realised, which was, respectively, 200, 400, 600 and 800 m. During all tests, the contact between the surface of the specimen and the ball was not interrupted. Only after all tests had been completed in accordance with the adopted experimental plan were the degree of wear and microhardness evaluated. The degree of wear was also determined in relation to the reference surface not subjected to the tribological test.

The assessment of surface wear (see Figure 3b) was carried out on the basis of the developed methodology for investigating changes in the geometric texture of the surface with the previously developed methodology—described in Section 2.2. The test section covered an area of 10 mm × 10 mm, from which four separate sections of 2 mm × 4 mm were then extracted with a single wear track.

The surface layers were examined using an FE-SEM SU-70 (Field Emission Scanning Electron Microscopy) microscope (Hitachi, Naka, Japan). The specimens for microscopic examination, after mechanical cutting, were mounted in conductive resin (Polyfast, Struers, Ballerup, Denmark) and mechanically ground and polished using a diamond and alumina suspension. The microstructures were chemically revealed using an etchant containing 1 cm³ of nitric acid and 100 cm³ of ethanol.

3. Results and Discussion

3.1. Measurements of Surface Parameters

Examples of surfaces after the complete hybrid machining operation are shown in Figure 4 and Figure 5.

The average results of the selected geometric texture parameters including the crack (on the corresponding friction path) are shown in

Table 2. Mean values of the surface geometrical texture measurements obtained during the experimental tests, where the expanded uncertainty of the parameter measurement was USa = 0.18 μm, USz = 3.49 μm, USp = 3.01 μm, USv = 2.52 μm, UVmp = 0.04 μm, UVmc = 0.84 μm, UVvc = 0.73 μm and UVvv = 0.09 μm.

Friction Path (m)	Specimen	$\overline{\mathit{S}\mathit{a}}$ (μm)	$\overline{\mathit{S}\mathit{z}}$ (μm)	$\overline{\mathit{S}\mathit{p}}$ (μm)	$\overline{\mathit{S}\mathit{v}}$ (μm)	$\overline{\mathit{V}\mathit{m}\mathit{p}}$ (mL/mm2)	$\overline{\mathit{V}\mathit{m}\mathit{c}}$ (mL/mm2)	$\overline{\mathit{V}\mathit{v}\mathit{c}}$ (mL/mm2)	$\overline{\mathit{V}\mathit{v}\mathit{v}}$ (mL/mm2)
0	D8	1.14	17.30	9.86	7.43	0.088	1.736	2.212	0.239
	D6	1.22	16.96	8.38	8.58	0.085	1.947	2.398	0.297
	G8	1.29	16.73	8.09	8.63	0.087	2.047	2.571	0.304
	G6	1.24	16.39	9.08	7.32	0.084	2.042	2.544	0.315
200	D8	1.30	19.60	10.2	9.37	0.089	2.008	2.510	0.297
	D6	1.34	16.69	7.93	8.76	0.085	2.045	2.490	0.297
	G8	1.37	16.79	8.77	8.02	0.092	2.024	2.601	0.279
	G6	1.32	18.69	8.87	9.82	0.092	2.069	2.573	0.315
400	D8	1.35	20.61	10.7	9.93	0.085	2.256	2.705	0.350
	D6	1.37	18.06	9.47	8.60	0.089	2.405	2.885	0.352
	G8	1.45	17.65	9.59	8.06	0.096	2.204	2.838	0.308
	G6	1.36	18.98	8.79	10.2	0.094	2.116	2.669	0.305
600	D8	1.50	27.34	14.9	12.4	0.094	3.288	3.481	0.459
	D6	1.42	21.61	10.9	10.7	0.090	3.030	3.355	0.389
	G8	1.47	19.97	11.4	8.60	0.099	2.504	3.140	0.330
	G6	1.37	16.53	8.30	8.23	0.095	2.564	3.158	0.345
800	D8	1.54	31.11	15.9	15.2	0.092	3.742	3.827	0.433
	D6	1.55	23.63	13.5	10.2	0.126	3.891	4.446	0.398
	G8	1.50	22.39	11.3	11.1	0.110	3.459	4.037	0.365
	G6	1.50	22.52	12.8	9.74	0.109	4.176	4.417	0.450

.

Hardness measurements showed a significant increase in surface hardness after burnishing compared to the reference surface hardness after the milling process—see

Table 3. Results of Vickers surface hardness measurements (HV1).

Specimen	HV1	S
D10 Reference milling	189	6
D8	295	15
D6	314	14
G10 Reference milling	189	6
G8	332	11
G6	346	12

.

The reference surfaces in all cases were characterised by similar indicator values in both groups. The differences in the recorded parameter values in this case were within the natural range of variation of ±5% of the mean value of the selected index (Figure 6).

3.2. Studies of Surface Changes After Milling and Burnishing

The wear tracks formed on the surfaces during the tribological tests resulted in a gradual increase in the height values of the surface geometrical texture indices as the friction path increased (Figure 7). It should be noted that the surfaces of specimens D8 and D6 burnished at a feed rate of f_m = 0.4 mm were not as plastically strengthened as those of specimens G8 and G6 burnished at a feed rate of f_m = 0.7 mm. When producing the surface of type D specimens, there was a sixfold reduction in the height of surface irregularities compared to a twofold reduction in the height of irregularities for type G specimens.

The easier and faster formation of wear tracks on the surface of specimens with a higher compression is to be explained by the separation of the harder, more plastically processed high roughness peaks (as illustrated by the greater changes in the values of the Sp indices in relation to the variations in the values of the Sv indices). The introduction of high stresses in the surface layer results in an increase in its hardness at the expense of a loss of its elastic properties. The particles falling off of the hardened surface during tribological tests acted as an additional abrasive contributing to faster surface erosion.

The changes in functional indices related to peak volume, core volume and surface valley volume during tribological testing are proportional to the realised friction path (Figure 8). This means that, in a similar surface image, the emerging wear track starts to dominate the tested surface section. Hence, the focus of the remainder of this study is on investigating the geometry of the wear track. Its width and depth, depending on the technological machining parameters, can be a better index of the resistance of the surface to abrasive wear.

3.3. Determination of Surface Wear

A cross-section of the surface that passed through all the wear tracks on a given specimen was selected for wear analysis (Figure 9a). Scans of the 11 mm × 8 mm surface were taken according to the previously presented methodology for measuring the surface geometrical texture. Using an automatic procedure in the MCubeMAP 8.1 software, the scratches were extracted. Their width, depth and surface area were then determined (Figure 9b).

The average values of the dimensions characterising the geometry of the wear tracks are shown in

Table 4. The average values of the wear track dimensions obtained during the test and the values of the determined degree of wear.

Friction Path (m)	Specimen	W (mm)	D (μm)	A (μm2)	Norm(W)	Norm(D)	Norm(A)	Sum(WDA)	DOW
200	D8	0.295	2.843	532	0.234	0.235	0.093	0.562	0.161
	D6	0.326	1.484	439	0.328	0.055	0.063	0.446	0.123
	G8	0.305	1.070	243	0.265	0.000	0.001	0.266	0.063
	G6	0.217	1.660	241	0.000	0.078	0.000	0.078	0.000
400	D8	0.357	2.998	850	0.421	0.256	0.195	0.872	0.264
	D6	0.388	2.400	936	0.515	0.176	0.222	0.913	0.278
	G8	0.322	2.570	704	0.317	0.199	0.148	0.664	0.195
	G6	0.233	1.480	287	0.048	0.054	0.015	0.117	0.013
600	D8	0.549	3.793	1726	1.000	0.361	0.475	1.836	0.586
	D6	0.440	2.992	1328	0.673	0.255	0.347	1.275	0.399
	G8	0.397	3.620	1185	0.543	0.338	0.302	1.183	0.368
	G6	0.300	3.850	1109	0.250	0.369	0.277	0.896	0.273
800	D8	0.533	5.164	2384	0.953	0.543	0.685	2.181	0.701
	D6	0.502	5.435	2614	0.860	0.579	0.758	2.197	0.706
	G8	0.431	5.850	2218	0.645	0.634	0.632	1.911	0.611
	G6	0.316	8.610	3370	0.299	1.000	1.000	2.299	0.740

. The observations initially made on the topographic images and the conclusions drawn from the analysis of changes in selected indices of the surface geometrical texture were confirmed when analysing the geometry and surface area of the wear tracks formed in the tribological test (Figure 9).

In order that the degree of wear (DOW) could be determined, the values for wear track width, depth and area were normalised according to the following relationship:

$${Norm}_{\left(x\right)}={\displaystyle \frac{x_i-x_{min}}{x_{max}-x_{min}}}$$

(2)

where x stands for the values of width W, depth D and area A, respectively.

The sum of the normalised values was determined as follows:

$${Sum}_{\left(WDA\right)}={Norm}_{\left(W\right)}+{Norm}_{\left(D\right)}+{Norm}_{\left(A\right)}$$

(3)

A situation in which a wear track with the maximum values of width, depth and area would have been observed during the experimental tests would have given a value of ${Sum}_{\left(WDA\right)}=3$. Such a situation would have to be considered the maximum degree of wear. However, such a situation did not occur during the experimental tests. Therefore, the following relationship was used to determine the actual degree of wear:

$$DOW={\displaystyle \frac{{Sum}_{\left(WDA\right)i}-{Sum}_{\left(WDA\right)min}}{3}}$$

(4)

The resulting DOW values range from 0 to 1. They are summarised in Table 4, while Figure 10 shows the changes in DOW in graphical form.

On the basis of the analysis carried out of the determined degree of wear, initial observations of changes in the state of the geometrical texture of the surface can be confirmed. The wear tracks created as a result of the tribological test in the initial phase change this state only slightly. It is only when the friction path of 400 m is exceeded on the surface of specimens machined at f_m = 0.4 mm and when the friction path of 600 m is exceeded on the surface of specimens machined at f_m = 0.7 mm that the actual degree of surface wear exceeding the value of 0.25 is observed.

At both milling feeds f_m = 0.4 mm and f_m = 0.7 mm and in the full range of the applied burnishing force F_b, a significant degree of wear occurred after a friction path of 800 m was exceeded. The geometry of the wear tracks and their surface area led to the determination of a wear degree of 0.75. In this case, the deep crack was the dominant element in the observed spatial geometrical state of the surface. In most cases, the surfaces of the specimens burnished with F_b = 200 N wore more slowly compared to those burnished with F_b = 600 N. These surfaces also did not have such strongly rolled sharp peaks of unevenness. In the course of the experimental tests, no technological parameter association was obtained for which an actual wear rate above 0.9 would have been observed during tribological tests.

The most favourable properties were obtained for specimens with a surface milled at a feed rate of f_m = 0.7 mm and burnished with a force of F_b = 200 N. This was a combination of high-performance burnishing with low force values. This is a treatment suitable, for example, for the spindle bearings of modern CNC machine tools. The prerequisite for obtaining a good surface finish is the combination of shaping milling conditions that produce regular irregularities with relatively smooth peaks and shallow valleys. Such irregularities during shaping can be reduced in the range of 50% of the initial Sa or Sz values (obtained after milling).

Experimental studies have shown that a higher burnishing force (inducing higher burnishing stresses) does not always lead to a higher wear resistance of the surface. Rather, high peaks of unevenness form in the surface, which break away and cause accelerated erosion of the surface. This mechanism is currently being investigated and it is planned to publish information on this subject in a separate article. In confirmation of this mechanism, an SEM image of specimen D8, which was one of the less resistant during the experimental tests, is included in Figure 11b. Such observations were not made for specimens showing high wear resistance (see Figure 11a).

4. Conclusions

On the basis of the conducted literature review and the realised experimental research, the following cognitive and utilitarian conclusions can be drawn for practical applications:

• In addition to the value of the burnishing force, the state of the surface geometrical texture after milling has a significant influence on the end result of hybrid manufacturing;
• In hybrid machining operations, it is not so much the height of the resulting irregularities and the effectiveness of their levelling by burnishing that is important, but the geometry of the irregularities. After milling, surfaces with small, regular irregularities with smooth peaks and shallow valleys were found to be the best in tribological tests. Such roughness can be plastically levelled out during burnishing. Rolled peaks that do not connect with the surface core should be avoided;
• The realisation of high-performance hybrid manufacturing (by milling and burnishing), in which irregularities of considerable heights are burnished at significant force values (often above 1 kN), can lead to good smoothness and reflectivity on the surface (i.e., low roughness). However, locally high surface consolidation at shallow depths does not necessarily lead to an increase in the resistance of the surface to abrasive wear. It is better to take care of the efficiency of the surface burnishing procedure by using a higher feed rate and a ball in the tip of the burnisher with a larger diameter;
• Experimental studies have shown that a higher burnishing force (inducing higher burnishing stresses) does not always lead to a higher wear resistance of the surface. Rather, high peaks of unevenness form in the surface, which break away and cause accelerated erosion of the surface.
• The observed effects of the favourable tribological properties at low burnishing forces require further investigation. The present study should be regarded as preliminary research, and the developed methodology for determining the actual degree of surface wear should be applied to a wider range of experimental studies. The residual stresses obtained during surface burnishing need to be better understood, and their value may be a better correlation factor between technological processing parameters and the increase in the resistance of the surface to tribological wear. Numerical simulation models can be particularly helpful for their assessment.