• Journal of Conservation Science
• /
• v.35 no.2
• /
• pp.145-152
• /
• 2019

The metal-manufacturing method and smelting temperature of ancient metal-production processes have been studied by analyzing the principal elements and microstructures of slag. However, the microstructure of slag varies according to the solidification cooling rate and types and relative amounts of various oxides contained within the smelting materials. Hence, there is a need for accurate analysis methods that allow slag to be distinguished by more than its composition or microstructure. In this study, the microstructures of slag discharged as a result of smelting iron sands collected from Pohang and Gyeongju, as well as the slag excavated from the Ungyo site in Wanju, were analyzed by using metalloscopy, scanning election microscopy-energy dispersine X-ray spectroscopy(SEM-EDS) and wavelength dispersive X-ray fluorenscence(WD-XRF). Furthermore, the microcrystals were accurately characterized by performing Raman micro-spectroscopy, which is a technique that can be used to identify the microcrystals of slags. SEM-EDS analysis of Pohang slag indicated that its white polygonal crystals could be Magnetite; however, Raman micro-spectroscopy revealed that these crystals were actually $ulv{\ddot{o}}spinel$. Raman micro-spectroscopy and SEM-EDS were also used to verify that the coarse white dendritic structures observed in the Gyeongju-slag were $W{\ddot{u}}stites$. Additionally, the Wanju slag was observed to have a glassy matrix, which was confirmed by Raman micro-spectroscopy to be Augite. Thus, we have demonstrated that Raman micro-spectroscopy can accurately identify slag microcrystals, which are otherwise difficult to distinguish as solely based on their chemical composition and crystal morphology. Therefore, we conclude that it has excellent potential as a slag analysis technique.

• Journal of Conservation Science
• /
• v.33 no.2
• /
• pp.75-83
• /
• 2017

This study produced smelting slag through the reproduction of an ancient iron manufacturing technique, with the aim of facilitating a comprehensive understanding of the process by analyzing the slag components. The research suggests an interpretation method using the ratio of the subcomponents relative to the main slag components as an alternative to existing methods. We investigated the component source within the smelting furnace from which the slag is derived by developing an understanding of the tendency between slags. Based on bivariate graph and triangular coordinate data analysis, it was found that a slag can be categorized according to its components. The groups were identified as the ore slag group(centered on the ore), and the clay slag group(centered on clay and granite soil). This research determined that it is possible to estimate the components derived from the slag, depending on which group they belong to or resemble, as shown in Figure 4~7. It was found that a comprehensive understanding of the ratio between the components was more accurate than a simple analysis of the contents, for the interpretation of ancient iron manufacturing processes. This is based on the fact that a higher ratio of $TiO_2$ was detected by the components analysis, and an analysis of all the slag showed that the value of $CaO/SiO_2$ ratio was lower than 0.4, which corresponds to the reproduction experiment condition in which flux was not used.

• Journal of Conservation Science
• /
• v.26 no.2
• /
• pp.171-182
• /
• 2010

The slag excavated from Gyesil-ri in Gongju, Yeonje-ri in Cheongwon and Beopcheonsaji (temple) site in Wonju are analyzed by X-ray Fluorescence Analyzer, metallurgical microscope, SEM-EDS etc., for chemical composition and microstructure to figure out the raw material and the iron manufacturing technique. First of all, as a result of principal component analysis, the total Fe-content of slag from Gyesil-ri is 39 to 44% and the modified rate is 15 to 21%, which is common in ancient iron slag. Yeonje-ri site is found the ancient iron-smelting furnace. The total Fe-content of slag from Yeonje-ri is 41 to 43% and modified rate is 18~30%, which is also the general value in the ancient slag. However only slag is excavated in the residential area at Beopcheonsaji site and there is no iron making relic. In addition, the result of principal component analysis contains that the total Fe-content of Beopcheonsaji site is 52 to 57%, and modified rate is 8 to 14%. It shows that the total Fe-content of Beopcheonsaji site is higher than relic from Gyesil-ri and Yeonje-ri and the modified rate is lower than other sites. This results mean that recollecting rate of Fe in Beopcheonsaji site is lower than other sites. Also, as a result of minor elements analysis, the slag from Gyesil-ri has the higher level of Ti, V and Zr than other sites and the microstructure are observed as magnetite and ulvospinel, so that the raw material of slag is iron sand. But the slag from Yeonje-ri and Beopcheonsaji site are identified to use iron ore. As a result of microstructure observation, fayalite, gray-columnar crystal, is found in the slag from Yeonje-ri and big wustite as main phase is observed in the slag from Beopcheonsaji site. This study show that the slag from Yeonje-ri is made of smelt ash produced during smelting works and the slag from Beopcheonsaji site is made of forging ash produced during forging work concerning the excavated location and the microstructure.

• Journal of Conservation Science
• /
• v.3 no.1 s.3
• /
• pp.13-18
• /
• 1994

Chemical analysis and metallographic observations of the iron slag were carried out in an attempt to estimate the old iron-making process. The slag containing $9.3\%\;TiO_2$ without Cu indicates that the ore used for smelting was sand iron, not rock ore. The phases identified in the slag were $ulv\ddot{o}spinel$, magnetite, $w\ddot{u}stite$, fayalite etc. This also supports the fact that the smelted ore was iron sand. The total amount of Fe and slag-making components$(=SiO_2+Al_2O_3+MgO+CaO)$ were $40.7\%\;and\;36.1\%$, respectively. These values were average ones found in the old slags formed in the ancient iron-making process. Assuming that $TiO_2$ in the ore combines with FeO, resulting in the formation of $ulv\ddot{o}spinel$, the estimated temperature of smelting was found to be about $1200^{\circ}C$.

• Journal of Conservation Science
• /
• v.33 no.4
• /
• pp.297-312
• /
• 2017

In the ancient iron-making process, a slag former was often added so that iron and other minerals in the ore could be smoothly separated. However, there are insufficient data for judging whether a slag former was added. Thus, in this study, we conducted a smelting experiment to understand the material characteristics of a steel structure that differed depending on the addition of a slag former. It was found that the steel structure produced in the first experiment had a total Fe content of 39.45-52.94 wt%, which decreased to 34.89-38.92 wt% in the second and third experiments. CaO compounds such as calcite, gehlenite, and hercynite appeared, in addition to iron oxides, after the addition of a slag former. As a result of an assessment of whether a slag former was added by comparing the ratio between the components, it was found that the ratio of $CaO/SiO_2$ was 0.42. From a comparative analysis of $Al_2O_3/SiO_2$ and $CaO/SiO_2$, it was judged that the ratio of $Al_2O_3$ and $SiO_2$ can be utilized as an index to judge similar systems of smelting process (ore, furnace wall, and fuel).

• Journal of Conservation Science
• /
• v.34 no.4
• /
• pp.273-281
• /
• 2018

$Ji{\check{u}}di{\grave{a}}n$ iron production site in China is a relic smelting site, which in the past produced pig iron. In this study, scientific analysis of the smelting furnace and collected slag was conducted to reveal some aspects of the ancient Chinese smelting technique. A 3D model of the smelting furnace showed a narrow lower part and an upper section which increased in diameter upwards. Although the smelting furnace relic does not include the upper part and its complete shape cannot be predicted, the remaining part suggests that the furnace had a larger diameter in the central part compared to the upper and lower parts. Most of the collected slag was completely vitrified. Long prismatic fayalite was observed in the matrix of some samples. The iron particles contained phosphorus, which could not be discharged during smelting work. In addition, as the $CaO/SiO_2$ ratio was 0.42 or lower in the results of the content analysis, no CaO slag former had been added. However, the ratio of $CaO/SiO_2$ to $Al_2O_3/SiO_2$ did not have a constant trend. This needs to be investigated in a further study.

• Journal of Conservation Science
• /
• v.34 no.1
• /
• pp.39-50
• /
• 2018

This study replicated traditional smelting methods to produce iron blooms from sand iron. The metallurgical properties of the slag and the iron blooms were analyzed. The sand iron materials used in the smelting experiments, which were based on ancient documents, were collected from Gyeong-Ju and Pohang. Analysis by WD-XRF and XRD showed that Gyeong-Ju's sand iron contains a high-titanium, with magnetite, and Pohang's sand iron contains a low-titanium, which magnetite and ilmenite were mixed. Analysis of the slag with XRD, and the micro-structure with metal microscopes and SEM-EDS, confirmed that the major compounds in the slag of the Gyeong-Ju's sand iron were fayalite and $w\ddot{u}stite$, and those in the slag of the Pohang's sand iron were titanomagnetite and fayalite. The differences in the main constituents were confirmed according to the Ti quantity. Finally, we observed the microstructures of the iron blooms. In the case of the iron bloom produced from Gyeong-Ju's sand iron, the outside was found to be dominantly a pearlite of eutectoid steel, while the inside was a hypo-eutectoid steel where ferrite and pearlite were mixed together. While, the major component of the iron bloom produced from Pohang's sand iron was ferrite, which is almost like pure iron. However, there were many impurities inside the iron blooms. Therefore, this experiment confirmed that making ironware required a process that involved removing internal impurities, refining, and welding. It will be an important data to identify the characteristics of iron by-products and the site through traditional iron-making experiments under various conditions.

• Journal of Conservation Science
• /
• v.33 no.6
• /
• pp.519-532
• /
• 2017

Studies on ancient ironmaking technologies are primarily based on archaeological surveys and scientific analysis data, and technological systems are examined by comparing the results of restorative experiments. In this study, to examine the ancient iron production technologies such as smelting and smithing in the Jungwon area, a restoration experiment was conducted based on archaeological data, and the iron and slag, etc. produced in the experiment were analyzed. Further, the changes in physicochemical properties due to the smelting of the raw material, specifically, iron ore were determined, and the smithing process, which involves fabrication of ironwares, was analyzed along with the characteristics of each step. In the case of smelting, increasing recovery rates and production of high-quality primary iron material were important for the following processes. For the iron bars produced through the smithing process, it was found that quality improvements made by reducing physical defects such as inclusions or gas holes were more important than the composition of the iron itself. The study also yielded comparative study data for various byproducts, such as smithing slag, which could be utilized in other ironmaking technology studies.

• Journal of Conservation Science
• /
• v.37 no.1
• /
• pp.34-42
• /
• 2021

This study analyzes six slags excavated from the iron making site in Gogi-ri, Namwon, Korea to understand the characteristics of the ruins, and to confirm the iron making process performed at the time. The chemical components of the iron making by-products from the Gogi-ri site were analyzed, and the findings indicate total Fe contents between 23.24% and 37.56%, which are lower than the typical total Fe content found in ancient iron making processes. The deoxidation agent contents of the slags ranged from 43.88% to 58.13%, which are higher than the typical deoxidation agent content of ancient iron making processes. The high content suggests smooth separation between iron and slags, and TiO₂ detected from the site suggests the use of materials with high titanium content in the iron making in the region. As for the microstructures of the slags, some slags have long pillar-shaped fayalites, while others have pillar-shaped wüstite along with ulvöspinel. Slags from the forging furnace show hammer scales created by both the earlier stages and later stages of forging work. The findings suggest that the iron making site in Gogi-ri, Namwon, Korea used to be an iron making facility where a full range of iron making process was carried out ranging from smelting to forging, and the ironmakers used a wide array of technologies to manufacture iron products.

• Journal of Conservation Science
• /
• v.35 no.4
• /
• pp.373-383
• /
• 2019

We characterized the smelting process and smelting furnace through scanning electron microscopy-energy dispersive spectroscopy, wavelength dispersive X-ray fluorescence, X-ray diffraction, and raman micro-spectroscopy with 13 relics including slags and furnace walls excavated from square-shaped building sites and pits of the Three Kingdoms site at the Ungyo site section I. Our results revealed that the principal components were FeO and SiO₂; and CuO, PbO, and ZnO were contained in small quantities. Furthermore, fayalite, magnetite, augite, copper, and cuprite were found. High contents of FeO or SiO₂ components seem to have been added to form fayalite to remove gangue in the smelting process. The relatively low content of S detected in the copper prills suggests that roasting was performed well. Cristobalite and mullite, which are minerals that indicate high-temperature found in the furnace wall, show that the smelting temperature was higher than 1,250℃. The findings of this study show a high possibility that the Wanju Ungyo site is smelting remains of copper ores, which are nonferrous metals, rather than iron. Various smelting byproducts excavated in this area in the future will help us better understand the copper smelting process that may have been performed since ancient times.