Viscosity and thermal properties of slag in the process of autogenous smelting of copper

Abstract

The temperature dependences of slag viscosity in the process of autogenous smelting of copper–zinc concentrated products in Vanyukov furnaces (VF) at the Sredneuralsky Copper Smeltery JSC have been revealed. Values of viscous flow activation energy have been determined. Phase transition temperatures in the processes of slag heating and cooling have been measured. The influence of calcium oxide additives on viscosity and on the crystallised slag liquidus point has been evaluated. The results obtained may be used for slag composition improvement with the purpose to reduce copper losses and simplify material preparation to subsequent depletion operation.

Keywords

Copper–zinc concentrate Autogenous smelting Viscosity Thermal properties

Introduction

The autogenous smelting of copper–zinc concentrates at the Sredneuralsky Copper Smeltery JSC is carried out in Vanyukov furnaces (VF). The adopted technology is characterised by a high specific productivity of the unit and a large ratio of the oxygen supply to the concentrate mass to be processed (^{Vanyukov et al., 1988}). Under given conditions, formation of the slag with increased Fe³⁺ content is a highly probable event that would undoubtedly have an effect on physicochemical properties of the melt, including its viscosity. The melting temperature and viscosity of the melt determine the completeness of matte–slag emulsion separation and copper losses with suspended sulphide particles. Optimisation of the slag viscosity is one of the reserves allowing improvement of the melting unit performance and saving non-ferrous metals losses.

Properties of the slag with higher iron oxides content were considered by ^{Vanyukov and Zaitsev (1969}, ¹⁹⁷³⁾, ^{Denisov et al. (1999)} and ^{Ji et al. (1997)}. As it was reported earlier by ^{Seki and Oeters (1984)}, the main factors determining the melt viscosity are the temperature and composition being expressed by the slag basicity coefficient. Extensive information concerning binary and ternary systems viscosity was given by ^{Verein Deutscher Eisenhüttenleute (1981)}. Viscosity of the multicomponent melts FeO_x–SiO₂–CaO–MgO–Al₂O₃ had been measured (^{Sheludyakov, 1980}; ^{Zhang et al., 2012}) in the separate narrow compositional ranges that are typical for specific manufacturers. The data on copper-smelting slag with increased iron (III) oxide and zinc oxide contents are limited (^{Vaisburd et al., 2002b}; ^{Huaiwei et al., 2012}). Moreover, nothing is known about thermal properties of such slag.

The goal of the present study was to find out viscosity and thermal properties of the slags of the Sredneuralsky Copper Smeltery JSC and to substantiate the slag composition providing copper losses minimisation.

Methods and Materials

The slag viscosity has been measured by means of a home-made oscillatory viscosimeter (^{Stengelmayer et al., 1985}). The instrument works in the mode of resonance oscillation and allows determination within the range 0·1 and 10 Pa s. Experimental set-up (Fig. 1) consists of a heating furnace, a spring-supported viscosimeter, a ring-type magnet, an active oscillator and a measuring cell. Experimental procedure was similar to the one described by ^{Musikhin et al. (1992)} and ^{Denisov et al. (2013)}. The slag samples under study were kept in molybdenum crucibles when cooling the melt starting at 1375°C with the average rate about 24 K min⁻¹. The surface of molten slag was blown round with a flow of gaseous argon during the whole test. Every time the test probe made of tungsten wire 300 mm in length and 1·5 mm in diameter was immersed into the slag to a depth of 10 mm. Smooth sinking of the probe in the melt under study was controlled by a special screwed hoisting device taking the point of touching the liquid slag surface as zero point. The frequency of the probe oscillations was 50 Hz. Analytical signals were received by a digital voltmeter connected parallel to the pickup coil. The temperature was measured by Pt–Pt/Rh thermocouple. Its hot junction was placed inside the melt and the opposite cold terminals were thermostated at 0°C. Calibration of the instrument had been done using reference liquids. As was recommended by ^{Stengelmayer (1973)}, they were glucose-containing water solutions of cadmium iodide and barium iodide with density and viscosity 2·7–3·0 g cm⁻³ and 0·04–12·5 Pa s, respectively. Viscosity of the reference solutions had been measured by the well-known Stokes method. Each slag sample was tested doubly in order to check the reproducibility of the results. It was shown that the reproducibility does not exceed measuring error. The relative error on the viscosity measurement was ±5%.

Figure 1

Experimental set-up for viscosity measurement

Thermal behaviour of the slag of existing commercial production has been examined by means of the thermoanalyser Netzsch STA 449C Jupiter, Selb, Germany using simultaneous measurements by thermogravimetry (TG) and differential thermal analysis (DTA) or differential scanning calorimetry (DSC) in alumina analytical crucibles. The rate of continuous heating up to 1200°C and subsequent cooling down to 850°C was 20 K min⁻¹. The flowrate of purging argon was 30 mL min⁻¹. The masses of the powdered slag samples in DSC and DTA tests were about 23 and 32 mg, respectively. Phase transition temperatures were estimated with the help of the calibration file, which was generated as a result of preliminary testing of the melting point of pure metals (In, Sn, Al, Au, Pd) constituting the set of certified reference materials supplied by NETZSCH Gerätebau GmbH. The standard features and setting of the Netzsch Proteus Thermal Analysis software were used for quantitative estimation of those heat effects that were observed on DSC and DTA thermoanalytical graphs. Experimental facilities and procedures provided the resolving ability of mass change and temperature determinations ±0·01 mg and ±1–5 K, respectively. Each run was repeated one time and the observed reproducibility was commensurate with the measuring error.

Test subjects were various slag blends formed in the process of oxidising melting of copper and zinc concentrated products in VF for copper production at the Sredneuralsky Copper Smeltery JSC. These specimens were picked out from the mixers. Slag compositions are shown in Table 1. Apart from the compounds mentioned below, the slag samples contain 0·21–0·27 wt-% Pb, 0·01–0·10 wt-% As and 0·06–0·09 wt-% Sb. Along with industrial specimens (# 1–# 8), the properties of the slag obtained by adding 8·2 (# 9), 9·9 (# 10) and 13·1(# 11) wt-% of calcium oxide to the sample # 8 have been studied. The used calcium oxide was prepared by a roasting of CaCO₃ at 950°C for 30 min.

Table 1

Chemical analyses of the slag samples

Sample #	Content/wt-%
Sample #	Fe	Fe₃O₄	SiO₂	Al₂O₃	CaO	Cu	Zn	S	MgO
1	38·0	7·7	33·2	3·7	3·2	0·80	4·55	1·0	1·0
2	35·5	3·3	33·5	4·2	3·1	0·70	4·92	1·0	1·1
3	34·9	4·4	33·8	4·0	3·1	0·62	5·05	0·8	1·2
4	36·8	3·9	33·0	4·3	3·2	0·71	3·64	0·9	1·4
5	37·9	9·1	30·3	3·6	2·8	0·97	3·74	1·1	1·1
6	36·2	6·2	31·9	3·9	2·4	0·71	3·82	0·9	1·1
7	36·5	6·6	31·5	3·9	2·5	0·77	3·89	0·9	1·1
8	36·7	3·5	31·3	4·0	2·6	0·76	3·47	1·0	1·1
9	31·9	3·7	27·2	3·5	8·0	…	…	1·0	1·0
10	31·8	3·1	27·1	3·5	9·9	…	…	0·9	0·9
11	30·8	2·4	26·3	3·3	13·1	…	…	0·9	0·9

Made by remelting of the sample # 8 together with calcium oxide in argon atmosphere.

Results and Discussion

The microstructure of the crystallised slag of copper–zinc concentrates smelting in VF at the Sredneuralsky Copper Smeltery JSC (Fig. 2) was analysed using the scanning electron microscope JSM-59000LV. The main phases of the slag are iron silicates, magnetite, fine sulphide inclusions and matte particles 100–300 μm in size (^{Selivanov and Gulyaeva, 2012}). Silicates are represented by the coarse crystals of fayalite Fe₂SiO₄, while the interspace is filled with pyroxene (Ca, Fe)SiO₃ and olivine–pyroxene eutectic. In terms of phase composition and structure, this microstructure is close to that of typical slag obtained by the same way in the process of autogenous smelting at the other copper-smelting plants (^{Vaisburd et al., 2002a}).

Figure 2

Micrographs of the slag # 8 before a and after b remelting with CaO. 1: fayalite; 2: magnetite; 3: matte particles; 4: pyroxene; 5: sulphides; 6: iron-calcium olivine

Increase of CaO content in the slag results in the structure refinement, silicate eutectic suppression and magnetite crystallisation in the form of inclusions of irregular and elongated shape (^{Selivanov et al., 2013}). The iron–calcium olivine Fe_1·5Ca_0·3Zn_0·1Mg_0·1Si_1·1O₄ appears instead of fayalite. The matte inclusions in the sample with higher calcium oxide content have irregular shape and a size of 100–150 μm; fine sulphide inclusions are observed at the borders of oxide phases.

The results of the molten samples viscosity measurement are within the range 0·23–0·48 Pa s at 1300°C and 0·26–0·58 Pa s at 1245°C (Fig. 3). Slag # 5 shows divergent viscosity values as it achieves 1·30 Pa s at 1245°C. This may be explained by higher magnetite content (9·1%). Analysis of the experimental observations has shown that the growth of magnetite content in the slag causes a rise of viscosity values.

Figure 3

Viscosity of the slag samples # 1–# 8 versus temperature

The temperature dependence of the viscosity η is represented in the form of Arrhenius-like equation 1 where A and B are the certain constants for the given slag composition, E_η is the viscous flow activation energy (J mole⁻¹), R is the universal gas constant, T is the temperature (K). The coefficients A and B of the equation (1) were calculated from the experimental data on viscosity values and are given in Table 2. The temperatures of the slag crystallisation beginning (T_η) have been determined from viscosity data as inflection points of lnη curves plotted against reciprocal temperature, and viscous flow activation energy has been calculated from the slope of these curves in the area of melt homogeneity. The estimated values of viscous flow activation energy and crystallisation temperatures of molten slag are within the ranges of 21·8–81·8 kJ mole⁻¹ and 1047–1235°C, respectively (Table 2). Viscosity characteristics of the slag under consideration are close to those of the slag obtained by VF processing at a Balkhash Copper Smelter as reported by ^{Vaisburd et al. (2002b)}. According to ^{Chen et al. (2013)}, the values of viscous flow activation energy in the system FeO–SiO₂ lie between 27·2 and 95·0 kJ mole⁻¹ and the ratio FeO_x/SiO₂ has prevalent influence on the values of viscosity.

Table 2

The values of A, B, activation energy, temperature range of homogeneity, and crystallisation temperature of the slag according to the results of viscosity measurements

Sample #	lnη = A+B·(E_η/R·T)		E_η/kJ mole⁻¹		T_cr/°C		Temperature range/°C
Sample #	−A	(B·E_η/R)·10⁻⁴	Experiment	Calculation	Experiment	Calculation	Temperature range/°C
1	4·580	0·590	49·1	48·2	1067	1102	1350–1192
2	3·283	0·290	25·8	30·5	1047	1056	1350–1247
3	3·262	0·311	25·8	23·7	1104	1065	1350–1274
4	4·189	0·536	44·6	44·5	1235	1208	1350–1244
5	6·320	0·884	73·5	76·5	1206	1252	1350–1288
6	3·027	0·263	21·8	43·6	1122	1135	1350–1274
7	6·720	0·915	76·0	50·9	1167	1118	1350–1274
8	7·600	0·985	81·8	80·8	1157	1166	1350–1274
9	2·660	0·159	13·2	15·0	1023	1049	1375–1246
10	3·467	0·277	23·0	17·7	1010	1046	1375–1171
11	4·961	0·379	31·5	34·0	1040	1005	1373–1283

On the assumption of the viscous flow activation energy dissipation only for making vacancies, the authors estimated the radius r_q of the latter from E_η values according to the formula by ^{Esin and Geld (1966)} 2 where σ is the surface tension, N_A is the Avogadro number, E_η/N_A is the energy of a single hole formation. The same value σ = 360 mN m⁻¹ as reported by ^{Vaisburd et al. (2002b)} was taken for all the slag samples. The hole radii calculated using equation (2) for the slag samples # 1–# 8 are within the range 0·89·10⁻¹⁰–1·74·10⁻¹⁰ m. These results are compatible with the oxygen ion size (1·32·10⁻¹⁰ m). The data obtained demonstrates at a first approximation that the monoatomic ions are the units of a viscous flow in the slag of autogenous smelting.

Proper selection of the model for compositional dependence of viscosity is a difficult task since the slags under consideration are multicomponent systems containing sulphide components in addition. ^{Bazán et al. (2006)} had studied fayalite slags and applied the index of basicity K_v to viscosity calculations. Taking into account specific composition of the copper-smelting slag containing zinc oxide with a marked alkalinity, one can express this K_v by component concentrations as the ratio 3

The estimated dependences of the slags # 1–# 8 viscosity on K_v value at the temperatures 1250 and 1300°C are described by the following equations with the R² coefficient of determination about 0·90 and 0·46, respectively. 4 5

In general, the raise of K_v for the slag under study from 1·48 to 1·71 results in the growth of η values. The levels of SiO₂ and Fe in the slag vary within ±1·8% around the mean values 32·3 and 36·6%, respectively, and the concentration of Fe₃O₄ changes within ±3·5%, relative to the mean 5·6%. That is why the Fe₃O₄ content has significant effect on the K_v values in the considered data array. Hence, viscosity of the slag of copper–zinc concentrates smelting increases considerably with the growth of available magnetite level.

Adding of calcium oxide (up to 13%) to the slag (samples # 9–# 11) decreases viscosity value of the latter (Figs. 4 and 5). This observation is in agreement with the data reported by ^{Okunev et al. (1986)}, ^{Ji et al. (1997)}, ^{Tarasov (2006)}, ^{Huaiwei et al. (2012)}, and ^{Selivanov and Tyushnyakov (2013)}. The slag samples with higher calcium oxide content are characterised by lower values of viscosity (down to 0·08 Pa s at 1300°C) and viscous flow activation energy (13·2–31·5 kJ mole⁻¹). Their crystallisation temperatures also decrease down to 1010–1040°C. The influence of CaO can be explained by the growth in the number of monoatomic ions in the melt, leading to destruction of the coarse silicon-oxygen complexes and their substitution with the less-complicated associates. The viscous flow activation energy is also dependent on the size of structural units of the melt and falls down with the breaking up of these units into smaller ones.

Figure 4

Viscosity of the slag # 8 with various CaO additions versus temperature

Figure 5

Viscosity of the slag at 1250 and 1300°C as a function of CaO content

It was shown by experiments that the melts with higher viscosity contain more copper. This observation may be attributed to imperfect separation of the liquid bath into matte and slag during the process of smelting and to some changes of the slag composition. Copper content in the slag correlates with its index K_v according to the following linear relation with R² about 0·96 (Fig. 6) 6

Figure 6

The effect of K_v index on the copper content in the crystallised slag (samples # 1 – # 8)

Proceeding from the results of viscosity measurements, one may conclude that the less copper contents in the slag are related to the compositions with 35–36% Fe, 31–32% SiO₂, and 7–8% CaO.

Statistical analysis of the experimental data on viscosity gave the following dependences of the crystallisation beginning temperature T_η and the viscous flow activation energy E_η on the slag composition with R² values 0·97 and 0·79, respectively 7 8

Application area of the equations (7) and (8) corresponds to trivalent iron fraction (Fe³⁺/Fe_tot) within 0·05–0·12 and to the overall content of zinc oxide and calcium oxide between 6·1 and 16·2%. The values of T_η and E_η calculated by the equations (7) and (8) (Table 2) are in a satisfactory agreement with the experimental data.

Results of thermal analysis of slag # 1 indicate no sample mass loss when heating until 900°C. Further heating up to 1200°C results in minor sample mass decrease (by 1·8%) caused by interaction of sulphide and oxide components of the slag. Multiple endothermic peaks with the onset at 991°C are observed in the heating curve (Fig. 7). The first thermal effect at 999°C corresponds to sulphide melting (T_sulph). Two other endothermic peaks appear at about 1038 (T_ox) and 1119°C (T_liq) because of the melting process of oxide components. The last one may be considered an indicator of liquidus of the slag. In the process of the sample cooling, liquid phase crystallisation starts at 1071°C (T_cr) and is followed by overlapping peaks at 1038, 953 and 919°C. In general, all the thermoanalytic curves (Fig. 7) for the other industrial specimens (slag # 2–# 8) have the similar shape.

Figure 7

Differential scanning calorimetry (DSC) curves of the slag # 1–# 8 for the heating/cooling rate 20 K min⁻¹

Analysis of the results of thermal measurements (Table 3) has shown that the melting temperatures of a matte component are within 915–991°C, meanwhile, the liquidus temperature of the oxide phases changes from 1103 to 1119°C. Oxide eutectic melts around 1040°C. The area of liquid slag crystallisation when cooling ranges from 1071 to 1017°C. The slag tends to subcooling (by 70 K on an average). The obtained data on melting temperatures differs slightly from FeO_x–SiO₂–CaO system phase diagram information reported by ^{Verein Deutscher Eisenhüttenleute (1981)} and ^{Kongoli et al. (2006)} because of more complicated composition. The phase transition temperatures depend mainly on the slag composition and on iron's state of existence. The growth of Fe/SiO₂ ratio in the slag promotes the increase of liquidus temperature. The same effect is true also for the rise of Fe³⁺/Fe²⁺ ratio.

Table 3

Melting/crystallisation temperatures of the slag according to the thermal analysis data

Sample #	Temperature/°C
Sample #	T _sulph	T _ox	T _liq	T _cr
1	991	1038	1119	1071
2	987	1044	1105	1034
3	938	1040	1103	1041
4	932	1047	1114	1043
5	956	1038	1118	1060
6	924	1045	1111	1009
7	933	1045	1115	1017
8	915	1044	1114	1054
9	…	954	1013	948
10	…	969	1007	970
11	…	950	1029	966

Thermoanalytical study of the sample # 9 with 8·2% of CaO has indicated (Fig. 8, DTA heating curve) an endothermic melting starting at 955 and finishing nearby 1013°C. Crystallisation of the cooled melt begins at 956°C. There are no traces of crystallisation of sulphide phases in the samples with higher CaO content. This observation may be explained by separation of coarse inclusions in the process of initial slag remelting and by diminishing of fine sulphide phases because of better sulphides solubility in the oxide melts of larger basicity. Adding of calcium oxide (up to 10%) to the slag of VF makes the samples’ melting temperature lower (Table 3). Excessive increase of CaO content (over 10%) may result in a rise of the slag melting temperature.

Figure 8

Differential thermal analysis (DTA) curves of the slag # 9–# 11 for the heating/cooling rate 20 K min⁻¹

The performed study has shown that formation of the slag with optimal properties (viscosity value 0·3–0·6 Pa s and liquidus temperature 1105–1120°C) allows smelting processing in VF at a lower temperature with the overheating above the liquidus about 50–100 K. The rise of magnetite content over 5–6% has adverse impact on the slag viscosity and promotes its inhomogeneity.

Conclusion

The viscosity and liquidus temperature of slag have been studied depending on the temperature and composition. Minimum viscosity values (0·3–0·6 Pa s at 1250°C) and crystallisation beginning temperatures (1105–1120°C) pertain to the compositions with low magnetite content and Fe/SiO₂ ratio within 1·05–1·16. The values of viscous flow activation energy for the slag of copper–zinc concentrates smelting lie in the range 22 and 82 kJ mole⁻¹. Calcium oxide additions up to 13% make viscosity fall down by 1·5–2·0 times.

Thermal analysis of the slag of Cu–Zn concentrated products processing in VF indicates the melting temperatures 915–999°C for sulphide components and 1103–1119°C for oxide phases. Oxide eutectic in the slag melts around 1043°C. The melting process in the slag with higher CaO content starts at about 950–969°C. At CaO concentration up to 10%, the liquidus temperatures reach 1007–1114°C. Continuing growth of CaO content in the slag (to 13%) makes the melting temperature somewhat higher. The slag crystallisation temperatures taken from the data of thermal analysis and viscosity measurements are close to each other.

The copper concentration in the slag varies from 0·62 to 0·97%. The higher percentage refers to the most thick-flowing slag having elevated liquidus temperature. The lowering of copper content and decreasing the slag viscosity and liquidus temperature may be achieved by adding high calcium oxide-containing components to the charging mixture for smelting.

References

Bazán

, Goñi

, Castellá

, Brandaleze

, Verdeja

and Parra

. 2006. Determination of viscosity of fayalite slags for kv model and measurements by means of inclined plane, Rev. Metal. (Madrid, Spain), 42, 84–90.

Chen

, Raghunath

and Zhao

. 2013. Viscosity measurements of “FeO”-SiO₂ slag in equilibrium with metallic Fe, Metall. Mater. Trans. B, 44, 506–516.

Denisov

, Belousova

, Istomin

, Bakhvalov

and Pastukhov

. 1999. Molten oxides structure and properties, 498. Yekaterinburg, UB RAS.

Denisov

, Istomin

, Denisova

, Ryabov

, Chumilina

and Kuchumova

. 2013. Viscosity of Bi₂O₃-B₂O₃ system melts, Rasplavy, 1, 43–49.

Esin

and Geld

. 1966. Physical chemistry of pyrometallurgical processes. Part 2, 704. Moscow, Metallurgia.

Huaiwei

, Fei

, Xiaoyan

, Bo

and Xin

. 2012. The viscous and conductivity behaviour of melts containing iron oxide in the FeO_t-SiO₂-CaO-Cu₂O system for copper smelting slags, Metall. Mater. Trans. B, 43, 1046–1053.

, Sichen

and Seetharaman

. 1997. Experimental studies of the viscosities in the CaO-FenO-SiO₂ slags, Metall. Mater. Trans. B, 28, 827–834.

Kongoli

, McBow

, Yazawa

, Takeda

, Yamaguchi

, Budd

and Llubani

. 2006. Liquidus relations of calcium ferrite and ferrous calcium silicate slag in continuous copper converting, in Sohn International Symposium Advanced Processing of Metals and Materials, Volume 1 – Thermo and Physicochemical Principles: Non-Ferrous High-Temperature Processing, 69–87, San Diego, CA, The Minerals, Metals & Materials Society.

Musikhin VI , Pastukhov

, Denisov

, Istomin

and Bakhvalov

. 1992. Viscosity of boron oxide-containing melt, Rasplavy, 4, 40–45.

10.

Okunev

, Selivanov

, Shin

and Kuznetsov

. 1986. Viscosity of CaO-FeO_x melts, Izvestia AN SSSR, Metally, 5, 36–39.

11.

Seki

and Oeters

. 1984. Viscosity measurements on liquid slags in the system CaO-FeO-Fe₂O₃-SiO₂, Trans. Iron Steel Inst. Jpn., 24, 445–454.

12.

Selivanov

and Gulyaeva

. 2012. Crystallization of iron-containing oxide-sulphide melts, in Crystallization – science and technology, (ed. Andreeta

), 271–302, Rijeka, InTech.

13.

Selivanov

and Tyushnyakov

. 2013. Effect of the oxidation level of iron on the viscosity of FeO_x-CaO melts, Russ. Metall., 2013, (9), 647–651.

14.

Selivanov

, Gulyaeva

, Zaripov

, Belyaev

and Selmenskikh

. 2013. Influence of calcium oxide on microhardeness and structure of high-ferruginous slags of autogenous melting, Tsvetn. Met., 12, 18–23.

15.

Sheludyakov

. 1980. Composition, structure and viscosity of homogeneous silica and silica-alumina melts, Vol. 157, Alma-Ata, Nauka.

16.

Stengelmayer

, Prusov

and Bochegov

. 1985. Improvement of viscosity measurements technique using oscillatory viscometer, Zavod. Lab., 51, (9), 56–57.

17.

Stengelmayer

. 1973. Graduation of oscillatory viscometers, Zavod. Lab., 39, (2), 239–240.

18.

Tarasov

. 2006. The influence of calcium oxide on metal losses with the slag in the bubbled bath, Tsvetn. Met., 7, 18–19.

19.

Vaisburd

, Berner

, Brandon

, Kozhakhmetov

, Kenzhaliyev

and Zhalelev

. 2002a. Slags and mattes in Vanyukov's process for the extraction of copper, Metall. Mater. Trans. B, 33, 551–559.

20.

Vaisburd

, Brandon

, Kozhakhmetov

and Kenzhaliyev

. 2002b. Physicochemical properties of matte-slag melts taken from Vanyukov's furnace for copper extraction, Metall. Mater. Trans. B, 33, 561–564.

21.

Vanyukov

and Zaitsev

. 1969. Slags and mattes of nonferrous metallurgy, Vol. 407, Moscow, Metallurgiya.

22.

Vanyukov

and Zaitsev

. 1973. Theory of pyrometallurgical processes, 504. Moscow, Metallurgiya.

23.

Vanyukov

, Bystrov

, Vaskevich

, Bruek

, Zaitsev

, Kirillin II , Komkov

, Mantsevich

, Miklin

, Sorokin

, Fedorov

, Tsesarsky

and Shubsky

. 1988. Liquid bath smelting, 208. Moscow, Metallurgia.

24.

Verein Deutscher Eisenhüttenleute . 1981. Slag atlas, 208. Düsseldorf, Verlag Stahleisen M.B.H.

25.

Zhang

G.-H

, Chou

K.-CH

, Xue

Q.-G

and Mills

. 2012. Modeling viscosities of CaO-MgO-FeO-MnO-SiO₂ molten slags, Metall. Mater. Trans. B, 43, 64–72.

Viscosity and thermal properties of slag in the process of autogenous smelting of copper–zinc concentrates

Abstract

Keywords

Introduction

Methods and Materials

Results and Discussion

Conclusion

References