Effect of oxide scale structure on shot-blasting of hot-rolled strip steel



The effect of oxide scale composition of hot-rolled strip (Q235) on shot blasting is studied in this article. The properties of the oxide scale on the strip surface change during storage. The shot blasting is an important on-line acid-less descaling technology. The effect of shot blasting is affected by many factors, among which the composition of oxide scale may play an important role. However, there are few studies on the relationship between the oxide layer content and the descaling effect.



The morphologies of oxide scales at different storage times are observed by scanning electron microscopy (SEM), and the compositions are analyzed by X-ray diffraction. These strips are then shot blasted and descaled with different amounts of abrasive, and the descaling effects are compared by SEM.



The results show that the eutectoid structure Fe3O4/Fe in the oxide scale will gradually transform into Fe3O4. In the case of short storage time, the content of the eutectoid structure is high, and it is difficult to remove the oxide scale. While the strip with a long storage time has no eutectoid structure Fe3O4/Fe and FeO, it is easy to remove the oxide scale during the shot blasting process. The composition of the oxide scale has a significant effect on the effect of shot blasting, and it provides significant guidance to the optimization of the descaling process parameters.


41 pressure water. The high-pressure water jets cause thermal changes, shocks, vibrations, and 42 scouring on the surface of the strip. The dynamic pressure of high-pressure water becomes the 43 hydrostatic pressure and invades the bottom of the oxide scale, causing the oxide scale to peel off 44 from the surface of the substrate (Choi J W & Choi J W, 2002). This technology is widely 45 applied in hot rolling process, but it can't be used to the cold rolling procedure since the energy 46 of water is too small to remove scales. 47 Abrasive water jet descaling technology uses high-pressure water to accelerate steel sand, 48 quartz sand and other discrete bodies, and sprays the mixed abrasive stream to the strip surface at 49 a certain angle through a nozzle to crush the oxide scale. Both the discrete body and water in this 50 method can be recycled, and the descaling effect is obvious. However, due to the large water 51 flow of the system, the high-pressure plunger pump requires higher cleanliness of the water, the 52 water circulation system is always in a high load state, and the nozzle wears severely under long-53 term service, so this technology can only be applied to narrow band steel descaling or surface 54 treatment of small parts (Meng, Wei, & Ma, 2016). 55 Tensioning descaling is a mechanical method of repeatedly bending the strip steel. After the 56 metal substrate is subjected to stress, a certain elastoplastic deformation occurs. The oxide scale 57 on the surface of the metal substrate is broken due to brittleness and the purpose of descaling is 58 achieved. Tensioning descaling is generally used in cases where the material is not seriously 59 hardened and the product quality requirements are not strict (Tongqing, 1998;Bakhmatov et al., 60 2014). Smooth-Clean Surface (SCS) technology is used in a closed space to automatically adjust 61 the roll gap of the grinding roller according to the thickness of the strip. At the same time, the 62 surface of the steel plate is continuously washed by circulating filtered water, and the ground 63 iron oxide is taken away to achieve surface cleaning. Finally, a 7 μm thick anti-rust layer is 64 formed on the surface of the metal substrate. This method is not suitable for cold rolling, deep 65 drawing and rotary deep drawing (Tamura et al., 2020). 66 In order to realize the application of on-line acidless descaling for broad steel strip 67 production, the Material Works Ltd. of the USA developed EPS (Eco-Pickled Surface) system. 68 In this system, the abrasive shot blasting was applied to the industrial fields and proved to be an 69 effective method to ensure the strip surface quality after descaling (Voges & Mueth, 2012; 70 Voges, Mueth & Lehane, 2008). However, the energy consumption and processing cost of the 71 system is so high that it cannot replace the pickling process yet.

72
Our research group is very interested in the non-acid oxide scale removal technology, and 73 proposed the oxide scale removal technology combining shot blasting with high-pressure water, 74 and designed the relevant experimental equipment. The most important research direction is to 75 optimize the process parameters of shot blasting to reduce system energy consumption and 76 processing costs. We have studied the effect of shot blasting speed on the descaling effect, and 77 the results show with the increase of the projectile velocity, the damage area of the oxide scale is 78 increased, and the damage area is composed of the direct destruction area and the indirect failure 79 area. (Wang et al. , 2018 Actually, the effect of shot blasting is affected by many factors, among which the 81 composition of oxide scale also has an important effect on the removal of scales. Parameters 82 such as steel type, rolling speed, rolling temperature, cooling speed and coiling temperature etc. 83 will affect the oxide scale composition (Zhou et al., 2011;Gong et al., 2009). The oxide scale 84 layer of ordinary carbon steel generally consists of three layers (Bonnet et al., 2003): the inner 85 layer is a solid solution of FeO and Fe 3 O 4 , the middle layer is Fe 3 O 4 , and the outer layer is Fe 2 O 3 . 86 During the hot rolling process, the main component of the oxide scale layer is FeO. According to 87 the Fe-O equilibrium phase diagram (Chen & Yeun, 2000, the eutectoid reaction of 88 FeO can produce a mixed product of Fe and Fe 3 O 4 below 570 °C. After laminar cooling and air 89 cooling, a large amount of FeO will transfer into precipitates by eutectoid reaction. After being 90 exposed to the air for a long time, the outermost layer of the oxide layer continues to be oxidized 91 to Fe 2 O 3 . Therefore, in the process of the exposure in the air, the oxide scale composition is 92 varying continually. However, there are few studies on the relationship between the oxide layer 93 content and the descaling effect, which is a key factor affecting the descaling effect, and is also 94 the research objective of this paper.

95
In this paper, two kinds of Q235 strip steels with different air cooling time are selected for 96 the research. Firstly, the difference in scale composition on the strip surface is obtained through 97 energy dispersive spectrometer and X-ray diffraction analysis. Then the shot blasting 98 experiments are carried out, and the descaling effect is observed by scanning electron 99 microscopy (SEM). Moreover, the influence of the variation of the scale's composition caused 100 by air cooling time on the descaling effect of shot blasting is analyzed, which provide important 101 guidance to the improvement of acidless descaling process in industrial production. The research 102 route is shown in Fig. 1. 103

104
The test samples are two Q235 strips stored at different times. Firstly, the oxide scale 105 morphologies and compositions are measured by energy dispersive spectrometer (EDS) and X-106 ray diffraction (XRD). Then, the descaling experiments are performed using the shot blasting 107 descaling experimental device developed by NERCFRE. The electron microscope is used to 108 observe the removal effect in the experiments.

109
The object of XRD inspection is the surface of the sample after shot blasting. The bottom of 110 the sample is fixed on the platform by means of bonding. The model of the XRD device is 111 Ultima IV, and the type of Tube is ceramic X-ray tube.

112
As for scanning electron microscope (SEM), the sample was cut into 10mm×10mm squares 113 and then embedded into the resin. The SEI mode is used to observe the surface morphology, and 114 the BSEI mode is used for element detection. The type of equipment used is ULTRA 55. The experimental samples were taken from the Q235 hot rolled strip of the practical 117 production line of a steel company. During the hot rolling process, the temperature dropped from 120 group of subsequent specimens from which were labeled as No. 1. The other was produced 121 within one week before the experiment, the sample and the group of subsequent specimens from 122 which were labeled as No. 2. The reason for using samples that have been stored for a long time 123 is to explore the descaling effect of samples with different oxide layer compositions, and the 124 composition of the oxide layer can also be changed in other ways. Table 1 shows the chemical 125 composition of the two samples. It can be seen there are little difference between them, and the 126 influence on the mechanical properties can be neglected. 127 2 Experimental Method. 128 2.1 Oxide scale's composition analysis. 129 The procedure was as follows: (1)The slot blasting descaling experimental facility. 141 The acidless descaling experimental facility designed by NERCFRE is shown in Fig. 2. 142 This device mainly includes six major units, which are uncoiler, 5-roller tension leveler, slot 143 descaling, high-pressure jet, sweeping-drying and coiler. The main process parameters include: 144 impact angle, impact line speed, particle size and abrasive weight.  c. Both the samples in step b were cut in the middle area to obtain 2 specimens with size of 153 20mm×10mm respectively, and the specimens obtained by cutting were divided into two groups. 154 One group of specimens was observed by a ZEISS ULTRA 55 scanning electron microscope for 155 the descaling effect on the front of the specimens. Another group of specimens was mounted 156 with the cut surface, and the descaling effect from the cut surface was observed. The cross-section morphologies of the oxide scale observed by SEM are shown in Fig. 3. 164 The energy dispersive spectrometer of the iron and oxygen elements at the outside, intermediate 165 and inside positions of the oxide scale by the ZEISS ULTRA 55 scanning electron microscopy 166 are shown in Table 2. The results of the phase analysis by X-ray diffractometer are shown in Fig.  167 4. 168 2 Experiments of the shot blasting experiments with a small amount of abrasive 169 The descaling effects from the front surface's scanning after the shot blasting with 2kg of 170 abrasive are shown by Fig. 5. 171 The oxide scales are layered and have a certain thickness. It is difficult to determine 172 whether the oxide scale is completely peeled off from the base body only from the front surface's 173 scanning. Therefore, it is necessary to observe the effect of descaling from the cross section, as 174 shown by Fig. 6. 175 The SEM results of No. 1and No.2 groups after shot blasting with 20kg abrasive are shown 176 in Fig. 7 and Fig. 8. 177 3 Experiments of the shot blasting experiments with a large amount of abrasive 178 The 4000 times magnifications of descaling effect from the section's scanning after the shot 179 blasting with 20kg of abrasive for No. 1 and No. 2 group are shown in Fig. 9 and Fig. 10. 180 Since Fig. 9 is the 4000-times magnification result of the SEM, the field of view is very 181 small In order to improve the reliability of the research, a larger view field was chosen and the 182 area scanning of energy dispersive spectrometer was conducted, as is shown in Fig. 11. 183 Similarly, a larger view field was chosen and the area scanning of energy dispersive 184 spectrometer for No. 2 group was conducted, as is shown in Fig. 12. As is shown in Fig. 3(A), for the No. 1 group, the thickness of oxide scale is relatively 190 uniform and is about 9.5μm, and the structure is compact and well combined with the basal body, 191 which indicates that the oxidation of the strip surface is uniform and adequate during the hot-192 rolling and long-time air cooling process. In Fig. 3(B), for the No. 2 group, the uniformity of the 193 oxide scale thickness is worse than that of No. 1 and the average thickness is about 12μm. It is 194 obvious that there are many defects in the structure of oxide scale. By the above comparison, 195 there are apparent differences of scale morphology with the rolling and cooling conditions 196 difference. And as the chemical composition changes, the density of the oxide scale gradually 197 increases. As shown in Fig. 6(A), for the No. 1 group, after the shot blasting with a small amount of 224 abrasive, the oxide scale at the pit's edge fall off completely and the basal body is revealed and 225 the peeled areas are large. There are not obvious cracks of the oxide scales in and around the pits, 226 but there are tiny gap between the scale layer and the basal body near the peeled areas.

227
As is shown in Fig. 6(B), for the No. 2 group, the peeled areas of the scale layer is small, 228 and the basal body is not completely revealed. However, there are obvious cracks of the oxide 229 scales in the pits. Thus, it can be deduced that compared with No. 1 group, the oxide scale of the 230 specimens of No. 2 group has lower hardness and better combination with the basal body. The 231 descaling effect of No. 1 group is better when the impact force of the projectile reaches a certain 232 level.  Fig. 7 (A) shows the descaling effect at 100x magnification in the backscattering mode. The 236 darker part represents the area where the oxide scale has not fallen off, and the lighter part 237 represents the area where the oxide scale has fallen off. It can be seen that most of the oxide 238 scale has been peeled and only a few remains after the shot blasting with a large amount of 239 abrasive. The 500 times magnification of the peeled areas is shown by Fig. 7(B), and it can be 240 observed that the pits of the basal body have become relatively smooth due to multiple hits.  Fig. 8 (A) shows the descaling effect at a magnification of 50 times, and Fig. 8 (B) is a 242 partial enlarged view of Fig. 8 (A). The relatively uniform color in the figures indicates there is PeerJ Mat. Sci. reviewing PDF | (MATSCI-2020:02:45849:2:0:NEW 30 Aug 2020)

Chemistry Journals
Analytical, Inorganic, Organic, Physical, Materials Science 243 only one kind of material in the surface. And the energy dispersive spectrometer results show 244 that the oxygen content is 30.28%, the iron content is 69.72%, which indicates that the layer is 245 the remaining oxide scale rather than the basal body. It can be obtained that the outer oxide scale 246 layer falls off during the shot blasting process, but the inner oxide scale layer still exists on the 247 substrate, which also confirms that the oxide scale is a layered structure. 248 3.2 Analysis of the descaling effects from the section's scanning 249 As is shown in Fig. 9, the oxide scale after the shot blasting with 20kg of abrasive for No. 1 250 group has been peeled cleanly without obvious residue, and the surface is smooth after a large 251 number of random hits. As is shown in Fig. 10, the oxide scale after the shot blasting with 20kg 252 of abrasive for No. 2 group has not been peeled completely, but the thickness is reduced 253 from12μm to 6μm, which means that the outer oxide scale falls off with the shot blasting, but the 254 internal scale layer still exits. In addition, obvious cracks appeared on the surface of the 255 remaining oxide scale.

256
As is shown in Fig. 11, where a larger view field was chosen compared with Fig. 9, the 257 result of the area scan can indicate the content of the element by the depth of the color. The 258 scanning area is shown by the green line frame in Fig. 11 (A), and the scanning result of oxygen 259 element is shown in Fig. 11 (B). It can be seen that there is no large amount of oxygen between 260 the mounting powder and the basal body, which indicate that the oxide scale has fallen off after a 261 large number of shot blasting and there is no oxide scale remaining.

262
As is shown in Fig. 12, where a larger view field was chosen compared with Fig. 10, the 263 scanning area is shown by the green line frame in Fig. 12 (A), and the scanning result of oxygen 264 element is shown in Fig. 12 (B). It can be seen that there is a significant area of oxygen 265 accumulation between the mounting powder and the substrate, which indicates that after a large 266 amount of shot blasting, the oxide scale still exists.  For oxide scale without eutectoid structure, in the case of only descaling by shot blasting, as 274 the thickness of oxide scale gradually decrease, the efficiency of descaling will be greatly 275 reduced, resulting in increased costs. Therefore, after the shot blasting and descaling, an 276 additional high-pressure water jet process can be added. Firstly, a large area of oxide scales is 277 removed by shot blasting. At this time, the binding capacity between the remaining oxide scales 278 and the basal body becomes weak, and then it can be completely removed by direct spraying 279 with high pressure water further.

280
For oxide scale with eutectoid structure, using shot blasting to remove oxide scale is less 281 effective. The method of combining shot blasting and pickling should be explored. By studying 282 the best process, it can reduce pollution emissions and production costs and improve production 283 efficiency. In this paper, two kinds of Q235 strips stored at different times were selected to analyze the 286 difference of surface oxide scale composition and the effect of shot blasting descaling, which 287 provided a basis for the optimization of shot blasting process. The main research contents and 288 conclusions are as follows:

289
(1) The EDS and XRD were used to observe and analyze the composition of the two Q235 290 steel scales stored at different times. It is found that the composition of the steel strip after hot-291 rolling is significantly different during long-term storage. During the storage of the strip, the 292 oxide scale will continue to be oxidized, and the eutectoid structure Fe 3 O 4 /Fe of the inner layer 293 will be oxidized to Fe 3 O 4 . The hot-rolled strip scale with long storage time will have no eutectoid 294 structure Fe 3 O 4 /Fe and FeO.

295
(2) The descaling experimental facility designed by NERCFRE was used to perform shot 296 blasting and descaling treatment. The scanning electron microscope was used to observe the 297 effect of a small number of shot blasting effects of two Q235 strip steels. Although Fe 2 O 3 and 298 Fe 3 O 4 have high hardness, they are easy to fall off during shot blasting, and the strips that have 299 not been stored for a long time are prone to scaly fracture due to the presence of Fe 3 O 4 /Fe 300 eutectoids. However, it is more firmly bonded to the basal body, and it is relatively difficult to 301 remove the oxide scale.

302
(3) The scanning electron microscope was used to observe the effect of a large number of 303 shot blasting effects of two Q235 strip steels. It is found that for strips that have been stored for a 304 long time, the main components of the oxide scale are Fe 2 O 3 and Fe 3 O 4 , which can be more 305 easily removed by shot blasting; while for strips that have been stored for a short time, the scales 306 contain eutectoids structure Fe 3 O 4 / Fe, shot blasting can reduce the thickness of the oxide scale, 307 but it is more difficult to completely remove it.