The Abrasive Wear Resistance of Coatings Manufactured on High-Strength Low-Alloy (HSLA) Offshore Steel in Wet Welding Conditions

Some marine and offshore structure elements exploited in the water cannot be brought to the surface of the water as this will generate high costs, and for this reason, they require in-situ repairs. One of the repair techniques used in underwater pad welding conditions is a wet welding method. This paper presents an investigation of the abrasive wear resistance of coatings made in wet welding conditions with the use of two grades of covered electrodes—an electrode for underwater welding and a commercial general use electrode. Both electrodes were also used for manufacturing coatings in the air, which has been also tested. The Vickers HV10 hardness measurements are performed to demonstrate the correlation in abrasive wear resistance and the hardness of each specimen. The microscopic testing was performed. For both filler materials, the coatings prepared in a water environment are characterized by higher resistance to metal–mineral abrasion than coatings prepared in an air environment—0.61 vs. 0.44 for commercial usage electrode and 0.67 vs. 0.60 for underwater welding. We also proved that in the water, the abrasive wear was greater for specimens welded by the general use electrode, which results in a higher hardness of the layer surface. In the air welding conditions, the layer welded by the electrode for use in the water was characterized by a lower hardness and higher resistance to metal–mineral abrasion. The microstructure of the prepared layers is different for both the environment and both electrodes, which results in abrasive wear resistance.


Introduction
The number of offshore structures that are operated in changing and harsh conditions increases each year [1]. The aggressive environmental factors are responsible for offshore damages, due to, e.g., fatigue, mechanical damages, erosion and corrosion [2][3][4]. Undersea structures can be damaged due to abrasion as a consequence of the movement of sand and other solid particles caused by sea currents and waves. In addition, the internal surfaces of fluid transport pipelines may be damaged due to abrasion and cavitation [5]. Apart from typical damage in the form of cracks, these phenomena can cause a reduction in the cross-section of the elements. These factors make it necessary to use protective coatings on the steel surfaces, which improve the corrosion and wear resistance [6,7].
These coatings can be deposited by thermal spraying during the production of offshore structures [8]. In some applications, the composite coatings can be used [9,10]. However, the most commonly deposited layers are manufactured by the use of ceramics and metals. The ceramic coatings Al2O3/TiO2, ZrO2/CaO can be applied by a thermal spraying process [11][12][13]. The arc thermal spraying is also used to deposit the metals and its alloys [14,15]. The metals can be applied by laser cladding [16], arc welding [17] and friction surfacing [18]. Information on thermal spraying directly in water is very limited. Liu et al. proposed a novel underwater repairing and remanufacturing process and characterized the impact of the water environment on the process of applying aluminum coating by the cold spray method on a steel substrate [19]. Another promising technology to repair damaged components and prepare coatings is in-situ underwater laser welding [20][21][22]. However, most of the coatings used in offshore structures are applied by arc welding processes during manufacturing [23,24].
The research works are focused on the improvement of wear and corrosion resistance of laid layers. One of the methods to reduce the fracture of coatings could be the use of heating/cooling controls and buffers [25]. The coating applications are made in anair environment. Increasing theinterpass temperature decreasesthe abrasive wear resistance, which is a result of the increase in the grain size [26]. Unfortunately, the offshore structures still need to be repaired, which must be carried out in the water environment (in-situ) as it is lower in cost than in the case of transferring the structures into air welding conditions. The most often used underwater repair method underwater is wet welding. This process is carried out in direct contact with water and can be performedusing different methods, e.g., by the self-shielded flux-cored arc welding (FCAW-S) method [27][28][29]. The FCAW-S process does not require the use of shielding gas, and it is easy to automate and train welders [23,[27][28][29][30]. However, manual metal arc (MMA) welding is still the most commonly used as an underwater welding process [31,32].
The water environment generates some essential problems during repair works. The first is a higher diffusible hydrogen content in deposited metal than in air conditions, which is the result of the dissociation of water in the welding environment [32,33]. The problems of hydrogen embrittlement increase when the structure is in sea water [34]. The second disadvantage associated with welding underwater is the high susceptibility of steel to form brittle structures in the heat-affected zone (HAZ), which causes high values of residual stress [35]. These factors, including a high cooling rate, are responsible for the cold cracking of high-strength low-alloy steel (HSLA) joints made in underwater conditions [35,36].
As the results of wear tests of the underwater welded joints properties are extremely rarely reported, in this study, underwater manual metal arc coating deposition on HSLA S355G10+N steel was carried out. The deposit appearance, microstructure and hardness of specimens were investigated and compared with the results of tests carried out on specimens welded in an air environment. In addition, tests on abrasion resistance were carried out for specimens manufactured in both conditions. To the best of the author's knowledge, in the literature, there are no researches published that provide a detailed discussion on the problem of abrasive wear resistance of coatings manufactured in the water environment.

Used Materials
For welding, S355G10+N steel plates (U.S. STEEL, Košice,Slovak Republic) with dimensions of 120 × 100 × 16 mm were chosen as a base material (BM). This is an HSLA steel grade dedicated to making offshore structure elements [9]. The chemical composition of the BM has been tested by emission spectroscopy with a spark excitation method. As a filler material, two grades of covered electrodes were chosen. The first was an ISO 2560-A: E38 0 R11 [37] Omnia rutile electrode with a diameter of 4.0 mm for general use, which provides welds with good plasticity, which decreases the susceptibility to cold cracking. The second electrode was a specially formulated wet welding Barracuda Gold electrode-the nearest equivalent is ISO 2560-A: E42 2 1Ni RR 51 [37]. The chemical compositions of the used materials are listed in Table 1.

Welding Procedure
For the tests, four specimens were prepared with the use of MMA. Two of them were welded in the water environment (0.25 m water depth) by two different grades of covered electrodes. Two specimens were made in the air by different electrodes. To cover all surfaces of the steel plate, the number of beads were different for each environment: for specimens welded in the air-12 beads-and for specimens made underwater-14 beads. The welding parameters were chosen according to previous investigations [32,35,38] and the manufacturer's data. Welding with Omnia electrodes was carried out by a welding current with the range 150-160 A (the manufacturer allows the range 130-160 A), and with Barracuda Gold electrodes the range 190-200 A (the manufacturer allows the range 170-220 A). These parameters produced heat input values in the range of 0.7-0.8 kJ/mm for Omnia electrodes and higher than 0.9 kJ/mm for Barracuda Gold electrodes, which limits the susceptibility to cold cracking of used steel. For specimens made in the air, beads were laid one after the other. The research plan developed at the beginning assumed conducting the experiment in conditions as close as possible to real conditions, used in industrial applications. During welding in the air of the tested steel grade, there is usually no need to use a special bead sequence. However, during welding in the water environment, due to the need to reduce the tendency of steel to cold cracking, a change in welding technique is recommended, which was found in our previous study [38]. Specimens welded underwater were characterized by a lower width of laid beads. This resulted in a higher number of beads for underwater welding conditions than for specimens manufactured in the air. The scheme of the specimens is presented in Figure 1.

Examination Procedure
After welding, steel plates were cut for specimens for hardness HV10 measurements (one specimen from each plate) and for abrasive wear resistance tests (two specimens from each plate). For specimens, for the abrasive wear test, after cutting, the upper surfaces of specimens were groundwith a surface grinder to ensure parallelism with the lower surface. The scheme of cutting a steel plate is presented in Figure 2. To determine quantitatively the abrasive wear resistance of Omnia and Barracuda Gold electrode layer MMA surface deposits, the tests of abrasive wear type metal-ceramic were conducted in accordance with standard ASTM G 65-00 Procedure A [39]. The 25 mm wide and 75 mm long abrasive wear resistance test specimens were cut from the middle area of Omnia and Barracuda Gold electrode layer MMA surface deposits. All specimens were weighed to the nearest 0.0001 g, as required by ASTM G65-00. During the tests, the rubber wheel apparatus made six thousand revolutions, and the flow rate of the abrasive (sand A. F. S. Testing Stand 50-70) was 335 g/min. The force applied to press the test coupon against the wheel was TL = 130 N (test load (TL)). After the abrasive wear resistance test, the specimen was weighed at a weight sensitivity of 0.0001 g. The mass loss of specimens of Omnia and Barracuda Gold electrode layer MMA surface deposits was reported directly and relatively in comparison to the mass loss of the reference Hardox 450 steel, which is commonly used [40]. Next, the densities of surface deposits were measured and the abrasive tests' results were reported. The deposits density was determined using hydrostatic RADWAG WAX 60/220 weight, based on the three measurements of specimens taken from each of the prepared coatings weighed in the air and in the water. The schematic diagram of ASTM G65-00 Procedure A abrasive wear resistance test and test stand are presented in Figure 3. The Vickers HV10 hardness measurements were taken in accordance with EN ISO 9015:2011 [41], using a Sinowon V-10 stand, with a measurement error of ±3HV10. Before measurements were performed, the tested cross-sections were ground by sandpaper with a granulation of up to 1000. Then, they were etched by Nital 4% to show the structure of the layer and HAZ and to calculate the depth that the upper surface was ground to. Then, the upper surfaces wereground at the surface grinder to ensure the parallelism with the lower surface. The hardness HV10 measurements were prepared on the upper surface of the specimen and on cross-sections in accordance with the two measurement layers. The schematic view of the hardness HV10 measurements is presented in Figure  4.
. The microscopic tests were done in accordance with the EN ISO 17639:2013 [42]. Observations were carried out with the use of an optical microscope. From each specimen, the welds and HAZ were observed. The specimens wereground, polished and etched by Nital 4%.

Abrasive Wear Resistance
The density of Omnia and Barracuda Gold electrode layer MMA surface deposits were measured and the abrasive tests' results were reported as volume loss in cubic millimeters, by converting mass loss to volume loss as follows [39]: Volume loss, (mm 3 ) = mass loss (g) : density (g/cm 3 The results of low-stress abrasion wear resistance to the metal-ceramic scratching when usingdry quartz sand as the abrasion material of the HARDOX 450 wear plate and MMA surface layer deposits of the Omnia and Barracuda Gold electrodes tested are presented in Table 2. The views of the specimen abrasion marks after the test are presented in Figure 5. The observed surfaces of investigated coatings (Figure 5a-d) have typical morphology for abrasive wear resistance test. It was stated that two mechanisms are responsible for the wear of the surface. In addition to the expected rolling mechanism [43], the effect of the grooving mechanism was found. This effect results from sharp-edge abrasive particles, which penetrated deep into the investigated surface. The effect of penetration is a larger amount of material removal from specimens, which initiated the grooving mechanism [44]. In contrast to the wear mechanism for Hardox 450 steel (grooving), the rolling mechanism was predominant in investigated coatings. Another aspect of determining higher wear resistance of Hardox 450 (C=0.26%, Cr=1.40%) steel is lower carbon and chromium content in investigated coatings (Table 1). For this reason, the layers are intensively strained and cut by abrasive grains [45]. Hardox 450 steel (samples H1, H2) had the greatest relative resistance to the metal-mineral abrasion. The surface of the arc hardfacing layers at the place of impact of the abrasive compared to Hardox 450 was characterized by a deeper metal wiping. The top layers padded with a covered electrode-Barracuda Gold-made both in a water environment and in the outdoor welding, compared to layers made in analogous conditions with an Omnia rutile electrode, are characterized by greater resistance to the metal-mineral abrasion, Figure 6. It should be emphasized here that in the case of both consumables used, the resistance to abrasive wear of layers made in a water environment was higher than that of layers made under normal conditions.

Hardness HV10 Measurements
Before measurements were carried out, the macroscopic tests were performed for each cross-section in accordance with the EN ISO 17639:2013 [42]. There were no imperfections detected during macroscopic testing. An example view of the prepared cross-section is presented in Figure 7. The macroscopic photographs were used to calculate the dilution for layers performed in the air and in the water. The dilution is a factor that complicates predicting the effects of multi-pass welding, which affects, among other aspects, hardenability [46]. The dilution was calculated i.e., D = Abase/A, where A is the total molten area, and Abase is the molten area of BM [47]. The scheme of dilution calculation is presented in Figure 8. The dilution values for specimens welded in the air were higher than those for specimens welded underwater. These values depend on the bead sequences used for the surface of specimens, which was described in our previous studies [38]. Coronado et al. [48] stated that higher dilution levels reduced wear resistance. The different dilution values result from an assumed methodology of bead sequences.
The hardness HV10 measurements show differences in all specimens. For the specimens welded inthe air, the coating made by electrodes for underwater welding was characterized by higher hardness. Different situations were observedfor measurements of specimens welded in the water environment. The coating produced by electrodes for underwater welding was characterized by a lower hardness. The investigated S355G10+N steel is classified as material from group 2.1. in accordance with EN ISO 15614-1:2017 [49]. The maximum hardness assumed by this standard for this group cannot exceed that of 380HV10. The hardness measurement showed that the coatings made underwater by the Omnia electrode did not fulfill the requirements of hardness by the standard EN ISO 15614-1:2017. The results of the hardness HV10 measurements are presented in Table 3. After welding in the air, the coating made by the electrode for underwater welding was characterized by higher hardness and higher relative abrasion resistance than the coating welded by the Omnia electrode. In the case of underwater welding, the layer (Barracuda Gold electrode) with lower hardness was characterized by a higher relative abrasion resistance. The literature [48] showed that there is no relationship between the hardness and relative abrasion resistance, which was proven in this investigation. For both environments, the Barracuda Gold electrodes produced the layer with the higher wear resistance. The Barracuda Gold layers were welded with higher values of heat inputs, which decreased the cooling rate. This providesa higher abrasive wear resistance for layers produced by the Barracuda Gold electrodes. The same factor was responsible for higher abrasive wear resistance of layers welded inthe air. The water, as a welding environment, increased the cooling rate, which generated brittle structures [27][28][29][30][31]35,38] with lower wear resistance.

Microscopic Study
The microscopic observations confirmed the earlier results. There were observed significant differences in the weld and in HAZ of specimens produced by different electrodes in different environments. The layers prepared in the water are characterized by the presence of dendritic structure in welds, which consisted of bright fine-grained pearlite arranged in columns. At the boundaries of dendrites, the acicular ferrite grew from these columns (Figure 9a,b). In the welds performed by the Omnia electrodes, cracks were observed (Figure 9b). This may be the result of Ni content in the Barracuda Gold electrode deposit, which improved the mechanical properties of weld metal [50]. The structure of the welds of coatings performed in the air is different. They are characterized by a partial disappearance of the dendritic structure (Figure 9c,d). This area is built by a fine-grained ferrite and pearlite structure, which results from the lower cooling rate in the air in comparison to the water. The weld produced by Barracuda Gold in the air is characterized by finer grain size, which improved wear resistance. This is in agreement with the results presented by Cui et al. [51]. The structure of HAZ was similar for coatings performed by both electrodes in the air. The structure indicated the presence of refined and tempered low-carbon martensite mixed with normalization structures with fine ferrite and pearlite (Figure 9d). The HAZ of underwater specimens is characterized by the presence of brittle bainitic and martensitic structures, which results in cracking observed for both used electrodes. In specimens prepared by Barracuda Gold, these cracks are located perpendicular to the fusion line ( Figure 9e). The cracks observed in specimens performed bythe Omnia electrodes are located perpendicular and parallel to the fusion line. The length of those cracks was much higher. It was stated in the literature [44] that the presence of the cracks influencesa lower abrasive wear resistance of performed coatings. The exemplary results of metallographic microscopic tests are presented in Figure 9.

Conclusions
The results of the performed investigations showed a significant influence of the welding environment on the hardness and wear resistance of the produced layers. We proved that relative abrasion resistance depends on the type of electrodes and the environment of welding. From a practical point of view, the most important finding is that there is no need to transport elements to the surface, and they can be repaired in-situ in the water. It will reduce the cost of maintaining and repairing of offshore infrastructure.
The performed examinations allow us to draw these conclusions: 1. Water, as an environment of the welding processes, generates higher HV10 hardness of steel coatings than air does. The microstructure of coatings performed in the water contains the brittle structures in the HAZ, which resulted from a high cooling rate. The structures for specimens manufactured in the air are characterized by the presence of normalized structures. 2. The hardness strongly depends on the welding environment, but no influence on the abrasive wear resistancewas found, neither in the air nor in the water environment. 3. The coatings produced by the electrodes for underwater welding (Barracuda Gold) are characterized by a lower volume loss in the metal-mineral abrasion conditions than layers welded by the general usage electrodes (Omnia). This is due to the smaller grain size of the weld metal produced in the air (Barracuda Gold), and the greater number of cracks for coating made in the water (Omnia).
4. The water environment of welding produces coatings with higher abrasive wear resistance than layers welded in the air, which resulted from different metallographic microstructures formed in coatings manufactured in other environments.
Funding: This research received no external funding.