The Process Development of Glass Cullet and Recycled Glass Aggregate for Improving Recycling Rate

10 Due to depletion of resources and the spread of environmental pollution, the 11 sustainability of raw materials is emerging as an important issue. Glass bottles are one 12 of the products that are easy to recycle, and many studies have been conducted to 13 improve the recycling rate. In this study, we attempted to develop the waste glass bottles 14 process that can be recycled into a glass cullet and recycled glass aggregate. To produce 15 a cullet from waste glass bottles(WGBs), color quality standards must be satisfied. 16 Therefore, we applied a multistage color sorter to the experiment. The recycled 17 aggregate glass must adjust the particle size. Thus, we experimented with the optimum 18 crusher selection test when applying the crushing process. And, we confirmed the 19 appropriateness of process by aspect ratio analysis of product. In addition, we 20 confirmed the trends in the data required to set the optimum design and operating 21 conditions of the selected vertical shaft impact(VSI) crusher using discrete element 22 method(DEM) simulations


Introduction 26
Increasing consumption patterns globally are resulting in depletion of resources and 27 the spread of environmental pollution; therefore, the sustainability of raw materials is 28 emerging as an important issue. Various studies have been conducted on the 29 sustainability of raw materials; nevertheless, many products are disposed in landfills 30 or incinerated due to economic constraints. Landfill and incineration can cause many 31 problems, such as air pollution and heavy metal leaching; a better future-oriented 32 method should be suggested [1][2][3]. 33 In 2019, the amount of waste glass bottles (WGB) generated was about 615,000 34 tons in Korea. The associated recycling rate is reported to be approximately 79.0% and 35 486,000 tons [4]. Since 2010, the recycling rate of WGBs of over 75% has been 36 achieved, but is about 10% lower compared with Germany and Japan, countries with 37 advanced recycling systems and strategies. In addition, since only manufacturers with 38 an annual delivery of more than 10 tons, sales of more than 1 billion won, and annual 39 imports of more than 300 million won in the previous year are obligated to recycled 40 producers, it is difficult to confirm an accurate amount of WGBs

Color sorter 119
Recycling WGBs into cullet is the most economical method, however, color 120 classification is essential for this process. To achieve the quality standard, an 121 experiment was conducted by applying an optical sorter which was manufactured to 122 identify the different color mixing rates for each color. The color sorter is 123 compartmentalized according to the supply method of the feed: as chute and belt types. 124 Fig. 5a presents a schematic diagram of the chute type color sorter used in this study 125 [26]. Fig. 5b shows the fundamental structure of the chute-type color sorter. The color 126 sorter supplies the sample uniformly using a vibration feeder, the quantity of light 127 transmitted is measured by a high-resolution CCD line camera, and then the color data 128 of the samples are collected in a color sorter. Based on the collected data, the product 129 that is not the designated input color for the sorter is separated into a 'defeat box' by an 130 air nozzle. The major experimental conditions for the color sorter are color separation 131 order, color sensitivity, feed rate, air pressure, and particle size. In this study, we 132

Crusher 137
For recycling WGB under 10 mm into a recycled glass aggregate, satisfying the 138 recycled aggregate particle size standard is essential. In this study, an experiment was 139 conducted using four types of equipment: a hammer crusher, a shredder, a roll crusher, 140 and a VSI crusher, which characteristically show the main force applied to the particles 141 (Fig. 6). The key specifications and experimental conditions of the crusher used in this 142 study are listed in Table 4. The shredder and hammer crusher were tested using the 143 crusher specifications, and a roll crusher was used at a rotation speed of 280 rpm and a 144 roll gap of 2.5 mm, which is obtained from the equation of Mineral Processing Design 145 and Operations (Second Edition) [28]. The rotation speed of the VSI crusher can be 146 controlled by a control box, and 1,000 rpm was used to test the WGB, which is more 147 easily crushed than ores. 148

4.1.Color sorting experiment result 151
To recycle mixed color WGB into a glass cullet, the color quality standard must be 152 satisfied. The color quality standard, in terms of difference color rates, is under 5% for 153 amber and green and is under 1% for clear. Since clear has the most stringent quality 154 standards, it was fixed to the end of the sorting order, which has the least amount of 155 WGBs to be removed. The order of green and amber were changed and tested as 156 condition A and B. The condition A involves the removal in the following sequence: 157 first amber, second green, third amber and green, while condition B is sequenced as 158 first green, second amber, third amber and green. The general color sorting experiment 159 removes the defeat color, which is usually a small portion. However, the sample used 160 in this study contained the desired color at a lower rate, therefore we removed the 161 desired color, instead. products, but did not satisfy them for, green. In addition, the recoveries from condition 170 green and clear measured in color sorter were distributed similarly, the removal of green 172 as the first action (as in condition B), also affected the clear, so we assessed that to be 173 of a low grade. Therefore, the color quality standard of the glass cullet can be satisfied 174 when separating the mixed-color WGB using condition A in the color sorter.

Crushing experiment result 182
Recycling WGB into recycled glass aggregate must satisfy foreign substances and 183 size distribution quality standard. According to KS F2576, the content of organic 184 foreign substances such as plastic, wood, paper, and vinyl must be under 1% of the 185 aggregate volume. In addition, the aggregate must not emit an odor or contain or 186 chemical substances harmful to the environment. The majority of the foreign substances 187 in the samples used in this study were removed by hand, eddy-current separation, 188 magnetic separation, and air separation in the pre-treatment process, as we aimed to 189 satisfy the foreign substance quality standard. The particle size distribution of recycled 190 glass aggregate must satisfy the prescribed KS F2527 standard [25], displayed in Table  191 5. We conducted crushing experiments on WGBs less than 10 mm in size using a 192 shredder, roll crusher, hammer crusher, and VSI crusher, and presented the particle size 193 distribution of the crushed product in Fig 8.  shredder, respectively. As comparison with the prescribed particle size distribution, 198 indicated that the products of the roll crusher and shredder did not meet the minimum 199 particle size distribution standard because they were not adequately crushed. Shredder 200 is generally applied to the dismantling of ductile and composite materials, and it was 201 judged that the shredder was not suitable for crushing brittle materials such as glass. 202 The roll crusher crushes by delivering a single compressive force, because of which the 203 crushing of the WGBs, which have plate-shaped particles, was not accomplished. The 204 over-crushing was observed in the hammer crusher, because the hammer impacts 205 directly into sample. Furthermore, owing to the abrasion of the hammer, we decided 206 that the hammer crusher was not suitable for crushing the WGBs. The VSI crusher is 207 equipment that crushes particles based on an impact force, similar to a hammer crusher, 208 but does not over-crushing the sample, since the force is delivered to the inner wall 209 using an impeller. We, therefore, determined that the crushing occurred optimally only 210 with the use of the VSI crusher which satisfied the particle size distribution quality 211 standard of recycled glass aggregate.

Aspect ratio analysis result 215
According to BS 812 [29], in the case of using crushed mix color WGBs as recycled 216 glass aggregate, the angularity or absence of round aggregate particles is a very 217 important attribute since it affects the convenience of handling the mixture of the 218 aggregate and binder. In Zingg's classification system [30], when the aspect ratio of the particle exceeds approximately 1.5, it is mentioned that the mixture was significantly 220 influenced by a rolling motion and the associated traction. In addition, BS 812 defined 221 elongated particles as those with an aspect ratio greater than 2.2. Consequently, in this 222 study, we measured the aspect ratios of the WGBs before and after crushing and 223 confirmed them on a graph of cumulative frequency. The aspect ratio was measured 224 using the ImageJ program, which can measure and analyze length, area, and 225 circumference. After capturing the image of the sample, it is imported to Image J and 226 the shadow is removed, and the clarity enhanced using a contrast function. Then, to 227 demarcate and divide the particle boundary accurately, it was measured and analyzed 228 in the binary mode. Fig 9 presents the ImageJ process. 229 Figure 9. Aspect Ratio measurement using ImageJ 230 Fig. 10 shows the aspect ratio results of the feed and VSI crusher product for each 231 particle size. From Fig 10, it can be seen that approximately 20% of the natural sand 232 particles had an aspect ratio value over 1.5. In the case of the feed, the aspect ratio of 233 over 1.5 for 10.0-5.0 mm, which is a relatively large particle, was about 18% similar to 234 → → that of natural sand, but the 5.0-2.5 mm and 2.5-1.2 mm particles were confirmed to be 235 about 50%. In the case of the product, the aspect ratio over of 1.5 for 10.0~5.0 mm 236 increased by about 22% compared to feed particles, but the 5.0~2.5 mm and 2.5~1.2 237 mm particles decreased by about 20%. In addition, we confirmed the aspect ratio of 238 over 2.2, which was defined by BS 812 as an elongated particle standard, was lower 239 than that of natural sand for all particle sizes. The box plot in Fig. 11 shows that the 240 aspect ratio of the product had a low value with an average distribution as compared 241 with natural sand for all particle sizes. Thus, an aspect ratio higher than that of natural 242 sand could be obtained through crushing by VSI crusher, so we ascertained that the 243 WGB was appropriate for recycling into a glass aggregate.

4.3.VSI crusher DEM simulation 250
The crushing process uses the most energy in the recycling process can be improved 251 through optimal design and operating conditions, increasing the economic feasibility. 252 We confirmed the trends in the data required to set the optimum design and operating 253 conditions of the selected VSI crusher using DEM simulations. We replicated the VSI 254 crusher using CAD, and then conducted simulations after importing it into EDEM, 255 which is a 3D particle dynamics interpretation program based on DEM. In this 256 simulation, we set the length and rotation speed of the impeller as experimental 257 parameters, and the experimental conditions are presented in Table 6. In addition, the 258 physical properties of the WGB and VSI crusher and the aspect ratio of 1.5 of the 259 particles were applied through the aspect ratio analysis. We measured the force applied to the particle and the total particle numbers in the Based on the above results, we measured the tendency of the force applied to the 273 particle in the simulation results 7 to 10 s. Fig 14 shows the results of the force applied 274 to the particle as particle to particle and particle to the inner wall. In all conditions, the 275 added force from the particle to the particle was higher than the force added through the inner wall to the particle. Thus, the VSI crusher helps evaluate that the crushing rate 277 for collisions among particles is higher than that between particles and inner walls. Fig  278   15 presents the results of the force applied to the particle according to the length and 279 rotation speed of the impeller. The length of impeller increased from 0.3, 0.5 to 0.7 m, 280 and correspondingly the force applied to the particle increased. The rotation speed of 281 impeller also increased from 1,000, 1,200 to 1,400 rpm, and the force applied to the 282 particle increased accordingly. The greater the length and rotation speed of the impeller, 283 the greater the force applied to the particle, but the length of the impeller was more 284 affected by the force than the rotation speed. Consequently, we confirmed that the VSI 285 crusher is more influenced by the impeller length than by the rotation speed. We 286 assessed the trends in the data obtained in this study to be meaningful. However, in the 287 future, to establish optimum design and operating condition settings, further studies 288 should be conducted.

Conclusion 298
In summary, we presented the optimum process for recycling WGBs into a glass 299 cullet and recycled glass aggregate. To produce a glass cullet, we tested the optimum 300 operating condition selection experiment using a color sorter. For recycling WGB as recycled glass aggregate, we conducted optimum crusher selection, aspect ratio 302 measurements, and a DEM simulation experiment. The findings of this study are 303 summarized as follows. As per various analyses, the sample used in this study was soda 304 glass, which consists of Na2CO3, SiO2, and CaO. To produce glass cullet, the color 305 sorting is essential and removing the glass in the order of ambergreenamber, green 306 is the optimal condition to satisfy the color quality standard. This result indicates that 307 because the RGB value of green and clear inputs were distributed similarly, when green 308 was removed preferentially, the ratio that needs to be removed is increased. The VSI 309 crusher was selected to be the best suited for producing recycled glass aggregate 310 through a crushing experiment. As a result of the aspect ratio analysis, an aspect ratio 311 higher than that of natural sand could be obtained through crushing, and therefore the 312 WGB is suitable for recycling into a glass aggregate. The trend of the number of 313 particles and force of the VIS crusher had a linear relationship with the length and 314 rotation speed of the impeller; of these, the length of the impeller was more affected by 315 the. We assessed that the simulation results provided useful information for establishing 316 the optimum design and operating conditions. In conclusion, this study provides 317 information on the WGB recycling process, and we believe that this result will 318 contribute to increasing the rate of recycling. 319

Availability of data and materials 320
All data generated or analyzed during this study are included in this article and its 321 Energy dispersive X-ray spectrometer OXFORD SwiftED3000 407