Figure 19 – uploaded by Digital library Bohemia

See full PDF downloadDownload figure

Fig. 17. Scheme of the polished section of the bloom from the smelt II.2. Above: distribution of the carbon content (dark), etched by nital. Below: segregation of phosphorus (dark), etched after Oberhoffer. Crossed areas: parts with sponge iron and isolated grains of ferrite. 2-5. — Figure 17 Scheme of the polished section of the bloom from the smelt II.2. Above: distribution of the carbon content (dark), etched by nital. Below: segregation of phosphorus (dark), etched after Oberhoffer. Crossed areas: parts with sponge iron and isolated grains of ferrite. 2-5.

Related Figures (26)

Fig. 1. Panoramatical view on the experimental bloomery at Brezno 1964, Furnaces of the Slav type are built-in, a Roman period shaft furnace behind. The bloomery process must be considered as the classical and original method of iron extraction. A very long and important part of the evolution of iron extraction is represented by this art, which — under European conditions — was the only practical way to obtain this useful metal. The product of the bloomery process was subsequently processed into various modifications of wrought iron and wrought steel. Research on this early industrial process is also of interest in the science of metallurgy.

Fig. 3. Profile of the Slavic furnace of the type Zelechovice prepared for the smelt II.2. On the right hand: the position of the metallic bloom in the cavity during the process.

Fig. 4. Roman period shaft furnace of the type Scharmbeck (the shaft is transportable). On the right hand the positior of the product after the smelt.

Fig. 5. Section of the Slavic furnace of the type Zelechovice I after the smelt. This smelt was not successful. The figure shows the lining of the shaft. 1—3, a—c controll points. construction, dimensions, and the composition of their refractory linings are well known.” Furnace no. and 147 XXV served as the prototype. The dimensions of the furnaces were: height 90 cm (shaft hearth 85 cm), throat diameter 35 cm, shaft bottom diameter 15 cm, length (cavity included) cm, tunnel length 36cm, tuyere diameter 4 cm, inclination of the tuyere axis 65°. These dimensions differed slightly after finishing (Figure 5). Furnace II was a little lower, and its tunnel and cavity were a little wider. The tunnel of this furnace was shortened before smelting no. 2 took place (Figure 3). Furnace IV was used for preliminary experiments with the building style and air circulation; later on it served as a drying furnace. Furnaces I and II were each equipped wit h three control holes, inclined at about 15° and positioned on the south side (Figure 5). Each hole was designed to allow measurements in three different positions: (A) at the shaft axis, (B) in the middle of the shaft radius and (C) at the furnace wall.

Fig. 6. Scheme of the small charcoal pile used for tests. 4 horizontal channels, 5 perpendicular shaft between the rods, 6 sand, 7 turf coat and branches, 8 pine logs 4.1, Burning charcoal. Pine charcoal is not available on the market; moreover, charcoal burnt in piles is more suitable than that from distillation retorts for metallurgical experiments.12? We therefore erected small piles on the site.

Table I — Manufacture of Charcoal The ore supplied to the Zelechovice-type furnaces was a sort of haematite, very similar to the ore samples found during excavation of the original bloomery. For the Scharmbeck-type furnace we used Kiruna D magnetite with a higher phosphorus content, since we wanted to study the transition of phosphorus into slag and reduced metal. 4.1.3. Dismantling the piles. After the piles were dismantled carbonized logs were observed which flared up in the open air. It was necessary to spray the coal with sand and water. Only the very ends of the lower logs were uncarbonized. We crushed the charcoal after cooling into pieces of about 4—5 cm, and later on reduced these pieces to 2 cm.

Table III — Composition of the ore for the Scharmbeck furnace 5.3. Ore dressing. Both sorts of ore were dried in air for about four hours on a simple brick-and- metal hearth. Large pieces of ore had to be crushed by hand (Figure 9) using a sledge hammer on a stone block. We originally wanted particle sizes up to 20 mm, and later on reduced the size to 10 mm. The crushed ore was then sifted. Table III shows the composition of the ore for the Scharmbeck furnace. It is clear that the pho phorus content of the Kiruna D ore is not constant; this sort of raw material was not very goo for studying the correlation of phosphorus content in the ore and in the reduced metal.

Table IV — Composition of the tested haematite square metal sheet. The third layer of wood was covered with an ore charge, and then log and ore layers were alternated up to a height of 65—70 cm. We placed in the pile a total of 40 kg of ore. After three-quarters of an hour another 20 kg of ore was put in; at the height of 35 cm, new log layers were added and the third charge of ore was spread on the surface (Figure 9). The operation took 6 hours and 15 minutes, about 250 kg of wood and 185 kg of ore (haematite) were used, and the weight of ore after roasting was 174 kg. This operation was repeated in roasting pile IT (rectangular, height 55 cm) which burned for 3 hours and consumed 74kg of ore and 150 kg of wood (weight loss 4kg). The roasted ore was magnetic (Fe,O,). During roasting the first phase of reduction took place. Tables IV and V summarize the roasting of the ore. The Kiruna D ore was used without roasting.

Various measurements were made during the smeltings: rate of gas flow, gas analysis and temperature (both the latter measured at short intervals). Fuel and ore charges were also weighed in advance.

Fig. 7. Building the charcoal pile.

Table VI — Consumption of raw materials 7.4, Smelt in furnace III (Scharmbeck type). Drying and preheating were finished at noon of the same day on which they had been begun. At 1 : 30 pm the first ore (Kiruna D magnetite) and fuel charge was put in. The shaft was filled up only to control point 1. We had intended to test this furnace by induced draught, but the process went on so slowly (maximum temperature 1042°C at control point 3 and 650°C at control point 2; CO content at 4 pm only 6%) that it became evident that this sort of ore could not be smelted in this type of furnace by induced draught. We therefore tried to introduce forced draught (21 liters/min). After this change the temperature increased to 1445°C at control point 3 and the CO concentration reached 23—25% shortly after 5 pm. At 6 : 00 a flame was seen at the top of the furnace and fluid red slag was observed at some control points, but pieces of unreduced ore appeared in the hole opposite the tuyere. Blowing was stopped at 6 : 33 pm and the furnace was allowed to cool overnight. On the next day the position of the slag ball on the shaft was determined (Figure 4). The slag was situated above the tuyere, and close to it were some pieces of unreduced ore. The bowl hearth was full of charcoal: as expected, no reaction took place in the bowl, and the slag did not go in. The furnace wall was only slightly attacked by the slag. Although all samples of the slag conglomerate were magnetic, no iron sponge could be found and no entrained metallic iron could be separated from the slag mass. Thus this smelt could not be considered successful; the following conclusions, however, could be drawn: in water. The data for the smelt are given in Table VI. The yield on the separated iron ore was about 5 kg, about 25% of the iron content of the ore. This yield is comparable with the productivity of the ancient bloomery process. But only about 2:5 kg of the iron was fit for forging due to our lack of the necessary skill. The zones of the iron sheet network in the slag indicate that the reheating process was not completed; perhaps 4—5 hours instead of 2 are necessary for this process.

Fig. 9. Preparing ore. Above: crushing and sieving. Below: roasting under reducing conditions.

Fig. 10. Above: chromatograph for analysing gases. Below: section of the Slavic furnace I after the smelt and cooling.

Or tne reVU, in the ore. Microscopic investigation: as already mentioned, only tips of the bloom (presumably those adjacent to the tuyere mouth) had the ferritic structure. There is a slow transition from the ferritic-pearlitic zone to a pearlitic or martensitic zone (Figure 19). The microhardness of the ferrite was very high (247—267 dphn). In other parts of the bloom there was a hard martensitic or bainitic structure of quenched high-carbon steel (microhardness 486—616 dphn). Ferrite needles (microhardness 234-321 dphn) and cementite needles (microhardness 1355—1810 dphn; Figure 19) were found in certain spots. In those spots which were spontaneously tempered by the heat of the body, the structure was pearlitic or sorbitic (microhardness 310—390 dphn; diameter

9.3. Mineralogical investigations of slags. The most important samples are those of the smelt in furnace I (specimens la—b, 22). They could be taken from the exactly known places of the con- glomerate in the cool furnace. The choice of the slag material from the furnace II (smelt 2) was

(length 0-01—0-3 mm) even in the fayalite areas were stated. Slag from the lower drops of the conglomerate is quite different. Its main component is wiistite FeO (50—60%), fayalite amounts to 40—50%. Wiistite occurs sometimes in allotriomorphic grains or in form of dendrites. Elongated fayalite (length 0-15—1-5 mm) could be also observed among wiistite. Instead of magnetite a glassy matrix filling the intercrystaline space was stated. The amount of this glassy mass did

Fig. 18. Microstructures of the bloom from the smelt II.2 (specimen 270). 1 white ferrite grains in the spongeous area (crossed on Fig. 17), dendrites of wiistite, fayalite, etched by picrate of sodium, 100. 2 decarburized tip of the bloom, ferrite and inclusions, etched by nital, 300 x. 3 transitional area with pearlite, nital, 300 x. 4 the same, ferrite, pearlite, etched by nital, 300 x.

Fig. 19. Microstructures of the bloom from the smelt II.2 (specimen 270). 1 carburized area, dark pearlite, white ferrite, etched by nital, 550 x , 3 dark needles and network of cementite, etched by picrate of sodium C,H,(NO,),ONa, 200 x, 3 dark martensite, white cementite, etched by nital, 300x, 4—5 martensite, dark spots of troostite, white cementite needles, 300 x . 6 troostite and two different areas of martensite, traces of cementite, etched by nital, 300x.

Fig. 20. Mineralogical investigation of the slag. 1 slag from the smelt in the furnace I, upper part (specimen la), opaque black magnetite, grey crystallized area of fayalite, white voids, 11x. 2 detail of the fayalite zone, black dendrites of magnetite, 132 x, both as thin sections. 3 slag from the same smelt, but from the bottom, dark grey fayalite and opaque wiistite, thin section, 11 x. 4 the same, polished block: light grey wiistite, grey crystals of fayalite, dark grey glassy matrix, 200x.

Fig. 21. Mineralogical investigation of the slag. 1 Slag from the smelt II.2: white iron sponge, light grey dendrite of wiistite, grey fayalite, dark grey glassy matrix with fine crystallized fayalite, polished block, 200 x. 2 Slag from the furnace III: white iron crystals, light grey wiistite, dark grey fayalite, black strips of glassy mass, polished block, 200 x . 3 Slag from the furnace III, inclusion of the wall fragment in the opaque wiistite, white voids, thin section 11x. 4 Detail of the inclusion, quartz grains, muscovite and black dendrites of the penetrating wiistite, thin section, 132 x.