**1. Introduction**

Several industries make extensive use of steel components, among which are two different microstructure designations that are distinctive: austenitic AISI 316 and martensitic UNS S46500 stainless steels [1,2]. Since turning and milling are the conventional processes to machine these materials, the presence of irregularities and defects is inherent. This unevenness on metal surfaces causes considerable energy dissipation (friction), surface damage, and fracture during the service life of these components [3]. To minimize these issues, a reduction in roughness and an increase in mechanical properties are required [4,5]. Accordingly, numerous final machining operations have been proposed as applicable solutions, among them, the ball-burnishing process [6,7].

This procedure confers extra mechanical properties on the treated pieces, maintaining low costs and reducing execution times [7–10]. Properties such as strain hardening are amplified on a metal surface due to the plastic deformation prompted by the displacement of an indenter at a given pressure [7,10]. At the same time, wear, corrosion, and fatigue resistance are improved due to the newly induced residual compressive state [7,8]. Furthermore, the surface appearance is enhanced because of the decrease in roughness [6,10,11]. However, to achieve these advantages, process (overloading) and material limitations (loss of ductility) [12], which depend on the microstructure [13–15], must be overcome by using a satisfactory configuration [10–13,16–18]. Industrial components, such as lasting valve seals, pistons, bearing bores, and shafts for pumps, are burnished to reduce friction and noise levels and increase their service life [17,18]. Nevertheless, the burnishing of reinforced martensitic stainless steel (such as UNS S46500) components has not been studied so far,

**Citation:** Torres, A.; Cuadrado, N.; Llumà, J.; Vilaseca, M.; Travieso-Rodriguez, J.A. Influence of the Stainless-Steel Microstructure on Tribological Behavior and Surface Integrity after Ball Burnishing. *Materials* **2022**, *15*, 8829. https:// doi.org/10.3390/ma15248829

Academic Editor: Francesco Iacoviello

Received: 18 November 2022 Accepted: 8 December 2022 Published: 10 December 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

enabling innovation in the improvement of this material's surfaces. UNS S46500 is a stainless steel characterized by the presence of Ni3Ti nanometric precipitates in a martensitic matrix [2]. The few previous works quote the process's ability to introduce a deep, highly compressive layer into steel surfaces of this nature (martensitic), with a maximum value location (subsurface level) that is mainly influenced by the applied load and, to a lesser extent, by the speed, feed rate, and number of passes [13]. High loads result in shear stress toward the surface [12]. However, an undue load leads to excessive shear stress and therefore premature degradation of the surface finish [11,12,16–18]. Thus, experimental approaches and simplified predictable models have been conceptualized in order to set the parameters for the process in the search for good surface quality (smooth finish) and an optimum surface residual stress state. For instance, a prior study established a correlation between the roughness and the compressive layers. The interrelation between martensitic wear volume and residual stresses showed a strong inversely proportional linear dependence [19]. In most cases, an inverse relationship between the skewness parameter (Ssk) and wear volume is also recognized [19]. On the other hand, models have achieved a reliable roughness prediction, but only allow for qualitative adjustment (inaccurate results) in terms of residual stresses [20].

These hits and misses address the study of the burnishing-induced plastic deformation phenomenon as a tribological interaction in which it is essential to consider the first basic integral parameter that governs the process: friction. It is the tribo-contact between the burnishing ball and the machined material (roughness, microstructure, and mechanical properties) that determines the intensity of the strain-induced behavior at the material subsurface [12]. The high friction generated by an increase in the load leads to an induced stress state in the leading bulge similar to that induced by uniaxial compression loading. In contrast, the rear zone behind the indenter reacts as if a uniaxial tension is imposed. The higher the friction coefficient, the shallower the maximum shear stress at the sub-surface. Consequently, the plastic strain is concentrated in a thinner surface layer. Nevertheless, overextended friction values could lead to surface decline (fracture, tensile residual stresses) [12]. When friction decreases, the depth of the maximum shear strain increases (reducing the efficiency of cold-work nanostructuring) [12] and could eventually promote residual stress relaxation, reducing crack propagation inhibition [21]. Thus, the burnishing tribo-interaction defines the geometry (by the plastic deformation degree) and the maximum residual tensor location (by the shear-stress depth). In this regard, the tribological interaction between the ball and the rough surface during the ball-burnishing process is tackled numerically through simulations. Amini et al. [21] developed a model that takes into account the alterations of the friction coefficient between the ball and an extruded ferritic AISI 1038 steel surface. Depending on the defined preload, a low friction coefficient could not spawn significant advancements in the roughness and compression stress state. By contrast, a high friction coefficient could lead to an intensification in the pile-up and a decline in the stress state. This means that a factual friction coefficient must feed the models in order to reproduce and enhance the final surface integrity required for industrial components [19,21]. Moreover, Amini et al. proved that the direction of the highest induced residual stress concentration depends on the burnishing route, regardless of the initial stress state produced by machining [21]. The utmost burnishing effect is made in the perpendicular direction to the process, which means that the burnishing process can induce anisotropic properties in the target piece, in agreement with its final application [19,22]. Consequently, each parameter needs to be established according to the use of the piece, prioritizing the geometric (roughness) or metallurgic (hardness and compression stress state) characteristics [13,22,23]. Therefore, both the micro- and macro-responses to the process must be investigated through the microstructure's influence on friction behavior.

Consequently, this study reveals that surface improvements (finish and residual stress state) also depend on the tribological interaction degree between the ball and a defined microstructure. Thus, this tribological interaction is now conceptualized numerically by the friction coefficient. Accordingly, a reinforced martensitic stainless steel matrix and an austenitic stainless steel textured surface are evaluated under the same milling and burnishing process conditions (in agreement with the machining conditions applied to the already characterized ferritic AISI 1038 steel [11,21,23–25]) in terms of friction and surface integrity using a scratch test procedure, 3D optical profilometry (surface finish), and X-ray diffraction (XRD) technique (residual stresses induced by cold working). The results show that under the same machining configurations, the induced surface integrity depends on the self-hardening coefficient due to the different tribo-contacts during the execution of the burnishing while providing reliable inputs for future integral modeling and process parameterization. Therefore, the interaction of the ball with an established macro-texture is not enough to generalize the process; it is necessary to consider the contact at the micrometric level to define the burnishing applicability.
