xiv balkan mineral processing congress

Proceedings of the

XIV BALKAN MINERAL PROCESSING CONGRESS Volume I

Edited by Sunčica Mašić

TABLE OF CONTENTS - VOLUME 1 PLENARY PRESENTATIONS MINERAL RESOURCES AND MINERAL PROCESSING IN BOSNIA AND HERZEGOVINA Nadežda Ćalić, Nedžad Alić, Miroslav Glušac, Rejhana Dervišević, Erna Mandžić

23

IS SUSTAINABLE MINING OF PHOSPHATE ROCK PROBABLE? James L. Hendrix

29

THE EFFECT OF VARIOUS COMMINUTION PROCEDURES ON THE FLOTATION OF SULFIDE ORES C T O’Connor, N Chapman, J Mishra, T Khonthu, N. Shackleton, J Wiese, K C Corin

36

INVESTIGATIONS ON SELECTIVE GRINDING OF MIXED DOMESTIC WASTES IN BALL MILLS Halit Z. Kuyumcu, Stefan Platzk 43

MATERIAL ANALYSIS AND MINERAL CHARACTERIZATION CHARACTERIZATION AND MINERAL COMPOSITION OF THE COMPLEX ORE CONCENTRATE FROM THE MINE FLOTATION Miroslav Sokić, Slobodan Radosavljević, Jovica Stojanović, Vladislav Matković, Dragana Živković, Nada Štrbac 53 THREE-DIMENSIONAL DISTRIBUTION OF GRAINED MATERIALS CHARACTERISTICS Tomasz Niedoba

57

INFLUENCE OF FILLER MODIFICATION ON THE WETTING AND ADHESION OF FIBER SEPIOLITE-EPOXY NANOCOMPOSITE SURFACES Birgül Benli, M.Fatih Can, Mehmet Sabri Çelik

60

COMMINUTION AND CLASSIFICATION GLASS SAND FLOTATION FOR IRON IMPURITY REMOVAL Andrian Valchev, Miroslav Marinov, Irena Grigorova, Ivan Nishkov

68

MICRO-FINE GRINDING OF POTASSIUM FELDSPAR USINGSTIRRED BALL MILL Kemal Bilir, Halil Ipek

73

GRINDING MEDIA FOR DRUM MILLS Lubomir Kuzev, Nikolay Hristov, Kremena Dedelyanova

79

ASPECTS REGARDING THE ACCELERATED AND STATIC SOLUBILIZATION ON THE GOLD AND SILVER ORES Marius Zlăgnean, Viorica Ciocan, Tomus Nicolae, Liliana Ciobanu 83 RESEARCH OF APPLICATION OF BRAMAC CRUSHER TO SHAPE OF GRAIN IN PROCESS OF IGNEOUS ROCK PROCESSING Ekrem Bektašević, Izudin Sjerotanović, Alen Baraković

88

INFLUENCE OF MECHANICAL ACTIVATION ON PHYSICAL CHARACTERISTICS OF FLY ASH Ljubiša Andrić, Ivana Jovanović, Zagorka Aćimović-Pavlović, Vladan Milošević, Danijela Golubović

93

8

MICRO-FINE GRINDING OF POTASSIUM FELDSPAR USINGSTIRRED BALL MILL Kemal Bilir, Halil Ipek

EskisehirOsmangazi University, Faculty of Engineering-Architecture, Department ofMining Engineering, BatMeselik, 26480 ESKISEHIR Abstract: The aim of this study is to investigate dry and wet grindability of potassium feldspar in stirred ball mill down to micro-fine sizes. The effects of parameters such asball diameter, stirring speed,grinding time andinterstitial filling ratio on micro-fine grinding were investigated. Experiments were carried out using 24 full factorial designs. The main and interaction effects on grindability of potassium feldspar were evaluated using Yates’ method. Experimental results have shown that stirred ball mill is very effective for the grinding of potassium feldspar with a median size in the range from 10 to 1 µm. Keywords: Potassium feldspar, stirred ball mill, micro-fine grinding, factorial experimental design. 1. INTRODUCTION Potassium feldspar is one of the basic porcelain body raw materials beside kaolin and quartz and it is used as a fluxing agent (-75 µm) and for forming a glassy phase (-10 µm) in porcelain body; it reacts with other raw materials (kaolin and quartz), and lowers their melting temperature [1-2]. As a result of this, while lowering firing temperature of porcelain, firing time also gets reduced. In porcelain production, while kaolin is used after slurring in water, feldspar and quartz are used after wet grinding in conventional ball mills to -75 µm.One of the most important reasons of grindingpotassium feldspar to -75 µmis to increase fluxing capability by increasing its surface area.Therefore, fluxing capability of potassium feldspar used may well be further increased by grinding down to microfinesizes(-10 µm) using stirred ball mills. Stirred ball mills are more advantageous especially in grinding -75 µm due to their easier operation, simpler construction, lower energy consumption and higher grinding rate compared to conventional grinding ball mills [3-10].Therefore,stirred ball millsin recent years have beenincreasingly usedfor micro-fine and ultra-fine (-1 µm) grindingin industries such as mining, ceramic and paint. The amount of energy obtained per unit time and unit volume is very high in stirred ball millsdue to the use of smallgrinding media with high stress intensities in themills providing aneffective size reduction in theproduct [11-18]. Stirred ball mills employ stirrers comprising a shaft with pins or disks of various designs to agitate the grinding media inside the mill. These mills can be used horizontally and vertically besides dry and wet grinding. The diameters of the balls used as grinding media are vary between a few hundred microns and a few millimeters. Grinding media can be metallic (carbon steel, chrome steel, stainless steeletc.) or non-metallic (alumina,ceramic, zirconium oxide, zirconium-silicate etc.) depending on the use of materials in the industry.The size of the feed

material to be ground may change from a few millimeters toa few microns. The purpose of this study is to investigate the effects of some operational parameters such asstirring speed, ball diameter, interstitial filling ratio and grinding time on dry and wet grindability of potassium feldspar in a stirred ball. 2. EXPERIMENTAL 2.1 Equipment The grinding experiments were performed in a vertical stirred ball mill manufactured by Union Process, Inc. (Figure 1). The net volume of milling chamber is 603.4 cm3. This laboratory type ball mill is designed for grinding media, from 3 mm to 6 mm. The mill has shafts with arms and runs at the RPMs from 100 to 650. It can be operated wet or dry grinding conditions.In order to control grinding parameters in the experiments, the stirred ball mill was connected with a variable frequency drive and personal computer. 2.2 Material and Method Potassium feldspar (-75 µm)used in grinding experiments was obtained from KutahyaPorcelain, Inc. Density of potassium feldspar used in experimental studies was measured by a pycnometerand was found to be 2.65 g/cm3(average of five measurements). Alumina balls with two different diameters were used as the grinding media in the experiments. Ball charging was determined to be 40% of effective volume of the mill tank. Fractional ball filling (J), fractional powder filling (fC) and interstitial filling ratio (U) values was calculated from Equation 1, 2 and 3 respectively [19].

73

J    mB  B  V    1 0.6 

fc 

 mp



 p  V   1 0.6 

U  f c  0.4  J 

(1) (2) (3)

Figgure 1. A labooratory scale batch b type verrtical pinnned stirred mill m used in exp perimental stuudies Experrimental condiitions are giveen in Table 1. Tablee 1.Experimental conditionss Mill ddiameter, cm Mill length, cm Total mass, g Ball T Ball ddensity, g/cm3 Gap bbetween the boottom and the end of the stirrerr, mm The fr fractional ball filling for U=0.7 75 Sample mass, g for U=1.0 00 Slurryy concentratioon (weight), %

8 12 552 3.9 6.35 0.22 112.5 150 70

After each experiiment, all thee media anddground samplles were remooved from thee mill chambber, and the grround sampleswereseparateed from the gr grinding mediaa by sievingg.Particle sizze distributioons of grounnd samples have h been meeasuredwith a laser diffraction size analyzer (M Malvern Masttersizer 2000)). Experriments were carried out using 24 full fa factorial designns. The maiin and interaaction effectss were evaluaated using Yaates’ method.R Results are preesented accordding to this methodology. m Detailed desccription of facctorial designn and Yates’’ method aree given below w. A facctorial designn is used wheen there are several factorrs of interest in an experim ment or the efffect of severaal factors arre to be sttudied in orrder to

74

determine th he main andd interaction effects. Thee effect of a faactor is definedd as the chang ge in responsee produced by a change in tthe level of th he factor. Thiss is called a main effect because it refers to thee primary facttors in the exxperiment. In such designss factors are varied toggether and all possiblee combinations of the leevel of the factors aree investigated [10]. ber of factors iin a factorial design d grows,, As the numb the number of effects thhat can be estimated alsoo 4 d has 166 grows. For example, a 2 factorial design factor-level combinationss. In this desiign, there aree four main efffects to be eestimated (A, B, C and D),, six two-facto or interactionss (AB, AC, BC C, AD, BD andd CD), four thrree-factor inteeractions (ABC C, ABD, ACD D and BCD), and a one four-ffactor interacction (ABCD).. In most situ uations the spparsity of effeects principle;; that is, the sy ystem is usuaally dominated d by the mainn effects and low-order l inteeractions. Thrree-factor andd higher interactions are uusually negliigible. If thee number of faactors is modeerately large (k≥ 4 or 5), a common praactice is to runn only a singlle replicate off the 2k design and com mbine the higher orderr a an estimatee of error [20-2 21]. interactions as Particle sizee (-10 µm) was used ass a responsee variablein factorial designn. Experimentss were carriedd out in randomized orderr. The factorrs and factorr i the grindingg experimentss are given inn levels used in Table 2. i the design Table 2.Facttors and factorr levels used in Factors or Variables Low High level level (-) (+) dB - Ball diam meter, mm 3 5 N - Stirring Speed, S rpm 300 600 T - Grinding g time, minutee 30 150 U - Interstitiaal filling ratioo,% 0.75 1 The responsse for the coorresponding experimentall runs is placced in the fiirst column (y). ( The firstt entries in thee next columnn (1) are obtain ned by addingg the pairs off column (y) together and d the secondd entries would d be subtractiing the top nu umber of eachh pair from the bottom num mber. In the same s way thee n (1). Finallyy column (2) is obtained from column mn (k-1) thee Column (k) is obtainedd from colum process is caarried out for as many timees as there aree factors. The first numberr in the colum mn (k) is thee t responses.. In addition to estimatingg total of all the effects, Yates’ method can provide the sum off squares of every effect iff analyses of variances aree required. These are obtainned by squariing the valuess in the final column and ddividing by th he number off experiments. TS AND DISC CUSSION 3. RESULT dy, Potassium m feldspar (-75 µm) wass In this stud ground (dry y and wet) to the miccro-fine sizess usingstirred ball mill annd the effectts of variouss parameters such asball diameter,stiirring speed,,

grinding timeand interstitial filling ratio were investigated. The main and interaction effects of the four factors (variables) on particle size (-10 µm)were evaluated using Yates’ method. Results of factorial design and Yates’ method are given in Table 3-5. Table 3 shows factors, factors levels andYates notation. Table 4 shows combined data of the Yates’ method and ANOVA (Analysis of Variance) to simplify the decision about the significance of factors in dry grinding conditions. The table value of F (1,3) for α=0.05 is compared with the calculated F value (Fc) for decision. If the calculated F value is greater than the value of F table (FT),the decision iseffective.If the reverse, i.e., the calculated F value is smaller than F table, the decision is not significant in this case. Table 3. Factor and factor levels used in the experiments Yates Factor effects Notation dB N t U mm rpm min % (1) a + b + ab + + c + ac + + bc + + abc + + + d + ad + + bd + + abd + + + cd + + acd + + + bcd + + + abcd + + + +

The main effects of each factor for dry grinding experiments on particle sizeare shown in Table 4 and in Figure 2. The most effective interaction value (notation bcd=94.5%) on the same graph (Figure 2) also is shown. The order of significance of factors according to response variable and Yates notation is bcd, c, b, aand d, respectively.The effect of the grinding time (notation c) and stirring speed (notation b) on grinding efficiency (based on particle size) is strong.These two main factors have positive effects and their total effects are 215.3 and 106.9, respectively (Table 4).According to factorsselected at this study, grinding time is one of the most important variables. Ball diameter and interstitial filling ratio factors are not significant and have negative effects. Total effects of these factors are 25.7 and -34.1, respectively.

dB mm 3 5 3 5 3 5 3 5 3 5 3 5 3 5 3 5

Factor levels N t rpm min 300 30 300 30 600 30 600 30 300 150 300 150 600 150 600 150 300 30 300 30 600 30 600 30 300 150 300 150 600 150 600 150

Table 4.Combined data of the Yates’ method and ANOVAfor dry grinding Yates -10 µm Notation (%) TE SS DF MS (y) (1) (2) (3) (4) (5) (6) (7) (1) 45.0 84.5 222.4 536.1 1038.1 ------a 39.5 137.9 313.7 502.0 -25.7 41.3 1 41.3 b 72.0 165.5 189.0 -7.1 106.9 714.2 1 714.2 ab 65.9 148.2 313.0 -18.6 -0.9 0.1 1 0.1 c 89.1 75.9 -11.6 36.1 215.3 2897.1 1 2897.1 ac 76.4 113.1 4.5 70.8 1.9 0.2 1 0.2 bc 65.5 139.7 -2.2 29.3 -74.3 345.0 1 345.0 abc* 82.7 173.3 -16.4 -30.2 30.7 58.9 1 58.9 d 34.7 -5.5 53.4 91.3 -34.1 72.7 1 72.7 ad 41.2 -6.1 -17.3 124.0 -11.5 8.3 1 8.3 bd 60.9 -12.7 37.2 16.1 34.7 75.3 1 75.3 abd 52.2 17.2 33.6 -14.2 -59.5 221.3 1 221.3 cd 70.2 6.5 -0.6 -70.7 32.7 66.8 1 66.8 acd* 69.5 -8.7 29.9 -3.6 -30.3 57.4 1 57.4 bcd 94.5 -0.7 -15.2 30.5 67.1 281.4 1 281.4 abcd* 78.8 -15.7 -15.0 0.2 -30.3 57.4 1 57.4 Error 173.7 3 57.9 * The higher order interactions used for Total 4893.2 15 estimate of error

FC (8) --0.71 12.34 0.00 50.05 0.00 5.96 1.02 1.26 0.14 1.30 3.82 1.15 0.99 4.86 0.99

U % 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 1 1 1 1 1 1 1 1

FT Decision (9) (10) ----10.13 NS 10.13 E 10.13 NS 10.13 E 10.13 NS 10.13 NS 10.13 10.13 NS 10.13 NS 10.13 NS 10.13 NS 10.13 NS 10.13 10.13 NS 10.13

75

Cumulative undersize, %

100 90 80 70 60 50 40 30 20 10 0

Feed Notation a Notation b Notation c Notation d Notation bcd

0.1

1

Particle size, µm

10

100

Figure 2. Particle size distributions of feed, notations and best result in dry grinding Table 5.Combined data of the Yates’ method and ANOVAfor wet grinding

Cumulative undersize, %

Yates -10 µm Notation (%) (y) (1) (2) (3) (1) 39.2 88.2 239.0 615.6 a 49.0 150.8 376.6 511.8 b 70.0 179.6 182.2 29.8 ab 80.8 197.0 329.6 63.6 c 83.7 68.9 20.6 80.0 ac 95.9 113.3 9.2 105.0 bc 100.0 134.5 32.8 -14.2 abc 97.0 195.1 30.8 -13.2 d 29.7 9.8 62.6 137.6 ad 39.2 10.8 17.4 147.4 bd 45.0 12.2 44.4 -11.4 abd* 68.3 -3.0 60.6 -2.0 cd 52.8 9.5 1.0 -45.2 acd* 81.7 23.3 -15.2 16.2 bcd 96.6 28.9 13.8 -16.2 abcd* 98.5 1.9 -27.0 -40.8 * The higher order interactions used for estimate of error

100 90 80 70 60 50 40 30 20 10 0

TE SS (4) (5) 1127.4 --93.4 545.2 185.0 2139.1 -27.4 46.9 285.0 5076.6 -13.4 11.2 -29.0 52.6 -57.0 203.1 -103.8 673.4 33.8 71.4 25.0 39.1 1.0 0.1 9.8 6.0 9.4 5.5 61.4 235.6 -24.6 37.8 Error 43.4 Total 9110.2

DF (6) --1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 3 15

MS FC (7) (8) ----545.2 37.68 2139.1 147.84 46.9 3.24 5076.6 350.85 11.2 0.78 52.6 3.63 203.1 14.03 673.4 46.54 71.4 4.93 39.1 2.70 0.1 0.00 6.0 0.41 5.5 0.38 235.6 16.28 37.8 2.61 14.5

FT Decision (9) (10) ----10.13 E 10.13 E 10.13 NS 10.13 E 10.13 NS 10.13 NS 10.13 E 10.13 E 10.13 NS 10.13 NS 10.13 10.13 NS 10.13 10.13 E 10.13

Feed Notation a Notation b Notation c Notation d Notation bc

0.1

1

Particle size, µm

10

100

Figure 3. Particle size distributions of feed, notations and best result in wet grinding

76

Considering the total effects of factors in Table 4, to increase the grinding efficiency, ball diameter and interstitial filling ratio should be low levels, and grinding time and stirring speed should be high levels. Table 5 shows combined data of the Yates’ method and ANOVA (Analysis of Variance) to simplify the decision about the significance of factors in wet grinding conditions. The main effects of each factor for wet grinding experiments on particle size (-10 micron) are shown in Table 5 and in Figure 3. The most effective interaction value (notation bc=100%) on the same graph (Figure 3) also is shown. The order of significance factors according to response variable and Yates notation is bc, c, b, a and d. The effect of the grinding time (notation c), stirring speed (notation b) and ball diameter (notation a) on grinding efficiency (based on particle size) is strong. These three main factors have positive effects and their total effects are 285,185 and 93.4, respectively (Table 5).The other (notation d) has a negative effect and its effect is also strong. According to factors selected at this study, grinding time is one of the most important variables. Considering the total effects of factors in Table 5, to increase the grinding efficiency, interstitial filling ratio should be low level, and grinding time, stirring speed and ball diameter should be high levels. 4. CONCLUSIONS A series of dry and wet grinding experiments revealed that grinding time and stirring speed are important factors in the case of dry grinding, while ball diameter, stirring speed, grinding time and interstitial filling ratio factors are important in the case of wet grinding. Experimental results have shown that stirred ball mill grinding is very effective for the production of potassium feldspar with a median size in the range from 10 to 1 µm. ACKNOWLEDGEMENTS The authors would like to acknowledge Eskisehir Osmangazi University, Scientific Research Projects Committee for the financial support of this project (ESOGU BAPProject # 201015018). NOMENCLATURE TE SS DF MS FC FT E NS J fc U k

Total effect Sum of square Degrees of freedom Mean square Calculated F value F value taken from Table Effective Not significant Fractional ball filling Fractional powder filling Interstitial filling ratio The number of factors

D L V Cs dB mB mP

B P N t dP

Mill diameter Mill length Mill volume Slurry concentration based on weight (%) Ball diameter (mm) Ball mass Powder Mass Ball density Powder density Stirring speed (rpm) Grinding time (min) Particle diameter (µm)

5. LITERATURE [1] Bayraktar, İ., Gülsoy, Ö.Y., Can, N.M. ve Orhan, E.C.: Feldspatlarn Zenginleştirilmesi. 4. Endüstriyel Hammaddeler Sempozyumu.pp.97-105, İzmir, 2001. [2] Hzal, M.: Potasyum Feldspatlarn Dünü, Bugünü ve Yarn. 2. Endüstriyel Hammaddeler Sempozyumu, İzmir, pp.31-39, 1997. [3] Bernhart, C., Reinsdh, E., Husemann, K.: The influence of suspension properties on ultra-fine grinding in stirred ball mills. Powder Technology 105, pp.357-361, 1999 [4] Choi, H., Lee, W., Lee, J., Chung, H., Choi, W.: Ultra-fine grinding of inorganic powders by stirred ball mill: effect of process parameters on the particle size distribution of ground products and grinding energy efficiency. Metals & Materials Int., Vol. 13, No.4, pp.353-358, 2007. [5] Szegvari, A. and Yang, M.: Versality of attrition milling (Wet or Dry Process; Batch or Continuous Mode). Seminar on Powder Production by Fine Grinding, the Pennsylvania State Univ., June 13, 7 p, 1995. [6] Schilling, R.E. and Yang, M.: New developments in attritors. Paint and Coating Industry, April, 6 p, 2001. [7] Just, A. and Yang, M.: Attrition dry milling in continuous and batch modes. The Powder and bulk Solids Conference/Exhibition–Chicago, IL, 4 p, 1997. [8] Goodson, R., Larson, F. and Sheehan, L.: Energy Input Monitoring During Attritor Milling. International Journal of Refractory and Hard Metals, June, 7 p, 1985. [9] Padden, S.A. and Reed, J.S.: Grinding kinetics and media wear during attrition milling. Ceramic Bulletin, Vol. 72, No. 3, pp.101-112, 1993. [10] Dowdle III, H.J.: Grinding Glazes. Ceramic Industry, October, 3 p, 1993. [11] Dikmen, S. ve Ergün, L.: Karştrmal Bilyal Değirmenler. Madencilik, 43, 4, pp.3-15, 2004. [12] Orumwense, A.O.: The effect of media type on regrinding with stirred mills. SME Annual Meeting, Feb. 28-Mar. 2, Salt Lake City, UT, 6 p, 2005. [13] Becker, J.E.: Attrition Mill Fine Grinding of Advanced Ceramic Powders. Conference Advanced Ceramics ’87, February 17-19, Cincinnati, Ohio, 7 p.1986. [14] Szegvari, A. and Yang, M.: Fine grinding of

77

high value added industrial minerals by attrition milling. Les MinerauxIndustrielsMateriaux des Annes 90, March, Quebec, Canada, pp.1-5, 1989. [15] Szegvari, A.: Fine grinding of ceramics with attritors, Ceramic Technology International 1994, 3 p, 1994. [16] Ma, Z., Hu, S., Zhang, S. and Pan, X.: Breakage Behaviour of Quartz in a Laboratory Stirred Ball Mill. Powder Technology, 100, pp.6973, 1998. [17] Celep, O. ve Alp, İ.: Karştrmal değirmenler ile ince öğütmenin refrakter altn cevherlerine uygulanabilirliğinin incelenmesi. Madencilik, 47, 3, pp.15-26, 2008. [18] Reed, J.S.: Principles of ceramics processing. Second edition, Chapter 7, John Wiley & Sons, Inc., 1995. [19] Austin, L.G., Klimpel, R.R., Luckie, P.T.: Process engineering of size reduction: Ball milling. SME, New York, 1984. [20] Montgomery, D.C.: Introduction to statistical quality control. Third Edition, Chapter 11, John Wiley & Sons, Inc., 1997. [21] Tamhane, A.C.: Statistical analysis of designed experiments: Theory and Applications. Chapter 7, John Wiley & Sons, Inc., 2009.

78