1 Introduction

The surface morphology of materials serves as a fundamental functional attribute governing their practical utilisation [1]. Therefore, standards are set regarding the quality and integrity of material surfaces, which significantly influence the overall performance of any given component [2]. Increasing surface roughness directly increases the surface area in contact with the ambient environment. In terms of absorption properties, rougher surfaces are preferred over smoother ones [3]. For example, when external agents, such as lubricants or protective coatings, are applied onto a target surface of a component to prolong its life, the protective layer should be deposited on a roughened surface to ensure better adhesion. To facilitate the acceptance of varnish or a certain thickness of the lubricant, achieving optimal roughness is necessary. In the case of lubrication, the optimum roughness provides a balance between the increase in friction due to greater surface roughness and the reduction in friction due to the presence of lubricant. Different values of optimal roughness are, nevertheless, considered in the case of cyclic loading of components, where the smoother the surface, the longer the fatigue life [4]. Obtaining optimal surface roughness is also important in the biomedical industry due to its impact on the integration of implants with bone tissue [5]. Increased osseointegration can be facilitated by intentional roughening of the implant because bone cells grow preferentially on roughened surfaces [6]. To achieve functionality improvement in various applications is necessary to reach different levels of optimal surface properties. There is a wide spectrum of methods for surface modifications to achieve the optimum topography and roughness for each application. Usually, when a certain roughness is required, technologies such as shot peening, laser beam, or plasma spraying are used [7]. Disadvantages of these procedures are the presence of a heat-affected area in laser peening [8] and plasma spraying [9], and the embedding of foreign particles in the form of abrasive particles into the surface during sandblasting [10].

To avoid the mentioned drawbacks, a potential solution can be found in the modification of material surfaces by pure water as a tool when the complete reduction in the heat-affected area in the machining zone is reached providing specific topography and changes in subsurface properties [11]. The advantage is also the possibility of machining a wide range of materials such as metals [12], composites [13, 14] as well as difficult-to-machine ceramics [15]. Water jet treatment offers the power of a high-velocity water jet (WJ) [16] acting with a stagnation pressure (ps). After passing through the nozzle, the accumulated high-pressure liquid is transformed into a high-velocity jet that acts on the target material. The result is erosive wear [17] of the material in the form of removal of small material particles; the effectiveness of this technique depends on the material properties [18], but especially on the hydraulic parameters of the liquid and the cutting speed [19].

The erosive impact of WJ can be further intensified. One of the ways to achieve this is the addition of abrasives in the high-velocity jet, i.e. abrasive water jet (AWJ) [20]. The abrasive particles are accelerated by the velocity of the water stream [21], which increases their kinetic energy upon impact with the target material. The material removal mechanism is then defined by combining the cutting mechanisms [22], fatigue, brittle fracture, and erosive wear [23]. However, the drawback is again the deposition of abrasive artefacts, which can remain stuck in the surface [24]; therefore, its additional cleaning is necessary [25].

The other way to increase the efficiency of water jetting while preventing surface contamination by particles is the application of impact pressure (pi) on abrasive-less water jets. Liquid shock defined by Cook [26] consists of two phases. In the first phase, the liquid behaves in a compressible manner. It evokes pressure on the target surface, the magnitude of which depends on its hydraulic properties and the elastic properties of the material [27]. The impact between a liquid and a solid surface causes a gradual increase in the contact periphery and internal compression of the droplet, up to the point where its surface tension is incapable of maintaining the integral shape. At this point, the second phase of rupture of the outer layer of the water droplet occurs followed by the transformation of the accumulated pressure into an incompressible high-speed stream of water flowing around the entire perimeter of the contact surface [28]. Under the same hydraulic conditions, the generated impact pressure causes higher surface damage than stagnation pressure [29]. For that reason, several technological concepts of generating repetitive impact pressure have been proposed in the form of interruption of the constant flow of water with a rotating disc [30] or self-resonating fluid oscillators [31]. Although these technologies offer the generation of impact pressure, they also contain disadvantages such as the technological complexity of the fabrication of self-acting and fluid oscillators or wear of the mechanical parts of the rotating disc [32].

The viable and competitive approach to creating an interrupted water jet providing individual water clusters is the pulsating water jet (PWJ) technology [33, 34]. It presents an option due to its decreased operational expenses, decreasing the production of reduced operating pressures, particularly in comparison with continuous water jet (WJ) [35] and abrasive water jet [36] (AWJ) methods [37], thereby necessitating lesser pump power [38]. The PWJ technique involves the acoustic assembly to introduce the high-pressure fluctuations into pressurised water jets in the high-pressure acoustic chamber [39]. As the water jet exits the nozzle at high speed, it steadily transitions into distinct groups of water droplets [40].

The repetitive impingement of water droplet on the target surface causes the cyclic loading that modifies surface and subsurface layers of material [41] Time for stress relaxation of the material is reduced, compressive stresses are induced and effects such as plastic deformation, recrystallisation, grain refinement and phase transformation happen [42]. Finally, it leads to failure of the material integrity in the form of roughening and the erosive removal of material [43]. In addition to surface erosion, subsurface micro-tunnelling processes can occur, which increase the depth range of erosion [44]. The creation of surface irregularities depends on the hydraulic conditions of the droplets.

PWJ technology offers the flexibility to fine-tune the water droplets hydrodynamic properties simply by varying the nozzle diameter or modifying the supply pressure [33] and flow rate [45]. The frequency at which water droplets strike affects their ability to erode surfaces, resulting in various stages of erosion [46]. These essential stages, initiated by the impact of liquid droplets, encompass the phases of incubation, acceleration, culmination, depletion, and termination [47]. Using these stages, it is possible to determine the specific erosion phase addressed in various studies of the influence of PWJ process parameters on the target material.

In the work of Srivastava [48], during the incubation phase, adjustments in supply pressure and traverse speed were employed to alleviate subsurface stresses, leading to alterations in the surface characteristics of the AISI 304 austenitic steel welded joint samples.

To modify the surface topography and subsurface properties the end of the incubation stage or acceleration erosion phases are usually used [49], where the gradual removal of the material surface layers starts. The modifications in surface topography can be carried out not only by changing the nozzle shape, i.e. using a circular or flat nozzle, but also by tuning the other technological parameters such as traverse speed [50], and supply pressure [51].

The variation in the traverse speed and supply pressure can, moreover, affect subsurface properties, i.e. increase the microhardness, particularly in the subsurface regions close to the impacted surface [52]. These findings have also been used in the targeted modification of the surface structure of titanium [53] and stainless steel [54]. The results showed that the surface roughness depends predominantly on the variation in the number of water droplet impacts per unit length (IPUL) hitting the material surface. IPUL is determined mainly by the ultrasonic frequency, the trajectory pattern, and the traverse speed.

The results achieved in studies conducted so far were carried out under atmospheric conditions. Despite achieving satisfactory results, variations in surface topography [55] were commonly observed across a large area of the treated surface, indicating non-uniformity and anisotropy. A submerged approach is currently investigated for drilling [56]. This technique leverages the buffering effect of the surrounding water body to soften the impact of the incoming stream of water droplets, thereby diminishing the lateral jetting effect and yielding a gentler treatment that produces less distinct erosion marks. This approach can be used to address this issue, involving submerged [57] conditions during PWJ treatment, when the surrounding water column cushions the incoming water droplet stream, reducing the lateral jetting phenomenon and resulting in softer treatment with less pronounced erosion marks [58]. It leads to the hypothesis that large-area treatment will exhibit more uniform hardness and roughness parameters. Contrary to atmospheric conditions, also cavitation bubbles [59] may influence the erosion process under submerged conditions.

The impact of water pulses at PWJ technology is a competitive option for modifying the surfaces of materials [53]. However, up to date, only a few studies for example [60] has addressed the possibility of modifying the target material using a combination of the PWJ method and submerged conditions under hydrostatic conditions of a liquid medium. According to a study by He [61], the water stream is formed gradually at a certain distance after exiting the nozzle. Since the water stream is exposed to the resistance of the surrounding environment, it causes the formation of small, separated water bubbles on its circumference. A further increase in the distance causes the complete breakdown of the central region of the steady water flow and it leads to the formation of only separated bubbles. In another study [62], a comprehensive analysis of submerged water flow is presented that details the basic results of the change in the pressure fluctuations in the water stream after exiting the nozzle when the operating pressure is changed. At the same time, the fluctuating pressure is used to drive the surrounding static water. It results in significant differences in water velocities and the shear effect causing the formation of acting cavitation bubbles. A review study dealing with cavitation [63] showed the use of hydraulic cavitation to peen structural parts at nuclear power plants, which improved fatigue strength and modified the tribological properties of structural parts.

Recent results of submerged water jets are also noted in the field of AWJ and surface treatment of titanium alloys. In Ref. [64], the surface was modified using a micro-AWJ under submerged conditions with a water level depth of 60 mm. The results showed a certain potential for changes in surface and subsurface properties with a change in supplied pressure. Another example of submerged AWJ and an immersion depth of 50 mm were used in Ref. [65] to improve the erosion resistance and achieve the change in surface properties for an operating pressure of 180 MPa, a constant traverse speed of 5 mm/s, and an overlap factor of 0.05 mm. Induced residual compressive stresses to a depth of 140 to 200 µm have been defined as a critical parameter for improving the integrity and durability of aircraft engine blades. Submerged abrasive peening combined with a cavitation process was introduced in Ref. [66]. Cavitation abrasive integrated water jet technology was used to modify the surface roughness and increase the subsurface microhardness. All these works demonstrate the importance of the surrounding environment, where the immersion of a high-speed water jet affects the material modification process.

This work is focussed on determining the possibility of modifying surfaces by applying PWJ under submerged conditions. A comparison of the resulting surface with surfaces created under the same water jet conditions acting under an atmospheric environment was performed. The effect of PWJ technology on the surface treatment of AISI 304L material was observed using a supply pressure of p = 70 MPa and a nozzle diameter of d = 0.8 mm with an excitation frequency of fs = 20 kHz. To attain the desired surface coverage, two trajectory patterns were chosen, at traverse speeds of v = 1, 2, 3, 4, and 5 mm/s for linear trajectory and v = 2, 4, 6, 8, and 10 mm/s, for perpendicular interleaved trajectory. These combinations of trajectories and speeds ensure an equivalent number of water cluster impacts per unit length (IPUL) of material, facilitating the comparison of conditions and trajectory types on the resulting surface. Two trajectories and their respective traverse speeds were employed to alter the surface under atmospheric or submerged conditions. Under submerged conditions, a surface water height of wd = 150 mm from the target material's surface was selected. The generated samples were then analysed using the areal surface roughness height parameters Sa, Sz, Sp, and Sv. Scanning electron microscopy (SEM) was used for surface quality evaluation and morphology assessment. The performed analyses allowed us to evaluate the suitability of the treated surface with a specific roughness for potential engineering and biomedical applications.

2 Materials and methods

2.1 Material

AISI grade 304L stainless steel was chosen as the experimental material for the surface treatment using the PWJ method. AISI 304L is a chromium–nickel-based austenitic steel selected providing a good combination of corrosion resistance and plasticity. The material was received as a 5 mm thick plate having dimensions of 150 × 1500 mm, from which samples measuring 20 × 200 mm were cut. A comprehensive overview of the mechanical properties and chemical composition of the experimental material is presented in Table 1 [67].

Table 1 Mechanical properties and chemical composition of AISI 304L [67]
Full size table

2.2 Experiment

The surface treatment was carried out using a PWJ technology, i.e. a hybrid ultrasonically excited, designed and manufactured at the Institute of Geonics of the Czech Academy of Sciences in Ostrava. Samples of the AISI 304L material used in this study were fastened on the clamping table to avoid any undesirable movement during the experiment. A surface area of 20 × 30 mm was treated using the PWJ method with a water supply pressure of p = 70 MPa generated by a Hammelmann HDP 253 high-pressure pump with a maximum operating pressure of 160 MPa. The high-pressure water was fed into an acoustic chamber, where pressure fluctuations were induced into the water using an oscillating sonotrode. The oscillating motion of the sonotrode was caused by its mechanical connection with vibrating piezoelectric ceramics, which converted the electrical harmonic signal generated from the Ecoson WJ UG 630-40 acoustic generator into mechanical motion [68]. The effect of the conversion of the initial generated signal frequency fs = 20 kHz is influenced by the resistance of the sonotrode to the water pressure in the acoustic chamber, which affects the impedance of the circuit and brings the technology into a resonant frequency [69]. Therefore, the chamber was designed with the ability to change its volume continuously by varying the acoustic chamber length, lc from 0 to 24 mm. The water under high pressure passes from the acoustic chamber through a nozzle exit of diameter d = 0.8 mm. After exiting the nozzle, it is transformed into a high-velocity water jet where the induced velocity fluctuations cause a gradual breakdown of the steady water stream into individual water clusters, which subsequently impact repetitively the surface of the AISI 304L stainless steel material. The distance between the nozzle outlet and the material surface is standoff distance z. To ensure sufficient time for the transformation of the CWJ into PWJ, i.e. into discrete water clusters, and to exploit the maximum erosion potential of the impact pressure [70], it is required to determine the optimum value of z. It was done by conducting pilot erosion tests, with a staircase trajectory movement, a constant traverse speed and a uniform increment in the standoff distance of 1 mm after traversing 10 mm in the longitudinal direction similarly as in previous work [71]. The optimum standoff distance corresponds to the widest groove reached. The selected technological parameters for the PWJ experiments are shown in Table 2.

Table 2 Technological parameters for the experiment
Full size table

The main experiments consisted of surface treatment by PWJ in two types of surrounding medium (see Fig. 1):

  1. (1)

    air with an atmospheric pressure of pa = 0.101 MPa and

  2. (2)

    submerged with a water level height of wd = 150 mm and a density of 998 kg/m3, creating a hydrostatic pressure of ph = 0.001467 MPa. The absolute pressure of the submerged condition is the sum of the atmospheric and hydrostatic pressure values, which is pr = 0.1025 MPa.

Fig. 1
Fig. 1
Full size image

Experimental setup of the PWJ head assembly under a atmospheric and b submerged conditions with a water level height of wd = 150 mm

Two types of surface treatment trajectories were applied (see Fig. 2) with 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 perpendicular interleaved trajectory pattern. The trajectory path, i.e. movement of PWJ head, was controlled by an ABB IRB 6640-180 robotic arm. In the linear trajectory, the PWJ head moves across the material surface only once in one direction. The perpendicular interleaved implies the formation of two mutually perpendicular passes. To ensure the same number of water cluster impacts per 1 mm of material (IPUL), the twofold traverse speed of the perpendicular interleaved pattern was set per one jet pass.

Fig. 2
Fig. 2
Full size image

Surface modification using a PWJ technology, under b atmospheric and c submerged conditions using d linear and e perpendicular interleaved trajectory over the surface of the target material

Both trajectory types require further specification and tuning of overlapping factor (OF), i.e. parameter describing the distances of the trajectory axis of individual passes of the PWJ head. Too large OF parameter leads to the formation of an untreated or undertreated surface, while too small overlapping makes the process time-consuming. OF optimisation is based on the erosion effect assessment on 20 × 20 mm surfaces on a sample treated by traverse speed 1 mm/s with overlapping reaching 100%, 75%, 50% and 25% of the original measured width of erosion groove at optimal standoff distance. A groove produced in the atmospheric conditions exhibited a width of 1.8 mm, and submerged conditions produced a groove width of 1 mm. The most optimal surface coverage without untreated lines was created for 25% overlapping for both used mediums, i.e. OF = 0.45 mm at atmospheric and 0.25 mm at submerged conditions (Fig. 2).

2.3 Roughness measurement

The modified surfaces of the samples were scanned by a non-contact method using a MicroProf FRT optical profilometer. The device uses white light separated into monochromatic components and focussed on the sample surface. The reflected light is detected by the spectrometer and converted to elevation data via calibration. The profilometer allows both profile measurement and scanning of the sample area. The sample surfaces were analysed using areal surface roughness parameters, where the scanned surface after application of S-filter, F-filter and L-filter provided an S-L surface on which the areal surface roughness height parameters, i.e. arithmetic mean height Sa, maximum height Sz, maximum peak height Sp, and maximum valley depth Sv defined in ISO 25178-2 [72] were analysed, see Fig. 3b). To obtain information on the homogeneity of the surface, such measurements were repeated at three different sample locations (Fig. 3a) where the results of the measurements were then averaged to produce graphs in the form of box plots showing quartiles and minimum and maximum values. The average measured values and the standard deviation were also numerically presented in the tables of the corresponding parameters in the results and discussion section. For comparison, the original untreated surface was optically analysed (by the method described in Topography observations) (Fig. 3c). Its original morphological properties averaged from the three sample location were also measured (Fig. 3d).

Fig. 3
Fig. 3
Full size image

a Schematic representation of the measured areas of the sample, b presented height parameters of surface roughness Sa, Sz, Sp and Sv, c optical analysis and d morphological parameters of the untreated surface

2.4 Topography observation

The PWJ-treated surfaces in both atmospheric and submerged conditions using linear and perpendicular interleaved trajectories were analysed after drying of all samples using a KEYENCE VHX-5100 digital microscope. A Tescan LYRA 3 XMU FEG/SEMxFIB scanning electron microscope (SEM) was used to observe topographic features and erosion mechanisms. The micrographs were taken with an accelerating voltage of 10 kV using a secondary electron detector to focus on the topographical information of the PWJ-treated surface.

3 Results and discussion

Photographs of the modified AISI 304L surfaces after both atmospheric and submerged conditions are shown in Fig. 4. In order to preserve the possibility of comparing the treated surfaces in atmospheric and submerged conditions without the influence of the surrounding medium, optical analyses were processed after the PWJ treatment process on the dried target sample. Visual assessment of surfaces shows that submerged conditions result in significantly lower surface roughness in both linear and perpendicular interleaved trajectories compared to atmospheric conditions. A comparison of two trajectories also shows a different effect on the resulting surface when a linear trajectory causes slightly higher surface roughness than perpendicular interleaved. A detailed analysis of surface topography and measurement of values of areal roughness parameters (Sa, Sz, Sv and Sp) were performed to quantify the surface quality and to compare the effectiveness of the experimental setup of the PWJ technology in atmospheric and submerged conditions. Results are discussed in the following paragraphs.

Fig. 4
Fig. 4
Full size image

Optical analysis of created surfaces of linear and perpendicular interleaved trajectories under atmospheric and submerged conditions

3.1 Evaluation of surface roughness parameters S a and S z

The results of the surface roughness parameters Sa and Sz for topography generated in atmospheric and submerged conditions using both linear and perpendicular interleaved trajectory with evaluated statistical deviation in measurement are listed in Table 3.

Table 3 Results of average areal surface roughness parameters Sa and Sz of linear and perpendicular interleaved trajectories in atmospheric and submerged conditions obtained from three sample measurements
Full size table

Using a low traverse speed of 1 and 2 mm/s for the linear trajectory and 2 and 4 mm/s for the perpendicular interleaved trajectory, corresponding to the similar number of impacts per unit of length (IPUL), high values of the Sa parameter (from tens to hundreds of microns) were observed under atmospheric conditions. A similar trend can be observed in the case of the Sz parameter reaching values from hundreds to thousands. According to the measured roughness values, the surface created by atmospheric PWJ at low traverse speeds can be classified as significantly damaged by erosion. Both visual and numerical differences in the obtained surface quality can be displayed in the 3D surface representation, as shown in Fig. 5.

Fig. 5
Fig. 5
Full size image

Comparison of the surface topography created using PWJ technology under atmospheric/air conditions: a linear trajectory, v = 2 mm/sz; b perpendicular interleaved trajectory, v = 4 mm/s. Comparison of the topography of surfaces created using PWJ technology under hydrostatic/submerged conditions: c linear trajectory v = 2 mm/s; d perpendicular interleaved trajectory, v = 4 mm/s

The Sa parameter, i.e. arithmetical mean height, is a statistically stable characteristic, although it possesses a limited predictive power of local surface topography changes. Despite its shortcomings, the Sa is often used in engineering practice for quality control of surfaces. Besides the Sa, the other area roughness parameters were used to assess quantitatively the effectiveness under atmospheric and submerged conditions of PWJ. The Sz parameter, i.e. the maximum height of the peaks on the measured sample area is a more sensitive parameter to even small changes in the areal roughness. The plot of dependence of roughness parameters Sa and Sz on traverse speed is in Fig. 6 for both trajectories and two surrounding environments.

Fig. 6
Fig. 6
Full size image

Influence of traverse speed and treatment trajectory type on areal surface roughness parameters Sa and Sz under a atmospheric, and b submerged experimental conditions

Figure 6a describes erosion behaviour for atmospheric PWJ. At low traverse speeds (1 and 2 mm/s), both Sa and Sz parameters correspond to a very rough surface formation with high irregularities for the linear trajectory. The highest values (approximately Sa = 161 µm and Sz = 1406 µm) were observed at the traverse speed of 2 mm/s, where the highest measurement deviations from the mean values are also recorded. Between traverse speeds of 2 and 3 mm/s, the sudden drop of the Sa parameter occurred from hundreds of microns to tens of microns; then with further increase in traverse speed, it stayed approximately constant. The graphical representation of the Sz parameter corresponds largely to the trend of Sa; nevertheless, the higher sensitivity to peaks and valleys created on the evaluated surface can be manifested as a higher scatter of Sz values for traverse speed between 3 and 5 mm/s. The described erosion behaviour can be attributed to the reduction in the number of water clusters impacting the surface (IPUL) and the material properties and its ability to resist erosion. Due to a larger value of IPUL at low traverse speeds, the material relaxation time is minimised, leading to accelerated material erosion. Therefore, a significant material loss occurs as the removal of material fragments. A significant order-of-magnitude jump in the measured Sa and Sz values ​​was observed between feed rates of 2 and 3 mm/s. Since the hydraulic and technological parameters of the PWJ process are the same, this phenomenon remains open to explanation only by the IPUL parameter. IPUL values of 20,000 and 10,000 caused the target material to be highly cyclically stressed, leading to early material fatigue where its mechanical properties were insufficient to withstand such loading over time. In the process of water erosion, the primary impact pressure is important, which causes the generation of initial unevenness in the weakest parts of the surface. Further impacts of water clusters caused that the water cluster was broken on the created unevenness, where part of the water got into the skid and flowed down the walls of the peak. In this way, there could be an acceleration of the water falling into the depression, which contributed to the deepening of the resulting unevenness. The reduction of IPUL to 6667 may have caused the number of water cluster impacts to be insufficient for extreme valley deepening and the mechanical properties of the material are better able to resist local failure. For this reason, it is possible to define a critical IPUL parameter, when the target material is able to resist extreme deepening of erosion, which depends on its mechanical properties. Contrary to it, the higher the traverse speed is the lesser material erosion occurs, i.e. the PWJ is not powerful enough to form high peaks and deep valleys and instead of material removal only roughening happens. Figure 6a shows also results for the perpendicular interleaved trajectory. The treatment caused a smooth progression of gradually decreasing both Sa and Sz parameters with increasing traverse speed without a step change. The highest observed values for the perpendicular interleaved trajectory (~ Sa = 185 µm and Sz = 1579 µm) were recorded for the lowest traverse speed (v = 2 mm/s). The numerical values listed in Table 3 confirm this trend.

Summarising the results of the average values of the areal surface roughness parameters Sa and Sz, the perpendicular interleaved trajectory showed a smoother downward trend compared to the linear trajectory, where a small initial increase in roughness values was followed by a sudden drop when the traverse speed increased from 2 to 3 mm/s. The linear trajectory at a lower traverse speed (v = 1 and 2 mm/s) caused the formation of a non-uniform surface with a hardly predictable evolution trend. The high fluctuation in surface roughness values was stabilised only at higher traverse speeds (from 3 to 5 mm/s). Comparing the measured values, higher surface irregularities were observed for the perpendicular interleaved trajectory (Sa = 185 µm) in contrast with the linear trajectory (Sa = 161 µm for the same IPUL value). Such behaviour of the surface roughness parameter is different than observed in a study [73] where higher surface irregularities were measured for the linear trajectory.

The variance in roughness parameter trends could stem from the processed material and its mechanical characteristics, which markedly influence the erosion progress [53]. The erosion resistance [74] of the material depends on the mechanical properties of the material [75] and its initial roughness [76]. However, technological input parameters such as the ultrasonic output power or the actual excitation frequency generated also enter the process and influence the erosion response. The experimental results imply that a perpendicular interleaved trajectory is more suitable for predicting the erosion state based on the PWJ process parameters due to the steady downward trend in the measured roughness values of Sa and Sz. The reason behind the observed phenomenon can be seen in the following: an initial PWJ pass (as in the case of a linear trajectory) forms a deeply non-uniform and irregular surface. All created irregularities are corrected by a secondary perpendicular pass of the water jet overlapping the initial erosion path. The erosive forces contribute to the flattening of the originally elevated but weaker peaks during the second pass. This process results consequently, in a reduction in peak height, leading to a decrease in measured surface roughness parameters. This trend was also confirmed in the work [54] that focussed on the surface modification of AISI 304L using a different initial frequency fs = 40 kHz.

The perpendicular interleaved trajectory at the PWJ process contributes to an increase in the homogeneity of the surface topography and leads to the stabilisation of the erosion process. However, except for the more homogeneous coverage of the eroded surface, the perpendicular interleaved trajectory causes higher measured roughness values than the linear trajectory with the same value of IPUL. This phenomenon could potentially be attributed to the supplementary impact of the water clusters on the target surface, consequently inducing deeper grooves or channels at points between the initially generated and rounded peaks.

Figure 6b describes erosion behaviour for submerged PWJ treatment where the water level height was wd = 150 mm above the material surface. The plot of the Sa and Sz parameters achieved using the linear trajectory shows a decreasing trend with increasing traverse speed, where the highest measured roughness value was observed at the lowest traverse speed.

It can be observed that there is a large difference in erosion evolution under atmospheric and submerged conditions. For example, in the case of Sa parameter and the linear trajectory, the lowest traverse speed of 1 mm/s results in 19 times higher value under atmospheric (159 μm) in comparison to submerged (8 μm) conditions. When evaluating the percentage decrease in Sa for the increasing traverse speeds, the difference was from 98 to 37% with an exception, when at a traverse speed of 4 mm/s, a violation of the continuously decreasing trend was observed resulting in the difference between atmospheric and submerged condition of only 1%, i.e. these conditions resulted in approximately the same value of Sa roughness parameter. It can be associated with possible changes/fluctuations in the technological conditions, such as ultrasonic frequency or power, which resulted in a more erosive effect on the target material. However, at submerged conditions, the dependence of the Sa parameter on traverse speed for the linear trajectory is more anticipated because there are no significant and unpredictable decreases/drops in the observed values. It can also be attributed to the much lower values of the measured roughness parameters related to the surface roughness generated on the samples under atmospheric conditions. Lower roughness values indicate lower material loss and a reduction in erosion potential predicting that no larger material fragments are being separated from the surface, as observed for experiments carried out under atmospheric conditions. The reduction in erosive behaviour is due to the resistance generated by the surrounding hydrostatic pressure of a water medium for the PWJ exiting the nozzle and impacting the submerged material surface. The energy of the PWJ is reduced and consumed in overcoming the surrounding static water to reach the material surface, causing a reduction in resultant velocity and, therefore, reducing the induced impact pressure on the target material surface. According to the authors of [77], the drag induced by the static water also causes a radial increase in the water flow with increasing distance from the nozzle exit. Thus, the lateral outflow of the jet is reduced as the radial distance from the jet axis increases. Water clusters act on a larger surface area at the impact point. This phenomenon consequently results in a diminished impact pressure on one side. On the other hand, it facilitates a more uniform treatment of the material surface, as evidenced by the quartile distributions depicted in the generated box plots. The results of the perpendicular interleaved trajectory also showed a significant reduction in the values measured under submerged conditions compared to the samples generated under atmospheric conditions (Fig. 6b). All measured values (Sa, Sz) were lower compared to the linear trajectory. This behaviour can be attributed to the smoothing effect of the second pass of the jet described above. The percentage reduction in roughness values of the surfaces generated by the perpendicular interleaved trajectory in submerged conditions has been calculated to be from 97 to 43% at a traverse speed from 2 to 10 mm/s compared to the atmospheric conditions. Similar results showing the potential for changes in surface and subsurface properties with a change in supplied pressure were presented in the work [64], where the surface of titanium alloy was modified using a micro-AWJ under submerged conditions with a water level depth of 60 mm. The surface roughness of Sa was reduced by 27% after the 70 MPa pressure process compared with the untreated sample. Increasing the pressure to 130 MPa caused the roughness increase, in which microcracks and channels with abrasive fragments left incorporated into the surface were observed. Figure 6b confirms the decreasing trend for both trajectory types and shows a significant decrease in the roughness values for the submerged conditions compared to the atmospheric conditions. In summary, the Sa and Sz parameters showed a significantly lower erosion potential of the impact pressure under submerged conditions, where the presence of a liquid medium, i.e. static water caused energy absorption of the high-speed jet. Nevertheless, in the case of homogeneity of treatment in terms of small changes of surface topography on large areas, the submerged treatment can be advantageous.

3.2 Evaluation of surface roughness parameters S p and S v

Another areal height surface roughness parameters, namely Sp and Sv, describing the surface topography were evaluated. Measured values are plotted in Fig. 7 and listed in Table 4 for atmospheric and submerged conditions and two treatment trajectories.

Fig. 7
Fig. 7
Full size image

Influence of traverse speed and treatment trajectory type on areal surface roughness parameters Sp and Sv under a atmospheric and b submerged experimental conditions

Table 4 Results of average areal surface roughness parameters Sv and Sp of linear and perpendicular interleaved trajectories under atmospheric and submerged conditions obtained from three sample measurements
Full size table

At atmospheric conditions, the parameter Sp, i.e. maximum peak height, for the linear trajectory and increasing traverse speed, is plotted in Fig. 7a. For the low traverse speed of 1 and 2 mm/s, the highest values of the Sp parameter were measured. A mild increase in Sp value was observed during increasing traverse speed from 1 to 2 mm/s. Then a sharp decrease in the measured values of the Sp parameter occurred. At the perpendicular interleaved trajectory, the highest measured maximum peak height was observed for the lowest traverse speed followed by a gradual decrease in values with an increase in traverse speed.

Apart from the fact that the perpendicular interleaved trajectory provided the Sp parameter higher than the linear trajectory (approximately 833 µm for perpendicular interleaved vs 605 μm for linear trajectory), perpendicular interleaved showed a more favourable downward trend contrary to the linear trajectory. The graphical representation of the parameter Sv (Fig. 7a), i.e. deepest valley, also shows the same behaviour as the Sp parameter, i.e. mild increase in values followed by a sharp decrease for the linear trajectory treatment, and decreasing trend for the perpendicular interleaved trajectory.

The results of the mutual comparison of Sp and Sv are interesting. For the linear trajectory, in some cases (v = 2 and 4 mm/s), the maximum peaks of the profile (Sp = 748 and 52 µm) are higher than the deepest valley depths (Sv = 657 and 39 µm). An opposite trend was observed in the study by the authors of Ref. [54], where the water erosion caused higher profile depths than the peaks. The results of the present study can be explained by the high-impact energy loss and, hence, the limitation of its erosive capacity on the impact acting on sharp peaks. At the same time, the local properties of the material may have also caused the formation of solid peaks, resulting in the disruption of water clusters against their walls, causing a weakening of the impact pressure, which is no longer able to erode the material to a greater depth. Moreover, in the case of perpendicular interleaved trajectory, comparing peaks and valleys, the values of Sp (Sp = 833 and 60 µm) were measured higher than Sv at certain traverse speeds (Sv = 746 and 40 µm for traverse speeds of 2 and 10 mm/s respectively). It could be due to the same effect of the breakdown of the impact pressure potential as in the previous case.

To sum up the results, the qualitative classification of the surface shows the predominance of valleys over peaks for certain traverse speeds. Significant prevalence in valley depths may be attributed to the tunnelling [44] process and the formation of local craters. On the other hand, examples of the formation of higher peaks may reflect the local sufficient resistance of peaks to further ongoing wear (or fracture). The results showed more favourable surface conditions for the perpendicular interleaved trajectory, as no significant deviations were achieved, as was observed in the case of the linear trajectory. Deep and sharp valleys with rounded peaks were dominantly formed. This shape provides a prerequisite for wear resistance since the peaks are not formed into a pointed and brittle shape prone to fracture.

The maximum peak height, Sp, of the surfaces generated by PWJ treatment under submerged conditions showed significantly lower values than under atmospheric conditions, similarly as in the case of the Sa parameter. The measured values are listed in Table 4 and plotted in Fig. 7b. For the linear trajectory, the highest Sp value was observed at a traverse speed of 1 mm/s, but a significant scatter in the measured values was observed at a traverse speed of 2 mm/s. In contrast, the values showed lower scatter and magnitude for perpendicular interleaved trajectory, which is attributed to the smoothing effect of the second pass. The results of the deepest valley depth, Sv, showed a similar decreasing trend (Fig. 7b) as in the case of maximum peak height, Sp parameter.

Submerged treatment resulted in an interesting result concerning the difference in the magnitude of the maximum value of the Sv parameter compared to the Sp value. The maximum value for Sv is shown by the perpendicular interleaved trajectory with a traverse speed of 2 mm/s. Moreover, this trajectory is also dominant for traverse speeds of 8 and 10 mm/s. This may be due to the micro tunnelling process in the material surface. The perpendicular interleaved trajectory can cause a greater deepening of the original valleys just by the second pass, or it may be due to the cleaning process of the valleys from the originally broken-off fragments from the first pass. An unconfirmed hypothesis may also be the effect of cavitation bubbles generated in the space above the material surface under submerged conditions, where the microjet formed after the imploding of the cavitation bubble resulted in deeper erosion valleys.

The submerged conditions caused the formation of a more homogeneous and regular surface with a better prediction capability than the atmospheric conditions. However, the summation of the results also predicts a higher increase in the surface area with dense distribution, which can also be used in the surface coatings application of lubricants and others. Likewise, the results support the idea of improving the conditions of ossification of bone cells into the implant surface, which is important in medical practice.

The reference samples were treated using continuous water jet (CWJ) with the same conditions, i.e. atmospheric and submerged conditions combined with two trajectories of water jet treatment, i.e. linear and perpendicular interleaved. A graphical representation of results for CWJ surface treatment technology is in Fig. 6. In all cases, the reference samples treated by CWJ did not exhibit any damage created on the material surface, confirming the importance and high effectivity of impact pressure pi over the stagnation pressure ps prevalent in CWJ [78] for erosion initiation similarly as in work [79].

3.3 SEM observation of erosion mechanisms

Surfaces treated by linear and perpendicular interleaved trajectories under both atmospheric and submerged conditions were analysed using SEM. Both trajectories under both environments were documented in micrographs with a view field of 356 μm. This view field was chosen to investigate a sufficiently large area to identify dominant material removal mechanisms. Since the submerged treatment was in centre of attention of the present study, the reached surface topography of the samples was additionally documented in more detail using a view field of 178 μm to focus on a specific area for the erosion phenomena description.

Figure 8 shows micrographs of erosion created using the PWJ method with the linear trajectory. The PWJ treatment under atmospheric conditions (left column of pictures) at low traverse speeds of 1 and 2 mm/s providing a high value of IPUL, caused the formation of surface topography of high roughness as described in the previous chapter. The deformation of the peaks was observed leading to the smoothening of the peak surfaces. In some cases, the formation of surface asperities on top of the peaks was also observed. The submerged conditions at these traverse speeds of 1 and 2 mm/s showed a much smoother surface. A cavities presence was observed despite the overall smoother surface. An erosion crater of approximately 120 µm in width was documented (see Fig. 8). The bottom of this micro-erosion crater showed an interesting feature of lateral flow inlets. It can be related to the tunnelling effect as described in Ref. [44]. The top edge of the crater showed signs of tearing of the material.

Fig. 8
Fig. 8
Full size image

SEM observation of surfaces treated using the linear trajectory technique

The surface treatment conducted using the PWJ method under atmospheric conditions and a slightly increased traverse speed of 3 mm/s led to the formation of a smoother surface; nevertheless, still it was possible to observe occasional erosion craters such as those presented in Fig. 8 depicting a microcavity at surface valley end having a width of approximately 70 µm. The deformation of the material formed the inlet cavity walls due to the input of the water jet and the process of wall extrusion and smoothing. Submerged conditions at the same traverse speed of 3 mm/s exhibited the generation of smaller and partially connected local voids creating a crack. The detail in Fig. 8 where a crack with a bridge, i.e. unremoved material overlaying the central part of the crack, was observed. Such penetrations could have been generated by the impact and drilling phenomenon of tiny liquid droplets. This indicates the process of lateral outflow penetration into the lower regions below the top surface of the material. The central bridge could have been formed in two ways, namely by the liquid going below the surface from one side, then deflecting its trajectory due to a change in the local material properties and penetrating out of the surface. The second possible process of cavity formation could be via the entry of a lateral jet from two sides. The subsurface interconnection of voids forms the bridge at the point where the lateral jets meet.

The increase in traverse speeds to 4 and 5 mm/s, i.e. decrease in values of IPUL to 5.000 and 4.000 impacts, resulted in mild roughening, i.e. no significant surface erosion in the sense of material removal under both atmospheric and submerged conditions occurred. Deformation of the extruded material into the periphery of the jet footprint was observed leaving traces of the straight path of the water jet on the surface (marked by an orange arrow).

Figure 9 shows micrographs of surface topography after the PWJ treatment with perpendicular interleaved trajectory observed for both atmospheric and submerged conditions. Low traverse speeds of 2 and 4 mm/s, i.e. high values of IPUL of 10,000 and 5000 impacts, respectively, result in the case of perpendicular interleaved trajectory the absolute value of number of impacts affecting the surface of 20,000 and 10,000 impacts. The surfaces generated using these parameters under atmospheric conditions achieved significant surface erosion leading to high values of areal surface roughness parameters. The hydrostatic pressure effect at the same traverse speeds under submerged conditions caused the formation of a more homogeneous structure, where the observed material removal may be caused by lateral jetting. This difference can be attributed to the second pass in the perpendicular direction, where the water flow was forced to seek new paths on the material surface. The first pass caused the generation of microcracks, eventually merging erosion pits (see a similar state for the linear trajectory with the same distribution of water cluster impacts and traverse speed of 2 mm/s). The material integrity in the lower parts of the erosion crater was weakened. In the second pass, the material was cyclically loaded in the perpendicular direction, which may have caused the upper parts to be pushed into the hollow spaces of the cavities and channels in the subsurface layers. This significantly increased the number of smoothened peaks observed.

Fig. 9
Fig. 9
Full size image

SEM observation of surfaces treated using the perpendicular interleaved trajectory technique

Increasing the traverse speeds to 6 and 8 mm/s caused the generation of micro-craters of approximately 40 µm in size with smoothed walls at the crater entrances at atmospheric conditions. The crater size for the perpendicular interleave trajectory is smaller when compared to the linear trajectory (~ 70 µm, see Fig. 8), which shows the influence of the second perpendicular pass and the distribution of the erosive effect on the target material surface. The shape and bottom of the erosion crater generated at a traverse speed of 8 mm/s are interesting due to the microcavities' presence. This type of crater may have been caused by the detachment of a fragment of material during the first pass from a local defect of the material, where the second perpendicular pass caused typical erosion in the form of microcavities, and channel formation mentioned above. The submerged conditions at these velocities caused the formation of local valleys with smoothed bottoms and upheaved crater walls. No significant cracks or voids were observed, which may be attributed to the increased traverse speed of the jet leading to a reduction in the number of impingements per unit area.

At a traverse speed of 10 mm/s, traces of an exposed surface structure of the material with occasional microcavities were observed for both atmospheric and submerged conditions. The presence of these rather rare micro-voids is not related to the effect of PWJ, trajectory and environment, but rather it is due to the presence of irregularities on the original surface of samples in the as-received state, i.e. before the application of PWJ (see Fig. 10).

Fig. 10
Fig. 10
Full size image

SEM evaluating the erosion of PWJ in both submerged and air-exposed environments

Figure 10 focuses on the comparison of the details for PWJ treatment under both atmospheric and submerged conditions. The linear trajectory, where the situation is not complicated by overlapping during the second pass of PWJ was selected to manifest typical topographic surface features at a print magnification of 5000 × with a view field of 35.6 μm. This SEM imagery was produced to elucidate better the differences in erosion mechanisms under two environments. Untreated material shows exposed grain boundaries in the as-received state. Three traverse speeds from 3 to 5 mm/s providing treatment with low IPUL value were chosen. It corresponds to the beginning of the roughening process to compare the erosion mechanisms during the incubation stage of PWJ treatment.

Treatment under atmospheric conditions (left column of pictures in Fig. 10) under the highest traverse speed of 5 mm/s led to damage concentrated on the grain boundaries with several cracks following the shape of the original valleys that occurred on the untreated sample. With a decreasing traverse speed, the number of cavities and cracks increased. Cracks started to propagate, and cavity deepening was observable. Pile-ups attributed to lateral flow were visible around the cavity entrances. The decreases in traverse speed to 4 and 3 mm/s resulted in the beginning of material removal; there was no trace of the original surface (before PWJ treatment). Deep cavities and interconnected cracks were apparent.

Submerged conditions (right column of pictures in Fig. 10) under a traverse speed of 5 mm/s exhibited many small areas exhibiting the original surface with visible grain boundaries. The surface is riddled with erosion pits having small diameters and clusters of erosion pits are observable. Some of these erosion pits show crack formation on their base. Other damage is mostly accumulated on the original exposed grain boundaries. The further decrease in traverse speed to 4 mm/s led to more damage accumulation along the grain boundaries. It leads to an interaction between surface irregularities and lateral flow that results in the encirclement of some grains. Further decrease in traverse speed to the value of 3 mm/s interconnected with an increase in the number of impacts acting on the surface (IPUL value of 6667) resulted in a more severe damage accumulation along the grain boundaries was observed; nevertheless, the damage was lower than that of a traverse speed of 5 mm/s and atmospheric conditions, which agrees with areal surface roughness parameters measurement.

4 Conclusions

The current research showcases the use of an ultrasonically-activated PWJ technique for the surface treatment of AISI 304L stainless steel. The study examined how the number of water cluster impacts per unit length—regulated by the jet's traverse speed—and the trajectory pattern employed for treatment, both in atmospheric and submerged states of the liquid medium, influenced texture expressed by various areal surface roughness parameters. The findings can be summarised as follows:

  • The nature of a continuous water jet does not allow the manifestation of impact pressure.

  • Pulsating water jet in submerged and atmospheric conditions cause erosion wear compared to continuous water jets.

  • A pulsating water jet creates a more homogeneous surface under submerged conditions in terms of low values of the maximum height of the profile.

  • Erosion activity depends on hydrostatic pressure. PWJ is intense in the atmospheric conditions. While in submerged conditions, the topography is determined by the exposure of grains, and in the case of experiments in atmospheric conditions, the erosive wear from the point of view of roughness profile parameters is more than 10 times higher.