UNIT OPERATIONS IN MINERAL PROCESSING ~ by: John M. Currie Senior Instructor tment of Chemical MAY 1973 $18.00 ISBN 0-918062-13-6‘No part of this Manual may be reproduced in any form without the prior permission Of the Publisher. aFOREWORD To assemble a Mineral Processing text of @ level appropriate for students of engineering technology has not been an easy assignment. It is the author's view that the basic scientific fundamentals which are a part of Mineral Processing must be given their rightful place. Enough emphasis should be given to fundamentals in order to permit the student to rationally relate the possible to the probable as well as to promote real understanding rather than mere acquaintanceship. Basically by means of the concept of unit operations, the course is oriented ‘toward the engineering and operating viewpoint. The program has been designed to teach principles rather than specific operational techniques and to acquaint the student with the latest concepts and future trends. Emphasis is laid on learning through solving problems of type encountered in industry. A serious attempt has been made to correlate the course, with the materials and ideas being presented by other Departments at the Institute to those who study Mineral Processing. The aim has been to make the course interesting and stimulating as well as utilitarian. I would like to express my appreciation to J.T. Denley for his invaluable help in proofreading both for clarity end ertors. To Joan LaBrash of fhe Cisrictlum Development Branch, British Columbia Department of Education. go my thanks for her cheerful under standing and professional assistance while setting out and typing the manuscript JM, Currie Burnaby, British Columbia May. 1973,[ Univeridea Francia op, HeeUE PO cnnFOREWORD To assemble a Mineral Processing text of a level appropriate for students of ‘ongineering technology has not been an easy assignment. It is the author's view that the basic scientific fundamentals which are a part of Mineral Processing must be given their rightful place, Enough emphasis should be given to fundamentals in order to permit the student to rationally relate the possible to the probable as well as to promote real understanding rather than mere acquaintanceship. Basically by means of the concept of unit operations, the course is oriented toward the engineering and operating viewpoint. The program has been designed to teach principles rather than specific operational techniques and to acquaint the student with the latest concepts and future trends. Emphasis is laid on learning through solving problems of @ type encountered in industry. A serious attempt has been made to correlate the course, with the materials and ideas being presented by other Departments at the Institute to those who study Mineral Processing. The aim has been to make the course interesting and stimulating as well as utilitarian. | would like to express my appreciation to J.T. Denley for his invaluable help in proofreading both for clarity and errors. To Joan LaBrash of the Curriculum Development Branch, British Columbia Department of Education, go my thanks for her cheerful under- ‘standing and professional assistance while setting out and typing the manuscript. JM, Currie Burnaby, British Columbia May, 1973GREEK SYMBOLS USED angular velocity surface tension, dynes per cm e density of solia; grams per cm? WS per ft? viscosity, puré liquid, poise viscosity, (bulk) of pulp contact angle density of fluid; grams per cm’ lbs per ft? summation standard deviation, population standard deviation, means, standard deviation, sample pulp density, grams per em’ bs per ft’ micronGREEK SYMBOLS USED angular velocity surface tension, dynes per om “density of solid; grams per cm’ lbs per ft? viscosity, pure Tiawk—peias viscosity, (bulk) of pulp contact angle density of fluid; grams per cm? lbs per ft* summation standard deviation, population standard deviation, means. standard deviation, sample pulp density, grams per cm’ lbs per eF micron viiiCHAPTER 1 | INTRODUCTION TO THE TECHNOLOGY OF MINERAL PROCESSING It is readily apparent that mineral substances are required in a reasonably high state of purity of concentration. One cannot apply on an economic basis, expensive extraction and refining processes to the small amounts of metals such as copper. lead. zine, nickel and molybdenum, etc. which occur in the earth's crust. It is necessary to ‘concentrate these valuable minerals into a small bulk which can then be subjected to in tensive and expensive treatments for the recovery of the pure metal. The bulk of the waste materials have first been discarded by the relatively cheap operations which con- stitute Mineral Processing Mineral Processing is the treatment of raw materials (minerals| from the earth's surface to yield marketable products. by methods that in general do not destroy the physical or chemical identity of the minerals. This definition rules out of nuneral proces: ‘sing Such operations as oil refining, leaching, heat treatment and other chemical procedures. ‘This course in Mineral Processing is designed to study the basic principles used in the concentration of minerals from their ores.” Where necessary, application of the various processes to specific operations will be reviewed. The term technology 1s used in the sense of application of technical skil!s along with economic justification of the process. A sound knowledge of basic sciences such as inorganic, organic and physical chemistries. physics and mathematics as well as the engineering crafts required in the hhandling of large tonnages of rack and fluids 1s of prime importance in the study of this technology. ‘The approach to the study of Mineral Processing will be by way of quantitative ‘and qualitative studies of Unit Operations. Mineral Processing consists of widely vary- ing sequences of steps. the basic principles of which are independent of the material being treated. in designing a treatment method, it is essential that one recognize the ‘steps involved. To these fundamental basic steps we apply the term unit operations, The understanding of specific unit processes of operations must, wherever pos: sible be based on fundamental physical and chemical laws. The results of such @ metnod of study are really only applicable to pure or near pure mineral specimens. Unfortunetely ‘pure mineral specimens are a rarity in actual practice, with the result that a large measure of empiricism must be applied to specific cases. Experience and judgement play {an important part in the application and scale-up of theoretical considerations. Mineral processing combines a series of quite distinct unit operations. In these basic steps we can include: 1. COMMINUTION ~ particle liberation 2. PARTICLE CLASSIFICATION by - (a) dimension — (screening) (b} differences in density and mass (hydraulic sizing)1-2 3. SEPARATION OF SOLIDS FROM LIQUIDS by:~ (a) thickening (b) filtration 4. SEPARATION OF SOLIDS FROM SOLIDS by:~ (a) differences in density (heavy liquid or heavy media) (b} flowing film concentration (based on solid sizes and shapes) (c) hindered settling (jigging) (4) flotation (e) agglomeration (f) -magnetic forces (g) electrostatic forces 5. LEACHING AND AGITATION 6. MATERIALS HANDLING — pumping, conveying, etc. ‘The necessary unit operations and their sequences are quite often different for Gifferent minerals. However, on examination of most treatment schemes it will be seen ‘that the majority of operations are made up of different combinations (in kind and degree) Of a reasonably small number of the above mentioned unit operations. The existence or consideration of a treatment scheme implies the production of a Product in sufficient quantity and quality and at a price of interest to purchasers. The goal ‘set for any mineral dressing operation is the production of a concentrate of the wanted (valuable) mineral, of a grade as high as possible at a cost as low as possil is goal is to be achieved concomitant with the maximum possible recovery in the concentrate, of values in the ore and the maximum possible rejection of waste as residue. All the various unit processes are synchronized to the end that the economic return per ton of ore is a mai mum. Benefits resulting from mineral processing prior to smelting or other treatment can be ‘summarized thus : 1. Since little or no waste is being shipped in the concentrate, quite consider- able freight savings can be realized, 2. Since the waste in the shipment has been considerably reduced, metal losses in slags or leaching residues are correspondingly reduced, 3. The reduced tonnage of materials to be treated, either by smelting of leach- ing, results in drastic reductions of treatment charges. 4, When concentration methods are both efficient and cheap, it is often possible 10 take advantage of low-cost, high tonnage mining methods to produce ores of lower grade than would normally be produced by high cost selective min- ing methods. This situation is met with in both underground (block caving) and surface (open pit) mining methods, 5. Efficient and low cost mineral processing plants make it possible to treat material which, without the availability of cheap concentrating methods, could never be classed as ore, ‘The problem which follows, illustrates the actual economic gain to be miade from processing a crude ore and shipping the concentrate, compared to shipping the ore direct, toa smelter for treatment, EXAMPLE A copper-gold ore assays: Au — 0.45 02. per ton Ag — 4.50 oz. per ton Cu — 2.20%Activator: flotation 9-10; cationic 9-13; anionic 9-13 Adsorption: 7-11, 7-12; Gibbs Equation 7-11. Atomic Bonding: covalent 7-4; ionic 7-4; metallic 7-5 unit cell concept 7-5 autogenous Grinding: 6-16 Ball Mill: 6-1 é Bond F.G.: 3-3; 3+8; work index 3-8; 3-12; 3-13 Bubble +: mineralization 8-1; thermodynamics 8-2; force of attachment 8-14 Capillary Forces: 7-15 Circulating Load: 5-15; 6-5 Classifier: Aikens 11-19; Cyclone 11-24; Dorr 11-19; Hardinge 11-19 Classification: 1-1; 11-1; mechanical 11-19; efficiency 11-23: operation variables 11-22 Comminution: 1-1; 3-1 Contact Angle: 7-18 Critical Velocity: 14-10; 14-12; 14-16; 14-18; 14-22; 14-23 relationship between pulp density and critical velocity 14-27 Crusher: cone 4-16; 4-17; 4-18; 4-197 4-20 gyratory - 3-11; 4-2) 4-9; 4-11; 4-12; 4-15; 4-22; 4-23 impact - 3-9 ; jaw = 2-5; 4-1; 4-2; 4-47 4-5¢ 4-10; 4-22; 4-23 Selection- 4-21; 4-24 Cyclone Classifier: 11-24; principles 11-24; velocity distributions 11-30; zero vertical velocity 11-30; separating forces 11-36; separating efficiency definition 11-40; d,, size, definition 11-44; 4,, size calculation 11-44; capacity 11-44; residence time 11-50; design variables 11-52; cyclone, Heavy Media: 11-24; selection 12-12; critical medium density 12-13Depressants: 9-10 Dilution, pulp: definition 11-10 Dithiophosphate: 9-3; 9-6 Drag Coefficient: 10-4; table of values 10-4 Dupre's Equation: 7-18 Electrical Double Layer: 7-17 Energy: free surface 7-19; pumping 14-34 Feed Size "FP": Bond Equation 3-5 Flotation: 7-1; Flotation Process Summary 7-2; collectors 9-1; 9-2; anionic collector 9-3; cationic collector 9-4; reagents 9-3; 9-7; 9-9; 9-10; collector attachment related to mineral surface condition 9-5; strength of collector attachment 9-6; time required 8-4; modifiers 9-9; free energy of surfaces 7-19; frothers 9-7 Flowing, Film Concentration: 12-23; application of 12-28 Flow Sheets: 1-5; 12-10 Frothers: 9-7; frother action 9-7; composition 9-8; application 9-9. Gaudin: 2-9 Gibbs Adsorption Equation: 7-11 Gravity Separation: 12-1; flowing film 12-23; jigs 12-15 shaking tables 12-31. Grinding, Autogenous: 6-16; primary 6-17 secondary 6-17 Grinding: 6-1; Ball mill theory 6-1; 6-3; media 6-1; 6-2; 6-7; power 6-2; liners 6-2; 6-4; 6-15; flow sheet 6-5; tonnage 6-6; ball size calculation 6-11; Rod mill 2-3; 3-10; theory 6-10; rod size calculation 6-12 Grinding Mill: ‘speed, power 6-12; critical speed 6-14; liners 6-15; ball Consumption 6-16; rod consumption 6-17. 5-3 Gyratory Crusher: 3-11 Grizz], Hal He: Hur Hy< RyHallimond Tube: 7-20 Heavy Media: 12-2; cone ad drum separator 12-2). medium Characteristics 12-2; 12-3; separating vessel 12-3; l2-f; media requirements 12-5; media gravity versus medium Viscosity 12-7; 12-8; feed requirements and preparation 12-8; flow sheet 12-10. Heavy Media Cyclones: 12-11; selection 12-12; critical medium density 12-13 Hindered Settling: 10-12; 12-15; 12-197 Humphreys Spiral: 12-29 Hydraulic Transport of Solids: 14-1; flow regimes 14°2; Fanning friction equation 14-1; Bernoulli's theorem 4-3; flow of water through steel pipes (U.S.G.P.M-) 4-9; (.G.P.M.) 14-41; multiphase flow 14-10; critical deposit velocity 14-10; pressure-head relationship (4e11, 14-33; Hazen-Williams flow formula 14-11; head losses in pulp flow 14-15, 14-23; power requirements io2o8, 14-29;, terminology of pumping and pipelines 14-30, 34-31, 14-32, 14-33; pumping efficiency 14-34; pipeline and pump design calculations 14-34 Hydrogen ica concentration: 9-10; 9-11 Impact Crushing: 3-9 gigs: 12-15; separating forces 12-15, 12-17; particle facceleration 12-19; operating variables 12-22; Johnson Concentrator: 12-28 Kick's Law of Crushing: 3-3 Liquid-Liquid Separation: 1-2 Liquid-Solid Separation: 11-15Mineral Characteristics useful in separatio Mineral Processing: definition: 1-1 Modifier, froth flotation: 9-9 Molecular structure of water: 7-6; 7-7 Newton's equation for.settlement of coarse particles in a fluid 10-5. Overflow, Thickener: 11-1 Particle: Size - definition 2-1; surface 7-3; surface hydration 7-7; surface, effect on flotation 8-6; surface effect on settling 10-2; calculation of surface area 2-11; 13-11; pH control: 9-10 Pipeline: design calculations 14-34; critical velocity in 14-10; 14-15; 14-23; Hazen Williams friction formula 14-11; tables of friction losses 14-9; 14-41; Product Size “P": Bond work index W; 3-5 80% passing size (size modulas) 3-9 Pulp dilution: 11-10 Pulp density tables: 14-39; 14-40; Pumps: characteristic curves 14-42; rubber lined sand 14-38; eure 14-28; 14-29; terminology 14-30; efficiency Quicksands: 10-14 Reagents, flotation: 9-1; 9-3; 9-7; 9-9; 9-10 Reduction Ration Rot and Reynolds Number Ng: 10-7 Rittingers Law of Crushing: 3-2 Rod Mill: 6-10Sampling: 13-1; binomial distribution 13-10; confidence levels 13-6; estimating sample sizes for mill products - using binomial distribution 13-11; using increment sempling 13-13; Beonetric mean 13-57 sampling practices in Mineral Processing 13-14; statistical theory 13-2; standard deviation 13-6; small samples 13-97 Schuhmann: 2-6; 2-97 2-10;'3-8 screen analysis: ‘Tyler Screen Scale 2-2; 2-3; 2-5; 5-27 graphs, crushed products 5-19; 5-20 Screen area requirements: 5-26 Sereening: $1; types 5-37 open circuit 5-22) closed otreuit 5-23; sieve bend 5-6; 5-7; trommel 5-97 vibrating 5-10; SUT; S299) operating variables, 5-14; efficiency 5-15 Sedimentation: 10-1 10-2 Settlement of coarse particle in flui Settlement of fine particle in fluid: 10-5 Settlement of intermediate size range: 10-6 Settling Pond: Tailings 11-1 Settling Rate: Castleman function Shaking Table: Wilfley type 12-31 smelter Contracts: 1-3 Stoke's Law: 10-6 Surface Tension: 7-12; of solutions 7-17; contact angle 7-18 Table: Wilfley: 12-31 tailing disposal: 11-1 thickening: 11-57 determination of thickener areas }i-t/ Keckecer 11-9) free settling area determination 11-10; compression depth determination 11-11; qrommiel Screen: 5-9 qyler Screen Scale: 2-2; 5-27Underflow, Thickener: 11-10 Velocity, critical flow: 14-10; 14-12; 14-16; 14-18; 14-22; 14-23; settling in a fluid’ 10-2; 10-5; 10-6; Waste disposal: 11-1; Water molecule: structure 7-6; Wilfley table: 12-31; Xanthate: 9-2; 9-3; 9-613 ‘This ore can either be shipped directly to a smelter for treatment or treated in & concentrator and the concentrate shipped to a smelter for treatment. In the concentrator, 95%. of the copper. 90% of the gold and 85% of the silver are recovered in the concentrate. ‘The concentrate assays Au = 4.88 02. per ton Ag — 45.75 o7. per ton Cu - 20% Fe - 30.0% S10, = 10.0% ALO, - 5.0% The ore or concentrate is shipped to @ copper smelter which pays for the metals ‘according to the following smelter schedule Gold: if 0.03 02. Au per dry ton or over, pay for 96.75% at net mint price. Silver: if 1.0.02. Ag per dry ton or over. pay for 95% at the average price for silver during the week following delivery at the smelter. There will be ‘4 minimum deduction of 1 02. of silver per ton treated. Copper: deduct from the wet copper assay 1.3% (or 1.3 units of 20 Ibs. per unit in smelter terminology|. Pay for the rest of the copper at the daily net export price quoted (here either London or New York price may be F ‘specified) for the week following receipt at the plant, less a deduction fof 2 cents per pound of copper paid for Charges: Base smelting charge $ 14.00 per dry ton; Zine ~ 7%s free, 30¢ per unit in excess; Arsenic — 1% free, 50C per unit in excess: Antimony ~ 1% free, 50¢ per unit in excess, Freight: $2.00 per ton of ore and $3.00 per ton of concentrate from mine 10 smelter ‘The analysis is based on a lot of 100 tons of crude ore, case | DIRECT SHIPMENT TO SMELTER Silver content = 100 * 4.50 02. = 450 02. Ag Smelter pays for 100 (4.60 ~ 1.001 = 35007. Ag Gold content = 100 * 0.45 o7 = 45002, Au Smelter pays for (0.96751 0.45 = 43.637 07. Au Copper content za} 42900) 100, = 4400 Ibs Cu 100. Smelter pays for {225,13} (2000) 1? = 1800 tos Cu 106Charges — freight — 100 x $2.00 = § 200.00 smelting — 100 x $14.00 Total Payments — gold ~ 43.537 x $38.00 silver ~ 350.0 x $2.60 ‘copper ~ 1800 $(0.45 ~ 0.02) = Total = NET RECEIPTS FROM DIRECT SMELTING = $1738.41 CASE It PROCESSING ORE AND SHIPPING CONCENTRATE 100 tons of ore are concentrated at a cost of $6.00 per ton and the concentrate is shipped to the smelter for treatment. In the concentrator, metal recoveries in the concen- trate are as follows: copper 95.0%, gald 90% and silver 85.0% Copper recovered = 95%, equivalent to (0.95) EI 2000 tbs, oF 41.8 Ibs. per ton of ore, The copper recovered from 100 tons of ore is therefore 4180 lbs. Tne concentrate is 25% copper and accordingly there will be produced an “500 Pay for (8.96) (2000) ( £8.36 tons concentrate. 25.0 = 1.3) ) 100 1962.6 Ibs of copper. Silver recovered = 85%, equivalent to (0.85) 4.50 oz. per ton = 3.825 oz. per ton of ore. ‘The silver recovered from 100 tons of ore is therefore 382.60 02. Contained in 8.35 tons of concentrate. Pay for 95% or (.95) 382.50 = 383.38 o2. (deduction = 2.2875 oz. Ag per ton concentrate). Gold recovered = 90%, equivalent to (0.90) (0.48) 02. per ton = 0,405 oz. per ton of ‘ore. The gold recovered from 100 tons of ore is therefore 40,50 oz. contained in 8.36 tons of concentrate. Pay for 96.75% or (0.9875) (40.50) = 39.18375 oz. Charges : = Mineral Processing — 100 x $6.00 $ 500.00 Freight — 8.36 « $3.00 25.08 Smelting = 8.36 x $14.00 117.04 Total 5642.12 Payments: = Gold 39.18375 x $38.00 Silver 363.38 * $2.60 Copper 3962.6 * (0.45 - 0.02) Total NET RECEIPTS FROM CONCENTRATING AND SMELTING $3495.57 GAIN applicable to CONCENTRATING $1757.16 A line diagram illustrating the various stages (in sequence of the treatment required to process an ore is called a FLOW SHEET. A typical simplified scheme for the treatment of a lead ore is shown opposite. |SCHEMATIC FLOW SHEET Dry Concentrate To Smelter Galena PbS Pyrite Fes, Limestone Caco, Quartz SiO, Silicate (Fe. Mn, Al) SiO,When considering @ treatment scheme for any ore, an assessment must first be made of its hardness and toughness. These are the qualities which determine its response to conminution and hence to liberation, Hardness and toughness are determined from impact crushing and grindability tests, From these tests, estimates are made of unit, ower consumption, replacements for wear through abrasion and attrition and maintenance costs. all of which are major cost items. Once the various minerals have been suitably freed from each other by comminu- tion, concentration of values can follow. For concentration of values to be successful the liberated particles must differ sufficiently in physical, chemical, electrical. magnetic ‘or other properties to respond to some applied driving force. The principal characteristics ‘and or properties utilized to make separations are given in the table which follows.PROBLEMS (a) Calculate the net smelter return on the following truck shipment of lead concentrate to the COMINCO SMELTER at Tadanac, 8.C. given: (o) tc Gross weight of shipment — 48,320 Ibs. Moisture Assays — Average Quotations: ‘Treatment we) = 14.1% Au 0.05 oz per ton 4a 179.80 02 per ton Po (wet) 42.60% 2n 40% 8 13.1% Si0, 2.8% Fe 46% cao 10% as 01% % 05% Au $35.00 per oz Ag 2.20 per oz Pb 13.50 cents per Ib. (New York) Zn 12.00 cents per Ib. (St. Louis) Prices quoted are in U.S. funds on which the premium ‘on conversion to Canadian funds is 8%. Base charge $16.00 per dry ton concentrate and credits and penalties as per lead ore and concentrate schedule. Compare the net smelter return with the gross value of the metals in the concentrate. Calculate the net smelter return on this concentrate if shipped to Kellogg. Ideho.2. Calculate the net smelter return on the following shipment of zine concentrate to the COMINCO ZINC PLANT at Tadanac, B.C. given: Gross weig Moisture Assays: Average Quotations : Treatment : ht of shipment — 69,280 Ibs. - 96% Au 0.05 oz per ton Ag 6.9 02 per ton Pb (wet) 28% zn 56.7 % s 318% SiO, 10% Fe 47% cao 06 % ca 0.36% Au $35 00 per oz Ag 2.20 per oz Pb 13.50 cents per Ib. (New York) Zn 12.00 cents per Ib. (St. Louis) ca 2.60 per Ib. Prices quoted are in U.S, funds on which there is @ premium of 8% on conversion into Canadian funds. Base charge $12 00 per ton and credits and penalties a8 per schedule for purchase of zinc concentrates.CHAPTER 2 UNITS OF PARTICLE SIZE MEASUREMENT AND PRESENTATION OF SIZE DISTRIBUTION DATA ‘The response of a particle to mineral processing is influenced to a large degree by its shape or. put in another way, by the ratio of its surface to its volume, For a given volume, of all geometric forms. a sphere has the minimum surface area, while a flat plate has @ maximum surface area for the same volume. The larger the surface is with relation to volume, the faster is the reaction rate and the greater is the surface energy. The shape cor more correctly, the surface developed by the particle shape. has an impact on mineral processing through Frictional retardation of particles Transfer of heat to oF from a particle Stoichiometric reactions Surface chemistry reactions Providing protection to any other mineral it may enclose Its effects on the fluid in which it is in suspension, i.e. on viscosity of the fluid Assessing and representing the surface area of an infinite number of shapes and sizes by some meaningful statistical method is very iniportant part of minosal proces: sing, At this time our purpose is not to investigate the various methods used in size measurement but rather to look at the interpretation of procedure to utilize measurements, PARTICLE SIZE ‘One defines a spherical particle exactly by giving its diameter, However, most mineral powders are composed of non-spherical particles, for which the size cannot be specified in terms of 2 single unique measurement. Particle size then, is not a concise ‘quantity and for any non-spherical shape it could have several values depending on the way in which the measurement was obtained. For instance; in microscopic sizing. size is an average of several dimensions (length and breadth); in sieving, size is equal to the aperture of a square mesh screen of a standard size which will yust retain the particle, jn sedimentation sizing, size is measured as the diameter of a sphere of the same specific gravity and settling rate as the particle in question, To go one step further. size can be Getined as the cube root of particle volume, @ quantity which 1s determimed by counting and weighing @ given number of particles. Since these definitions of particle size all give slightly different results for a given particle or powder, it is important 10 keep in mind the methods used wien compar ing data or in using data obtained partly by one method and partly by anotherSIZE SCALE In order that comparisons are valid as well as to facilitate exchange of informa- tion, sizing is based on a standard scale, From experience it has been found that the ‘most satisfactory scale is one in which the successive sizes form a geometric progres- sion. The reference point is 74 microns (0.074 mm), which is the aperture on a 200 mesh woven wire screen. The ratio of the successive sizes in the standard scale is V2, wt ‘means that the area of the opening of any screen in the series is twice that of the screen just below it but only one half of the area of the screen next above it in the series, The standard scale in use is that described in ASTM E-11-61. This specification is met to satisfaction by the Tyler screen scale. There is however, a slight difference in that Tyler sieves are identified by a mesh designation and U.S. sieves by an opening in millimetres ‘or microns. The standard size scale follows. TABLE 1 Tylor Standard Seren i oom ate 6.730 3 4.760 4 3.360 6 2.380 ts 8 1.680 10 1.190 1“ 0.840 840 20 595 595 28 420 40 35 237 297 48 210 210 6 149 149 100 105 105 150 ora % 200 053 53 270 037 37 400 2% 185 13 93 85) SubSieve a6 (Range 33 23 ~ 16 12MINERAL CHARACTERISTIC EXPLOITED Color. Lustre Radio-activity Specific Gravity and Mass. Shape Surface Activity Chemical Reactivity Ferromagnetisi Surface Conauctivity Texture, Hardness. PROPERTY UTILIZED. optical response ‘gamma ray emission Ditterential displace- ment in a fluid due 10 difference in buoyant effects frictional force developed in a stream flow adsorption of specific ‘eations oF anions to modify surtace tension dissolution by appropriate reagents magnetic riod lity of electrons fon the particle surface difference in the response to crushing grinding. screening classification ‘TWPE OF OPERATION hhand sorting of ore oF automated device triggered by reflected light automated device triggered by scintillation counter Gravity separation of and gravels by heavy media. 11g. table, spiral flowing film table separators Froth flotation and agglomerate tabling to recover aerophilic minerals Hydrometallurgy- leach processes 10 recover gold silver. certain base metals magnetic telds used 10 recover wanted mineral High voltage separator air separation of astwsios fibres: separation of diamonds from quart? and otner waste.CHOICE OF METHODS The treatment method chosen must follow a set of general principles and at the ‘same time be right for a specific ore. Ore types can be conveniently classified as mas- sive, intergrown and disseminated. Coal and many bedded iron ores are of the first type ‘and processing operations are simple since separation of valuable and waste is accomp- lished by a reasonable amount of crushing. The values in ores which are classed as intergrown can be freed only partially by crushing and require a certain amount of grind- ing to complete the separation. Disseminated ores are those where the values are ‘sparsely'distributed through a waste rock matrix, all of which must be reduced to a fine size to liberate the values for separation. Now is as suitable a time as any to emphasize the importance of carrying grind- ing to the point of just unlocking the values and no further; overgrinding is a wasteful process both in power and materials and what may be even worse, it makes efficient recovery mote difficult to accomplish,2-3 PRESENTATION OF SIZING DATA Particle size distribution refers to the manner in which particles are quantitatively ributed among the various sizes; in other words a statistical relation between quantity and size. This data is best presented for use in the form of graphs plotted according to ‘one or two of several general methods. The most common type of weight frequency plot is illustrated in figure 1, where both cumulative weight percent retained (ordinate) and si (abscissa) are on an arithmetic plot. The data used in figure 1, follow. TABLE 2 SCREEN ANALYSIS OF A ROD MILL PRODUCT Teas Welt] weight | Cortaive cretina sire | Tyler Seroon | Sample (gms) | Percent | Weight Percent | Opening Size_| “Mosh Size | Retained on | Retained | Retained Inches this Screon_| this Scraen | this Screen 0.131 6 093 8 i on 066 180 | (10 08 05 08 | = ows | 100 | 4 18 78 | 80 on 0326 so | 20 18 758} IS iy | on 0232 595 2B 8s as | 2ae | on o16s 420 35 110 11.0 35.0 | on ore 297 48 144 aa | 494 [on | ‘one2 | 2t0 65 ns ws | 619 | om 0058 149 100 10.3 10.3 Peace | on coai | 105 160 78 78 800 | on 0029 ” 200 - = 4 / [Pass 160 | 0028 150 20.0 20 | 1000 | ToTaL 100.0 | 100.0 | 100.0ois ejoneg 0 GANIVAIU IN3DUSd LHOIM ISNIVOW .x. 3ZIS FTD1LUVE | 40 107d OUaNHLY ae | 4Particle Size 2-5 ‘The value of @ cumulative direct plot lies in its application to finding percentages that will pass or be retained by openings other than those used in the sieve tes’. as well as the reverse-finding what size opening is required to pass or retain @ designaied per centage, This is done as follows : ‘After the screen or other sizing test has been made, the percentages retained are plotted against the corresponding sizes and a smooth curve dravn in. Note the vertical ‘at the extreme right represents the pan. Next on any vertical line, mark off the number of spaces indicated by the signifi: cant (inch) figure shown at the base of the vertical which represents the screen opening, In the example of figure 1, the vertical line marked 0.093 was chosen and 93 divisions were used on the percentage scale. A diagonal line joining this point (83) und the zero or ppan point on the bese line is then drawn. The horizontal projection of the intersection of the diagonal and any vertical line will then, indicate the size of the opening represented by the vertical line. To find what percentage will be retained by, or pass an opening of say 0.028 proceed as follows: Locate the horizontal line representing 28 in the percentage scale and at its intersection with the diagonal, drop 2 perpendicular to the base. The point at which this vertical cuts the curve represents the percentage retained or in this case 17% coarser ‘and 83% finer, To find what opening will retain 41% and pass 5% proceed as follows: Through the point representing 41% retained on the curve, erect a line perpendicular to the base. The intersection of this line with the diagonal projected horizontally to the percent. ‘age scale shows that the required size is 0.014 inches. TABLE 3 SCREEN ANALYSIS OF JAW CRUSHER PRODUCT wy 2) ie a 8) re WEIGHT PERCENT verb Screen | Retained by | CORSE | Fine than this Number | this Screen | itis Size | Size Cumulative Ail Pass 100.00 2 190 38 10 3 15.58 82 62 4 18.98 53.54 6 14.68 18 66 8 930 39.36 10 72 2214 i 580 26.34 | 20 575 2058 28 1s 16.14 % 1 3.90 12.24 48 ] 3.00 9.24 6 | = 210 92.65 74 100 200 | (ata. 5 i) 170] 85.86 3.44 200 | 85 3741 259 20 6 | (9805 | 134 5 4 | 9848 18 \ 34 96.80 | 120 | 30 s10 | 50 | 2 | gas | 68 | eo | een | 50. 14 | 6 | 36 08 99:73 | 27 7} 9880 20 | og}. 985 | 14 wi | Bee - {ae “Figure 2, using the data of Table 3, illustrates a plot of weight percent retained ‘on each screen (the ordinate, on an arithmetic scale) against the particle size (abscissa, ‘on a logarithmic seale). Figure 3, again using the data of Table 3, illustrates another type of plot. Here the cumulative weight percent finer is plotted as ordinate on an arithmetic scale against the particle size as abscissa, on a logarithmic scale. Figure 4, again using the data of Table 3, gives what is termed the SCHUHMANN, PLOT. Cumulative weight percent finer is plotted as ordinate against particle size as abscissa, both on a logarithmic scal Also plotted on figure 4, is the GAUDIN PLOT. In this plot weight percent retained on each screen is plotted as ordinate against corresponding particle size as abscissa, both on a logarithmic scale. In the Schuhmann plot, it will be seen that the distribution in the finer sizes, (approximately below 10 mesh) is represented by @ straight line, the equation of which has the form Y= 100 Wy, we where, Y= cumulative percent weight finer x = particle size, microns k = size modulus (which indicates the maximum theoretical size of particle in the sample) =) : distribution modulus (or Slope of the line) + The utility of the Schuhmann type plot is demonstrated by an examination of the curve in figure 5. This sample, representing the feed to a ball mill, was sized down to 74 microns or 200 mesh. The data as given, do not give any information about the particle sizes below 200 mesh. By extrapolation, based on the Schuhmann plot being a straight line, it is possible to make an assessment of the percentages of particle sizes down to at least 6 microns. In figure 5, the value of “m” was determined as 0.55 by scaling and 0.52 by calculation. 0.9, at x= 210e y = 1400} y = 37.66% from screen analysis, rT 210)" 1400. 37.66 = 100 [ = 100 (0.15)" *ose 1oMGOwd YBHSNUD MVP Vv 40 NOLINBHUISIG 3ZIS STOLLVd ONIMOHS V1VG 40 107d ADNANDIY AHOIIM Z IUNOIS (o1e98 ommaie60)) suoson wy — .x,, 3218 J1OUNE ses sez in om wee sa 6 oy ex pourmey wosieg 16!om“b 3UNDId NI_G3SN VIVO BNYS 40 1074 3ALLYINWND -€ BUNDLE (oje08 ouumi29601) suo wy ~ .X, 3ZIS ZTOUNYE ose over vez, ost 365 362 in om owe oar £6,(Qo1d pep) — G3NIVITE AHOIIM INZDWRd . M- 29 sum — 6X. 3ZIS STIUANVE 00°08 0004 01 os ‘onpoud woai6 @ 20} siumsv0o ove pue y 904m y “B01 + 4 60 4 = m4 607 atv = ka 00.08 4980 uo pourmes 1u09s0d 14610K, 101d NIanYD) oO ‘1 SUNDIA NI-G3SN Viva aWYS JO 101d 3ALIVIANND € 3uNDIS (og weuynyas) — YaNId AHDIIM 3ALLVINAND ANIOURd «A~001 oot oL ‘suoi9}M — 4X,» 3ZIS FOUN saa 0334 Tun TIve 48} 107d NNYWHNHOS °s auNE Is lez ws vee.0e ana ve woelog si wu 100105 sug % MUD Pouyerey a % ‘sua pourerey "a ozs sp ue s0ulg weaved 1y6oy oxne|RUNDJog 37.66 = log 100 + m log 0.16 1.57588 = 2.00 + m (1.17608) = 0.42412 = m (-0.82391) [w= 052; From examination of the curve it is seen that the greater the value of “m". the ‘more uniform will the product be and the smaller will be the spread of the bulk of material into the very fine and very coarse sizes. This is an important criterion of the suitability of metal powders in powder metallurgy. ‘Among the more useful applications of plots of sizing or screen analysis are the following: 1. Comparative efficiencies of crushing or grinding units for plant purposes are studied by relating the work done and the product sizes when an ore is crushed in different types of crushers or ground in different types of mills. Particle surface areas can be calculated from sizing analysis. The results can be reported as square centimetres of surface area per gram of sample. 9. Estimating the power required to crush and/or grind an ore from a given feed size to a given product size. The base from which such a calculation is made is the screen size at which 80% (by weight) of the feed to the crushing unit will pass and the screen size at which 80% of the product from the crusher will pass. This data, combined in Bond's formula for conminution, makes possible accurate forecasts of plant power requirements. 4, The calculation of the sizing efficiency of a classifier or cyclone can be closely estimated from screen analysis of the feed and products from such an operation. The degree of subdivision or particle size of @ substance is usually the decisive factor in determining its surface to volume ratio. For example, suppose we have one gram of pure crystalline silica, whose density is given as 5 and we consider the particles to be cubes of side I The surface area of an individual particle is equal to 6 |? em?, ‘The weight of an individual particle is equal to I? 8 gram. 1 ‘The number of particles per gram is equal to to ( —— ) 6 1? em? = —°_ em? (Qyerers When 55 in om? per gram, at various values of | 70 gram per cm, the following results are obtained for surface area TABLE 4 LENGTH (= 1) of Particle om mm Micron Cm? per Gm. 10"! 1.0 1000 22.2 107 1 10 220, 107 (0001 O41 222000. 10-6 (00001 0.01 2,222000. 10-7 000001 0.001 22,2200.PROBLEMS 1. Following are the data from a crushing test conducted on @ sample of copper ore. The material — FEED — wes crushed in a JAW CRUSHER whose discharge opening was set at ¥s inch. Feeo TWLER ScneEN SCREEN APERTURE WT GMS. RETAINED NUMBER MICRONS THIS SCREEN So eee, Cs Sore eee 7565 | | 10" 13634 ! a8 : f 9498 | ‘mon 1385 i + 10" i 1680 1270 { = 10 ! It 1403 PRODUCT 455 6795, i 12802 1680 ik 756 6502 {a} Make ® Schuhmann plot for each of the two samples — FEED end PRODUCT. (b) From the graph determine the 80% passing size for each material (c) From the graph determine the size modulus K for each material {d) Calculate the equation of the line which represents the particle size distribu: ton of the productmae room (a) (by ‘The size analysis of the screen underflow from 2 Rod Mill circuit at an iron ore con- ccentrator is follows: We. percent retaipied thi: plus 6 mesh 8 10 4 2 36 28 60 35 83 48 87 65 at 100 84 160 76 200 Bt 335 ae 400 75 pass 400 24.9 Make a frequency plot (arithmetic scale) of the weight percent cumulative re- tained against particle size. Estimate the 80% passing size. 1) Make 2 log-log plot of the cumulative weight percent finer against particle e. Estimate the 80% passing size. 2) What is the value of *m* for this distribution? 3) From the graph estimate the weight percent of material finer than 16 microns. Date from the operation of a cyclone making a separation at very fine sizes to pre- pare @ suitable flotation feed is given below. The underflow in this case 1s the flota tien feed T MARY CYCLONE E propuct | FeeD ‘UNDERFLOW | _ OVERFLOW [Wr RETAINED | WT% RETAINED | wT% RETAINED + 325 mesh | 42.8 66.7 ; = + ou | 73 wt | 14 35 i 14.7 13.7 1 126 ra 84 23 23 6 63 os GBS 2 | 33 orf 106 - 2 | wo | 19 | 366 (a ee 1000 Make 2 log-log plot of cumulative weight percent finer against particle size for ‘each product. (Schuhmann Plot) Estimate the value of K for each product. Find the equation which represents the distribution of particles in each product, Froth Flotation — 50th Anniverary Volume. p. 506 “Preparation of Flotation Plant Feed” ~ Smith & Sougstad.4 A plant produces metal powder by pressure precipitation with hydrogen from solu- tions of the metal. Only that portion of the precipitated metal between 44 microns ‘and 10 microns in size can be used by the metal powder industry. A sizing analysis, Of the precipitated powder is given below. WT% RETAINED THIS SIZE (on 200 mesh ALL PASS. on 325 mesh 70.0 ‘on 400 mesh 85. on 28 microns 12.4 on 18 microns 5.0 on 13 microns 22 pass 13 microns 19 TOTAL 100.0 Calculate the percentage of the precipitated product which lies between 10 microns ‘and 44 microns. Compare this with the percentage obtained directly from the graph.CHAPTER 3 om COMMINUTION Abasic requisite preceding any other operation we may study in mineral proces-~ sing is SEVERANCE of the mineral species in the ore, to the extent that the various species ‘can be separated into valuable and waste fractions. The general term applied to the reduc- tion of large pieces of rock into small fragments is COMMINUTION. Comminution is usually carried out into two related but separate steps — CRUSHING and GRINDING. CRUSHING is size reduction in the coarser range; that is to sey. material coarser than 10 mesh sizes. Crushing is carried out in heavy, slow moving equipment in which the breaker faces are mechanically prevented from making contact with one another. Since the machines, called crushers, are generally dealing with large size fragments, the magni tude of the force applied is usually very large. GRINDING is size reduction in the fine range, that is, material finer than 10 mesh (1680 microns or 1.680 mm). in grinding, the only thing which prevents contact between the grinding media is the material being ground. Since grinding deals only with smaller sized fragments, the magnitude of the forces applied to individual particles is much mach tess than that applied in crushing. However, the unit pressures involved in grinding can be equal to or even greater than pressures applied in crushing, Comminution is a process in which mechanical kinetic energy (of translation) is transformed by impact into (a) strain energy and (b) heat (through internal molecular fric- tion), in the material which is struck, If the material is struck with enough force in either a single massive blow or in several smaller blows, the critical strain in the material (or its elastic limit) is exceeded and the material breaks. During and after impact, the original energy can be accounted for as follows.'"” 4. Kinetic eneray of translation of both the impacted and impacting objects Kinetic energy of vibration of the components of the impacted system. Potential energy stored as strain energy in the system components, Heat generated by friction during deformation or by damping of the wave motion. 5, New surface energy of the fractured material, Charles and de Bruyn’ have shown that transfer of kinetic energy of impact to strain energy in the impacted body is dependent on the contact time between rock and hammer. The longer the contact time between the two, the less is the percentage of ‘applied eneray which remains in the rock as strain. They found that the maximum value Of the ratio of strain energy absorbed to the impact kinetic energy was 0.50. Concomit- tant with this finding is the suggestion that the maximum transfer of energy would be ‘effected by the use of a large number of small high velocity hammers rather than a single large one.‘‘An empirical proposition '?¥is put forward as the basis for a relationship between size reduction of a brittle solid and the energy required to bring about the reduction. That is (3-1) infinitesimal energy change constant infinitesimal size change object size = constant Equation (3-1) states that the energy required to make a small change in the size of an object is directly proportional to the size change and inversely proportional to the object size raised to a power n Three theories of comminution energy requirements have been put forward by Rittinger, Kick and Bond respectively. All three can be related to the general expression given in equation (3- 1) but at the same time it must be said that not one of the three theories is applicable over @ wide range of sizes. The Rittinger and Kick theories while tenable in some cases, were never of much use as practical tools. The “Third Theory of Crushing’ proposed by F.C. Bond 4" in 1962 is, on the other hand, based on a detailed com- pilation and study of numerous laboratory and plant crushing data and provides the techni- cian with @ reasonably accurate measure of power requirements, > Rittinger postulated that the energy required for size reduction of a solid was pro- portional! to the new surface area created. Expressed in symbols: Eq = Ko,-o4) (3-2) where E, = energy input per unit volume according to Rittinger K = constant final specific surface Ha, ©, = initial specific surface 0, = new surface produced Referring to equation (3-1) 6€ = -C % if we tet: initial particle size = x, final particle size after crushing = x. and assign a value of n = 2 E, x R 2 cage ca, nt gee 4 reef xe nd E, +c a]** ans ege te [ani] 3-3) fe (3:3) Which is similar in form to equation (3-2) since specific surface is inversely pro- portional to particle size, considering a given weight of materialick proposed that the energy feared to produce simiar changes o form of gener 'star votes masse uncon ofthe voles owes of bss Sean a eof energyappied to anit volume pradces four — 8170 ite Fate ae one of these sie ts four ~ V1 size ais il eae nit of energy. Ateratvely one could sy hat anther single oi a eine units of materiel would result in sixteen — VHB amt of ena raga refrtng to uation (3-1 Con at if we let initial particle size final particle size and assign a value of n= 1 (4) where Ey = energy input per unit volume according to Kick Ithas been shown by various investigators that these two ews apply to specific ‘cases of comminution. Kick’s theory holds for grinding particles less than one micron in Sie or for energy required in metal turning. On the other hand Rittinger’s interpretation apparently 18 valid for coarse particles. In these circumstances. Bond proposed @ new ceeroach to energy ~ size reduction relationships, one, which in his view. was applicaole Teolont operations. Rittinger concerns himself solely with new area. Kick is interested « +e Rime but Bond "points out comminution is concerned with both surface and volume The absorption of an evenly applied load is definitely proportional to the volume but break age starts with cracks on the surtace. He believes that work previously done on feed particles must be taken into consideration. This is because any subdivided materia! has povigusly had work done on it and TOTAL WORK INPUT must be considered. Bond postulates that the bulk of the necessary work input is used up in particle deformation and is released as heat through internal friction. Local deformation beyond the critical strain results in the formation of a crack tip, usually on the surface. Once the wreck nas been formed. energy in the stressed rack flows to the crack which then extends hrough the rock, causing splitting. Little or no additional energy need be applied 10 brit Tle materials once the crack tip has been formed. Accordingly it is considered that useful ‘work input is essentially consumed in the formation of crack tins. Table 1 gives the terms of reference used by Bond in his development of 2 “Third Theory of Crushing.” ‘The mathematical development of Bond's Third Theory of Crushing followsFe Fe Fp wo wus Y Ye TABLE 1 NOMENCLATURE Impact crushing strength in foot-pounds per inch. "° Diameter of cubes or spheres. Feed inches or microns 8 pct of feed passes Cutoff size of scalped feed. Equivalent feed size of scalped feed, Percentage of feed passing product size P. Ball mill grindability at product size. !° 10 Rod mill grindabi ity at product si Effective creck length in centimeters per short ton, Slope of percentage passing line on log-log plot. For normal material m= 1y2™ Product size = size 80 pct of product passes inches or microns. Grindabitity test sieve openings in microns. Reduction ratio = F/P, = Specific gravity. Work input in kw-hr per short ton to any comminution operation. Work index = Ws for P = 100 microns. Total work input in kw-hr per ton represented by a product size P. Useful work input. Mechanical efficiency. Percentage passing any diam X. = Percentage passing cutoff size Fo of normal feed with 80 pet — size F. When a particle (cube) of side “D" is broken, the average strain energy absorbed bby the cube is proportional to its volume D3. When a crack tip is formed on the surface of the particle, the strain energy in the particle flows to the surface. Average strain energy is now proportional to surface or D2.13, Thus both surface and volume factors have affected the breaking of the rock; when equal weight is given to these two factors, the energy absorbed by the cube of side D is the goomerric mean of the two conditions \DPEDF = Ds? e 4, The number of cubes of dimension O which are contained in a unit cube will 1 vary as + "va i 488 5, Therefore the energy required to break a unit cube = (O*/2) = be This is the same as saying that the total work, useful in breakege, which has been applied to a stated weight of homogeneous rock is inversely proportional to the square root of the diameter of the product particle. Thus if we desig: nate product size as "P”, K = constant and total work input as Wy (to size P) we can write aa wh, (osizer)= & if ve (3-6) From the mathematical approach used in deriving Bond's formula, equation (3-6) it is possible to derive working formulae to apply the theory to operations, This is done as follows For any value of F and P, where W is the work done in K.W.H. per ton of feed to break from feed size F to product size P, both expressed in microns, we write: x Op = (3-6 (0) We = (We = (3-6 0 \ K W= Web = ie = Lf (3-7) ee WE = BO Divide equation (3-6 (a)) by equation (3-7)from which WF (Wp = oF (3-8) or W = iWppt MEAP (3-9) \ In terms of reduction ratio, designated Fr, rr - Ean ie - Substituting for \ F and \® in equation (3-8), we get: Wie hp = (3-10) By definition in Bond's theory. the term WORK INDEX, written 2s W;, is equal to the K.W.Hr. per ton of feed, required to break a material from infinite size to a size of 100 microns, That is for this particular case P = 100 microns and W; = (Wy)p. and start ing with equation kK (, > (3-6) HP \e from which K = (We)p FP « i005 4, (3-6) from which K = W, TOO and (yp \Fo= Wi, TOO 4, 700 (i) te from equation (3-8), (Wklp andy, STO Fe or abe) Al any i \F-P 0 } | Equation (3- 11) states that if we know the value of W for any given size reduction from F to P, then we can calculate the work index W;, Conversely if the work index is known for an ore. then the energy input W, required to break (with the same efficiency) can be found by re-arranging equation (3-11)(3-12) (3-12 (a) since (yp We can resrite equation (3: 10). ye YTD _ y iF (3-13) ‘The student should note that equations (3-11, 12. 13) are in tens of K.WW.Hr. and microns. if F and P are in inches, then the value of 100 is replaced by 0.003937 and \ TOOIST = 0.0627. If the breakage characteristics of a material remain constant over all size renges ‘and the mechanical efficiency of all machines used is the same, then the value of the work index calculated for all these conditions should be the same. Bond thinks the work index remains fairly constant but he admits a possibility of there being variations in going from say 24 inches to 4 inches and from 1600 microns to 400 microns or from 400 to 100 microns and 100 microns to 25 microns. The work index expresses primarily. the resistance ‘of a material to breakage; variations in the value of Wi, can be attributed 10 aifferences in breakage characteristics as well as to differing machine efficiencies. ‘An example of a simple calculation for work index is given. To reduce a feed of 4600 micron size to a product of 400 micron size in a ball mill. it required 3 K.lA.Hr. per ton of feed. Calculate the work index W, we ee (an i ee (i (_\T___ | FH Sat me We] 13140) (2) ‘40 ~ 20 12K.W.Hr. per Ton NOTE: that the total work required previous to this breakage (that is to 1600s was 3 K.W.Hr. per ton and that total work to break to 400 microns ‘amounts to 6 K.W.Hr. per ton. Can you explain this?Referring again to equation (3- 1), it is possible to derive Bond's equation'®) if we assign a value of 1.6 to n, that is a value equal to the arithmetic average of the values. used in Rittinger’s and Kick’s calculations, g x, 8 2 ox dé = -C te H wt ie oan (h- 4) (3-14) ‘And since Bond considers the total work Wotai fequired to break material from a theoretically infinite size to a size modulus P, it is possible to re-write equation (3- 14) 2s follows: mf tld Eg = Mtotal = K 7 B = Mtotal \z 4) Experimental verification of Bond's theory, that is the relationship between energy and size reduction, is very difficult. In the first place, it is experimentally impossible to break particles to one uniform size. In the family of particles produced in a comminution ‘operation, it is impossible to describe accurately as Bond has done (as we shall see later), the weight relationships of sizes in a given product by a single number or size modulus, This is because there are an infinite number of size distributions to which the ‘modulus may’pertain. Further, the energies required to produce each of the size distribu- tions will all be different. This is borne out by the use of the Schuhmann'*? plot to des- cribe a size dist " v= vol] (3-15) In this equation K is the size modules denoting the theoretical maximum sized particles in the family and M is @ constant (or slope of the curve) determining the weight relationship of the various sizes\, (On the basis of Schuhmann’s equation and a great deal of experimental work, Charles‘®” has suggested that equation (3- 1) might better be written thus: E = axti-n) (3-16) where € = energy of comminution K = size modulus 11 = slope of a log-log plot of energy versus size reductionCharles has obtained experimental values for n in the range of 1.32 — 2.4 ‘Sine this range covers Bond's and Rittinger’s values of n of 1.6 and 2.0, there is validity in the argument that these hypotheses are applicable to specific cases of crush- ing and grini 0. The practical test of the correciness of Bonds hypothesis, as well as the value of the work index W;, lies in the comparison of a large number of plant and laboratory results. In general, these results show that the work index under a wide range of sizes and opere- tions remains substantially constant or in some cases varies in a predictable manner. It is for this reason that Bond's theories have become a very practical tool for the designer. As referred to above, the product size P, referred to in Bond's theory. can be termed the size modulus which is defined by Taggart as the 80% passing size on a screen ‘analysis plot. That is to say, P is the particle size such that theoretically or actually 180% of the weight of the particles in the family of particles will be smaller then this size P. Bond has described two methods by which he has determined in the laboratory, the value of the work index W). 1. IMPACT CRUSHING TESTS According to Bond, the work required to crush a ton of scalped (sized) feed at @ reduction ratio (Rr) of = 6 is given by: w KWH, 0218¢ (vF - VE }o 746 (3-17) ton oe NV where: $ = specific gravity of solid C = impact strength in ft.- Ibs. per inch of thickness P= 80% passing size of product in inches (-")= 43-12) ate (Ve 0.1626 C 3c (3-18) svt ‘and 100 = 0.003937 inches 0.1626¢. [sstap 13-20)Equation (3-20) is used to calculate the work index when impact crushing tests ‘are carried out on selected rock specimens. r b . Impact by a he F Impact by 308 hammer a F(t Ibs.) Impact Strength = —F (ft Ws.) _ fe af acs Sue = 2, ROD AND BALL MILL TESTS Bond has specified (for the purpose of test work at Allis-Chalmers laboratories) GRINDABILITY of an ore at any particular mesh size tested as the net production in grams of screen undersize (size P| per revolution of the test mill used for grinding. This term is represented by Gp for a rod mill and Gg for a ball mill and: For a ROD MILL Kin 2 Wp = - (3-21) P “net ton of 0.8% material passing OR size P For a BALL MILL wp = fate - 2 (32) net ton of 0.820 material passing °B size P Where Wp is @ laboratory measurement. For # product only 80% of which passes P, we can write: W = 0.80% (3-23) from equation (3-6) Wp = w [100 P substituting in (3-23) 100. 0.80 WpThe valu of 00), for various meshes is given below. This value multilied by the Wp value obtained from a grindability test gives the work index Wj 8 Examples of W, value calculated from PLANT DATA 1, GYRATORY CRUSHER — UTAH COPPER, Reference: Taggart 4.28 / 180. horsepower motor. 1200 t.p.hr. Work = 1250). 748 _ 9 993 K.luHR, 1200 TON ‘The crusher feed was @ maximum size of 64 inches. A sizing analysis established an 80% passing size of feed of 12.4 inches or 316.000 microns. The crusher was Set to discharge a maximum of @ and a sizing analysis established an 80% passing size of product of 6 or 152,400 microns we é \F=\P NV 100 0.093 STOO __ \316000 - \ TEEADOY 100 0.093 (5621 (39.2) 562 - 392 12, KAW.Hr. per Ton Laboratory crushing tests established the value of WW; to be 11.5 on samples of the same ore,2, BALL MILL ~ CHINO COPPER. Reference: Taggart 5-43 Power consumption: 162 h.p. Feed Rate: 16.7 tons ore per hour eit eet TON NEW FEED The balll mill feed (F) was found by screen analysis to have an 80% passing size of 770 microns. ‘The ball mill product (P) was found by screen analysis to have an 80% pass te ing size of 129 microns. w¥E P 00 V = 7.24 ve Ta = 13.9 K.WLHr, per Ton Laboratory tests were about 6% loner than this figure, SAMPLE CALCULATION OF WORK INDEX © CASE! The following results were obtained in a laboratory experiment designed to find the work index (Wj) of an ore, 1000 grams of copper zine ore were ground in a laboratory mill for 12 minutes. The Power input to the empty mill (contained ball charge only) was 226.9 watts. The power input to the milf when grinding the sample was 283.2 watts. Feed to the mill was 80% ‘minus 1630 microns. Product from the mill was 80% minus 79 microns. Find: (2) Net K.WHr. per ton of ore. required to grind ore from 1630 microns to 79 ‘microns. (>) K.WLHe. per ton of ore to grind from infinite size to 100 microns—that is, the work index Wj, SOLUTION: (a) Net watt input = 283.2 - 226.9 = 56.3 watts. Net power used= 56.3 x 12 watt-minutes (56.3) 12 x w = 11.28 « 10°? Khe Woight of material ground = 1000 gms. = 2.20 Ibs. Thorefore to grind 2000 Ibs. or 1 ton to the same size, the power required W is given by: ww ~ 2000 22 11.28 128 i Wate, 000 (a(b) @ CASEI ‘A test conducted by Allis-Chalmers laboratories on # sample of limestone gave an average impact value “C” equal to 10.8 ft. Ibs. per inch on 10 tests Find: {a) Work Index W; of the rock, if limestone has an $.G. of 2.60 {b) The horsepower required to crush 438 tons per hour of limestone from a feed size of 80% passing 28 inches to 8 product size of 80% passing 4% inches. 2.59 (Av. Impact Strength) Sp. Gravity (2.59) (10.8) (2.59) (108). [0-7 Kae por ton : Zo 10.7 KWH. per ton ; ro) (b) Work (WW) required per ton is given by: wey Sea 2 LH, per ton VF P (a) Work index W; for Feed size “F" = 28 = 711,000 4 r = 84? oe = 108,000 328 = 842-328 5s 10.7618) [10 : 22.) kW. per ton _ 842 328) ne = UOTE. (1 Jae HP te perton on (a8 0.267 H.P.Hr. per ton land to crush 438 tons per hour requires (438) tons {. 267) H.P. hour our von - nT HE. MOTOR REQUIRED 125 HP. (b) MY elPROBLEMS o 1. A certain British Colunbia copper ore has been found by laboratory investigation to have a Work Index value of 16,0. Feed to the ball, mill circuit has an average 80% passing size of 6400 microns (approximately 3/8" ) and the product has an average. 80% passing size of 121 microns. If mill of 18,000 tons per day is planned, calculate the horsepower required in the ball mill section. 2. It is planned to build a 4,000 t-p.d. concentrator at a new lead-zinc property. The ore is to be first crushed at the open pit to a size of 80% passing 3 inches and then transported to the mill by conveyor belt. The following table gives the relevant ‘operating data proposed for crushing and grinding at the mill. It has been found that, the work index of 13.0 is applicable over the entire size range. aa ————— OPERATION | CRUSHING | ROD MILLING | BALL MILLING circurt 1 DRY-OPEN” | WET- OPEN WET- CLOSED Hours per Day" ies wl 24 le gh Tons perHour =| = gs} 186.8 105.8 Feed: 80% size 3 34 14 mesh Product: 80% size | 19.080 p NM, | M8 Reduction Retio- Rr | wo ag w no | no | ao Note that crushing provides for two 7-hour shifts, eighth and sixteenth hours being for cleanup. Rod and Bali Milling provides for three 8-hour shifts of continuous operation Calculate: (2) Power input to the mill crusher (b) Power input to the Rod Millis) {ec} Power input to the Ball Mill(s) 3, A lead zinc ore is being ground from a feed size of 80% minus 3 mesh to a product size of 80% minus 100 mesh for flotation treatment and power consumption is 7.2 K.W.H. per ton of ore ground. Because of the changing nature of the ore with the increased depth of mining, the metallurgical recovery can only be maintained by finer grinding, Test work has indicated that by crushing the feed to 80% passing 4 mesh and tuming out @ ball mill product which is 80% minus 325 mesh, recoveries will be satisfactory. Calculate the power required to grind one ton of ore under the new conditions, accord- ing to the Rittinger. Kick and Bond theories. 4, Investigations into the possibility of recovering the copper in an old tailings dump revealed that the copper-containing mineral, chalcopyrite, consisted mostly of minus. 10 micron size grains locked in a silicate gangue. In order to get even a medium recovery of the chalcopyrite by flotation, extremely fine grinding of the tailings dump ‘material was indicated. Screen analysis of the dump material showed the 80% pass- ing size to be 196 microns. On one test where recovery was promising, the screen analysis of the flotation feed showed the 80% passing size tobe 23 microns, If you were designing a grinding plant to treat 400 tons of dump material per hour and the work index of the ore is known to be 15.6 K.WH. per ton, calculate the installed horsepower necessary.