Impacts of Surface Texture and Nature of Friction on Energy-Force Efficiency of Surface Plastic Deformation during Burnishing

Abstract

Burnishing, the plastic deformation of the workpiece surface due to sliding contact with a tool called burnisher, is a finishing operation widely used in various industries. In this work, impacts of the initial surface roughness Ra of the workpiece being burnished, the nature of friction in the contact zone, and the clamping force on the stability and energy efficiency of burnishing have been investigated. Experiments have been conducted with and without lubricant, represented by low-viscosity deep-hydrogenated fraction of sour oils, at initial surface roughness Ra of 0.8 and 1.25 μm and variable (100–200 N) clamping force. A key process indicator, which largely controls mechanics of burnishing, the temperature in the tool-workpiece contact zone has been measured using natural thermocouple method. Microhardness of the workpiece surface after burnishing has also been measured. It has been shown that changes in the temperature of the tool-workpiece contact zone are proportional to the changes in the squared tool clamping force. This dependence appeared to be universal and equally applicable to burnishing with and without lubrication. Based on the analysis of the experimental data, a new criterion of the burnishing efficiency has been developed. The new criterion simplifies the choice of optimum operational parameters and helps in preventing adverse impacts of structural phase transformations in the workpiece surface layer that unavoidably lead to reduced product quality and operational reliability and in reducing tool wear, which is critically important in the case of dry burnishing. The obtained results show that the nature of friction accompanying the surface plastic deformation has a significant impact on the stability and energy efficiency of the burnishing process. While the clamping force is equally important for burnishing with and without lubrication, the initial roughness Ra has an impact on dry burnishing only. Application of minimum quantity lubrication (MQL) under experimental conditions typical for industrial burnishing is found to be favorable. In particular, it was shown that MQL not only enhances the stability of burnishing process and but also increases its energy efficiency by more than 20%.

1. Introduction

Burnishing is the plastic deformation of the workpiece surface with a sliding tool called burnisher, which is widely used in various industries as a finishing operation. In the burnishing process, surface roughness decreases, hardness of the surface layer increases, with compressive residual stresses being formed in it, and the surface texture is created [1]. In contrast to grinding, burnishing is free of damages to the workpiece surface by fragments of abrasive grains. During burnishing, the workpiece rotates and the deforming tool (indenter), which is in contact with the workpiece surface and fixed in an appliance, has a longitudinal motion. Wide burnishing technology, where the indenter contacts the workpiece along the lateral surface without the longitudinal motion, is described in [2,3].

Being one of the common surface plastic deformation technologies, burnishing is used to machine surfaces with an initial roughness Ra of 0.6–0.9 μm to achieve Ra of 0.03–0.06 μm and to produce the mirror surface. Natural and synthetic diamonds, mineral ceramics, hard alloys, and high-speed tool steels are widely used as burnishing tool materials [1].

Contact phenomena and surface conditions arising from the interaction of tool and workpiece of different hardness in the contact zone are described in detail in [4,5,6,7]. Lubricating fluid in the contact zone creates a layer separating the tool and workpiece surfaces that has two main functions: lubrication and cooling. In the case of elastic bodies, numerical models of contact mechanics [8], which are based on boundary elements methods and analytical dependences [9], are used to calculate the critical conditions for the displacement of lubricant from the contact zone.

The modern machine-building industry consumes a huge amount of fluids for lubrication and cooling. A significant fraction of lubricants and coolants is used in mechanical processing that allows achieving physical, structural, and shape modifications of the material being machined. Properties of these fluids have greatly improved over the past decades due to the development of chemical technologies for their production, which resulted in a transition from natural lubricants and coolants to complex multicomponent synthetic ones. There exist three main risk factors associated with the use of modern lubricating and cooling fluids such as carcinogenicity (adverse health impacts), incomplete disposal with the formation of residues (environmental pollution), and low flash point (risk of fires).

Currently, there exist three main ways of implementing green mechanical processing technologies such as dry processing, minimum quantity lubrication (MQL), and the use of ecologically friendly lubricants. Recently, the application of an MQL to reduce the contact zone temperature and improve the surface quality has become a focus of intense research activity.

In work [10], mineral oil, which is probably not the best choice for implementing green processing technologies, was used. In paper [11], burnishing of titanium alloys was investigated. The surface integrity was evaluated using a set of parameters including roughness, microhardness, topography, microstructural changes, and tribological properties. Vegetable oils were used as a lubricant and a cryogenic cooling agent. Cryogenic cooling was performed using two nozzles with a diameter of 2 mm at a pressure of 6 bar, while the temperature was measured by a thermal imaging camera. It was found that that Ra values were quite similar but MQL and dry burnishing produce superior quality surface with uncoated tools. A comparison of dry burnishing, MQL, and cryogenic cooling shows that cryogenic cooling in combination with coated tool delivers best results. However, no estimates of production costs were given in [11].

In article [12], the authors studied burnishing with MQL, dry burnishing, and the presence of the additional cooling agent represented by CO₂ gas (supercritical CO₂, SCCO₂ method). Two burnishing tools made from carbide with a diameter of 10 mm and 16 mm were used to burnish SS400 carbon steel. A nickel-chromium thermocouple and a thermal imaging camera were used to measure the temperature. In dry conditions, the burnishing tool of 10 mm in diameter shows a 10% higher thrust force than that in MQL conditions. SCCO₂ + MQL approach exhibits a 48.5% and 43% of reduction in thrust force compared to the dry condition and MQL, respectively. The study revealed that the SCCO2 + MQL outperforms both dry and MQL burnishing in terms of burnishing force, temperature, and surface roughness [12] and that the difference between dry and MQL processing is small. Articles [13,14] present CAMQL (cold air with minimum quantity lubrication) studies where the optimal proportion of cold air needed to minimize temperature during processing was found.

In the recent work [15], their authors employed a new burnishing tool, explored the use of MQL in the form of spray and droplets, and studied the impacts of multiple factors on the surface properties of workpieces made of 17-4 PH stainless steel. They found that minimum surface roughness of 0.05 μm and maximum surface hardness of 405 HV had been achieved in the cases when diamond spheres with 4 mm and 3 mm radii, respectively, were used at optimum burnishing conditions. It was indicated that the droplets penetrate into the contact zone very effectively, then evaporate and seem to have only a little harmful impact on the environment.

Different opinion on the environmental and health impacts of cutting fluids has been expressed in [16], where the burnishing of stainless steel has been studied. In particular, it was stated that mist particles formed from conventional lubricants/coolants can have a significant impact on the health of machine tool operators and that “millions of workers get affected by working under different kinds of cutting fluids”. It was concluded that conventional lubricants, as well as nanoparticles, liquid nitrogen (LN2), and other agents, require a thorough assessment of their environmental safety. It was also found [16] that after diamond burnishing in the cryogenic cooling environment, the surface roughness was reduced by 33–50% and 34–51% compared to MQL and dry environments. After machining in cryogenic cooling conditions, moderate 5–7% and 6–10% increases in surface hardness in comparison to MQL and dry environments, respectively, were observed.

Occupational health and environmental pollution risks associated with harmful aerosol pollution produced during metalworking have been studied in [17], where aerosol mapping was used to assess particle number and mass concentration in the engine machining and assembly facility in winter and spring. The greatest particle mass concentrations in [17] were found around poorly insolated metalworking facilities. A significant reduction in mass median aerodynamic diameters of metalworking fluids, which dropped to 2.5 mg/m³ in the 1970s, to 1.2 mg/m³ in the 1980s, and to 0.5 mg/m³ in the 1990s, with no substantial changes seen in the 2000s and 2010s [18], is also an important factor due to epidemiological evidence of adverse impacts of ultrafine particles [19,20], which also have strong (and harmful) oxidative potential [21], on public health and environment. These considerations lead to a perfectly logical conclusion that any new or existing element of metalworking industry must be thoroughly investigated with regards to not only production efficiency but also occupational and environmental safety.

However, despite the critical importance of the implementation of green technologies in metalworking, maintaining and improving operational parameters of products, including roughness, remain the key priorities. While both dry and MQL burnishing are obviously better in terms of occupational and environmental safety than conventional processing, which utilizes large amounts of potentially toxic lubricants or/and cooling fluids, their relative efficiency in terms of production is yet to be fully understood, and a criterion that allows for a direct comparison of the efficiency of burnishing technologies is yet to be developed.

In this work, we have studied the impacts of the roughness of the surface being burnished and the nature of friction (dry contact surfaces vs. surfaces separated by a lubricant layer (MQL)) on specific plastic strain energy. It is important to note that in contrast to conventional burnishing involving a large amount of fluid, burnishing with MQL is not accompanied by heat removal from the contact zone. Thus, the fluid in the contact zone act as a lubricant only.

During burnishing, energies of plastic deformation and friction are converted into heat [22,23]. While the plastic deformation energy is associated with the useful work of reducing surface roughness, the energy expended on overcoming friction forces associated with the adhesion of the contacting surfaces is wasted. So, in accordance with the energy conservation law, the total dissipation energy is spent on the plastic deformation of surface microroughness and on overcoming the friction force. In the case when there exist two processes in which equal energy is spent on the plastic deformation, the process in which a larger amount of energy is wasted on overcoming the friction force occurs at a higher temperature. Since both energies are ultimately converted into heat, the temperature in the contact zone can be used as an indirect measure of the efficiency of the burnishing process. On the other hand, adhesion itself, not accompanied by plastic deformation, is unlikely to produce heat because, in this case, when heat is released during the formation of adhesive bonds, it should be absorbed upon rupture. The classical work of Beer and Bowden proved that even under the mildest friction regimes, microprotrusions are crushed and that the plastic deformation is not limited to protrusions alone and captures subsurface layers. Under plastic deformation, typically ≈10% of the work transforms into the latent stored energy.

Conventional instruments for measuring temperature, the main indirect indicator of burnishing mechanics, include a thermal imaging camera and thermocouples, which allow for accurate temperature measurements at the point of contact between the tool and the workpiece surfaces. In a series of experiments performed in this study, the temperature in the contact zone was measured by the natural thermocouple method [24], and, based on the experimental data, a new criterion characterizing the efficiency of burnishing under different conditions has been developed. The newly developed criterion simplifies the choice of optimum operational parameters and helps in preventing possible negative impacts of structural phase transformations in the surface layer of the workpiece that unavoidably lower both the product quality and operational reliability and in reducing tool wear, especially in the case of dry burnishing.

2. Equipment and Experimental Technique

Burnishing was performed with the lateral surface of a hard alloy cylinder of 6 mm in diameter, the axis of which is perpendicular to the axis of the workpiece, as shown in Figure 1.

The following parameters:

• V (RPM)—burnishing speed equal to the circumferential speed of movement of the indenter relative to the workpiece surface;
• S (mm/rev)—longitudinal feed of the tool (indenter);
• P (N)—indenter clamping force.

were used to control the burnishing process.

A workpiece with a diameter d (mm) (Figure 2) was fixed in a three-jaw chuck of a lathe (Figure 3a), which allows for the rotation of the workpiece with adjustable rotational speeds n (rpm). A tool with a mechanical fastening of an indenter (Figure 3b) in a holder was installed in the tool holder of the machine support and, in the process of burnishing, was moving along the workpiece axis with an adjustable speed Vp. The kinematic parameters of burnishing are related to each other via the following expressions:

$$V=\frac{\pi \cdot d\cdot n}{1000},$$

(1)

$$V_p=S\cdot n$$

(2)

To study the effect of lubricant on fastening, 50 samples were made. The samples were structurally similar to the parts produced by JSC AVTOVAZ (Figure 2).

The geometry of workpieces, their hardness, and material, as well as the operational parameters of the burnishing process, were chosen to be comparable with the parameters of the serial JSC “AVTOVAZ“ burnishing technology (

Table 1. Materials and operational parameters burnishing used in the production of Lada Niva Travel car parts at JSC “AVTOVAZ”.

Name	Intermediate Propeller Shaft Flange	Propeller Shaft Mounting Flange	Rear Axle Shaft
Material	Steel 43	Steel 43	Steel 38
Hardness	HB 165—215	HB 165—215	HRC 53
Diametric line (mm)	36	32	34.8
Length of burnished part (mm)	14	14	19.5
Burnishing velocity V, (m per min)	74.5	74.5	54.7
Feed S (mm per rotation)	0.15	0.15	0.12
Burnishing force P (N)	150	150	200
Tool radius R (mm)	4	4	3
Annual work program in pcs	900,000	180,000	265,000

).

The chemical composition of steel St 43, which is the analog of AISI 1043 Steel (UNS G10430), is shown in

Table 2. Chemical composition of steel 43 and steel 38.

	C	S	P	Mn	Cr	Si	Ni	Cu	V	Mo	W	Fe
steel 43	0.38–0.43	≤0.01	≤0.02	0.5–0.8	2.8–3.2	0.8–1.2	0.8–1.2	≤0.15	0.05–0.15	0.35–0.5	0.8–1.2	Bal.
steel 38	0.34–0.42	≤0.035	≤0.035	0.5–0.8	1.3–1.6	0.8–1.2	≤0.3	≤0.3	-	0.3–0.56	-	Bal.

(GOST 1050-2013 “Metal products from nonalloyed structural quality and special steels. General specification”).

Material of the working part of the indenter was sintered hard wolfram-cobalt alloy VK8 (equivalent to HG30), which consists of WC (92%) and Co (8%) and has the following properties: flexural strength 1666 N/mm², density 14.5–14.8 g/cm³, and hardness HRA not less than 88.0. The design of the indenter is a rod with a diameter of 6 mm.

Operational parameters that had been varied during the experiments are shown in

Table 3. Experimental parameters.

Parameter	Range
Burnishing speed V (RPM)	630
Tool feed rate S (mm per rev)	0.13
Diametric line (mm)	39
Burnishing force P (N)	100	250
Tool radius R (mm)	3
Lubricant	Absent	RZh8u
Initial roughness, Ra	0.8	1.25

. The RZh8u lubricant produced by Lukoil (Russia) is a low-viscosity deep-hydrogenated fraction of sour oils and was used in small quantities (MQL) for lubrication only. The flash point of RZh8u lubricant, determined in a closed crucible, is not lower than 120 °C, its kinematic viscosity at 20 °C is 6.0–8.0 mm²/s, and density is not more than 0.850 g/cm³. The mass fraction of sulfur in the lubricant did not exceed 0.02%.

Burnishing was carried out by the unworn part of the lateral surface of the cylindrical carbide tool by turning and vertically displacing it along the axis in the holder. The lubricant was applied in a thin layer, without abundant watering, with a sponge. The force was created by pressure in a closed hydraulic system using a cylinder with two rods. The tool was fixed on the first rod. The second regulating rod regulated the pressure in the hydraulic system with the help of a threaded connection. The indicator was a pressure gauge whose membrane was used s to dampen self-oscillations. The natural thermocouple method was used to measure the temperature of the contact zone. The electrodes were connected to the workpiece and the indenter. The calibration curve (see supplementary file) was used to convert the values of μV to t °C, with the error being no more than 1–2% [25]. To register thermo EMF, a digital multimeter ZT-C2 (ZOYI, Wuxi, China) with a measurement range of 200 mV and a resolution of up to 100 μV was used. For roughness measuring, a MarSurf PS 10 profiler (Mahr GmbH, Esslingen, Germany) was used.

3. Results and Discussion

Experiment results after the primary statistical processing are shown in

Table 4. Experiment results.

Initial roughness Ra 1.25 μm	Dry
	P, N	t °C
	100	200
	150	260
	200	275
	220	315
	250	550
	RZh8u Lubricant
	P, N	t °C
	120	140
	150	180
	180	200
	200	230
Initial roughness Ra 0.8 μm	Dry
	P, N	t °C
	100	185
	150	210
	200	250

. We repeated each experiment in Table 4 five times for the first six samples and three times for the last six ones. The relative errors for the tool pressing force P and for temperature t measurements were 5% and 10% (in °C scale), respectively.

At the first stage of the experiments, a correlation analysis was carried out in order to determine the influence of different factors, such as the tool pressing force P, the initial surface roughness Ra and the nature of the friction forces (dry burnishing vs. MQL with RZh8u lubricant) on the temperature in the contact zone. The results of the correlation analysis are presented in

Table 5. Parameter cross-correlations matrix. Marked correlations are significant at p < 0.05. N = 12 (casewise deletion of missing data).

Variable	Force P, N	Ra, μm	Surface/Lubricant	Temperature, °C
Holddown pressure P, N	1.00	0.23	−0.09	0.76
Ra, μm	0.23	1.00	0.41	0.20
Surface/Lubricant	−0.09	0.41	1.00	−0.43
Temperature, °C	0.76	0.20	−0.43	1.00

.

Based on the results of the correlation analysis (Table 5), we found that the clamping force has a dominant effect on the temperature rise because the corresponding Pearson’s correlation coefficient K = 0.76 is significant. In contrast, the influence of the initial roughness Ra is insignificant (K = 0.2)). It is also important that the application of a lubricant to the surface increases the burnishing efficiency by reducing the fraction of energy spent on friction (K = −0.43)

Impacts of friction nature (dry friction vs. friction through a separating layer of lubricant) were investigated by the method of group analysis through nonparametric criteria. The surface condition factor “dry” or “lubricated” is categorical, and, therefore, the statistical significance of the influence of this factor was assessed using the Mann–Whitney U-test for two independent groups of experiments. U-rank test [26,27,28,29] is a nonparametric analog of the Student’s t-test for comparing two mean values of continuous distributions. The U statistic (adjusted for ties) is accompanied by a Z value (normal distribution variate value with mean μ = 0 and standard deviation σ = 1) and the corresponding p-value (tests are significant at p < 0.05.). The results of the nonparametric analysis are shown in

Table 6. Mann–Whitney U-test for the significance of the division into the groups “dry surface”-“lubricant RZh8u”.

Variable	Rank SumDry	Rank SumRZh8u	U	Z	p-Level	Z Adjusted	p-Level
Temperature, °C	33.50	11.50	1.50	2.08	0.04	2.09	0.04

.

Based on the U and Z values of the statistics of the Mann–Whitney criterion (Table 4) with an error of less than 3.7%, one can conclude that the nature of friction accompanying surface plastic deformation has a significant impact on the efficiency of the burnishing process. A clear illustration of this effect using the group Box-plot graphs is presented in Figure 4. The analysis shows that in the case of dry friction, there exists a much larger scatter in temperatures measured in the contact zone, with the greatest deviations (more than 200 degrees C) from the upper 25% boundary of all measured temperatures. The median of measured temperatures for lubricated burnishing is more than 70 °C smaller than the median of all measured temperatures for dry burnishing.

In the third stage, the regression analysis was performed to determine the influence of clamping force and initial surface roughness on the temperature in the burnishing zone. Based on the results obtained at previous stages and due to the fact that the factor of the nature of friction (with and without lubrication) was categorical, the regression analysis for these two groups of experiments was carried out separately. The main goal of the analysis was to establish a quantitative measure of the efficiency of the burnishing process with and without lubrication.

The model under examination was given by the following equation:

$$T=a_0+a_p\cdot P+a_{pp}\cdot P^2+a_{Ra}\cdot Ra+a_{pRa}\cdot P\cdot Ra,$$

(3)

where

$a_0,a_p,a_{pp},a_{Ra},a_{pRa}$ are regression coefficients.

The statistical significance of the coefficients of the model (3) was determined by the Student’s coefficient t with a confidence interval of p-level < 0.005.

Statistical significance of coefficients of the regression model (3) and their comparative analysis for the conditions of dry friction and friction with lubrication determine the influence of various factors such as P (coefficients $a_p$, $a_{pp}$), initial roughness $Ra$ (coefficient $a_{Ra}$) and joint effect of P and $Ra$ (coefficient $a_{pRa}$) on the plastic deformation temperature.

The values of the significant regression coefficients (confidence level of 95.0%) of the model (1) and their statistical properties for dry burnishing and burnishing with lubricant are given in

Table 7. Significant regression coefficients (1) for dry burnishing.

	Estimate	Standard Error	t-Value df = 6	p-Level	Lo. Conf Limit	Up. Conf Limit
$a_{pp,}°\mathrm{C}/N^2$	0.0049	0.0013	3.8084	0.0089	0.0018	0.0081
$a_{Ra}, °\mathrm{C}/\mathsf{\mu }\mathrm{m}$.	112.6845	49.9404	2.6242	0.0394	7.6132	217.7559

and

Table 8. Significant regression coefficients (1) for lubricated (MQL) burnishing.

	Estimate	Standard Error	t-Value df = 2	p-Level	Lo. Conf Limit	Up. Conf Limit
$a_{pp}, °\mathrm{C}/N^2$	0.0033	0.0004	8.5939	0.0133	0.0017	0.0050
$a_0, °\mathrm{C}$	96.6760	11.2153	8.6200	0.0132	48.4205	144.9314

, respectively.

The tool clamping force exerts the greatest influence on the temperature in the contact zone at given burnishing conditions. As can be seen from Table 7 and Table 8, the temperature change is proportional to the squared clamping force. The adequacy (consistency) of regression models (3) with the coefficients given in Table 5 and Table 6 appears to be quite high, as evidenced by R = 0.88 for processing on dry burnishing and R = 0.99 for MQL burnishing. Since model (3) is interpolational, its adequacy is strictly ensured only in the range of factors shown in Table 4.

The initial roughness of the workpiece surface has a definite impact only on dry burnishing ($a_{Ra}>0$) as shown in Figure 5. In the case of MQL, the effect of the initial roughness is insignificant ($a_{Ra}=0)$. The effect of the clamping force is important l in both dry and MQL conditions. The degree of its influence is expressed via a quadratic dependence ($a_p=0,a_{pp}>0$). The dependence of the temperature in the contact zone on the clamping force for MQL is shown in Figure 6.

The $a_0$ coefficient should be specifically mentioned. Its non-zeros value for MQL indicates a more stable nature of friction regardless of external factors compared to dry burnishing, for which $a_0=0$. The comparison of observed and predicted temperatures for the two burnishing modes according to model (3) with the corresponding coefficients given in Table 5 and Table 6 is shown in Figure 7.

Based on the obtained regression dependencies, one can easily estimate the efficiency of lubrication for burnishing parts of the “shaft” type. As we mentioned earlier, a considerable fraction of work during dry burnishing is spent on overcoming dry friction forces, which leads to additional heating of the contact zone. Without accounting for the relatively small effect of thermal softening of the workpiece surface material, we define the energy efficiency of the process as the increment in the clamping force spent on increasing the temperature in the contact zone by 1 °C.

The squared clamping force is, according to Table 7 and Table 8, the most significant factor affecting the burnishing temperature in both dry and lubricated conditions. Since the clamping force determines the energy loss at burnishing, the ratio of coefficients at the squared clamping force in the regression model determines the relative efficiency of different burnishing modes in terms of energy spent to achieve the same surface roughness By analyzing model (3) with coefficients given in Table 7 and Table 8, we have determined the relative efficiency of the dry and MQL burnishing processes. The relative efficiency of these two processes in terms of energy consumption is given by the following Expression (4):

$$k_T=\sqrt{\frac{{\left(a_{pp}\right)}_{\mathrm{dry}}}{{\left(a_{pp}\right)}_{coolant}}}=\sqrt{\frac{0.0049}{0.0033}}=1.215$$

(4)

As can be seen from (4), burnishing with MQL is 1.22 times more efficient than dry burnishing.

Regression models based on Table 7 and Table 8 were analyzed for statistical significance of factors according to Fisher’s and Student’s criteria. Only significant factors were included in the models. It is important to note that temperature in the contact zone, a key factor of energy-efficient burnishing, depends on the square of the clamping force for both dry and MQL burnishing. This may be explained by the synergistic effect of the shares of total energy spent on overcoming friction forces and on plastic deformation of microroughnesses. Each share of energy is, in the first approximation, proportional to the clamping force and the area of the indenter workpiece contact zone, which is, in turn, also controlled by the clamping force.

The analysis of this phenomenon leads to a new criterion of the burnishing efficiency in the form of the squared root of the ratio of the coefficients at the squired clamping force in Equation (3) given by Formula (4). According to its meaning, the criterion compares useful work of the surface plastic deformation under different burnishing conditions. It is important to note that while the regression dependence for burnishing without lubricant contains the initial roughness Ra, that for burnishing with lubrication is independent of Ra. This indicates that in the experimental range of Ra, which is typical for industrial burnishing, the MQL provides a stable layer separating the tool and the workpiece and enhances, in this way, the stability of the burnishing process.

4. Surface Microhardness after Burnishing

An Axiovert 40 MAT (Carl Zeiss, Germany) microscope, including a Thixomet Pro material structure quality assessment program, an object micrometer, and an HVS-1000 digital stationary hardness tester (Oceanus, China) using the Vickers method, were used to determine the microhardness after burnishing with and without lubricant.

From samples No. 6 and No. 50, one sample was cut in the transverse direction of the shaft (Figure 8) for metallographic studies (Figure 9). The samples were pressed into the plastic on a SIMPLIMET 1000 (Buehler, Switzerland) casting press, and microsections were prepared on a TegraPol-11 + TegraForce-1 + TegraDoser-5 (Struers, Denmark) grinding and polishing machines.

Based on the results of the analysis of sample #6 (burnishing with lubricant, Figure 9a), the following information was obtained: microstructure-ferrite and perlite with a ferrite grain size of 9 points. The deformation of the microstructure to a depth of 20–25 µm was observed on the surface. Sample #50 (dry burnishing, Figure 9b: microstructure) ferrite and perlite with the ferrite grain size of 9 points. The distributions of microhardness from the surface to the sample for samples No. 6 and No. 50 shown in Figure 10 are quite similar.

As seen in Figure 9, the microstructure of the surface layer of samples is represented by ferrite with pearlite (dark fragments). This structure is the most common among carbon hypoeutectoid steels, with a carbon content of 0.3–0.4%. Perlite fragments do not change their size as they approach the surface, which shows that the residual deformation in the surface layer is small. This conclusion is confirmed by the analysis of microhardness (Figure 10), which is in the range of 200–215 HV and does not significantly change with depth.

5. Conclusions

In this work, impacts of the initial surface roughness Ra of the workpiece being burnished, the nature of friction in the contact zone, and the clamping force on the stability and energy efficiency of burnishing have been investigated, and a new criterion of the burnishing efficiency, expressed in the form of squared root of the ratio of the coefficients at the squired clamping force in Equation (3), and described by Formula (4) has been derived.

The new criterion simplifies the choice of optimum operational parameters and helps in preventing adverse impacts of structural phase transformations in the workpiece surface layer that unavoidably lead to reduced product quality and operational reliability and in reducing tool wear, which is critically important in the case of dry burnishing.

The present work leads us to the following conclusions:

• The nature of friction accompanying the surface plastic deformation has a significant impact on the stability and energy efficiency of the burnishing process. While the clamping force is equally important for burnishing with and without lubrication, the initial roughness Ra has an impact on dry burnishing only;
• The change in the key process indicator, the temperature in the tool-workpiece zone, is proportional to the changes in the squared tool clamping force. This dependence is universal and equally applicable to burnishing with and without lubrication;
• The nature of friction accompanying surface plastic deformation has a significant impact on the efficiency of the burnishing process. In particular, nonparametric analysis illustrated by the Box-plot graphs in Figure 4 shows that in the case of dry friction, there exists a much larger scatter in temperatures measured in the contact zone and the greatest deviations (more than 200 degrees C) from the upper 25% boundary of all measured temperatures. The median of measured temperatures for lubricated/MQL burnishing is more than 70 °C smaller than the median of all measured temperatures for dry burnishing;
• Under experimental conditions typical for industrial burnishing, lubrication/MQL not only enhances the stability of the burnishing process but also increases its energy efficiency by more than 20%.