Assessment of surface irregularities created by controlled liquid droplet on the surface of stainless steel AISI 304L

Engineering Science and Technology, an International Journal 47 (2023) 101558 Available online 1 November 2023 2215-0986/© 2023 Karabuk University. Publishing services by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Assessment of surface irregularities created by controlled liquid droplet on the surface of stainless steel AISI 304L Gabriel Stol´ arik a , Dagmar Klichov´ a b , Jakub Poloprudský c , Akash Nag d , Sergej Hloch a , * a Faculty of Manufacturing Technologies TUKE with a seat in Preˇ sov, Slovak Republic b The Czech Academy of Sciences, Institute of Geonics, Ostrava-Poruba, Czech Republic c Institute of Physics of Materials CAS, Brno, Czech Republic d Faculty of Mechanical Engineering, Vˇ SB-Technical University of Ostrava, Ostrava-Poruba, Czech Republic ARTICLE INFO Keywords: Droplet erosion Surface modification Pulsating water jet Surface roughness Areal parameters ABSTRACT Surfaces created by the erosive action of water droplets have not been sufficiently explored because of their stochastic structure, which depends on hydraulic parameters. This article details the complex analysis of surfaces created by the controlled distribution of water droplets on the surface of AISI 304L using an ultrasonically stimulated water jet. The traverse speed of the jet controlled the distribution of the water droplets. The surface topology was modified in an interleaved mode when the trajectories were parallel for all samples. Additionally, the second layer treated the other half of the preprepared samples with a perpendicular trajectory with respect to the parallel trajectory. The 3D surface reconstruction was performed using a noncontact MicroProf FRT measuring instrument. Several profile roughness parameters (Ra, Rz, Rv, Rp, Rsk, and Rku), areal surface roughness parameters (Sa, Sz, Sp, Sv, Ssk, Sku, Sdr, Sk, Spk, and Svk) and surface isotropy values were measured for the surfaces generated using variations in the traverse speed of the jet. The measurement results show a decreasing trend of surface roughness upon an increase in the traverse speed, and a better surface finish was also measured for the cross-hatch trajectory compared to the linear trajectory. The results also show that the generated surfaces have deeper valleys with truncated peaks, which are suitable for use in a wide range of technical and medical applications. This study shows the potential utilization of controlled liquid droplet impingement for surface preparation in various application domains. 1. Introduction Manufacturing processes are used to modify or transform any raw material or workpiece of no or less value into a finished or semi-finished product having a specific shape, size and surface finish. Different external forces change the surface conditions [1,2] differently, affecting the final product’s surface integrity [3]. In addition to the properties of subsurface layers, an aspect of surface integrity is surface roughness. Surface roughness [4] of the final product is an important feature for several aspects of mechanics, such as contact friction and deformation. Until now, roughness profile parameters have been used to identify and detect real surface geometry [5]. Machining with conventional technologies (e.g., turning and milling) is considered a primary production process [6], which provides the possibility of manufacturing a part mainly from the point of view of dimensional accuracy [7]. Secondary production processes aim to change a material’s surface and subsurface properties according to its application domain [8]. Modifications are performed mainly to remove residual tensile stress in welds [9] or change the surface topography of the implants for biomedical applications [10]. These secondary processes are often performed using technologies such as shot peening, laser texturing, and plasma spraying [11,12] which have certain disadvantages, such as pushing the substrate material into the surface with shot peening [13], the generation of a thermally affected area with laser texturing [14], and the reduction of the durability of the coated layer under high fatigue strength with plasma-sprayed samples [15]. In order to reduce these shortcomings, one possible nonconventional method, waterjet (WJ) technology, is currently being employed as a surface modification tool, which utilizes continuous stagnation pressure (p s ) to modify a material surface [16,17]. To increase the efficiency of the WJ modification process, abrasive grains are usually added to the water flow, resulting in the generation of an abrasive WJ [18]. However, this method also causes the embedding of * Corresponding author. E-mail address: [email protected] (S. Hloch). Contents lists available at ScienceDirect Engineering Science and Technology, an International Journal journal homepage: www.elsevier.com/locate/jestch https://doi.org/10.1016/j.jestch.2023.101558 Received 19 July 2023; Received in revised form 23 September 2023; Accepted 19 October 2023 Engineering Science and Technology, an International Journal 47 (2023) 101558 2 abrasive particles into the surface [19], which causes a change in its chemical composition and necessitates an additional operation for removing these particles, such as shot blasting [20]. To eliminate particle embedment, a pulsating WJ (PWJ) can be used, which utilises a high-frequency hammer effect with water droplets [21]. In this method, the material is subjected to repetitive water droplet impacts, which generate an impact pressure (p i ) several times higher than the p s used in continuous WJ (CWJ) under the same flow conditions [22]. The periodic application of p i induces stresses that exceed the ultimate strength of the material, causing the formation of elastic–plastic deformation in the surface topology, changing its properties in the form of subsurface stresses and surface roughness [19]. Surface roughness in the case of a material exposed to the repeated action of water droplets can vary depending on the degree of erosion [23]. A rotating disc [24] with opening, self-resonating nozzles [25] or an ultrasonic sonotrode is used as an inverter of a CWJ into water droplets. The disadvantage of rotating discs is the low frequency and, thus, the need for a longer exposure time. The disadvantage of self-resonating nozzles is the necessity of high flow rates in order to manifest oscillations arising during a sharp transition due to a change in the geometry of the opening in the nozzle [26]. A technologically more demanding but more acceptable option is the use of an ultrasound-stimulated WJ. This technological modification is able to generate in a short time a number of droplets that correspond to the nominal frequency of the acoustic generator [27]. A huge advantage of this technology is the high variability of the generated droplets depending on the hydraulic parameters, such as the pressure and diameter of the water nozzle. Another advantage is the possibility of flexible distribution of water droplets along a defined trajectory by varying the traverse speed of the jet. A specific feature of a surface eroded by droplets is the stochastic composition of irregularities detected under different technological conditions on different materials [28]. Surface roughness largely depends on the stage of erosion. Erosion stages are commonly classified into three main groups related to repetitive, multiple droplet impingements: incubation, acceleration, and culmination [29]. Each stage has distinct characteristics regarding surface roughness. The incubation stage can be further divided into pre-incubation and incubation stages. The pre-incubation stage can be considered a material state where no detectable or visible erosion occurs after any process. However, changes in the subsurface properties are due to the transmission of compressive stress into the material, which results in grain reorientation and refinement [30]. Next, detectable surface changes or surface roughness are achieved in the incubation stage, which lies in the Ra =1 – 5 µm range [8]. This stage is associated with a number of impingements lower than the threshold number required for the initiation of material disintegration. The material surface in this stage can be used for applications requiring a higher fatigue life. With an increase in impingements beyond the material threshold, detectable erosion and a higher magnitude of surface roughness occurs. Here, the magnitude of the erosion and surface roughness increase with an increase in the number of impingements—the surface roughness changes into an erosion crater or groove with a measurable depth. With a further increase in the number of impingements, the erosion rate decreases or does not increase much. This phenomenon is observed because of the resistance of the water layer already present in the formed erosion crater to subsequent droplets, which restricts the efficient interaction of water droplets with unaffected areas of the material [31]. Therefore, it can be observed that depending upon the intensity of the droplets and their distribution over the material, this can lead to achieving different material properties used for different purposes. In previous studies, the controlled use of the water hammer effect was used to strengthen welded joints in AISI 304 stainless-steel materials [32]. The efficiency of this process was determined using the surface roughness characteristics along with microhardness and induction of compressive residual stresses. The spread of water droplets controlled by the exit nozzle’s shape also influences the material’s surface roughness when tested in peening applications [33]. Further, the trajectory or path, along with the distribution of the intensity of the water droplet, can also be used for surface roughening operations. Previously, this hypothesis was tested on aluminum [34] and titanium alloys [35]. In both cases, surface roughness values were largely dependent on the intensity and volume of the droplets, along with the treatment strategy. The results showed that the subsurface properties, measured in the form of microhardness, were also enhanced, along with achieving the desired surface roughness. Therefore, it can be observed that PWJs can be used in various domains by utilizing the potential of water droplets. A porous structure is typical for surfaces created in this way. However, the erosion depth can have a higher impact due to the high intensity of the affecting droplets [36]. However, the determining factor for successfully utilizing this technology is the generation of the desired surface roughness. Surface roughness plays a significant role in various applications involving wear, fatigue, corrosion resistance, and biocompatibility [37]. For example, the cytocompatibility or adhesion of human cells to implant surfaces can be modified or accelerated upon varying the surface roughness. In addition, the mechanical efficiency of the materials undergoing high cyclic loading depends on the surface roughness of the material. Further, the wear or corrosion resistance of the material can also be enhanced by modifying the surface roughness characteristics. Therefore, successfully preparing the surfaces for these applications using PWJ requires an indepth analysis of the surfaces generated using this process. Recently, a publication has been published regarding the profile and areal surface roughness parameters, which provide comprehensive information on surface quality [38]. Height, spacing, and hybrid parameters are used to describe the surface texture very well. However, the Nomenclature symbols and abbreviations f f Actual frequency, [Hz] f s Starting frequency, [Hz] l c Acoustic chamber length, [mm] OF Overlapping factor, [mm] P Output power, [W] p Pressure, [MPa] Q Flow rate, [l/min] Ra Arithmetic mean height, [µm] Rku Kurtosis, [-] Rp Mean peak height, [µm] Rsk Skewness, [-] Rv Mean pit depth, [-] Rz Maximum height, [µm] Sa Areal Arithmetic mean height, [µm] Sdr Developed interfacial area ratio, [%] Sk Areal Core roughness depth, [µm] Sku Areal Kurtosis, [-] Sp Maximum areal peak height, [µm] Spk Areal reduced peak height, [µm] Ssk Areal skewness, [-] Sv Maximum areal valley depth, [µm] Svk Areal reduced valley depth, [µm] Sz Maximum areal height, [µm] v Traverse speed, [mm/s] w Erosion groove width, [mm] z Standoff distance, [mm] G. Stol´ arik et al. Engineering Science and Technology, an International Journal 47 (2023) 101558 3 results represent the surface quantitatively, where we obtain a value for the extent or spacing of points on the surface. This is essential information, but the specific number only provides an intuitive understanding of the surface’s appearance and how it interacts with its surroundings. However, a multiparameter evaluation of the surface topography is needed to interpret the surface’s functional properties, and this can be conducted using functional roughness parameters to evaluate the surface, which takes the surface analysis one step further. These parameters are based on the basic profile and area parameters used in this area of research, where they help to understand how a surface will perform. The Sdr parameter defines the complexity of the surface. This functional parameter describes the surface properties, such as sealing or adhesion. The main motivation of this study is therefore focused on a thorough and detailed analysis of the surface structure created by the PWJ hybrid technology from the point of view of examining several profiles and areal roughness parameters. The results of this work can help to classify the surface created with this method, mainly from the point of view of its functionality and its possible use in technical practice. Therefore, in this research study, different surface topographies were generated with PWJ using traverse speeds of v =1, 2, 3, 4, and 5 mm/s for the linear trajectory and v =2, 4, 6, 8, and 10 mm/s for the crosshatch trajectory with a supply pressure of p =70 MPa and a nozzle diameter d =0.8 mm on AISI 304 stainless steel. The generated surfaces were evaluated regarding the profile surface roughness parameters, including height parameters Ra, Rz, Rp, and Rv, and the probability density waveform parameters Rsk and Rku. Further, the surfaces were evaluated using areal surface roughness parameters, including height parameters Sa, Sz, Sp, Sv, Ssk, and Sku; functional parameters Sk, Spk, and Svk; and hybrid parameter Sdr. Additionally, the morphology of the surfaces was observed using SEM analysis. The main aim of the study was to understand and analyse precisely and thoroughly the surface capabilities that can be produced using PWJ technology. 2. Materials and methods 2.1. Material In this study, low-carbon austenitic stainless-steel AISI 304L was used as the workpiece material. This material was selected for investigation, as it is used in various components exposed to corrosive environments with chromium-based passive layer coatings. Attracting these protective coatings with the AISI 304L surface is important for extending the components’ life span in such environmental conditions. Therefore, optimal surface conditions that prolong the adhesion of these coatings on the material surface need to be determined. The mechanical and chemical properties of the used sample are shown in Table 1 and Table 2. For these experiments, a specimen with dimensions of 120 ×30 ×5 mm was clamped to the clamping table using fixtures and a 20 ×30 mm area was treated with each experimental condition. 2.2. Experiment The experimental treatment was carried out using a hybrid technology consisting of a PWJ constructed at the Institute of Geonics of the CAS. The high-pressure WJ was generated using a Hammelmann HDP 253 pump with a maximum flow rate of Q =67 L/min at a maximum pressure of p =160 MPa. The high-pressure water was transferred using piping to the cutting head directly into the acoustic chamber inside the nozzle head. In the acoustic chamber, pressure fluctuations were added into the pressurised water through an oscillating sonotrode, whose movement was provided by a mechanical coupling with piezoelectric ceramics. These piezoelectric ceramics were stimulated by electric signals sent from an Ecoson WJ UG 630-40 ultrasonic generator at a fundamental frequency of f s =40 kHz. After passing through the acoustic chamber and water waveguide, the pressurized water passed through a StoneAge circular water nozzle, which converted it into a high-velocity WJ [40]. Shortly after exiting the nozzle, the WJ was split into discrete clusters of water droplets that impacted the surface of the target material. The movement of the nozzle head was controlled using an ABB IRB 6640-180 robotic arm and its selected program for movement in the x-, y-, and z-axes (Fig. 1). Before the main experiments, it is necessary to optimise the technology so that the maximum magnitude of erosion is achieved for the selected operating parameters of supply pressure p =70 MPa and nozzle diameter d =0.8 mm. The first step was to ensure the optimum conversion of electrical signals into mechanical motion, which was significantly affected by the impedance and, consequently, the resonant and antiresonant frequency. The aim was to tune the system towards the impedance region and resonant frequency of the selected pressure and nozzle diameter, where the pulsations were generated with the lowest possible output acoustic power [41]. In PWJ technology, the resonant frequency and impedance region are tuned using the length of the acoustic chamber, technologically designed to continuously change its length from 0 to 24 mm by rotating a threaded nut. Since the chamber serves as a reservoir of high-pressure water into which ultrasonic vibrations are induced using the sonotrode, adjusting its length makes it possible to directly change the amplitude and thus find the impedance region with the correct resonant frequency. During this adjustment, the output acoustic power (P) and the actual generated frequency (f a ) were monitored on the control display, and then the optimal level of chamber length was evaluated from the graphical progression of Fig. 2, where the results clearly show the optimal acoustic chamber length to be l c =12 mm. The process of setting up the PWJ technology further included determining the optimum standoff distance between the nozzle and the surface of the target material. This setting is particularly important in terms of optimising the distance required for the CWJ to decay and form a series of well-defined water clusters with maximum erosive effect. The water clusters are formed after exiting the nozzle, and while travelling towards the sample, they are transformed into discrete shapes with a specific energy. Therefore, the optimal standoff distance should be determined to estimate the distance where these water clusters show the highest effect when interacting with the target material to ensure that the maximum potential of the technology to erode the surface is exploited. Preliminary erosion tests were carried out to determine the optimal standoff distance in which the nozzle head follows a stair-like trajectory with a horizontal movement of 10 mm and a vertical incremental step of 1 mm after each horizontal travel. This trajectory was carried out at a uniform traverse speed and, ultimately, resembled a step-like character. In this pilot experiment, the nozzle head trajectory started with a height of 1 mm above the workpiece, and after traversing 10 mm in a linear feed, the standoff distance was increased by 2 mm after each step. Thus, an erosion groove was created at each standoff distance. After using optical evaluation of the resulting erosion grooves, it was possible to determine the optimum standoff distance with the maximum erosion rate as z =77 mm. Surface treatment of materials requires that a certain surface area of the material be affected. Therefore, to cover a certain surface area, the single treatment lines should overlap and be close to the preceding one to cover the entire area. Upon repeating this process, it is thus possible to influence the surface area of the target material. However, in this case, Table 1 Mechanical properties of the AISI 304L [39]. Material Characteristics Tensile Strength R m (MPa) Yield Strength R p 0.2 (MPa) Elongation Ratio A (%) Hardness H (HV) Density ρ (kg/ m 3 ) Young’s Modulus E (GPa) 721 239 43 210 7900 193 G. Stol´ arik et al. Engineering Science and Technology, an International Journal 47 (2023) 101558 4 Table 2 Chemical composition of the AISI 304 L [39]. Material Composition (in %) C Mn Si S P Cr Ni Cu Mo N Ti Fe 0.027 1.591 0.201 0.020 0.035 18.259 8.259 0.526 0.356 0.074 0.002 Rest Fig. 1. Schematic representation of the experimental setup. Fig. 2. Technological adjustment of the acoustic chamber length. G. Stol´ arik et al. Engineering Science and Technology, an International Journal 47 (2023) 101558 5 the spacing between the grooves is important for several reasons. More spacing between the lines could cause an uneven influence on the surface or create places where the surface is eroded and, consequently, where it is not. This would cause a wavy and extremely inhomogeneous material surface. The other extreme is that if the overlapping factors of the paths are too low, it will cause too much erosion, which could be compensated by increasing the traverse speed. However, the process would still be unnecessarily time-consuming, creating a disadvantage from an economic point of view. Therefore, lateral displacement (i.e., overlapping factor) was used for the main experiments. This methodology was based on measuring the original groove width (w) created when finding the optimum standoff distance and creating ten lines sideby-side with a specified spacing. The spacing distances for these experiments were 100 %, 50 %, and 25 % of the measured groove width w =1.8 mm. After these experiments were performed, the surface without these raw lines was achieved at 25 % overlap, which means a lateral displacement of 0.45 mm (Fig. 3). After optimising all of the constant technological parameters (supply pressure p, nozzle diameter d, starting frequency f s , acoustic chamber length l c , standoff distance z and overlapping factor OF), the main experiments were conducted by varying the traverse speed of the nozzle head, which influences the erosion rate of the material by determining the number of cumulative impacts impinging per unit length of the material. The number of impingements directed to a certain place can be calculated by dividing the frequency by the traverse speed. Another variable taken into consideration in this study is the trajectory of the nozzle head. This factor takes into account different ways a surface can be treated using a WJ. The present study considers two treatment trajectories: linear and cross-hatch. In a linear trajectory, the jet passes the material in only one direction, as shown in Fig. 4a. While for a crosshatch trajectory, the jet covers the entire surface in one direction and then covers the entire surface again with a 90◦turned orientation, as shown in Fig. 4b. Traverse speeds of v =1, 2, 3, 4, and 5 mm/s were used for the linear trajectory experiment. However, doubled traverse speeds were used for the cross-hatch trajectory, i.e., v =2, 4, 6, 8, and 10 mm/s. These traverse speeds were selected to keep the treatment time the same for both treatment strategies. All of the technological and geometrical parameters and their respective levels used during the experiments are mentioned in Table 3. 3. Measurement 3.1. Surface roughness Each generated sample surface was scanned using a MicroProf FRT optical profilometer, which is based on the principle of chromatic distance measurement. White light is focused on the surface by a measuring head with a strongly wavelength-dependent focal length. The spectrum of the light reflected on the surface generates a peak in the spectrometer. The wavelength of this peak is used to determine the distance to the sample surface. Upon moving the lens one “x”-axis unit over the material being scanned, it is thus possible to obtain the surface parameters in a given line, which represents the two-dimensional knowledge of the surface being measured (on the “x”- and “z”-axes). The result of the measurement is the profile parameters of the structure, which are further defined according to ISO 21920–2. In this work, a twodimensional measurement of the sample was performed in eleven side-by-side lines, with regular spacing between them (Fig. 5 Left), and their average calculated value was used for the evaluation. For a more comprehensive analysis, the surface was also evaluated using a three-dimensional method (on the “x”-, “y”-, and “z”-axes), where the measurements result in areal surface parameters defined according to ISO 25178–2. The movement of the lens on the two axes, “x” and “y,” in this case, determined the size of the scanned area, which had dimensions of 3 ×3 mm (Fig. 5 Right). In order to gain knowledge concerning the homogeneity of the structure, such measurements were repeated at five sample locations, and then their average calculated value was used for the evaluation. The overall analysis of the measured data was performed using Mountains software in accordance with the mentioned standards. 4. Results and discussion 3D surface roughness maps were created to compare the generated surface topography visually (Table. 4). The characteristics of the generated surfaces are further illustrated using the Abbott-Firestone curve. This curve is commonly and widely used to describe the behaviour of the bodies in contact with each other, for example, the amount of abrasion or the ability to retain lubricant. At the highest point of the surface, the relative amount of material is 0 %, and at the deepest point, the relative amount of material is 100 %. However, a curve alone is not Fig. 3. Methodology for the determination of the overlap factor (OF) from the groove width (w). G. Stol´ arik et al. Engineering Science and Technology, an International Journal 47 (2023) 101558 6 sufficient to describe the properties mentioned because one curve shape may correspond to surfaces with different properties. Furthermore, the Abbott-Firestone curve does not distinguish whether the surface has grooves in a particular direction or whether the surface roughness is completely random. The curve does not account for surface anisotropy. Therefore, for a proper and intensive understanding of the surfaces generated after the action of the surface treatment by PWJ, various selected profile and areal surface roughness parameters are used to describe the surface in a quantified form (Table 5). An evaluation of the quality of a machined or generated surface is crucial and requires the utilisation of a range of standardised R-parameters for determining the suitability of the material to be used for a particular application. Therefore, relying on a single parameter only provides a limited perspective of surface quality, potentially leading to an inaccurate assessment of a workpiece’s overall quality. The parameters chosen for monitoring should align with the control needs of the surface’s operational aspects. The frequently used roughness parameter, arithmetic mean height (Ra), has a limited predictive capability, necessitating the use of other profile height parameters. For instance, the maximum height, Rz, provides information on the mean value, from all section lengths, of the per section sum of the largest peak height and largest pit depth. The parameters Rp (mean peak height) and Rv (mean pit depth) indicate whether the surface structure has higher protrusions or depressions. Other height parameters include skewness, Rsk, and kurtosis, Rku. The skewness serves as a measure of the height values’ symmetry from the centre line, while the kurtosis offers a specific measure comparing the concentration of height values with the inflation of other values. The profile height parameter Rz has a higher predictive capability for assessing surface roughness mainly because it corresponds to local roughness and comprises two other separate parameters that can be used to constrain and isolate height and depth roughness. This parameter generally comprises the averages of the five highest peaks and five lowest valleys along the measured profile length. In this experiment, the sample was measured in five 3 ×3 mm areas (Fig. 5), and each separate area was divided into 11 measurement lines. From the values measured in this way, basic statistics were obtained, from which mean values were plotted versus the traverse speed for both trajectories (Fig. 6). The graphical waveform for the linear trajectory (Fig. 6a) shows a clear decreasing trend in the height of the Rz parameter, with a gradual increase in the traverse speed. The highest measured value was observed at a traverse speed of v =1 mm/s, with an average value of Rz =139.97 µm, where the quartile plot shows that this value is composed, for the most part, of values lower than the average calculated value (quartile shown in blue in the plot). A gradual decrease in Rz parameter values was observed as the traverse speed increased to v =2, 3, 4, and 5 mm/s with their average values of Rz =73.39, 50.81, 21.63, and 13.04 µm, respectively. In all cases, the quartile of lower measured values from the mean value was equally prevalent, except for the case of a traverse speed of v =3 mm/s. However, such a deviation can be considered less significant, as it did not significantly disturb the overall downward trend and can, therefore, be attributed to slight changes or inhomogeneities in the local material properties or the output values of the acoustic generator, which may have slightly disturbed the erosion rate. The same decreasing trend in Rz was also observed when the material surface was affected by the cross-hatch trajectory (Fig. 6b), with higher traverse speeds of v =2, 4, 6, 8, and 10 mm/s with their measured mean values of Rz =90.15, 26.91, 12.95, 12.46, and 10.56 µm, respectively. In this case, the representation of the quartiles indicates a high range of measured values higher than the mean value only in the case of traverse speeds of v =2 mm/s and 4 mm/s (shown in orange on the graph). The more pronounced quartile of higher measured values is probably attributable to the use of a different type of trajectory, where the material is already eroded in two directions. Overall, however, it can be noted that the parameter Rz for the cross-hatch trajectory is much more homogeneous, especially in terms of comparing the spacing of the measured values and, hence the representation of the deviation in the box plots (average of the measured values). At the same time, with a Fig. 4. Schematic of the influence of the (a) linear trajectory and (b) double trajectory on the sample surface. Table 3 Technological parameters of the experiment. Run p (MPa) f s (kHz) Lc (mm) d (mm) z (mm) v (mm/s) Trajectory i/mm/pass OF (mm) 1 70 40 12 0.8 43 1, 2, 3, 4, 5 Linear 40,000, 20,000, 13,333, 10,000, 8,000 0.45 2 2, 4, 6, 8, 10 Cross 20,000, 10,000, 6,667, 5,000, 4,000 G. Stol´ arik et al. Engineering Science and Technology, an International Journal 47 (2023) 101558 7 cross-hatch trajectory, it is also possible to see much lower measured maxima and minima compared to a linear trajectory. Hence, when subjected to a cross-hatch trajectory, the initial assessment involves anticipating a more uniform surface. The characteristics of the crosshatch trajectory can precisely explain the erosion pattern observed. In the first (longitudinal) pass, the sample’s erosion occurs, marked by the removal of certain particles (evidenced by erosion caused by a linear trajectory at a velocity of 2 mm/s). Subsequently, the second pass further impacts the material, resulting in increased erosion. However, the erosion depth does not increase due to the interaction of water clusters with the peaks formed after the initial pass. The force exerted by the water clusters was attenuated by these existing peaks, thereby lacking sufficient energy to remove additional material or create deeper valleys. Additionally, this phenomenon smoothed out sharp and weak peaks, resulting in a more even surface after the cross-transition. This effect is also reflected in the measurement of a smaller average Rz value and reduced variations and extremes in the measured values. The height parameter Rz is composed of two parameters, namely, the mean peak height, Rp, and the mean pit depth, Rv. The mean peak height, Rp, is the mean value from all section lengths of the largest peak height of each section length. The trend of this parameter is shown in (Fig. 7a,b). The course of the profile has the same decreasing tendency as the parameter Rz, which means that with an increasing traverse speed of v =1, 2, 3, 4, and 5 mm/s, lower values of the parameter were measured, such as Rp =64.26, 29.06, 18.93, 8.61, and 5.85 µm, respectively (Fig. 7a). As with Rz, the largest quartile differences were measured for the Rp parameter at the lowest traverse speed of 2 mm/s with the lower values predominating from the mean value. The same decreasing trend with an increasing traverse speed was also observed for the cross-hatch trajectory, where the measured values for traverse speeds of v =2, 4, 6, 8, and 10 mm/s were Rp =37.60, 8.97, 5.88, 5.95, and 4.50 µm, respectively (Fig. 7b). The lower measured mean values of the crosshatch trajectory compared to the linear trajectory is mainly attributed to the smoothing process of the peaks produced in the first pass by the subsequent second pass, as described above. The same decreasing trend with increasing traverse speed was also prevalent for the second evaluation term of the parameter Rz, namely, the deepest valley Rv (Fig. 7c, d). The linear trajectory with an increasing traverse speed of v =1, 2, 3, 4, and 5 mm/s produced average measured valley depths of Rv =75.71, 44.33, 31.88, 13.02, and 7.19 µm (Fig. 7c). Significantly lower measured values can also be observed due to the same surface smoothing for the cross-hatch trajectory for traverse speeds of v =2, 4, 6, 8, and 10 mm/s generating valley depths of Rv =52.54, 17.93, 7.07, 6.51, and 6.06 µm (Fig. 7d). However, comparing the mean peak height Rp and the mean pit depth Rv, an interesting result is that at each traverse speed and trajectory type used, the profile height is lower than the value of the profile depth. This trend may be due to the main nature of the surface formation by PWJ, which is the removal of material from the sample surface by the impact of water clusters. Water, as a tool, does not have a precise shape definition and impacts a certain area of material, and it results in the erosion of the material in the form of small fragments, which leads to the non-formation of high peaks. However, the surface is compressed and flat after the material fragment is torn off. Such compression and shaping of the material surface account for the increase in the subsurface hardness of the material, which has been confirmed in previous results Fig. 5. Schematic of the measured areas on the sample: (left) representation of the location of the lines for obtaining the profile parameters; (right) area for determining the areal parameters. G. Stol´ arik et al. Engineering Science and Technology, an International Journal 47 (2023) 101558 8 [34]. The results show that the subsurface hardness increased after the PWJs action for both linear and cross-hatch trajectories, which may also explain the lower erosive efficiency and material removal achieved in the current study using the cross-hatch trajectory. The first pass of the cross-hatch trajectory partially compresses the surface, improving its ability to resist further water impacts that are generated by the second pass. Another parameter considered in the evaluation of the profiles is the skewness, referred to as Rsk. This profile is divided into two basic groups: positive and negative. The essence of the evaluation is to measure the symmetry of the profile about the center line. A profile with flattened peaks and deep valleys has a negative Rsk value; conversely, a profile with high peaks and filled valleys has a positive value. A prediction for the evaluation of Rsk can also be made from the previous comparison of Rp and Rv, where it was observed that the valleys of the profile parameter Rv dominate over the peaks of Rp. The specific measurement results confirmed this prediction, which is also shown in Fig. 8a,b. Negative values of Rsk were measured for both the linear and cross-hatch trajectories, indicating that the resulting surface had fewer or truncated peaks and more or deeper valleys. A perfectly symmetric surface or a surface having an equal influence of peaks and valleys has an Rsk value equal to 0. In this study, after the linear trajectory, the measured values were Rsk =-0.34, 1.12, −1.15, −0.96, and −0.34 for traverse speeds of v =1, 2, 3, 4, and 5 mm/s (Fig. 8a) and for the crosshatch trajectory Rsk =-0.60, −1.43, −0.33, −0.28, and −0.49 for traverse speeds of v =2, 4, 6, 8, and 10 mm/s (Fig. 8b). These measured Table 4 Summary of 3D surface roughness maps and Abbott firestone material ratio curve for all the examined surfaces generated after the action of PWJ. (continued on next page) G. Stol´ arik et al. Engineering Science and Technology, an International Journal 47 (2023) 101558 9 values show that in both cases of linear and cross-hatch trajectories, there was a negative character formation with valleys dominating. The character of such a surface can be well exploited in practice, mainly because less presence of brittle and high peaks does not weaken the product, which provides a small initial contact area generating high contact stress areas, but the greater part is made up of bulkier shapes with occasional valleys. Therefore, higher loads are needed to damage such a surface, delivering better wear and abrasion resistance. In addition to the Rsk parameter, the kurtosis, Rku, was also used to evaluate the probability density of the peaked profile. The parameter Rku describes the sharpness of the formed surface, and its density can be evaluated into two groups. A Rku <3 means that the profile is formed with fewer peaks but with a higher number of valleys and is called platykurtic. The other is Rku ˃ 3, when the profile is made up of a higher density of peaks for the most part and few low valleys and is called leptokurtic. The results of the Rku parameter measurements (Fig. 8) show that both linear and cross-hatch trajectories resulted in a leptokurtic profile. The measured values for the linear trajectory were Rku =3.16, 5.77, 7.09, 6.93, and 4.52 for traverse speeds of v =1, 2, 3, 4, and 5 mm/s (Fig. 9a) and for the cross-hatch trajectory were Rku =3.52, 7.57, 4.45, 5.09, and 5.02 for traverse speeds of v =2, 4, 6, 8, and 10 mm/s (Fig. 9b). In practice, the surface formed can be used mainly in applications where the application of necessary lubricants is required, since more frequent gaps are created and, thus, the applied lubricant can be applied in a higher volumetric amount. In conclusion, after combining the evaluation of the Rku and Rsk roughness parameters, it can be concluded that this method of surface treatment using PWJ can generate a surface profile that is sufficiently strong (tendency to abrasion resistance) but, at the same time, sufficiently porous (dense distribution of valleys), which can be used for further applications in the field of biomedicine, where high demands are placed on the surface of the implants, as they highly influence the rate of fixation and the lifetime of Table 4 (continued) G. Stol´ arik et al. Engineering Science and Technology, an International Journal 47 (2023) 101558 16 Fig. 17. Influence of the traverse speed on Spk for (a) linear trajectory and (b) cross-hatch trajectory; on Sk for (c) linear trajectory and (d) cross-hatch trajectory. Fig. 18. Graphical parameter progression of Svk versus traverse speed: (a) linear trajectory; (b) cross-hatch trajectory. Table 7 Average measured functional and hybrid areal surface roughness parameters with measurement uncertainty. Linear Trajectory Cross-Hatch Trajectory 1 mm/s 2 mm/s 3 mm/s 4 mm/s 5 mm/s 2 mm/s 4 mm/s 6 mm/s 8 mm/s 10 mm/s Sdr (%) 65.90 ±1.70 31.02 ±0.81 22.09 ±0.44 12.71 ±0.38 11.68 ±0.21 37.29 ±2.96 11.48 ±0.43 10.46 ±0.08 9.42 ±0.11 8.11 ±0.25 Sk (µm) 70.94 ±5.02 19.28 ±1.15 11.47 ±0.22 5.25 ±0.09 4.05 ±0.03 39.65 ±5.38 6.25 ±0.31 4.12 ±0.03 3.71 ±0.04 3.08 ±0.04 Spk (µm) 20.18 ±0.79 10.94 ±0.50 6.30 ±0.27 2.89 ±0.22 1.77 ±0.04 17.82 ±2.31 3.07 ±0.28 1.78 ±0.01 1.50 ±0.02 1.28 ±0.01 Svk (µm) 50.03 ±4.46 26.93 ±1.92 19.09 ±1.65 5.58 ±0.27 2.61 ±0.19 42.81 ±4.11 8.09 ±0.87 2.69 ±0.07 2.20 ±0.06 1.97 ±0.05 Table 8 Average surface isotropy of the surfaces with measurement uncertainty after the action of PWJ with both trajectory and traverse speeds. Linear Trajectory Cross-Hatch Trajectory 1 mm/s 2 mm/s 3 mm/s 4 mm/s 5 mm/s 2 mm/s 4 mm/s 6 mm/s 8 mm/s 10 mm/s Isotropy (%) 72.83 ± 1.91 78.81 ± 2.68 82.06 ± 0.89 85.23 ± 1.30 88.26 ± 0.95 83.58 ± 1.15 88.14 ± 1.09 90.35 ± 1.58 92.59 ± 2.35 96.74 ± 0.61 G. Stol´ arik et al. Engineering Science and Technology, an International Journal 47 (2023) 101558 17 results than profile roughness measurement due to its more accurate assessment of the overall roughness characteristics, including variations in height, spacing and spatial distribution of the features compared to profile measurements, which capture information only along a single line or path and does not fully represent the overall roughness. 3. The overall surface roughness values showed a decreasing trend with increasing traverse speed using both trajectories within the selected experimental domain. This behaviour of the roughness values with increased traverse speed is due to the lower water droplet interaction density over a given surface area (Ra =22.57 µm, Sa =25.56 µm and Ra =1.37 µm, Sa =1.36 µm for v =1 mm/s and 5 mm/s, respectively for linear hatch trajectory). 4. Different surface roughness parameters values measured, such as Ssk, Sku, Sdr, Sk and Isotropy, advocated the potential of using PWJ for the preparation of surfaces required for various applications involving wear, fatigue, corrosion resistance, and biocompatibility using both trajectory options. 5. The surface morphology study showed common hydrodynamic erosion surface characteristics such as pits, microchannels, cracks, and surface upheavement distributed evenly throughout the treated surface. The present study showed the potential of PWJ to be used as a tool in future by controlling the water hammer droplet phenomenon for various surface preparation and treatment applications supported by surface roughness measurement values which are considered benchmarks for accessing the quality of the generated surfaces. However, further optimization of different trajectories and technological conditions, along with wear, wettability and cytocompatibility tests on the generated surfaces, will be carried out in future studies for better insights and its correlation with the interaction of the water droplets and the material surfaces. Funding This work was supported by 01/TUKE/2023. This work was supported by the Slovak Research and Development Agency under contract no. APVV-22-0391 and the Scientific Grant Agency VEGA 1/0377/22. The experiments were carried out at the Institute of Geonics with the support of the Institute of Clean Technologies for Mining and Utilization of Raw Materials for Energy Use—Sustainability Program, Reg. No. LO1406, financed by the Ministry of Education, Youth and Sports of the Czech Republic, with support for long-term conceptual development from the research institution RVO: 68145535. Fig. 19. Surface irregularities isotropy: (a) minimal measured value for linear strategy v =1 mm/s; (b) maximum measured value for cross-hatch strategy v =10 mm/s. Fig. 20. Graphical progression of Isotropy versus traverse speed for a) linear trajectory and b) cross hatch trajectory. G. Stol´ arik et al. Engineering Science and Technology, an International Journal 47 (2023) 101558 18 Fig. 21. SEM images of areas treated with PWJ at a decreasing density of droplets per area, comparing two trajectory strategies (linear hatch and cross-hatch). G. Stol´ arik et al. Engineering Science and Technology, an International Journal 47 (2023) 101558 19 Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. References [1] E.S. Gadelmawla, M.M. Koura, T.M.A. Maksoud, I.M. Elewa, H.H. Soliman, Roughness parameters, J. Mater. Process. Technol. 123 (1) (2002) 133–145. [2] M. Lieblich, S. Barriuso, J. Ib´ a˜ nez, L. Ruiz-de-Lara, M. Díaz, J.L. Oca˜ na, A. Alberdi, J.L. Gonz´ alez-Carrasco, On the fatigue behavior of medical Ti6Al4V roughened by grit blasting and abrasiveless waterjet peening, J. Mech. Behav. Biomed. Mater. 63 (2016) 390–398. [3] A. Perec, F. Pude, M. Kaufeld, K. Wegener, Obtaining the selected surface roughness by means of mathematical model based parameter optimization in abrasive waterjet cutting, Strojniˇ ski Vestnik-J. Mech. Eng. 63 (2017) 606–613. [4] H.S. Kirols, D. Kevorkov, A. Uihlein, M. Medraj, The effect of initial surface roughness on water droplet erosion behaviour, Wear 342–343 (2015) 198–209, https://doi.org/10.1016/j.wear.2015.08.019. [5] A. Perec, Desirability function analysis (DFA) in multiple responses optimization of abrasive water jet cutting process, Rep. Mech. Eng. 3 (1) (2022) 11–19. [6] M. Uddin, A. Basak, A. Pramanik, S. Singh, G.M. Krolczyk, C. Prakash, Evaluating hole quality in drilling of Al 6061 alloys, Materials 11 (2018) 2443. [7] G. Krolczyk, P. Nieslony, S. Legutko, I. Samardzic, Investigation of the physical parameters of duplex stainless steel (Dss) surface integrity after turning, Metalurgija. 54 (2015) 87–90. [8] J. Poloprudský, A. Nag, T. Kruml, S. Hloch, Effects of liquid droplet volume and impact frequency on the integrity of Al alloy AW2014 exposed to subsonic speeds of pulsating water jets, Wear 488-489 (2022) 204136. [9] R. Williams, Surface modification of biomaterials: Methods analysis and applications, Elsevier, 2010. [10] X. Cheng, J.W. Fisher, H.J. Prask, T. Gn¨ aupel-Herold, B.T. Yen, S. Roy, Residual stress modification by post-weld treatment and its beneficial effect on fatigue strength of welded structures, Int. J. Fatigue 25 (2003) 1259–1269. [11] L. Shen, L. Wang, Y. Wang, C. Wang, Plasma nitriding of AISI 304 austenitic stainless steel with pre-shot peening, Surf. Coat. Technol. 204 (20) (2010) 3222–3227. [12] V.K. Jain, A. Sidpara, R. Balasubramaniam, G.S. Lodha, V.P. Dhamgaye, R. Shukla, Micromanufacturing: a review—part I, Proc. Inst. Mech. Eng. B J. Eng. Manuf. 228 (9) (2014) 973–994. [13] J. Havlikova, J. Strasky, M. Vandrovcova, P. Harcuba, M. Mhaede, M. Janecek, L. Bacakova, Innovative surface modification of Ti-6Al-4V alloy with a positive effect on osteoblast proliferation and fatigue performance, Mater. Sci. Eng. c. 39 (2014) 371–379. [14] A. Bharatish, H.N.N. Murthy, B. Anand, C.D. Madhusoodana, G.S. Praveena, M. Krishna, Characterization of hole circularity and heat affected zone in pulsed CO2 laser drilling of alumina ceramics, Opt. Laser Technol. 53 (2013) 22–32. [15] D.K. Gupta, D.S. Duvall, A silicon and hafnium modified plasma sprayed MCrAlY coating, Superalloys 1984 (1984) 711. [16] H. Fu, Y. Liang, Study of the surface integrity and high cycle fatigue performance of AISI 4340 steel after composite surface modification, Metals (Basel). 9 (2019) 856. [17] A. Momber, Hydrodemolition of concrete surfaces and reinforced concrete, Elsevier, 2011. [18] A. Perec A. Radomska-Zalas A. Fajdek-Bieda, Experimental research into marble cutting by abrasive water jet, Facta Universitatis FU Mech Eng 20 1 145. [19] D. Arola, A.E. Alade, W. Weber, Improving fatigue strength of metals using abrasive waterjet peening, Mach. Sci. Technol. 10 (2006) 197–218, https://doi. org/10.1080/10910340600710105. [20] H. Yang, X. Wu, G. Cao, Z. Yang, Enhanced boronizing kinetics and high temperature wear resistance of H13 steel with boriding treatment assisted by air blast shot peening, Surf. Coat. Technol. 307 (2016) 506–516, https://doi.org/ 10.1016/j.surfcoat.2016.09.029. [21] M.M. Vijay, J. Foldyna, J. Remisz, Ultrasonic modulation of high-speed water jets, 1994. [22] J.E. Field, ELSI conference: Invited lecture liquid impact: Theory, experiment, applications, in: Wear, 1999: pp. 1–12. https://doi.org/10.1016/S0043-1648(99) 00189-1. [23] M.Y. Amegadzie, E.D. Moreau, B. Christensen, I.W. Donaldson, A. Tieu, M.M. Vijay, K.P. Plucknett, Ultrasonic pulsed waterjet surface peening of an industrial aluminum-based metal matrix composite, Surf. Coat. Technol. 426 (2021), 127795, https://doi.org/10.1016/j.surfcoat.2021.127795. [24] C. Yang, S. Hou, J. Xu, Y. Zhang, Y. Zheng, E. Fernandez-Rodriguez, D. Zhou, Multicomponentwater effects on rotating machines disk erosion, Water (Switzerland). 12 (3) (2020) 757. [25] D. Li, Y. Kang, X. Ding, X. Wang, Z. Fang, Effects of area discontinuity at nozzle inlet on the characteristics of self-resonating cavitating waterjet, Chinese J. Mech. Eng. (English Edition) 29 (4) (2016) 813–824. [26] M. Zeleˇ n´ ak, Z. ˇ Ríha, P. Jandaˇ cka, Visualization and velocity analysis of a highspeed modulated water jet generated by a hydrodynamic nozzle, Measurement (lond). 159 (2020) 107753. [27] J. Foldyna, L. Sitek, B. ˇ Svehla, ˇ S. ˇ Svehla, Utilization of ultrasound to enhance highspeed water jet effects, Ultrasonics Sonochem. 11 (3-4) (2004) 131–137. [28] J. Poloprudský, A. Chlupov´ a, I. ˇ Sul´ ak, T. Kruml, S. Hloch, Surface and subsurface analysis of stainless steel and titanium alloys exposed to ultrasonic pulsating water jet, Materials. 14 (2021) 5212. [29] A. Nag, P. Hvizdos, A.R. Dixit, J. Petrů, S. Hloch, Influence of the frequency and flow rate of a pulsating water jet on the wear damage of tantalum, Wear 477 (2021) 203893. [30] P. Siahpour, M.Y. Amegadzie, A. Tieu, I.W. Donaldson, K.P. Plucknett, Ultrasonic pulsed waterjet peening of commercially-pure titanium, Surf. Coat. Technol. (2023) 129953. [31] F. Heymann, On the time dependence of the rate of erosion due to impingement or cavitation, Erosion by Cavitation or Impingement. (1967). [32] M. Srivastava, S. Hloch, L. Krejci, S. Chattopadhyaya, N. Gubeljak, M. Milkovic, Utilizing the water hammer effect to enhance the mechanical properties of AISI 304 welded joints, Int. J. Adv. Manuf. Technol. 119 (3-4) (2022) 2317–2328. [33] M. Srivastava, S. Hloch, L. Krejci, S. Chattopadhyaya, A.R. Dixit, J. Foldyna, Residual stress and surface properties of stainless steel welded joints induced by ultrasonic pulsed water jet peening, Measurement (lond). 127 (2018) 453–462, https://doi.org/10.1016/j.measurement.2018.06.012. [34] G. Stol´ arik, A. Nag, J. Petrů, J. Svobodov´ a, S. Hloch, Ultrasonic pulsating water jet peening: influence of pressure and pattern strategy, Materials 14 (2021) 6019. [35] G. Stol´ arik, J. Svobodov´ a, D. Klichov´ a, A. Nag, S. Hloch, Titanium surface roughening with ultrasonic pulsating water jet, J. Manuf. Process. 90 (2023) 341–356. [36] S. Hloch, K. Souˇ cek, J. Svobodov´ a, M. Hromasov´ a, M. Müller, Subsurface microtunneling in ductile material caused by multiple droplet impingement at subsonic speeds, Wear. 490–491 (2022) 204176. https://doi.org/10.1016/j. wear.2021.204176. [37] J.F. Senge, A.H. Astaraee, P. Dłotko, S. Bagherifard, W.A. Bosbach, Extending conventional surface roughness ISO parameters using topological data analysis for shot peened surfaces, Sci. Rep. 12 (2022) 5538. [38] D. Klichova, A. Nag, J. Poloprudský, J. Foldyna, F. Pude, L. Sitek, S. Hloch, Utilising of water hammer effect for surface roughening of Ti6Al4V, Int. J. Adv. Manuf. Technol. 126 (11-12) (2023) 5633–5647. [39] D. Lehocka, F. Botko, J. Klich, L. Sitek, P. Hvizdos, M. Fides, R. Cep, Effect of pulsating water jet disintegration on hardness and elasticity modulus of austenitic stainless steel AISI 304L, Int. J. Adv. Manuf. Technol. 107 (5-6) (2020) 2719–2730. [40] A. Nag, S. Hloch, A.R. Dixit, F. Pude, Utilization of ultrasonically forced pulsating water jet decaying for bone cement removal, Int. J. Adv. Manuf. Technol. 110 (3-4) (2020) 829–840. [41] A. Nag, G. Stol´ arik, B. Svehla, S. Hloch, Effect of Water Flow Rate on Operating Frequency and Power During Acoustic Chamber Tuning, in: Advances in Manufacturing Engineering and Materials II: Proceedings of the International Conference on Manufacturing Engineering and Materials (ICMEM 2020), 21–25 June, 2021, Nový Smokovec, Slovakia, Springer, 2021: pp. 142–154. [42] B. He, S. Ding, Z. Shi, A comparison between profile and areal surface roughness parameters, Metrol. Measure. Syst. 28 (2021) 413–438. G. Stol´ arik et al.