Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Optimization of primary gyratory crushing at Highland Valley Copper Rosario, Persio P. 2003

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_2003-0537.pdf [ 16.7MB ]
JSON: 831-1.0081159.json
JSON-LD: 831-1.0081159-ld.json
RDF/XML (Pretty): 831-1.0081159-rdf.xml
RDF/JSON: 831-1.0081159-rdf.json
Turtle: 831-1.0081159-turtle.txt
N-Triples: 831-1.0081159-rdf-ntriples.txt
Original Record: 831-1.0081159-source.json
Full Text

Full Text

OPTIMIZATION OF PRIMARY GYRATORY CRUSHING AT HIGHLAND VALLEY COPPER BY PERSIO P. ROSARIO A thesis submitted in partial fulfilment of the requirements for the degree of Master of Applied Science in The Faculty of Graduate Studies Department of Mining Engineering We accept this thesis as conforming to the required standard The University of British Columbia Vancouver, B.C., Canada October 2003 © Persio Pellegrini Rosario, 2003 Abstract This thesis presents the work done in a collaborative research project between the University of British Columbia and Highland Valley Copper. The research was aimed at understanding gyratory crusher liner wear in the overall context of the crushing process. Wear measurements were taken for in-service crushers during the research period using a novel laser profile measurement device. Data from the wear measurements was correlated with crusher production information such as current draw and throughput. This work resulted in enhanced knowledge of crushing chamber characteristics and their impact on crushing performance. In addition, an innovative and powerful way to evaluate crusher liner profiles was developed and new mantle profiles were designed. ii Table of Contents Abstract i i Table o f Contents i i i List of Tables , v i List of Figures v i i Nomenclature x i Acknowledgements.: x i i 1. Introduction 1 2. Research Objectives 3 3. Literature Review. 4 3.1 Crusher Machines 4 3.1.1 History 4 3.1.2 Crushers types and operation principles., 6 3.1.3 Gyratory crushers application 15 3.1.4 Gyratory crushers performance 20 3.2 Liner wear • 30 3.2.1 Mechanisms of wear 31 3.2.2 Liner materials 32 3.2.3 Wear measurements 34 4. Highland Val ley Copper Project 40 4.1 Operations Background 40 4.2 Crusher Operations at H V C 47 5. Experimental Approach 54 i i i 5.1 Equipment 54 5.2 Data Collection 64 5.2.1 Crusher operational data 64 5.2.2 Liner information 64 5.2.3 Chamber profile data 70 5.3 Data Analysis 72 5.3.1 Crusher Operational Data , 72 5.3.2 Chamber Profile Data 82 6. Results and Discussion 93 6.1 Improvement in Measurement Process 93 6.2 Wear Determination 99 6.3 Correlation of Operational Data and Liner Characteristics 103 6.3.1 Issues during the end of concave life 108 6.3.2 Mantle and crushing performance 114 6.3.3 Liners that provided optimum performance 115 6.4 Liner Management , 117 7. Conclusions 127 7.1 Achievements 127 7.2 Future Work Opportunities 129 8. Reference List 130 Appendix A Drawings of the new support structure 137 Appendix B Description of the measurement program 139 Appendix C Example of a measurement drawing result 148 iv Appendix D N e w mantle parts dimensions List of Tables Table 3-1 Increase in crusher throughput by changing feed characteristics (Burkhardt, 1982) : 22 Table 5-1 Example of a partial table result for a concave profile measurement generated bytheLPD 57 Table 5-2 Liner rebuild/installation: labour and costs (Wolff, 2002) 65 Table 5-3 Mantle types 65 Table 5-4 Crusher 4 liners detailed information 68 Table 5-5 Crusher 5 liners detailed information 69 Table 6-1 Crusher 4 - Measurement information 95 Table 6-2 Crusher 5 - Measurement information , 96 Table 6-3 Summary of liners information for 6 recent concave life periods 124 Table 6-4 Liner costs (parts, rebuilt and installation) and current total cost per ton 124 Table 6-5 Specific downtime per liner and current total liner downtime 125 Table 6-6 Liner costs (parts, rebuilt and installation) for one concave life and projected cost per ton 125 Table 6-7 Specific downtime per liner and projected total liner downtime 126 vi List of Figures Figure 3-1: Blake Jaw Crusher (Flavel, 1982) . 5 Figure 3-2: Gates Gyratory (Flavel, 1982) 5 Figure 3-3: Common jaw crusher's mechanisms (Zandee, 1989) 8 Figure 3-4 Cross section of a Gyratory Crusher (Courtesy o f Svedala-Metso) 12 Figure 3-5: Typical cone crushers (Zandee, 1989) 14 Figure 3-6 Effect of Feed Size on A G M i l l Tonnage (Dance, 2001) 16 Figure 3-7 Effect of Feed Size on S A G M i l l Tonnage (Dance, 2001) 17 Figure 3-8: Medium-size Crusher Product Effect on M i l l (Dance, 2001) 18 Figure 3-9: Comparison between manual and automatic control o f C S S (Moshgbar, 1995) 25 Figure 3-10 Crusher Feed & Product % Course - Constant Setting (Dance, 2001) 26 Figure 3-11: Straight versus non-choking concaves (Westerfeld, 1985) 28 Figure 3-12: C L P liners (Svensson and Steer, 1990) 29 Figure 4-1 H V C Simplif ied mil l ing flowsheet ( H V C , 2000) 42 Figure 4-2 Camera Image and Boundaries Recognition (Dance, 1998) 45 Figure 4-3 Two 84" 0 Mantle Cone Crushers Test Comparison (Flavel, 1982) 48 Figure 4-4 Automatic Crusher Control Graphic (Dance, 2001) 50 Figure 4-5 Gap measurement results and analysis performed at H V C 52 Figure 5-1 Major components of the L P D 55 Figure 5-2 L P D Installation schematic 55 Figure 5-3 Schematic o f the laser measurement 58 vi i Figure 5-4 Calibration procedure schematic 59 Figure 5-5 Concave profile generated by CDI software 60 * Figure 5-6 Mantle profile generated by CDI software 60 Figure 5-7 Laser profile final output.. 61 Figure 5-8 Original support / New support 62 Figure 5-9 Partial example of crusher-mantles report 66 Figure 5-10 Report from Technical Development Dept 67 Figure 5-11 Measurement Spots 71 Figure 5-12 Example of the complete data file of Crusher #5 (records 5075 to 5714 are hidden to facilitate visualization) 72 Figure 5-13 - Example of a graph with product quality plots 75 Figure 5-14 - Selected liner information for Crusher 5 from October 2001 to January 2002 : 77 Figure 5-15 - Example of mantle position plot with average line 78 Figure 5-16 - Example of the representation of data problems in product quality (circles indicate areas of noisy data where the weighted average of product quality was not calculated) 79 Figure 5-17 - Example of current draw plots 80 Figure 5-18 - Current draw adjusted plot 81 Figure 5-19 Deviations in laser measurements ;. 84 Figure 5-20 - Section view of two chambers and their slices 86 Figure 5-21 Chamber volume by height 90 Figure 5-22 - CSS versus vertical mantle position 91 viii Figure 6-1 Time spent in measurement tests (tests 5 to 8 were performed using the new support) 93 Figure 6-2 Chamber Volumes graph (using M P equals to A M P W ) for measurement 7.. 97 Figure 6-3 C S S versus M P graph for measurement 7 •' 98 Figure 6-4 Concave and mantle wear rate by slices for measurement 7 98 Figure 6-5 Average concave-wear rate 100 Figure 6-6 Simulated concave wear after 8 megatonnes 101 Figure 6-7 Graph #1 - Concave C002 at crusher #4 (29 June 2001-7 Feb. 2002) 104 Figure 6-8 Graph #2 - Concave C004 at crusher #4 (9 Feb. 2002-25 Sep, 2002) 105 Figure 6-9 Graph #3 - Concave C003 at crusher #5 (5 Jan. 2001-27 Sep, 2001) 106 Figure 6-10 Graph #4 - Concave C005 at crusher #5 (30 Sep. 2001-7 June 2002) 107 Figure 6-11 Example of 3 measured chambers that resulted in "good operation" 110 Figure 6-12 Example o f 3 measured chambers that resulted in "bad operation" 110 Figure 6-13 Cross-section view of two similar but not identical chambers I l l Figure 6-14 Comparison between chambers with different radius 112 Figure 6-15 - The difference between the discharging areas resulted from the wear of the concaves 114 Figure 6-16 Number o f mantles used per concave and their total tonnage 117 Figure 6-17 Comparison of chamber condition for under size mantles 120 Figure 6-18 Comparison of C S S - M P relationship for under size mantles 120 Figure 6-19 Comparison of chamber condition for over size mantles 121 Figure 6-20 Comparison of C S S - M P relationship for over size mantles .....122 Figure 6-21 Suggested replacement policy 123 ix Figure B-l L o c a t i o n o f the o r i g i n o f the " t r a c k " coordinate system 141 F i g u r e B-2 Schemat ic o f the tr igonometr ic relat ionships between the o r i g i n a l coordinate system and the " t r a c k " coordinate system 142 F i g u r e B-3 Snapshot o f the A u t o C A D d r a w i n g w i t h the measured prof i les and the o r i g i n a l parts 143 F i g u r e B- 4 G r o u p o f l ines and arcs that represents a mantle o r i g i n a l prof i l e ; 144 Nomenclature AG Ful ly Autogenous Grinding AMWP Average Mantle Position for the Week CDI Conveyor Dynamics, Inc CLP Constant Liner Performance CSS Closed Side Setting FMV Feed Material Variables GPS Global Positioning Systems HB Brinel l Hardness HVC Highland Val ley Copper LPD Laser Profiler Device MDV Mechanical Design Variables MOV Machine Operating Variables MP Mantle Position MTPH Metric Tonnes Per Hour OSS Opened Side Setting ROM Run-of-mine SAG Semi Autogenous Grinding TPH Tonnes Per Hour Acknowledgements The author would like to express his gratitude to HVC for all their support for the work presented in this paper. In particular, Arnie Adams for his confidence in the author since the beginning of the project. The author would also acknowledge all the other participants of the HVC project team for their kindness. The author is deeply thankful to his thesis supervisor Dr. Robert Hall and his co-supervisor Dr. Daan Maijer for all the guidance and support. In addition to their support, the author appreciates the bond that has been developed. Of course, without my wife's support and patience, and the love shown by her and my son, this thesis would not be completed. I am also thankful to my family, who always demonstrated trust in my capacity even from far away. The author would also like to acknowledge the financial support from the Natural Sciences and Engineering Research Council of Canada. xii 1. Introduction Canadian mining operations are facing competitive pressure from offshore high grade, low labour cost mines. To remain competitive, companies have introduced larger and more complex equipment. Complexity has been added through the addition of electronics and in some cases partial or ful l automation. Larger haul trucks, primary crushers and process plants are now commonly used in open pit mines. The increase in mine equipment size and operational improvements has resulted in higher production rates. For the successful integration o f these operations, it is important that increased production levels from the mine still provide a consistent product in terms of grade and size to the mi l l . Many companies are looking for ways to better evaluate the relationships between blasting, primary crushing and mil l ing efficiencies in order to develop new processing strategies to be applied in the optimization o f the overall process. A t the present, it is known that the optimization of primary crashing provides a great opportunity to enhance the overall operational efficiency in high tonnage mines. Increasing throughput and product quality at the primary crushing phase improves the productivity throughout the rest o f the comminution process (Burkhardt, 1982). In addition to reduced processing costs through the gains obtained in the crushing and mil l ing stages, maintenance costs are reduced by better machine availability and by enhanced reliability o f its components. 1 This thesis presents research aimed at understanding the influence of liner profiles and liner wear on gyratory crusher performance. Once a better understanding about these relationships is achieved, the mines may be equipped with better tools to optimize processing operations. 2 2. Research Objectives The primary objective of this research is to improve the understanding of liner wear in primary gyratory crushers. In other words, how chamber geometries and their modification impact on crushing capacity and product quality. In pursuit of the primary objective the following secondary objectives are targeted: Determine an efficient methodology to monitor liner profile wear over time. This work involves the assessment of a prototype laser-based methodology recently introduced in the supporting mine. - Establish an approach to accurately determine the dimension of crusher closed side setting (CSS) replacing the current "bucket test" test methodology. An online methodology is the ultimate goal. - Develop a database of wear information linked to other monitored crushing parameters such as: current draw, product size distribution and production rate. - Evaluate the liner profiles currently available at the mine and, if necessary, develop new profile designs. The .concurrent enhancement of crushing performance and the extension of liner lives is desired. 3 3. Literature Review 3.1 Crusher Machines 3.1.1 History Many centuries ago, weights were raised and dropped onto heavy rocks to crush them in order to enable minerals processing. The crushing process has evolved with the addition of different power sources to the process, beginning with the use of animal and waterpower. However, development of "modern" crushing machines only took place during the 19th century (Utley, 2002). The first crushing machine appeared in English mines in the early 1800's. During the Industrial Revolution, the "Cornish Rolls" device was developed, and though a very limited device, it started the process of miriimizing handwork (Flavel, 1982). Two kinds of crushing machines, not so different from the ones in operation nowadays, were invented in the second half of the 19th century. The Blake Jaw Crusher was the first in 1858 and the Gates' Gyratory Crusher was patented in 1881 (Figure 3-1 and Figure 3-2 respectively). 4 Figure 3rl: Blake Jaw Crusher (Flavel, 1982) Figure 3-2: Gates Gyratory (Flavel, 1982) The industrial revolution in the early 1900's promoted a growth in mining volumes. This growth resulted in an increase in crusher size as well as the invention of other types of crushers (the hammermill and the single sledging roll crusher). In 1919, the first 1.5 m (60 in.) gyratory crusher was manufactured by Traylor Engineering which remained the largest gyratory crusher for 40 years. In 1969 the same company introduced the largest machine to date at 1.8 m (72 in.) of feed size opening. Although the reasons are unclear, only one unit was made (Utley, 2002). In addition, as mining complexity increased, the number of comminution phases grew which led to the development of other types of crushers. Cone crushers were first developed in the mid 1920's by Edgar B. Symons to supply the demand for efficient fine crushing machines. Cone crushers are basically a small-scale gyratory crusher with chamber modifications and higher operation speed (O'Bryan & Lim, 2002). 5 3.1.2 Crushers types and operation principles Jaw crushers and gyratory crushers are the most commonly used machines for primary crushing due to their capability and robustness to handle great volumes and high strength materials. Primary crusher feed, the run-of-mine ( R O M ) ore, may contain lumps as large as 1.5 m across (Utley, 2002). Usual reduction ratios for primary crushing are around 8:1 (Major, 2002). Other types of crushers that can be employed in the primary comminution phase are: rotary breakers ( M M D sizer), impact crushers, high-speed roll crushers, and hammer mil ls. These machines are usually used for ores with specific characteristics such as: low compression strength, low abrasion index, and/or high clay content. In order to achieve the desired particle size necessary to process the ore by methods such as flotation or leaching, other stages of crushing and/or grinding are commonly applied after primary crushing and though several different processes exist for finer crushing, cone crushers are typically part of these processes. Thus, cone crushers can be used for secondary and tertiary crushing phases as wel l as for auxiliary phases in grinding mi l l applications. In order to fulf i l the requirements for such a variety o f applications several types of cone crushers are available. They are: the standard cone crusher, the horizontal impact crusher, the high pressure grinding rolls, the waterflush cone crusher, the disk crusher 6 (Telsmith Gyrasphere), the short head cone crusher, the Metso Gyradisk and the vertical impact crusher. Since this research is focused on optimization of the primary gyratory crusher operation, more attention is given to the description of this kind of machine. Since the jaw crusher was the basis for the development of the gyratory crusher and the cone crusher was derived from the gyratory crusher, they are also covered in this work. Jaw Crusher In the Jaw crusher, two planer surfaces alternatively crush the rocks imitating the animal jaw movement. One of the surfaces remains fixed and the other, the swing jaw, moves according to an eccentric drive - directly or indirectly, depending on the machine type. The way the swing jaw is pivoted and some other construction characteristics determine the different types of Jaw crusher. Details about the different types of Jaw crushers are shown in Figure 3-3. The dimensions of the rectangular receiving area are commonly used to describe them; for example, a 2.1 by 3.0 m (84 by 120 in.) jaw crusher has a 2.1 m width and 3.0 m gape. Jaw crushers are available in a wide range of sizes and capacities - from 50 to 1500 tonnes per hour (tph). 7 (?)TOGGI.E (T)SWING JAW SPRING Figure 3-3: Common jaw crusher's mechanisms (Zandee, 1989) There are several types of jaw crushers, such as the Blake double toggle, the Single toggle, the Dodge, the Universal, and the Telsmith (Gaudin, 1939). The types most commonly found in mining operations are the Blake double toggle and the Single toggle. The Single toggle has limited application for high abrasives ores because in this type of jaw crusher the swing jaw moves elliptically resulting in greater liner wear (Utley, 2002). The Blake double toggle is commonly used in primary crushing in both open-pit and underground operations. The double toggle drive mechanism and the positioning of the pivot point give a minimum displacement of the swing jaw at the inlet region as well as provide a strong breaking action for large ROM rocks. 8 In the Dodge crusher, as opposed to other jaw crushers, the location of minimum swing jaw movement occurs in the discharge region providing a more uniform product size. The Dodge type is the simplest jaw crusher but shows efficiency limitations in large-scale designs; hence, this type is restricted to laboratory-scale work (Wills, 1997). Although the capacity is smaller than the capacity of a gyratory crusher, a jaw crusher has the advantages of low cost, simplicity of operation and maintenance, and low head clearance, therefore they are broadly used in underground primary crushing (Major, 2002). Gyratory Crusher In the gyratory crusher, rock flows through a chamber formed by two inverted conical surfaces assembled one inside the other. The inner surface, the mantle, is movable and sits on a shaft called the mainshaft; the outer surface, the concave (or concaves) is fixed on the main frame of the machine. The mainshaft is guided by a concentric sleeve at the top and an eccentric sleeve-assembly at the bottom; known as the eccentric. A motor-pinion-gear set propels the eccentric, which in turn drives the mantle in a gyratory movement. Crushing occurs by the circular approaching and receding movement between the surfaces. Eccentric dimensions determine the displacement (or linear moving distance) of the mantle, also called the throw of the crusher. As a result of the assembly of the eccentric 9 at the bottom of the mainshaft, the maximum displacement of the mantle occurs at its bottommost region. Both mantle and concave are cast using abrasion resistant iron alloys and are designed to be replaced over time as they wear. They are also called mantle and concave liners. Gyratory crusher capacities can range from 350 to 10,000 metric tonnes per hour (MTPH) (Utley, 2002). The radial receiving opening (in inches) is the characteristic generally used to determine the size of a gyratory crusher, i.e. a 48 gyratory crusher has a receiving opening measuring 1.22 m (48 in.). Some manufactures add the largest diameter of the mantle to the size description, for example a 60-89 Superior (Metso) has an 1.52 m (60 in.) gap and the largest recommended diameter for the mantle is 2.26 m (89 in.). Since its invention, several different configurations of gyratory crusher have been developed mainly related to the design and support mechanism of the mainshaft. The different types of gyratory crushers include: the long-shaft spider-suspended type, the fixed-shaft type, the short-shaft gearless type, the short-shaft spider-suspended, and the hydraulic supported short-shaft; the latter being the one most commonly manufactured today. The hydraulic supported short-shaft gyratory crusher has the bottom extremity of the mainshaft supported by a hydraulic piston allowing a limited vertical displacement of the mantle. This vertical movement of the mantle, also found in spider-suspended types of 10 gyratory crushers, serves to compensate chamber wear. The hydraulic system gives an additional advantage as it also serves as a quick relief system that is used when the machine becomes blocked; this usually happens when the machine receives tramp material or is operated "too tight", i.e. with a small CSS. Another important aspect about the design of gyratory crushers is that the free movement of the mainshaft inside the eccentric minimizes attrition between the rocks and the surfaces; i.e. once the mainshaft, and therefore the mantle, is free to turn on its axis inside the eccentric, horizontal attrition over its surface is negligible and the main type of mechanical reduction is through compression; compression predominant crushers are most suitable for crushing extremely hard and abrasive rock (Utley, 2002). Figure 3-4 shows a cross-section of a gyratory crusher and the main components mentioned before. Barry Wills (1997) refers to the cross-section representation of the gyratory crusher not only for a better illustration of its mechanism but also to correlate its operation principle with jaw crushers. He explains that at any cross-section, the gyratory crusher can be compared to two double toggle jaw crushers operating at opposite phases. Similarly K. Gauldie (1954) used the jaw crusher operational principle to explain the gyratory crushing. He wrote: "The gyratory crusher may be regarded as a jaw crusher in which a large number of elementary, V-shaped jaws operate in succession. Each of these elementary jaws is advanced and retracted in each revolution and each contributes its share to the total output of the machine." 11 Figure 3-4 Cross section of a Gyratory Crusher (Courtesy of Svedala-Metso) The most advantageous characteristics of the gyratory crushers are as follows: 12 - the round shaped chamber provides higher capacities than jaws with the same gap, - high capacities and the possibility of direct dump from haul trucks (Utley, 2002), low maintenance cost per ton processed and high availability (Utley, 2002), - suitable for crushing hard ores up to 620 MPa (90,000 psi) compressive strength (Utley, 2002), - tend to offer more flexibility with respect to moderating feed rates (Major, 2002), a more even power draw when compared to jaw crushers (Gaudin, 1939), - low starting power peaks (Zandee, 1989), - more uniform wear on liners than jaw crushers (Zandee, 1989), and, - allow setting adjustment even when operating in automatic mode (Zandee, 1989). Cone Crusher The cone crusher, or reduction gyratory crusher, is basically a small scale gyratory crusher where the size of the mantle (head) is proportionally larger when compared to a gyratory crusher, and its outer surface flares out from top to bottom. This configuration results in a much flatter crushing angle than the one found in a gyratory crusher (Major, 2002) and provides an increased area of discharge to optimize throughput (Gaudin, 1939). The rotational speed is different too; cone crushers operate with speeds 2 to 3 times greater than the normal gyratory crushers. 13 The two most common types of cone crashers are the Symons and the Hydrocone, as shown in Figure 3-5. The Water-flush crusher is a design of cone crusher modified to allow the addition of water to the feed material. Figure 3-5:Typical cone crushers (Zandee, 1989). In the majority of cone crashers, a mechanism to release tramp material is provided. In some cases the concaves are lifted, while in others the head is dropped momentarily - by means of coiled springs arrangements in the first case and hydraulic support in the latter. 14 Cone crushers provide consistent product size, which is an important advantage for their selection in the quarry industry. This aspect, and the reduced costs associated with this smaller piece of machinery, explains the greater number of studies found regarding cone crushers than for gyratory. 3.1.3 Gyratory crushers application Gyratory crushers are the usual choice for primary cmshing in high tonnage open pit operations. This trend is even stronger when the ore to be processed is hard and abrasive and the downstream phase requires relatively coarse material such as the feed for grinding circuits equipped with SAG mills (Major, 2002). The gyratory crusher plays an important role in the link between the mine and the mill. The flow of material from mine to mill involves blasting, loading, hauling, crushing, milling and processing. Fragmentation and comminution occur during the blasting, crushing and milling steps of the process. In the overall context of the comminution process, the cost increases as the ore goes from blasting to crushing to milling (Wills, 1997). Highland Valley Copper (HVC) has been closely analysing the parameters involved in the comminution process as a whole and has been very active in assessing the relationship between the mine and the mill operations. Experiments have been conducted and served to confirm that there is a direct correlation between the amount of fines in the 15 mill feed and its throughput. Figure 3-6 and Figure 3-7 show examples of the correlations that were observed at HVC in both SAG and AG mills during tests conducted by Dance. As shown in both graphs the mill production rate (—Tonnage) is closely related with the amount of fines (•Fines), (Dance, 2001). 60 10 24 hours' 1400 1200 • % Coarse ^ 'At - T o n n a g e ^ %' L * r V * J 400 200 Figure 3-6 Effect of Feed Size on AG Mill Tonnage (Dance, 2001) 16 Another test conducted by Dance at HVC served to confirm that the medium-size crushing product, also called critical size, plays an important role in ,mill productivity. With the use of an image analysis system for particle size measurements (discussed in detail in section 4.1), and tracking the flow of the material from the crusher until the semi autogenous grinding (SAG) mill, Dance confirmed the negative effect of the critical size in mill throughput. The graph in Figure 3-8 confirms that the amount of medium-size in the crusher product (•) and in the subsequent mill feed (•) are inversely correlated to the SAG mill throughput (Dance, 2001). 17 60 46 i 40 336 #30 £ | 26 20 16 10 * Crusher Product ° Mill Feed (+24 hours) —Mill Tonnage (+24 hours) '24 hours' t 2200 2100 J. 2400 2300 t 1700 1600 Figure 3-8: Medium-size Crusher Product Effect on Mill (Dance, 2001) The effect o f the critical size mi l l feed is so significant that doubts related to the design of previously accepted comminution flowsheets have arisen. Major (2002) underlines the fact that it had been common for operations to select and implement circuits containing gyratory crushers and S A G mills only; leaving out cone crushers in their comminution flowsheets. However, as he claims, the new trend seems to be the return o f the use o f cone crushers even in flowsheets containing S A G mills in order to crush "recirculating pebbles" (Major, 2002). Major and Dance's work appears to suggest that the best way to solve the problems involving primary crushing product (grinding demands) would be the addition of more secondary crushers into the flowsheet. 18 However, analysing different facts reported by Dance and other authors who have worked on improving various crushing processes, there seems to be room for the alternative approach o f optimizing primary crushing performance. This has been the focus of several authors (Flavel et al, 1988), (Svensson and Steer, 1990), (Burkhardt, 1982), (Dance, 2001). Flavel (1988) listed examples of successes obtained by several research programs and operations that achieved gains in grinding efficiency through improving crushing product quality. Some o f these examples are listed below: - Edmiston and Keller (1975) from Sierrita mine, Arizona, reported that the performance optimization of the crushing process, resulting from detailed analyses of the crushing parameters, increased the capacity o f the concentrators from 65,336 to 78,040 tonnes per day. Excel l and Fitzpatrick (1978) from Broken H i l l Proprietary Co . mine at Whyalla, Australia, reported a 20% increase in grinding m i l l throughput attained by changing the cone crushing settings which also enabled a 15% improvement in crushing and screening plant throughput. - "In 1961 and 1962^ Bergstrom, et al hypothesized that, based upon research findings by Boliden A l l i s (previously Allis-Chalmers), efficient crushing processes could be used to significantly reduce overall comminution energy usage." Svensson and Steer pointed out that in mining operations, many times, inefficient crushing is easily "masked", i.e. coarser crushing product imperceptibly flows directly to 19 the grinding mills. In addition, they claim that on average crashing plants in the mining industry are less developed technologically than the ones in the aggregate industry (Svensson and Steer, 1990). 3.1.4 Gyratory crushers performance Taggart (1927) discussed the capacity of gyratory and jaw primary crashers and concluded: "capacity depends primarily upon character of ore, size o f feed and discharge setting. Throw, speed, angularity o f jaws, and character o f crashing surfaces have a material effect" (Taggart, 1927) Detailed information about the factors influencing crashing performance is the object of Bearman and Briggs ' work. They stated: "crushers operate within a performance envelope encompassing throughput, product size and shape, and power consumed". Though their work is based on cone crushers, the similarities between these types o f crushers and gyratory crushers suggests that the results may be transferable to an analysis using gyratory crashers. In the analyses o f crushing performance, Bearman and Briggs group the variables into three different categories: "Mechanical Design Variables ( M D V ) which do not change with time, Machine Operating Variables ( M O V ) which can be changed with time by the user, and Feed Material Variables ( F M V ) which may change significantly over short periods o f time and are very difficult to control" (Bearman and Briggs, 1998). 20 In the same work, Bearman and Briggs underline the effects o f liner material wear, wear profile, feed size distribution^ feed type, and feed rate as time dependent factors that may be included in M O V or F M V and that significantly impact crusher performance over time. Accordingly, following the above-mentioned categories some examples o f gyratory crusher performance variables are: - M D V : eccentric throw, fulcrum point position, speed (gyrations per time unit), and original chamber design 1. - M O V : feed rate and closed side setting (CSS). - F M V : mechanical properties o f the ore, feed size distribution, and choke feed level. It should be noted that crusher feed is an important parameter to be assessed. Fortunately, different approaches are available in a number o f works (Tunstall and Bearman, 1997), (Burkhardt, 1982), (Dance, 2000). "Well-fragmented material passes easily through the primary crusher, maximizing crusher productivity and minimizing power consumption, liner wear and mechanical breakdowns" (Tunstall and Bearman, 1997). The amount o f fines generated by blasting has a definite impact in crushing performance (Burkhardt, 1982). Table 3-1 shows results 1 If considering wear profile, this parameter might be categorized as M O V (this variable is detailed later in this section). 21 from tests done by Burkhardt that indicate that as the ratio of fines increases in the feed, crusher throughput rises. Furthermore, it can be concluded that primary crushing performance is intimately linked to the degree of fragmentation achieved by blasting (Tunstall and Bearman, 1997). In fact significant efforts have gone into optimizing blast design for crusher feed quality control (Tunstall and Bearman, 1997) (Dance, 2000) (Valery etal , 2001). Table 3-1 Increase in crusher throughput by changing feed characteristics (Burkhardt, 1982) 1.1 m Gyratory Crusher -125 mm Open Side Setting -Biotite Test No. Feed (mm) TPH Percent Change 1 -533+125 (No fines) 655 2 -508 + 0 (20% -125) 763 +16 3 -483 + 0 (40% -125) 887 +35 Since blasting costs are smaller than crushing and mil l ing costs, crushing performance can be improved by increasing the quality o f the feed (Wil ls , 1997), (Burkhardt, 1982), i.e. increasing blasting effectiveness is a good way to optimize the overall process. Following this suggestion, in 1998 H V C conducted a field trial to evaluate different blast patterns. A n area was divided in two halves and subjected to different blast patterns, one following the standard design and the other following a design with a higher powder factor. The material was tracked and sent to the same mil l ing line (C line). The results were contrary to what was expected. The finer blasted material did not generate higher tonnage at the mi l l , instead the inverse occurred (Dance, 2000). 22 By analysing what happened during the blast pattern trial it was realized that the finer blasted product rapidly slipped through the crusher producing a courser product than the one achieved with the coarser starting material from the standard design. Following the trend in the mining industry, HVC had been operated the primary crushers to always get maximum tonnage and avoiding a bottleneck condition, but this test served to bring more attention to the reality of the trade off between tonnage and product quality. After thorough analysis, HVC realized the importance of the gyratory crusher in the overall process: "unless we can maintain the finest crush possible, any gains to be made in blasting finer could be lost before it reaches the mill" (Dance, 2000). The blasting trial served to exemplify product quality degeneration from loose operation of the crusher, i.e. attention was given to one variable (feed size distribution) but others were forgotten (feed rate and choke feed level). Obviously one way to achieve choke fed crushing is by control of the feed rate. This is done with apron feeders ahead of the crushers. Using this crushing configuration, the feed rate can be considered an operating variable and its optimization is possible. Taggart described the optimum feed rate as being one that brings the machine as near as possible to its capacity (Taggart, 1927). Burkhardt accepts Sheppard and Witherow's recommendation that constant choke feeding increases the amount of finer product and adds that the maximization of gyratory 23 crasher product quality can be achieved by operating the machine at small CSS and choke fed condition (Burkhardt, 1982), (Sheppard and Witherow, 1938). Choke fed condition is also recommended for cone crashers, and "means that the volume above the crashing chamber is always full of material which flows into the crashing chamber at a pace decided by the crusher" (Svensson et al, 1996). The accepted reasons supporting the choke fed condition for cone crushers are listed below (Bearman and Briggs, 1998). "develops a uniform wear profile; gives a more consistent product sizing; maximises throughput; extends the liner life; improves the cubicity of the product" The gap or CSS is obviously an important variable, and is categorized as an MOV. Being part of the control of the machine CSS adjustment should be done carefully to optimize crasher performance. Work on the control aspect of cone crushers for the aggregate industry has been done in an attempt to predict and control the output product size (Moshgbar et al, 1995), (Bearman and Briggs, 1998). This work has relied on a mathematical model developed by Whiten to predict the rock breakage during crashing (Whiten, 1972). Using Whiten's 24 model Moshgbar developed a differential equation to describe the crusher product output as a function o f wear and therefore gap variation. Figure 3-9 shows the results using this model to simulate crushing with no control o f the gap, manual control of the gap and automatic control o f the gap. In this figure D S is the wear effect on the gap, % C S is the percentage volume of product with the required optimum size and %OS is the percentage volume oversized. NO COMPENSATION II MANUAL COMPENSATION III AUTOMATIC COMPENSATION Figure 3-9: Comparison between manual and automatic control of CSS (Moshgbar, 1995) The simulation results show that with no adjustment, the gap gradually increases, as expected, due to crusher wear, and the product quality degrades with time. These results demonstrate that manual adjustment o f the gap setting at predetermined intervals 25 improves the product quality, but at the cost of downtime. Finally, automatic control provides the best product control and the most consistent gap setting (section 3.2 contains more details about wear effect over gap adjustment procedures). Tests conducted at HVC demonstrate that by operating the crusher with a fixed setting the variation in feed size distribution results in similar fluctuation in product size distribution. Figure 3-10 shows the effect of feed size on product sizing for the crusher operating with a constant setting (Dance, 2001). 20 16 12 H £ •Feed ° Product • •4- **Vw o % • • 4 V "4 hours' Figure 3-10 Crusher Feed & Product % Course - Constant Setting (Dance, 2001) Chamber design, or cavity design, has been considered a key parameter in crushing performance by both the manufacturers and users of cone and gyratory crushers (Gaudin, 1939), (Burkhardt, 1982). In addition, some of them also claim that with the right design 26 of the liners, wear may be minimized and made more uniform along the profile (Bearman and Briggs, 1998), (Westerfeld, 1985), (Svensson and Steer, 1990). Westerfeld discussed the advantages (including the ones mentioned above) of the design of curved concaves also called non-choking concaves2. He demonstrated that this design of the concaves offsets the choke point to a higher position compared to the position achieved with straight concaves. Thus it minimizes excessive level of stress at the bottom of the chamber, which is responsible for localized and rapid liner wear in this region (Westerfeld, 1985). The drawings in Figure 3-11 were used by Westerfeld to explain the differences in material flow between the two configurations of crushing chambers. It is assumed that the two crushers have the same eccentric throw and the same discharge setting. The difference between the chambers is provided by the different profile of their concaves: the crusher on the left side (1) has a straight concave and the crusher on the right side (2) has a curved profile. Crusher 1 represents a straight chamber and crusher 2 represents a non-choking chamber. To analyse the crushing action in these two chambers, it is first assumed that both crushers are completely filled with a friable material and second, that the crushing action occurs in steady steps as numbered. Thus, at each step (or complete gyration of the 2 "Even though the design is called non-choking it does not afford absolute insurance against choking" (or blockage) "inasmuch as a choke point exists in the crushing chamber" (Westerfeld, 1985). 27 mantle) a volume of material is compressed in one region and then moves to a lower region. As can be seen from the figures, in crusher 1 the volumes successively decrease from region 0 to the bottommost region 19. So, region 19 has the highest probability of packing the material and therefore the choke point in a straight chamber is at its discharge level. On the other hand, for crusher 2 the volumes successively decrease only from region 0 to region 14. Thus, this region has the highest probability of packing the material and as shown in the figure the choke point of a non-choking chamber is located above the discharge level. Figure 3-11: Straight versus non-choking concaves (Westerfeld, 1985) 28 Following Westerfeld's work, Svensson and Steer described the mechanisms of the crushing process inside the chamber with the aid of slices and areas to introduce the constant liner performance (CLP) crushing chamber concept that has been used in cone crushers. They suggest that the advantage of this enhanced chamber design is that it "controls" the liner profile wear in such a way that "the feed opening and the capacity is maintained almost constant throughout the life of the liners" (Svensson and Steer, 1990). Figure 3-12 shows this type of chamber when new and at the end of its life. Figure 3-12: CLP liners (Svensson and Steer, 1990) 29 3.2 Liner wear As described in the previous section liner wear directly impacts the gap dimension as well as the chamber profile (Bearman and Briggs, 1998), (Svensson and Steer, 1990), (Westerfeld, 1985). Hence, liner wear is an important variable in the overall crushing operation as it is intimately related to product quality (size consistency and throughput) and cost; The rate of wear and its distribution among the different regions of the chamber results in profile modifications affecting liner life. Moreover, not only the replacement cost of the liners but also the costs associated with the variations in crusher performance can be attributed to liner wear. In addition, lack of an effective real time measurement system for the wear during operation complicates the application of a systematic adjustment of the gap over time. This lack of accurate gap information may result in either running the crusher too tight, causing a reduction in throughput, or running the crusher with a gap that is too wide, resulting in poor product quality. As a result, unintentionally operating the machine too tight may accelerate the deterioration of the crusher's drive components and once more impact costs. 30 3.2.1 Mechanisms of wear The mechanisms of crushing wear, the variables involved in it, as well as efforts in modelling its behaviour have been the focus of other works in the application of cone crushers (mainly in the quarry industry); (Delalande, 1986a), (Moshgbar et al, 1994), (Moshgbar et al, 1995), (Bearman and Briggs, 1998). Since system kinematics for cone and gyratory crushers are quite similar, the results from these studies are of great value for an investigation of wear mechanisms in gyratory crushers. During crushing, the rock particles are in rolling, impact and sliding contact with the liners of the machine (Moshgbar et al, 1994). Hence the wear of the liners is inevitable and is caused by gouging/ploughing (Bearman and Briggs, 1998). Following a more comprehensive tribological study, it has been determined that the system has a dominant open three-body abrasion wear mechanism, i.e. the type of wear "associated with the abrasion of one or two surfaces of moderate separation by abrasive particles which can move relative to each other as well as rotating and sliding over the abraded surfaces"(Moshgbar et al, 1994). In addition, as the three-body abrasion can be divided into gouging, high-stress and low-stress regimes, Moshgbar concluded that in cone crusher systems there are basically two wear regimes: low-stress at the top of the crushing chamber changing to high-stress as it gets closer to the bottom region where the main crashing zone is located. This theory is 31 supported by the commonly observed larger loss of material in the main crashing zone (Moshgbar et al, 1994). There is general agreement between several researchers about the variables that affects wear rate and the reasons for uneven-wear profiles (Parks and Kjos, 1991), (Bearman and Briggs, 1998), (Moshgbar et al, 1995). Of main interest are those listed below: - material properties of the liners, - properties of the feed (i.e. chemical composition, strength and moisture content), - operational parameters e.g. CSS, power, feed rate and crasher chamber. 3.2.2 Liner materials Cone crasher liners are usually made of an austenitic 12% Manganese steel namely Hadfield steel (Moshgbar et al, 1995), the same material had been exclusively used for both concaves and mantles of large gyratory crashers since their invention until the late 1960's when the first upper row concaves made of martensitic cast steel were introduced. Nowadays concaves and mantles can be cast from several different ferrous materials. The alloys most used are: martensitic steels, martensitic Cr-Mo steels, Ni-Cr (Ni-Hard) white irons and austenitic manganese steels (Parks and Kjos, 1991). Chrome white irons are frequently preferred for gyratory crasher concaves because of their high abrasion resistance, their high yield strength that minimizes plastic deformation in service ("growth"), and their cost effectiveness (Esco, 2003), (Parks and Kjos, 1991). 32 Austenitic manganese or Mn-Cr steels and the martensitic Cr-Mo steels are the alloys most recommended for gyratory crusher mantles (Esco, 2003). During the design and/or selection of alloys to be applicable in sacrificial wear components in the mining industry, the relationship between abrasion resistance and toughness must be considered. Generally, materials showing high abrasion resistance are hard and brittle. An exception is the austenitic manganese steel which combines both requirements, originally austenitic manganese is relatively soft with 200 HB (Brinell hardness) but under certain work conditions its hardness can increase to more than 500 HB which improves its abrasion resistance while retaining its desirable toughness (Parks and Kjos, 1991); (Diesburg and Borik, 1974). Suitable work hardening conditions as mentioned before seem not to be consistently achievable for the concaves under normal operation conditions in gyratory crushers. This limits the application of austenitic manganese steels for concave applications. The martensitic steels and white irons commonly exhibit high hardness values (typically from 500 to 600 HB) even at normal work conditions and have been commonly used for concave liners. Additionally, martensitic alloys have longer wear life than austenitic manganese steels, though their impact toughness is much lower than the one exhibited in austenitic manganese steels (Parks and Kjos, 1991). For material assessment, several laboratory wear tests have been developed (pin-on-drum, dry-sand rubber wheel, jaw crusher, and impeller drum) to assess abrasive wear 33 (ASTM, 1996); (Blickensderfer et al, 1985); (Wilson and Hawk, 1999). The US Department of Energy has compared results for different materials from laboratory wear tests to field test data developed using a "Planar array field test". Basically, a plate is designed with various steel samples and placed in the conveyance section of a mineral processing crushing circuit where it will be exposed to impact and sliding material at different angles and velocities (Tylczal et al, 1999). Although interesting from a material evaluation point of view, this paper does not provide a wear rate model that can be applied to gyratory crushers. It does suggest that laboratory tests are beneficial in evaluating material wear, but care must be taken to ensure that laboratory test mimics the wear process occurring in the actual application. For example the crusher test, which is a gouge test, does not correlate well with the "Planar array test" primarily due to the field test being an abrasive wear test. 3.2.3 Wear measurements As previously discussed, crusher liner wear influences product quality and is closely linked to the overall effectiveness of the process, therefore impacting crashing total cost. Moreover, the lack of accurate ways to measure the wear during operation, limits the efficacy of automated adjustment of the CSS resulting in undesirable variability of product quality over the lifetime of the liners. Gyratory crushers are large-scale types of equipment, which contain robust moving parts that gyrate eccentrically. In addition, the crushing feed usually contains large pieces of 34 abrasive tough rocks that are directly dumped at the top region of the machine. Currently it is impossible to assess liner deterioration in real time for gyratory crushers. The intrinsic characteristics of these types of machines and of the feed they process can provide an idea of the limitations in achieving a desirable online wear measurement. In the following sections descriptions of some examples of liner wear measurement techniques are given. These procedures have been applied by research labs and/or by industry and they serve to better describe the difficulties that are involved in this task. Dril l ing Holes Parks and Kjos describe the periodic use of profile measurements on worn liners as a way to monitor wear behaviour and assist in the selection of mantle profiles to be used over time (Parks and Kjos, 1991). They describe the technique of drilling holes of small diameter at the seams in between the concave parts (through the soft backing-material until the drill reaches the shell) to perform posterior measurements of the holes with the aid of pieces of wire. The limitations of this procedure are quite straight forward: the evaluation is only possible at discrete points, observation of scrap concave parts reveals different thicknesses close to the edges, mantles are mounted in a maximum of three pieces resulting in a low number of radial seams, large inaccuracy caused by the nature of the procedure, and time consuming. 35 Mines use a similar technique with the main difference being that holes are drilled directly into the liners (Adams, 2003). Although this technique can be better applied to mantles and may avoid the error incurred when measuring thicknesses close to the edges, it suffers from most of the disadvantages of the original procedure. In addition, it may create stress concentrators and lead to the deterioration of liner life. Ultrasonic Thickness Gauge There have also been attempts to measure the thickness of worn liners with the use of ultrasonic gauges (Parks and Kjos, 1991), (Adams, 2003). Difficulties and limitations mentioned by Parks and Kjos include the necessity of selecting sections with known parallel wear surfaces where the measurement is made to avoid loss of back reflection. Adams (2003) reported that HVC performed some tests in the past but the results were inaccurate and most of the problems were related to reflections and/or originated from the backing materials. Experiments to Measure Cone Crusher Chambers Delalande first showed interest in determining the optimum operational period for cone crusher liners following an observation of the degeneration in product quality over time and an apparent opportunity for cost reduction when shortening the life of the liners (Delalande, 1986a). In a later paper, Delalande reported that to continue the investigation it was necessary to get an understanding of the dynamics of wear across liner lives, he 36 then described three different methodologies to obtain the chamber profile of the crusher at the closest plane of use (Delalande, 1986b). Since the methods were designed for laboratory scale tests there are obvious limitations for easy adaptation of the methods to large primary crushers. However, the functional principles may serve to provide the foundations for other novel approaches for large-scale gyratory crushers. The three methods are briefly described in the following: 1- One method to obtain the chamber profile utilises a profiler device, this apparatus contains a horizontal bar (fixed at the top of the crusher) that support a second articulated bar which lies inside the chamber, a measure device similar to a tape measurement which slides over the second bar, and an electrical system that reads and records the inclination of the bar and the distance measurements over time. Basically the functioning of this apparatus is in reading the length and the inclination of the articulated set of measuring bars as its extremity travels touching the surfaces of the two opposite walls (concave and mantle). 2- Another method is called profile by casting. In this procedure a cylindrical latex bag is inserted inside the chamber and kept in a fixed position with the aid of a metallic tripod fixed at the top of the crusher and a rod linked to the tripod and kept inside the bag. The bag is filled with a pre-mixed two-component-resin (approximately 50% in volume) and the mixture is let to expand and solidify. The cast is cut in a predetermined manner, removed from the crusher and then the profile can be analysed. 37 3- The third method is called profiler by direct sketch. In this method a wood-board is patterned in the approximate shape of the crushing chamber and placed in the crusher. Once the board is in place, a compass with a fixed opening is drawn along the profile of the concave or mantle and its shape is traced onto the board. Sacrificial Sensors An online methodology to monitor the wear in cone crusher liners for the quarry industry has been the object of research and development for several years. The methodology would complement the implementation of a full condition monitoring system in order to achieve several operational benefits such as the optimum utilisation of the liners, product quality enhancement with the use of an automatic adjustment of the gap and feed rate, as well as the reduction of maintenance downtime (Moshgbar et al, 1995); (Yaxley and Knight, 1999). In this wear measurement methodology sacrificial sensors of approximately 0.5 mm diameter are embedded in different regions over the liners. The sensors wear away at the same rate as the metallic liners. Each sensor sends a signal that corresponds to its current length which enables the system to give an accurate representation of the wear at real time. Different prototype sensors have been developed using capacitive, resistive and conductive principles, and for each configuration laboratory and field tests have been performed. It was concluded that the most promising configuration is the one that applies multiple surface mount resistive sensors. Although the field tests showed problems such as short-circuiting of the sensors and signal spikes it is expected that with further 38 development the reliability and accuracy of the multiple surface mount resistive sensor may be improved and this measurement methodology may be used in industrial applications. 39 4. 4.1 Highland Valley Copper Project Operations Background HVC is a Teck Cominco and BHP open-pit mine operation located near the town of Logan Lake in the southern interior British Colombia, Canada. HVC is one of the largest copper mining operations in the world. In 2002, HVC mill achieved 50 million tonnes of total throughput (on average 137,000 tonnes per day), the largest throughput in 20 years of operation (Teck Cominco, 2003). At the beginning of 2000, ore reserves totalled 387 million tonnes at a grade of 0.417% copper and 0.009% molybdenum - gold and silver are present in small quantities that become noteworthy in the copper concentrate. On average, the strip ratio is 1:1 resulting in 270,000 tonnes mined per day in two pits simultaneously. Valley, the main pit, is located 3 km northwest of the mill plant and contains approximately 74% of the reserves; the remaining ore comes from the Lornex pit, located 1 km southwest of the mill plant (Richards, 2000). Mineralization is bornite and chalcopyrite for both pits. Valley pit shows a higher ratio of bornite to chalcopyrite and the reverse occurs in Lornex; Valley pit has a lower molybdenum sulphide mineralization than the Lornex pit (MacPhail, 1992). 40 Being a high tonnage and low grade mine, H V C ' s existence relies heavily on economies of scale. H V C has always applied innovative technology to increase productivity and to reduce operating costs. Some examples o f such advanced technologies in the history o f H V C are the application of a computer based truck dispatch system, movable in-pit crushers and conveyors for ore transportation in the Val ley pit, Global Positioning Systems (GPS) location technology on drills and shovels, shovel weighing system and fragmentation image analysis system. The overall mi l l ing process can be visualized in Figure 4-1. The R O M ore from the Lornex pit is trucked and directly dumped into a fixed 1.52 by 2.26 m (60 by 89 in.) Metso Superior Gyratory crusher; the crusher is driven by a 520 k W (697 hp) motor and is equipped with a heavy duty hydraulic hammer at its top to deal with over size material and to help clear blockages. The Lornex crusher is designated Crusher No . 1. 41 Figure 4-1 HVC Simplified milling flowsheet (HVC, 2000) The ROM ore from Valley pit is trucked and dumped into two 1.52 by 2.26 m (60 by 89 in.) Metso Superior Gyratory crushers, No. 4 and No. 5. These crushers are semi-mobile and located deep in the Valley pit. Differently from No. 1, the in-pit crushing layout includes a dump hopper and a 2.44 m variable speed inclined apron feeder which avoids truck direct feed and enables feed rate control. Previously, all three gyratory crushers reduced the ore to app. 250 mm (~ 10 in.). However, for the past few years, the crushers have been operating to reduce the ore size 42 to app. 150 mm (~ 6 in.) with the use of CSSs ranging from 127 to 140 mm (5 to 5.5 in.). As commonly described at the mine, the crushers were operating "loose" in the past but now they have been operating "tight". A network comprised of several kilometres of conveyor belts, feeders and surge piles is used to deliver the crushed product to three stockpiles located just beside the mill plant. Although there are limited crossovers between the different crushed products in the conveying phase, some blending of the grinding feed is possible if desired. Three variable speed apron feeders and two hydra stroke feeders are used to transfer the material from the three stockpiles (1, 2 and 3) to five grinding lines (A, B, C, D and E). The grinding lines were built at different times and consequently, there are substantial differences between them. A and B lines are similar, each of them is comprised of one primary SAG and two ball mills. C grinding line is similarly equipped with one SAG mill and two ball mills. The mills in C are larger than the mills in A/B lines and line C utilises a cluster of ten cyclones instead of seven for the A/B lines. Each of the other two grinding lines, D and E, consists of one fully autogenous mill (AG) and one ball mill. Each AG mill is equipped with one discharge grate and one vibratory double deck screen for the removal of critical size rock particles. A 2.1 m (7 ft) Symonds short head cone crusher is used in closed circuit to crush the oversize material from the screen in each line, and the ball mill operates in close circuit with a cluster of ten cyclones. 43 After comminution, flotation cells are used to produce a concentrate that contains both copper and molybdenum. Regrind circuits, comprised of ball mills and cyclones, are additional components of the flotation cells. In the final processes, molybdenum is separated from the bulk copper-molybdenum concentrate by flotation and leaching and the final products are arranged for shipment. As mentioned before, several advanced system are applied at HVC. The systems relevant to this research are discussed next. HVC utilises an image analysis system to monitor the size distribution of crusher feed, crusher product and the feed of the grinding lines. The system consists of several video cameras mounted in strategic locations and a PC-based fragmentation analysis system developed by WipWare Inc. called WipFrag. The software captures and digitises images of the material and isolates individual fragment boundaries as shown in Figure 4-2. The results of this fragmentation recognition are used to calculate particle areas, volumes, masses and the size distribution by weight. 44 Figure 4-2 Camera Image and Boundaries Recognition (Dance, 1998) The system has some limitations such as the inability to recognize fine particles (smaller than 15mm) and the need for controlled lighting conditions (in some location this is solved by the addition of halogen lamps). However, these restrictions do not compromise the objective of its application at HVC, as WipFrag outputs are mainly used as control signals. Even though the output accuracy is not the same as from lab screening analyses, WipFrag output has proven to be repeatable and reliable for its designed application (Simkus and Dance, 1998). The use of WipFrag output as a control signal is possible because, in this case, relative changes in the distribution are more important than the comparison of the signal to a "standard" sieve analysis (Simkus and Dance, 1998). HVC relies on state-of-the-art systems to monitor ore properties. At HVC, all the drills are equipped with GPS-based navigation and blasthole guidance systems as well as material recognition system from Aquila Mining Systems. The Aquila material 45 recognition system provides rock characteristics by the analyses o f dril l parameters such as rate of penetration and vibrations. Since 1997, H V C has been utilizing the Citect process control system from Citect Pty Ltd, Australia. Using Citect, operational data is gathered automatically from equipment instrumentation and then processed and recorded at small time intervals. Citect facilitates the search for real time and historical information as well as enables the overview o f the entire operation from several workstations at the mi l l . The system comes with a detailed graphic user interface and enables the creation of tailor made reports. Using Aquila , Citect, Dispatch, WipFrag and other technology systems H V C has achieved the capacity to track ore properties throughout the crushing and grinding processes. More details of a study conducted at the mine with regards to the systems mentioned can be found in (Simkus and Dance, 1998). 46 4.2 Crusher Operations at HVC As described before, every day a massive amount of material is milled at HVC (on average 137,000 tonnes). However, only three crushers are responsible for all the ore processed from the two pits. This condition by itself underlines how crucial the availability, maintenance and proper operation of the crushers is. Maintenance problems like unscheduled shutdowns may affect the profitability of the entire operation, while improper operation conditions at the crushers decrease the final throughput of the mill. The dynamic nature of the crushing process is fundamental to a desired flow of material from mine to mill. Though fragmentation and comminution occurs during blasting, crushing and milling steps of the process, there is a better opportunity with regards to costs in optimizing the first two processes since the cost increases as the ore goes from blasting to crushing to milling (Wills, 1997). HVC has been very active in enhancing the product quality resulting from both blasting and crushing, aiming for an overall improvement in performance of the comminution process. As discussed in section 3.1.4, in 1998 HVC initiated this optimization development by the investigation of the effects in mill productive by application of enhanced blasting designs. This work resulted in the recognition of the key importance of the primary crushing process in the overall comminution process at the mine. 47 In 1999, drawing on advances in cone crusher technology, HVC initiated the development of an automatic control system for the crushers working in the Valley pit. Flavel demonstrated that the use of automatic CSS controls in a cone crusher substantially improved both the capacity and .the quantities of finer sized product (Flavel, 1982). Figure 4-3 shows the difference between two similar sized cone crushers operating with and without automatic CSS regulations. In this case, the crushers are used in the secondary crushing process and they are equipped with screens to sieve the feed. 1 1210TPH-A<>fc£V fc — 358 TPH 852 TPH 519 TPH 500 ! 15 HP AUTOMATED SET 20-22 MM C S S NET PRODUCTIVITY • Vi" (13 MM) 435 TPH • 13 MM 417 TPH - 13 MM 775 TPH POWER R A T E 255 : 45 HP FIXED SETTING 22 MM C S S XJ DISCHARGE % PASSING 992 832 65.9 48.9 35.4 27.1 IN. 2 1 75 OS 0.38 0.25 0.586 51 25 18 13 9.5 64 99.6 79.9 532 35.3 26.4 205 0.491 J . 200 TPH 336 TPH • 13 MM 183 TPH - 13 MM 383 TPH KW/T Figure 4-3 Two 84" 0 Mantle Cone Crushers Test Comparison (Flavel, 1982) As shown in Figure 4-3, the machine equipped with automatic CSS control achieves a net production of 775 tph and has 48.9% of the discharge product passing -13 mm (-0.5 in.). On the other hand, the machine with fixed CSS achieves a net production of 383 tph and 48 has 35.3% of the discharge product passing -13 mm (-0.5 in.). Flavel explains that when using a fixed CSS the average operating power drawn is usually restrict to app. 50 percent "of that connected to guard against crusher stalling and minimize mechanical damage" and the application of the automatic control allows the machine to normally operate at much higher average power rate. As shown in Figure 4-3, there is a considerable difference between the operating power drawn for the two machines. The control system of the crushers at HVC aims to maintain a constant choke fed operating condition and to keep the product size distribution within a predetermined quality range. The design and layout of the semi-mobile crushers, containing the dump hopper and the variable speed apron feed, enables the application of such an automatic control system. A fuzzy logic-based control algorithm is the basis of the system. Based on a group of operational parameters, the algorithm adjusts the apron feeder speed and the vertical mantle position at 30 seconds intervals. The operational parameters that serve as inputs for the system are as follows: - dump hopper level, - crusher pocket level, - crusher motor power, - product size distribution, and, - tonnage. 49 Figure 4-4 shows the system graphical interface available for the operator and gives an example of real measurements and outputs. In this example, it is possible to observe that the dump level was at 46%, the pocket level at 36%, the tonnage was 4248 tph, the motor power drawn was 77.0 amps and the product size distribution showed 30%-37%-30 % moreover, the system outputs were: 31.1% for the apron feeder speed and 109 mm (4.3 in.) for the mantle position. Figure 4-4 Automatic Crusher Control Graphic (Dance, 2001) When a higher tonnage is requested in detriment of quality product, for example, when one of the crushers is down and/or the level of the stockpiles are low, a special system mode can be activated where the high tonnage is given priority and crushing is loose. 50 On the other hand, when the two crushers are in full operation and the frequency of the trucks is low, the control favours improved product quality and the crusher operates with a "tighter gap" (small CSS) . Motor power and oi l temperatures frequently serve as "health" parameters of the machine and, when extreme conditions are perceived, the system lowers the mantle and decreases the speed of the feeder until normal conditions are restored. Being a crucial determinant of crusher operational conditions, the vertical mantle position adjustment varies only within predetermined limits when the crusher is set for automatic control. To determine the mantle position limits, frequent direct measurement of the gap is necessary. In practice, what must be verified, as frequently as possible, is the relationship between a set of mantle positions (for example 50, 100, 150, and 200 mm) and their corresponding C S S dimension. This assessment is valuable because it serves as a tool to forecast which range of mantle vertical positions would provide the acceptable range of CSSs. A s mentioned in the literature review, the wear that liners are subjected to modifies the chamber shape and the gap over time (refer to section 3.1.4), therefore accurate forecasts of C S S versus mantle position are virtually impossible unless online measurements are available. 51 H V C has been assessing the relationship between C S S and mantle position every week using a measurement procedure known as the "bucket test". In this procedure, several metallic buckets (filled with sand) are thrown into the crusher while it is operating empty. The process is repeated for two or three predetermined mantle positions. By evaluating the bucket size before and after, the gap variation as a function of mantle position can be estimated. Typical results o f this procedure are shown in Figure 4-5. Although effective, the method requires the crusher to be down, lacks accuracy, and does not provide chamber profile information details. Crusher 4 Mantle Position vs C S S Date 19-Aug-02 High Limit: 4.9 inches (5.0-CSS) Mann*: instaled 2S-July-02 Low Limit 3.9 inches (5 5-CSS) ConcavM: Instaled 8-Feb-CC Wear Eat 0.16 inches/day Comments: (move mantle up this much per day) Figure 4-5 Gap measurement results and analysis performed at HVC 3 The units presented in graphs and in HVC's reports shown in this work follow the U.S. customary system because this is the system commonly used in the mine. 52 In order to create a better alternative to the "bucket method" for measuring the gap, and to develop an accurate method to collect comprehensive wear data, H V C purchased a prototype laser profiler device developed by Conveyer Dynamics, Inc. The equipment, its application, and measurements results are discussed in the next section (5.1). 53 5. Experimental Approach 5.1 Equipment The equipment used to measure crusher wear and the chamber shape was a laser profiler device (LPD). HVC purchased a prototype device of this type from Conveyer Dynamics, Inc. (CDI). It should be noted that the LPD is the first of its kind used to measure a crusher chamber profile. Figure 5-1 shows the major components of the LPD and Figure 5-2 shows a schematic of its installation inside the crusher. Basically the LPD is comprised of: - a support structure to mount it to the crusher, - a track for the laser to run on, an additional structure containing five calibration bars, - an actuator motor to drive the laser up and down the track, - a time of flight laser with a mirror for reversing the target direction, and, - software and a computer to collect and process the measurement data. 54 Figure 5-1 Major components of the LPD LPD S u p p o r t Figure 5-2 LPD Installation schematic The laser is a DME 2000 Distance Measuring Device from SICK Optic-Electronic Inc. The device optically measures the distance to a target (opaque) by transmitting a modulated red laser beam and measuring the time of flight of the beam. The laser is used in its proximity mode with a maximum range of 2 m (6.7 ft). The output signal is 4-20 mA with a resolution of 1 mm, a repeatability of 0.8 mm and an absolute accuracy of +/-5 mm. CDI assures that the absolute accuracy is improved to 1mm by the use of the calibration bars to yield a correction factor. The output signal of the laser is sent to an analog input of the actuator motor for transmission to the PC (Nims, 2001). The actuator motor is a SilverMax® "E" from Quicksilver Control Inc. This actuator motor communicates as a slave to a PC. The PC polls the actuator motor for the contents of the position/distance data. On average, a pair of coordinates containing the laser position on the track and the distance to the target is obtained approximately every 3 mm along the track (this can vary depending on the velocity used). The user interface, CDI Laser Scanner Program, is written in Visual Basic. This program executes all the necessary functions to run the tests, to acquire data and to provide final profiles. The laser device provides an excel spreadsheet for each test performed. The excel file contains data representing several points from the surface of the liner given as pairs of 56 coordinates. Table 5-1 shows an excerpt o f the spreadsheet (first 5 points from a concave measurement). Table 5-1 Example of a partial table result for a concave profile measurement generated by the LPD Actuator Position (Y L) (mm) Laser Distance (X L) (mm) 3.78 1291.48 7.14 1287.92 11.34 1287.17 14.7 1280.42 19.32 1279.74 A s shown in the example, each point has its actuator position on the track ( Y L ) in the first column and the laser distance to the target ( X L ) in the second column. These coordinates are not Cartesian X , Y pairs o f coordinates. Figure 5-3 illustrates the measurement process. 5 7 ( H o m e / S ~ t a r t ) Laser position on the t r a c k • Y u \ \LASER 1 L i n e r MIRROR L Target distance >Y—~——— Figure 5-3 Schematic of the laser measurement During the crusher profil ing procedure four types of measurements are taken: 1 s t - shooting the calibration bars with the laser beam perpendicular to the track; 2 n d - shooting the bars after levelling the mirror; 3 r d - shooting the concaves of the crusher; and, 4 t h - shooting the mantle after rotating the mirror 90 degrees. The first two measurements are performed to calibrate the L P D prior to acquisition o f the liner measurements. The CDI software uses these two calibration tests to calculate the angular dimension formed by the actuator and the vertical centre line of the crusher, or inclination angle of the track ( a ) . This angle is used to map the laser coordinate system ( X L , Y L ) to a Cartesian system. The calculation is made possible by comparing the results from the first two types of measurements when the laser shoots the same bar; at first 58 h a v i n g the laser m i r r o r base perpendicular to the track and second h a v i n g the laser m i r r o r base leve l led hor izonta l ly . F i g u r e 5-4 i l lustrates the procedure. Figure 5-4 Calibration procedure schematic O n c e the i n c l i n a t i o n angle o f the track, a , is calculated the software generates the prof i les o f the mantle and the concaves as p o l y l i n e s i n an A u t o C A D format. F i g u r e 5-5 and F i g u r e 5-6 s h o w an example o f a concave and a mantle prof i le , respectively. 59 Figure 5-5 Concave profile generated by CDI software j j £ fc E « Ym< lrnat fgmat loots Qlaw Dimension HaSy XchmeeWoiks Window H<* D c * H i » a < ^ i * a a a ^ i « " i a F » a > » » i - - 8 L . i a 3 i ^ t f c g t a i g t i i J ! i i y | T > j i © II 'J a S o 3 I « C i | | [ a * « | I l I i a l A s °* & «3> o gg a * r Oe Q <~ Q o - / - C J Et r B r A «** 73 Figure 5-6 Mantle profile generated by CDI software The profiles are transferred to a section-view drawing of the crusher and aligned with the original liner profiles in order to compare the wear. Figure 5-7 shows an example of the 60 final output o f the process containing three profiles from different days of measurement juxtaposed. Figure 5-7 Laser profile final output A s expected, during the initial use of a prototype device, some issues needed to be addressed. One major problem with the L P D was the time necessary to perform the 61 measurements, in particular the set up time. It was identified that the excessive set up time was related to the removal of the crusher spider cap in order to install the L P D support. The set up time was significantly reduced by the redesign and construction o f a new support structure. Figure 5-8 shows a picture o f the original support on the right and a picture o f the new support on the left. The actual drawings for the new support are included in Appendix A . Figure 5-8 Original support / New support The new support structure was designed such that: - it is installed around the mainshaft and there is no need to remove the spider cap (necessary with the original structure), - it is lighter than the original, dismissing the use of the crane for its transportation, - it provides a new mechanism of connecting the track to the support, minimizing the time spent during set-up, 62 - there was low costs involved in its construction and its design allowed an in-house construction. Thus, one could be built for each crusher. Another difficulty was related to the generation o f two separated profile drawings, one for the mantle and one for the concave in a horizontal position. To get the two profiles together fol lowing their real inclination as well as to solve some issues with the original software it was decided to develop a new program. 63 5.2 Data Collection 5.2.1 Crusher operational data Crasher operational data from the Citect system was collected during periodic field visits. The data was collected for crushers 4 and 5 from the 1 s t of January 2001 to the 30 t h of September 2002 and included the following items: - Crusher Bow l Level Mantle Position - Motor Current Draw (Amps) Production (mtph. at the conveyor) - Feed Size Classes (course, medium and fine) - Product Size Classes (course, medium and fine) Data points were obtained for one-hour periods. The Citect system processed the records for each 30-second interval to produce an hourly average of the 120 records. In addition, the maximum and minimum values within the hour were also recorded and collected. 5.2.2 Liner information In order to assess the impact of liner life on maintenance costs, the costs involved in liner replacements were collected. The costs reported by the maintenance department for Feb/2002 are listed in Table 5-2 and wi l l serve as a basis for liner management analyses. 64 Table 5-2 Liner rebuild/installation: labour and costs (Wolff, 2002) C o n c a v e s - Installation Labour/Parts C o s t (Cn) Liners - 4 R o w $79,500 Supp l i es $9,750 Labour (360 m a n hrs) $13,680 Total* (72 hrs down time) $102 ,930 Mant les- Rebuilt Labour/Parts C o s t (Cn) $30,550 $4,000 $5,500 Installation (48 m a n hrs) $1,850 Total* (12 hrs down time) $41,900 Liners Supp l ies Labour (144 m a n hrs) For more than three years, HVC has been using the same type of concave liners for Crushers 4 and 5. The concaves are supplied as a four-row set made of high chrome white iron by Penticton Foundry Ltd. Information and drawings for several different mantle types are available at the mine. As shown in Table 5-3, basically three sizes are available: under-size, standard-size, and over-size. In addition, as different designs, diameters and configurations (two or three pieces) exist, a total of ten types are listed. Although these ten types of mantle have been used at HVC, only the eight types shown in bold were used during the period analysed. Table 5-3 Mantle types Mant les M a x i m u m Diameters Vendor/Type Under-s i ze S tandard Over -s i ze m m (in.) m m (in.) m m (in.) E s c o / 2 p c s 2184(86) 2254 (88.73) 2286 (90) E s c o / 2 p c s (Type II) — 2235 (88) 2286 (90) Esco/ 3 p c s — 2216 (87.25) 2328(91.65) Frog Switch(Transwest)/ 2 p c — 2226 (87.62) 2302 (90.625) Columbia/ 2pcs(with ribs) — 2224 (87.56) — 65 Mantle and concave information, such as: period of use, type, alloy and number of reused parts was collected for the period of this analysis. The first source for this information was reports provided by HVC maintenance department (Figure 5-9 shows an example of these reports). In addition, information available from the Technical Development Department was cross-referenced (Figure 5-10 shows one example of these reports). Moreover, direct consultation with HVC personnel was initiated when any inconsistencies or lack of information occurred. #5 CRUSHER MANTLES Date Installed Date Removed # Weeks Vendor Type of Material DMT (Mill) Comments 08/03/01 08/31/01 4.0 Esco 0/S 1,253,851 Installed Mantle #3 09/01/01 11/22/01 12.0 Transwest o/s 3,B4B,B42 Installed Mantle #4 11/23/01 12/20/01 4.0 Esco 3 Pee 1,101,089 Inst. #2 cAw2,603,488T 12/21/01 01/03/02 2.0 Columbia Ribbed skins 822,553 Inst. #1c/W1,4B3,347T 2,305,900 Total Tons 01/04/02 01/31/02 4.0 Esco 2 Pee. 1,219,630 lnstJ5 cAw 1,421,6GB T 2,641,238 Total Tons 02/01/02 03A4/02 6.0 Esco 2 Pee. 1,268,140 Installed Mantle #2 03/15/02 03/20/02 0.8 Esco 2 Pee. 32,810 Installed Mantle #1 03/21/02 03/28/02 1.0 Esco 2 Pee 174,812 Inst.#2 cAv 1,268,140 T 1,442,952 Total Tons 03/29/02 04/11/02 2.0 Esco 3 Pee. Std. 401,486 Installed Mantle #5 04fl2/02 04/26/02 2.0 Esco 2 Pee 334,723 Inst. #1 cA<v 32,81 OT 367,533 Total Tons 04/27/02 05/09/02 2.0 Esco 2 Pee. 518,609 Installed Mantle #2 05/10/02 1.0 Esco 2 Pee. 149,117 Installed Mantle #1 Please Note: The mill Tonnages may not reflect an accurate figure - may be missing some Met. Tonnage data. Figure 5-9 Partial example of crusher-mantles report 66 Crusher 4 Mantle/Concave History 1 I Crusher 5 Mantle/Concave History Date Hew Mantle Mew Concaves | 1 Date Hew Mantle New Concaves 22-Jun-01 Esco 3-piece Standard 87.25" OD CZ18 alloy 30-Jun-01 . Frog Switch 2-piece Standard 87.56" Manganese 1.3Mtonnes HEW 06-Jul-01 Esco 2-piece Oversize 90"CZ18alloy 26-Jul-01 3-piece Standard 87.25" OD CZ18 alloy 0.65 Mtonnes Figure 5-10 Report from Technical Development Dept Combining all the information available comprehensive tables were prepared for each crusher. The results for Crushers 4 and 5 are shown in Table 5-4 and in Table 5-5, respectively. 6 7 Table 5-4 Crusher 4 liners detailed information Mantle Install. Removal Hours Cum. Vendor Type Material Code Date Date in use Tonnage Mantles M302u 6/29/01 7/26/01 610 1,598,086 F S . 2-pce Std 87.56" Mang. M503u 7/26/01 9/7/01 965 2,562,607 Esco 3-pce Std 87.25" C Z 1 8 M104n 9/8/01 10/18/01 844 1,978,516 Esco 3-pce Std 87.25" C Z 1 8 M505n 10/18/01 11/15/01 618 1,317,435 Esco 3-pce Std 87.25" C Z 18 M106u 11/17/01 11/29/01 262 568,610 Col. 2-pce Ribbed Std Mang. M507n 11/29/01 12/27/01 614 1,115,941 Esco 2-pce Std 88" Mang. M408n 12/28/01 2/7/02 907 2,027,699 Esco 2-pce O/S 90" Mang. Total 4820 11,168,894 M509u 2/9/02 2/21/02 257 517,944 Esco 3-pce Std 87.25" C Z 1 8 M410n 2/22/02 5/2/02 1573 3,531,753 Esco 3-pce Std 87.25" CZ 18 M311n 5/2/02 6/21/02 1144 2,733,731 Esco 3-pce Std 87.25" C Z 1 8 M112n 6/24/02 7/25/02 714 1,349,298 Esco 2-pce Std 88" T. II Mang. M213n 7/27/02 9/3/02 855 1,797,474 Esco 2-pce O/S 90" T. II Mang. M414n 9/4/02 9/25/02 408 862,988 Esco 2-pce O/S 90" T. II Mang. Total 4951 10,793,188 Concaves C002 6/29/01 2/7/02 4820 11,168,894 Penticton Standard W.I. C004 2/9/02 9/25/02 4951 10,793,188 Penticton Standard W.I. 68 Tab le 5-5 C r u s h e r 5 liners detailed information Mantle Install. R e m o v a l Hours C u m . V e n d o r T y p e Material C o d e Date Date in u s e T o n n a g e Mant les M 2 8 1 n 1/5/01 2/24/01 948 2,222,258 E s c o 2-pce U/S 84" M a n g . M 5 3 1 u 2/25/01 3/1/01 92 188,265 E s c o 2-pce Std 88.73" M a n g . M 1 8 2 n 3/2/01 5/25/01 1,715 2,970,699 E s c o 3-pce Std 87.25" C Z 1 8 M 2 8 3 n 5/26/01 6/21/01 566 1,071,485 Co l . 2-pce R ibbed Std M a n g . M 5 5 1 n 6/22/01 7/5/01 315 686 ,248 E s c o 3-pce Std 87.25" C Z 1 8 M 1 0 1 u 7/6/01 8/2/01 621 1,443,101 E s c o 2-pce O/S 90" C Z 18 M 3 5 1 n 8/3/01 8/30/01 614 1,339,742 E s c o 2-pce O/S 90" C Z 1 8 M 4 5 2 n 9/1/01 9/27/01 549 1,312,926 F.S. 2-pce O/S 90.6" M a n g . Total 5,420 11,234,723 M 4 5 2 u 9/30/01 11/22/01 1195 2,501,162 F.S. 2-pce O/S 90.6" M a n g . M 2 5 3 u 11/23/01 12/19/01 605 1,257,902 E s c o 3-pce Std 87.25" C Z 18 M 1 0 6 u 12/19/01 1/3/02 346 955,560 C o l . 2-pce R ibbed Std M a n g . M 5 0 7 u 1/3/02 1/31/02 6 1 9 1,443,466 E s c o 2-pce Std 88" M a n g . M 2 5 4 n 2/1/02 3/14/02 922 2,335,759 E s c o 2-pce O/S 90" C Z 18 M 1 5 5 n 3/15/02 3/20/02 34 22,911 E s c o 2-pce O/S 90" C Z 1 8 M 2 5 4 u 3/20/02 3/28/02 196 423 ,797 E s c o 2-pce O/S 90" C Z 18 M 5 5 6 n 3/29/02 4/11/02 309 777,223 E s c o 3-pce Std 87.25" C Z 1 8 M 1 5 5 u 4/11/02 4/26/02 337 640,472 E s c o 2-pce O/S 90" C Z 18 M 2 5 7 n 4/26/02 5/9/02 282 595,668 E s c o 2-pce O/S T . II 90" M a n g . M 1 5 8 n 5/10/02 6/7/02 618 1,218,575 E s c o 2-pce O/S T. II 90" M a n g . Total 5,463 12,172,495 C o n c a v e s C 0 0 3 1/5/01 9/27/01 5420 11,234,723 Penticton S tandard W.I. C 0 0 5 9/30/01 6/7/02 5463 12,172,495 Penticton S tandard W.I. 69 As can be seen in the tables only two materials have been used for the mantles, austenitic manganese steel ("manganese") and martensitic chrome moly steel ("CZ 18 alloy"). In addition to the liner information previously mentioned, dimensional details about the original liner profiles and some parts of the crushers were collected. For all types of liners available, drawings containing their profile were supplied by the vendors. Although original dimensions and part details were not supplied by the manufacturer of the crusher (they refused to provide it), historical measurements taken by the maintenance department were used to generate a section-view drawing of the crusher. More details about this drawing and its use in the wear determination procedure will be covered in the next section. 5.2.3 Chamber profile data Chamber profile data collection was initiated on June 14 2001 using the LPD (described in section 5.1). The original plan was to obtain one set of measurements for each crusher every other week, following a predetermined positioning arrangement. For every other measurement, the LPD was positioned at the 4 o'clock and 10 o'clock regions (relative to the control cabin), Figure 5-11 shows a positioning diagram. The two locations used for measurements were chosen based on practical observations which suggested that these locations present different concave profile wear, i.e. distinct high-localized wear regions. 70 C o n t r o l 4 O'clock Figure 5-11 Measurement Spots Due to changes in the crusher maintenance shutdown schedule by the mine, the data collection period increased from every second week to every third week. However, due to operational issues, situations arose where measurements could not be taken. This resulted in an actual frequency of measurements per crusher of three measurements every two months on average. 71 5.3 Data Analysis 5.3.1 Crusher Operational Data For each crusher the complete set of data from Citect along with basic liner information and measurement dates were all grouped into a single spreadsheet. Figure 5-12 shows examples o f records as they are listed in this comprehensive data file. Figure 5-12 Example of the complete data file of Crusher #5 (records 5075 to 5714 are hidden to facilitate visualization). HBeord Date & time U W Cancan B O M FaadStas Praam S E S " M b ton. | Max. M r . Cum. ton. Nbr. Cum. ton. i 1 J i r " ^ i um Cour. f a t Mod. Cow. fla Mad. (m/d/yy h:mm:ss) (mtph) (tons) (tons) (amps) w w W w (inches) (*> m w 5066 8/3/01 2:00:30 0.0 0.0 z Z coos 9,582,055 0.0 0.0 ~ o i r 3.0 ~ST3~ 39.7 " ft6F 21.0 38.8 323 5067 5066 S0S9 5070 8/3/01 3:00:30 8/3/01 4:00:30 8/3/01 5:00:30 8/3/01 6:00:30 - 0.0 0.0 0.3 1.4 0.0 0.0 8.4 18.8 - - C0Q3 C003 C003 C003 8,582,055 8,582,055 8,502,055 8,582,057 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 3.0 3.0 3.0 37.3 37.3 37.3 37.3 39.7 39.7 39.7 39.7 0.09 009 0.09 0.09 21.0 21.0 21.0 21.0 38.8 38.8 38.8 38.8 323 32.3 32.3 323 Start M351n M351n 0.3 1.7 5071 5072 8/3/01 7:00:30 8/3/01 8:00:30 - 2.1 2.6 58.1 18.3 4BB7.5 5386.3 M351n M351n M3S1n M351n 38 6.5 C003 C003 C003 coos 8,582,059 8,562,062 41.3 61.0 537 60.3 81.0 77.0 111.0 101.0 0.0 0.3 r . i 18.1 3.0 3.0 ni 331 57.3 57 3 57.3 37.3 39.7 , 39.7 39.7 397 0.72 2.18 2.18 2.10 21.1 21.6 24.8 31.0 39.0 39.3 32.9 26.7 32.0 31.2 34.6 35.9 5073 5074 8/3/01 9:00:30 8/3/01 10:00:30 1976 1 2568.0 1,933 4.551 8,584,036 8,586,606 '. ~. . . -5715 8/30/01 3:00:30 ~~Z~ 2049.5 4337 2 M351n 1,338,589 C003 9,920,644 70.9 184.0 12.3 7.4 55.7 5.55 33.2 25.0 384 5716 5717 5718 8/30/01 4:00:30 8O0/01 5:00:30 8/30/01 6:00:30 j - 1042.4 34844 M351n 1,339,631 C003 coos coos 9,921,686 n n o . , n T 62.0 172.0 B1.0 0.0 9.8 ' 3.7 0.0 CS 9.B 7.4 60.7 63.3 69.5 27.3 22.7 21.3 1.94 0.40 0.40 27.6 19.6 18 7 31.3 33.4 34.3 29.2 h f -111.1 0.1 1027.2 B.4 M351n MSS1n 1,339,742 1,339,742 9,921,797 9,921,796 48.1 0.0 24.5 24.7 5719 8/30/01 7:00:30 ~~z 0.0 0.0 M351n 1,339,742 coos 9,921,798 L O.D 0.0 0.0 7.4 69.5 21.3 0.39 18.7 34.3 247 5720 5721 8/30/01 0:00:30 \~Z O . D " ' 0.0 M351n 1,339,742 coos 9,921,798 0.0 0.0 0.0 7.4 69.5 21.3 0.29 18.7 34.3 247 8/30/01 9:00:30 0.0 0.0 M351n 1,339,742 coos 9,921,798 0.0 0.0 0.1 7.4 69.5 21.3 0.00 18.7 34.3 24.7 5722 8/30/01 10:00:30 Meat 0.0 0.0 M351n 1,339,742 coos 9,921,798 0.0 0.0 0.0 7.4 69.5 21.3 0.00 18.7 34.3 24.7 5723 8/30/01 11:00:30 End 0.0 0.0 M351n 1,339,742 coos 8,821,788 0.0 0.0 1.B 7.4 69 .5 21.3 0.00 1B.7 34 3 24.7 5724 5725 8/30/01 12:00:30 B/30/01 13:00:30 - 0.0 0.0 0.0 00 - — C003 C003 9,921,798 9,921,798 0.0 0.0 o.o 0.0 0.0 0.0 7.4 7.4 69.5 69.5 21.3 21.3 0.00 0.00 18.7 18.7 34.3 34.3 24.7 247 5726 B/30/01 14:00:30 -- 0.0 B 0 •• - coos 9,921,798 0.0 0.0 0.1 7.4 69.5 21.3 0.00 18.7 34.3 24.7 . •• In the table, each row contains a complete set of information for each hour o f operation as extracted from the Citect system. In addition, relevant data obtained from other sources 72 are as follows: liner installation/removal dates, liner identifications and measurement dates. The throughput column contains average-hourly values of throughput in metric tonnes per hour that were calculated and recorded by the Citect system from values gathered by a weightometer installed at the discharge conveyer belt. This average value represents the total crusher throughput for the one hour period and is used in the calculation of the mantle and the concave cumulative tonnage. The current information available in the Citect system is used as an indirect measure of the power draw assuming the voltage to be constant. Both the average per hour and maximum value per hour are listed in the table. The remaining columns list the bowl level, the size distribution of the feed as percentages of course, fine and medium categories, the mantle vertical position as well as the size distribution of the product in a similar manner as described for the feed size. A graphical representation of the production data was chosen to investigate correlations between the variables. The period for this analysis was the total time comprising two concave lives for each crusher. Thus, four comprehensive graphs have been used to present this information (the four graphs are given in section 6.3). 73 Ideally, the analyses should include all the parameters listed in the complete spreadsheet, however some of them have been omitted. Feed size distribution was left out of the graph due to knowledge of high likelihood of distortions in its values. Experience shows that the WipFrag system used for feed size distribution classification may give erroneous information during the day because of interference from sun light in the image acquisition system (Reddick, 2002). Bow l level was also excluded since its values vary significantly within a one-hour period and the average value does not represent a meaningful parameter. In addition, it was assumed that because of the automatic control employed on the crusher, this variable has already been optimized, i.e. dump level permitting the feed rate is adjusted in order to maintain the crusher as choked as possible. In the graphical analyses, production rate is investigated by plotting its maximum value per hour only. The rationale behind this decision is that the average value per hour does not properly indicate i f the machine is operating " loose", i.e. high throughput has been the target instead of adjusting C S S below the maximum limit. O f primary interest in the analyses is the determination o f extreme conditions o f product quality such as high quality and poor quality production occasions. Crushing product quality is usually related to the amount of course material and sometimes also related to the amount o f fines therefore these two percentage values are incorporated in the analyses leaving the percentage of medium size out (Dance, 2001). 74 To better investigate using product quality, the fine and coarse size distributions are plotted underneath each other in the same graph. In addition, the Y-axis direction o f coarse distribution is inverted so that when both the plot lines move upward or downward they indicate good and poor quality product trends respectively. A n example is shown in Figure 5-13. Figure 5-13 - Example of a graph with product quality plots In order to accommodate all the different parameters plotted together in a comprehensive graph for long periods (approximately 6000 hours per graph) SigmaPlot technical graphing software from SSPS® Inc., U S was used. Each parameter is plotted using an individual Y-axis and a single-common X-ax is corresponding to the time. 75 Figure 5-14 shows part of the graph for crusher #5-concave C005 and serves to exemplify how liner information is included in the graphs. Each period corresponding to one mantle life is represented as rectangle on the graph and contains its sequential number at the top. The cumulative tonnage achieved by each mantle is plotted on the bottom of the graph. The measurement dates correspond to long vertical lines. 76 C R U S H E R #5 - Concave CQ05 OcK)1 Q c W I Qct/01 Nov/01 Nov/D1 Dec/D1 Deo01 Jan j02 Figure 5-14 - Selected liner information for Crusher 5 from October 2001 to January 2002 The vertical mantle position, or shaft position, can vary between the 0 (bottommost) and 10 inches (topmost). A n example plot of the historical mantle position data is shown in Figure 5-15 (dotted line). In order to mitigate the effect of short term upsets on the 77 analyses (i.e. mantle position variation during idle operation) a weekly average value plot was produced using the same Y-axis and the result can be seen in Figure 5-15 (solid line). 10 Mantle Position r7?01 30/7V01 6/8^01 13/8/01 20/8/01 Figure 5-15 - Example of mantle position plot with average line The average values are calculated per each week by weighting the mantle position values to the hourly production for each mantle. The formula is given below: h '=168 X Ph' * Mh' W... = -h'-x w /?'=168 h'=l where Mw is the weekly average mantle position, Ph. is the hourly average production rate, Mh. is the hourly average mantle position, and, h' is equal to 1 for the first hour with Ph, > 0 after mantle installation or start o f a new week. 78 Weekly weighted averages are similarly produced for some of the other parameters. In addition, the calculation of the weighted average product quality value is enhanced with Excel conditional functions to cancel the average calculation during periods of high percentage of noisy data within the week period. WipFrag noisy data can be identified from situations containing several consecutive repeated values during normal crushing (i.e. consecutive meaningful positive hourly average production rates, greater than 1500 tph). Thus, periods containing problems with the data gathering system are shown in the graph as a missing part of the average line, as can be seen in three different week periods on the graph for Crusher 4 graph, shown in Figure 5-16. This was done to avoid misinterpretations during the analyses. Figure 5-16 — Example of the representation of data problems in product quality (circles indicate areas of noisy data where the weighted average of product quality was not calculated). 79 For current draw (amps) not only the hourly and weekly averages are plotted but also the maximum values per hour are included in the graph. A s shown in Figure 5-17, average amps are plotted in a line format and maximum amps are plotted in a scatter format. The maximum amps scatter plot indicates the occurrence o f high amplitude "spikes" which may confirm overload conditions at the machine i f the average value is also high. Figure 5-17 - Example of current draw plots Occasionally a bias in the motor current reading is observed. This is generally due to a pending motor failure. This phenomenon was recorded for two occasions within the total period analysed, one occasion for Crusher 4 (10/7/2002-1/25/2002) and the other for 80 Crusher 5 (1/5/2001-4/13/2001). When this occurred, the raw current draw values were adjusted to be in accordance to the rest of the period facilitating visual analyses. In order to adjust the values, a comparison between values observed for idle operation of the crusher is performed. The comparison is made between a known period containing normal motor operating characteristics and a period when the motor shows distortions. A calculation of the difference between the average idle values for these two periods is performed and the result is used to adjust the values to be plotted. The value used for the adjustment in the graph for Crusher 4 was -30 amps and for Crusher 5 the adjustment was -60 amps for the mentioned periods. These occurrences are shown on the graphs in lighter tones and one example is given in Figure 5-18. Figure 5-18 - Current draw adjusted plot Once the complete graph containing al l the parameters for the period of a concave life is plotted, a broad picture of the crushing characteristics over time is achieved. Thus, 81 periods containing significant results may be easily identified. The periods of main interest in this analysis are the ones showing overload conditions at the machine, identified by the current draw plots or conditions showing high quality product and mantle's high cumulative tonnage together. Moreover, after identifying these periods, the analysis may be enhanced by gathering information from measurements performed within the chosen periods. A s the dates for the measurements are also included in the graphs they are easily identified. Cross-referencing the two sets of information serves to provide an understanding of the influence of the crusher chamber profile on operational parameters and vice versa. The graphs are also useful when the analysis is performed in the opposite direction, that is, when interesting results obtained from wear measurements (for example when atypical wear o f the mantle is observed) determine regions of interest on the graphs. 5.3.2 Chamber Profile Data A s mentioned in section 5.1, issues related to the prototype L P D were previously addressed. O f major concern were apparent problems in the measurement device, in both the hardware and the software components o f the L P D which were affecting the accuracy of the results. Thus, an investigation was conducted by performing several laboratory tests with the equipment as wel l as carrying out field checks in order to compare L P D results and actual dimensions of the crusher. 82 The investigation resulted in the detection of some key issues that were responsible for a degradation in the accuracy of the results, as listed below: a misalignment between the mirror base and mirror vernier causing an error in the calculation of the inclination angle of the track; - the occasional appearance of noisy data in measurement results; - a bug in the software resulting in a systematic increase of the length of the profile drawings; inconsistencies between the positioning of the profile drawings (mantle and concave" profiles) and the actual original positioning of the liners in relation to each other and to the other parts of the crusher; and, systematic deviations between distances measured using the laser and real distances, as shown in the graph in Figure 5-19. 83 I Nominal Distance (mm) ^ ^ ^ ^ Figure 5-19 Deviations in laser measurements More details concerning the investigations of the L P D accuracy as well as information regarding the tests conducted can be found in reports by the author (Rosario, 2001), (Rosario, 2002a) and (Rosario, 2002b). Fol lowing this investigation, it was decided to write a new program to replace the major functions o f the original software provided with the equipment. This new program utilizes the raw data generated by the equipment, i.e. the resulting spreadsheets from the two calibration tests, the mantle test, and the concave test. This new program was also used to correct past measurements. A description o f program functions as well as the rationale applied in the calculations is available in Appendix B. 84 The new program satisfies several objectives such as improving accuracy and providing more comprehensive results. The new program provides several new features, such as the calculation of CSS by mantle position (more details about the enhancement of the measurement procedure is given in section 6.1). In addition, as this program incorporates the data from the original liners (drawing information previously transformed in a numeric format), wear calculations such as wear areas and wear rates for selected regions are made available. Furthermore, a tool to simulate the replacement of different liner profiles is also available (more details about simulations capabilities of the program are given in Appendix B). As discussed in the previous section, the achievement of meaningful, or "standardized", information from the measurements was of great interest in this analysis to complement the graphical analyses of crushing operational parameters. This new program utilises a rationale of "dividing" the crushing chamber in small slices to perform calculations resulting in new and enhanced information. This enhanced set of results facilitates the correlation between the measurements and operational data. The slicing technique that is applied in the program is shown in Figure 5-20. In this figure, two potential crushing chambers are given to illustrate a possible difference in the bottom region of the crusher. 85 86 Although the entire profile is obtained by the measurement process 4, only the bottom region of the concave (corresponding to the two bottommost concave rows) is analysed by dividing this region into eighty 25.4 mm-high slices (1.0 in.). The slices are used in the calculation of: - Chamber volumes by slices (in litres): a. with the mantle position (MP) equals to 0 mm (at the bottom); and, b. with M P equals to the average position for the week ( A M P W ) . - Radial distances between mantle and concave per slice (in millimetres): a. with M P equals to 0 and the mantle at 0 degrees of throw (in the middle); b. with M P equals to 0 and the mantle at its maximum displacement of throw; c. with M P equals to 0 and the mantle at its minimum displacement of throw. d. with M P equals to A M P W and the mantle at 0 degrees of throw (in the middle); e. with M P equals to A M P W and the mantle at its maximum displacement o f throw; and, f. with M P equals to A M P W and the mantle at its mimmum displacement o f throw. Min imum distance among the 80 slices for result " e " , i.e. C S S . 4 Due to the lack of reference targets inside the crusher chamber, the top region of the profiles (where minimum wear occurs) is used to orientate the positioning of the measured profiles in reference to previous measurements and to the original drawings of the parts. 87 Min imum distance among the 80 slices for result "f ', i.e. opened side setting (OSS). (A l l the results above are also generated using the profile o f a new mantle). - C S S results corresponding to the entire range of M P s , from 0 to 254 mm (10 in.) by 12.7 mm (0.5 in.) steps. - Concave-wear by slice, i.e. the radial difference in mm between the new profile and the measured profile o f the concave. - Mantle radial-wear by slice (mm), i.e. the radial difference in mm between the new profile and the measured profile o f the mantle. Concave-wear rate by slice (mm per mi l l ion tonnes of throughput). Mantle-wear rate by slice (mm per mi l l ion tonnes of throughput). In addition, the program generates three graphs, listed as follows: chamber volume by slices, - C S S results versus M P ranging from 0 to 254 mm (10 in.), and liner-wear rate by slice. Among the information available from the measurements and that provided by the new program, the first two graphs are of great value in complementing the graphical analyses of crushing operation parameters. These two graphs summarize the chamber profile information and describe the impact of M P adjustment on crusher chamber dimensions and the choking condition o f the chamber. More detail about these graphical results is given next. 88 "Choking" condition of the chamber Several authors correlate crushing performance to chamber profile and more specifically to the choking condition of the chamber (more details in section 3.1.4). In order to assess this relationship, the variation of chamber volume with height was included in the program as a graphical result. Similar to the approach discussed in the Literature Review, this type of graph helps in checking the chamber choking condition and in determining a choking point or region. Figure 5-21 gives an example of the graph generated by the program with the information for two hypothetical chambers plotted on it. This graph gives a visualization of the main geometric characteristics of the chambers. The characteristics of the chambers contained in this example and the differences between them can be extracted from the graph and are listed below: in chamber "A" the slice-volumes decrease following two different patterns (a linear and rapidly decrease rate from slice # 80 to slice #35 and a slowly decrease rate from slice # 35 to slice #13); - in chamber "B" there is a single linear pattern for slice-volume decreasing from slice #80 to slice #13; - chamber "A" has a defined choke region at slice #35; - in chamber "B" the choke point is at the very bottom of the concave; - in chamber "A" the volume drops from 40 to 28 litres from slice #35 to slice #13; 89 - in chamber " B " the volume drops from 64 to 28 litres from slice # 35 to slice #13; - chamber " A " is a non-choking chamber; and, - chamber " B " is a "straight" chamber. Figure 5-21 Chamber volume by height CSS information of the chamber The C S S and OSS data from a given chamber are calculated by the program based on the measurements. To calculate these numbers the program requires the input of a mantle position value (MP) . Thus, to determine C S S and OSS at the "moment" of the measurement the weekly averaged M P for the date o f the measurement is input. In 90 addition, a ratio of the C S S and OSS values for each wear measurement is calculated and used in the analyses. The graph available from the program describes how the C S S changes with vertical position of the mantle. On this graph, the relationship between C S S and a series of simulated mantle positions is given (from 0 to 254 mm (10 in.) with 12.7 mm (0.5 in.) increments). To perform the calculations the program uses the measured profiles for the mantle and the concave. Figure 5-22 shows an example of this type of graph containing the information for two hypothetical chambers (C and D ) . 10 -9 _ 8 aT 1 1 o c =- 6 -c o 2 5 (0 o Q. 4 o c 3 2 2 1 n —B— Chamber C —•— Chamber D 1 1 u ( I I I I 1_J w ) 1 2 3 4 5 6 7 8 9 10 1 CSS (inches) Figure 5-22 - CSS versus vertical mantle position 91 A s shown in Figure 5-22, chamber information may be compared by their range of C S S and the relationship between C S S and mantle position. From this graph it is possible to see that: - chamber " C " initial C S S is 140 mm (5.5 in.), result given M P equal to zero; - chamber " C " CSS-range is within 83 and 140 mm (3.25 and 5.5 in.); chamber " C " C S S and M P relationship is linear; - chamber " D " initial C S S is 203 mm (8.0 in.); - chamber " D " CSS-range is within 127 and 203 mm (5.0 and 8.0 in.), - chamber " D " C S S and M P relationship is non-linear; 92 6. Results and Discussion 6.1 Improvement in Measurement Process Measurement Time A s mentioned in section 5.1, the redesigned support structure o f the LPD resulted in a significant reduction in set up time. This gain combined with other developments, such as the reorganizations of the electrical boxes and cables as wel l as improved efficiency developed through experience by the maintenance personnel, reduced the original measurement time approximately 45 % (from app. 2 h and 35 min to app. 1 h and 25 min). Figure 6-1 compares the measurement time for eight tests; four before the use of the new support and four after. 2 2.0 1 2 3 4 5 6 7 8 Tests Figure 6-1 Time spent in measurement tests (tests 5 to 8 were performed using the new support). 93 Measurement Results With the new program and the measurement raw data collected during the period of the analysis, 40 complete sets of measurement results were produced. Table 6-1 and Table 6-2 give a summary of the measurements for Crusher 4 and Crusher 5, respectively. In addition, these tables show other information about the crusher at the dates of attempted measurements: the status of the measurement attempt, the average mantle position during the week of the measurement, the cumulative tonnage of the concaves and mantles as well as the mantle code; indicating which specific mantle was in place at the time of the measurement. 94 Tab le 6-1 C r u s h e r 4 - Measurement information M e a s . Spot A M W P Date Status C o n c a v e T o n . Mant le T o n . 3 10 6.35 12-Jul-01 C o m p l e t e C002 737,689 M302u 737,689 5 4 8.66 26-Jul-01 C o m p l e t e C002 1,598,086 M302u 1,598,086 7 10 2.82 9-Aug-01 C o m p l e t e C002 2,443,596 M503u 845,510 9 4 4.96 23-Aug-01 C o m p l e t e C002 3,296,167 M503u 1,698,081 11 10 7.33 7-Sep-01 C o m p l e t e C002 4,160,693 M503u 2,562,607 13 4 4.21 20-Sep-01 C o m p l e t e C002 4,899,695 M104n 739,002 15 - n/a 4-Oct-01 N o M e a s C002 n/a M104n n/a 17 10 7.09 18-Oct-01 Partial C002 6,139,209 M104n 1,978,516 19 4 5.21 1-Nov-01 C o m p l e t e C002 6,782,754 M505n 643,544 21 - n/a 15-Nov-OI N o M e a s C002 n/a M505n n/a 23 4 7.51 29-Nov-OI C o m p l e t e C002 8,025,255 M106u 568,610 2 5 10 5.08 13-Dec-01 C o m p l e t e C002 8,600,272 M507n 575,018 27 - n/a 27-Dec-01 N o M e a s C002 n/a M507n 1,115,939 29 4 2.22 10-Jan-02 C o m p l e t e C002 9,817,709 M408n 676,513 31 - n/a 24-Jan-02 N o M e a s C002 n/a M408n 1,387,899 33 10 5.56 7-Feb-02 C o m p l e t e C002 11,168,894 M408n 2,027,699 35 4 7.86 21-Feb-02 C o m p l e t e C004 517,944 M509U 517,944 37 10 2.53 7-Mar-02 C o m p l e t e C004 1,075,752 M410n 557,807 39 4 1.67 21-Mar-02 C o m p l e t e C004 1,809,700 M410n 1,291,755 41 - n/a 5-Apr-02 N o M e a s COM n/a M410n 2,152,528 4 3 10 4.65 18-Apr-02 C o m p l e t e C004 3,332,757 M410n 2,814,813 45 - n/a 2-May-02 N o M e a s C004 n/a M410n 3,531,753 47 4 4.07 16-May-02 C o m p l e t e C004 4,788,456 M311n 738,759 50 4/10 5.91 30-May-02 C o m p l e t e C004 5,553,485 M311n 1,503,788 53 4 8.12 21-Jun-02 C o m p l e t e C004 6,783,428 M311n 2,733,731 5 6 10 5.37 11-Jul-02 C o m p l e t e C004 7,572,752 M112n 789,324 58 4 5.59 25-Jul-02 C o m p l e t e C004 8,132,726 M112n 1,349,298 62 10 3.57 22-Aug-02 C o m p l e t e C004 9,412,297 M213n 1,279,570 65 - n/a 12-Sep-02 N o M e a s C004 n/a M414n n/a 67 4 4.40 25-Sep-02 C o m p l e t e C004 10,793,188 M414n 862,988 95 Table 6-2 Crusher 5 - Measurement information M e a s . Spot A M W P Date Status C o n c a v e T o n . Mant le T o n . 2 10 8.95 5-Jul-01 C o m p l e t e C 0 0 3 7,138,954 M 5 5 1 n 686,248 4 10 5.95 19-Jul-01 C o m p l e t e C 0 0 3 7,818,430 M 1 0 1 u 679,476 6 4 8.07 2-Aug-01 C o m p l e t e C 0 0 3 8,582,055 M 1 0 1 u 1,443,101 8 10 3.88 16-Aug-01 C o m p l e t e C 0 0 3 9,290,633 M 3 5 1 n 708,578 10 n/a 4.78 30-Aug-01 Mant le C 0 0 3 9,921,798 M 3 5 1 n 1,339,742 12 4 5.94 14-Sep-01 C o m p l e t e C 0 0 3 10,618,545 M 4 5 2 n 696,747 14 4 8.14 27-Sep-01 C o m p l e t e C 0 0 3 11,234,723 M 4 5 2 n 1,312,926 16 — 2.31 11-Oct-01 N o M e a s . C 0 0 5 n/a M 4 5 2 u n/a 18 - 3.51 25-Oct-01 N o M e a s . C 0 0 5 n/a M 4 5 2 u n/a 20 n/a 6.58 7-Nov-01 Mantle C 0 0 5 1,692,802 M 4 5 2 u 1,692,802 22 - 7.65 22-Nov-01 Partial C 0 0 5 2,501,162 M 4 5 2 u 2,501,162 24 4 5.96 6-Dec-01 C o m p l e t e C 0 0 5 3,167,164 M 2 5 3 u 666,002 26 10 5.38 19-Dec-01 C o m p l e t e C 0 0 5 3,759,065 M 2 5 3 u 1,257,902 28 10 6.19 3-Jan-02 C o m p l e t e C 0 0 5 4,714,625 M 1 0 6 u 955,560 30 4 5.05 17-Jan-02 C o m p l e t e C 0 0 5 5,431,364 M 5 0 7 u 716,739 32 10 6.69 31-Jan-02 C o m p l e t e C 0 0 5 6,158,091 M 5 0 7 u 1,443,466 34 4 2.41 14-Feb-02 C o m p l e t e C 0 0 5 6,881,548 M 2 5 4 n 723,457 36 10 4.27 28-Feb-02 C o m p l e t e C 0 0 5 7,850,215 M 2 5 4 n 1,692,124 38 4 3.86 14-Mar-02 C o m p l e t e C 0 0 5 8,493,850 M 2 5 4 n 2,335,759 40 10 3.20 28-Mar-02 C o m p l e t e C 0 0 5 8,940,557 M254u* 423,797 42 4 7.86 11-Apr-02 C o m p l e t e C 0 0 5 9,717,780 M 5 5 6 n 777,223 44 - 1.41 25-Apr-02 N o M e a s C 0 0 5 n/a M155u* n/a 46 10 8.57 9-May-02 C o m p l e t e C 0 0 5 10,953,920 M 2 5 7 n 595,668 4 9 4 6.56 29-May-02 C o m p l e t e C 0 0 5 11,790,386 M 1 5 8 n 836,466 51 - 7.68 6-Jun-02 N o M e a s C 0 0 5 n/a M 1 5 8 n n/a For each measurement used in the analysis, the complete set of results provided by the program were grouped in an Excel file (details in section 5.3.2) and the generated mantle and concave profile drawings were plotted together in an AutoCAD drawing containing their original profiles. These drawings were used for a visual evaluation of the wear (Appendix C shows an example of one measurement drawing result). 96 In addition to these drawings, three graphs were generated for each measurement. The first graph gives the chamber volumes by slices for two conditions: the chamber formed with mantle and concave as measured and the chamber formed i f a new mantle was installed (Figure 6-2 shows the result for measurement 7). The second graph plots the relationship between C S S and a series of simulated mantle positions (Figure 6-3 shows the result for measurement 7). The third graph plots mantle wear rate versus slice numbers and concave wear rate versus slice numbers together (Figure 6-4 shows the result for measurement 7). £ ja E 3 z © o CO 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 20 40 — e — C h a m b e r a s m e a s u r e d —•— C h a m b e r with a new mant le 60 80 100 Sl ice Vo lumes (litres) 120 140 F igure 6-2 C h a m b e r Vo lumes graph (using M P equals to A M P W ) for measurement 7. 97 C S S (inches) Figure 6-3 CSS versus MP graph for measurement 7. Figure 6-4 Concave and mantle wear rate by slices for measurement 7. 6.2 Wear Determination Concave-wear As described in section 5.3.2, the new program enabled the inclusion of new features related to wear. Moreover, the analysis of the concave-wear rate by slices resulted in the determination of an average wear rate per slice for the bottom of the concave for the two different measured regions, i.e. the wear occurring at the 4 o'clock and the 10 o'clock position. Figure 6-5 shows the average concave-wear for the bottom of the concave. To achieve the results plotted in Figure 6-5, some measurement information was excluded in the calculation due to issues with the data collected. Measurement results containing noisy data and measurement results for short concave-lives were excluded to enable a more accurate wear rate result. This was necessary due to the fact that the measurement accuracy of +/- 3 mm limits the wear rate calculation accuracy for the initial period of concave life. For example, consider the two cases that follow: - in measurement 3 in the early concave life ( 737,689 of concave tonnage), the radii difference in slice 40 of 5.85 mm +/- 3mm results in a wear rate between 3.9 and 11.9 mm/Mton (approximate accuracy +/- 4) - in measurement 62 in the end of concave life (9,412,297 of concave tonnage), the radii difference in slice 40 of 39.68 mm +/- 3mm results in a wear rate between 3.9 and 4.5 mm/Mton (approximate accuracy +/- 0.3). 99 The criteria used to reject measurements for the wear rate calculation was to not use measurements that give accuracy ranges greater than 0.75 mm/Mton. F igure 6-5 Average concave-wear rate 100 F o l l o w i n g the determinat ion o f the average wear-rates, it was possible to estimate the amount o f wear o c c u r r i n g at the b o t t o m o f the concave for any g i v e n tonnage. F i g u r e 6-6 gives the result for a s i m u l a t i o n u s i n g 8 mega-tonnes o f throughput for the t w o measured spots. Figure 6-6 Simulated concave wear after 8 megatonnes. T h i s result correlates w e l l w i t h pract ica l observations at H V C . C o m m o n l y , the highest wear rate is observed close to the seam between the t w o bottommost r o w s o f the concave parts at the 4 o ' c l o c k p o s i t i o n . In addi t ion, o n several occasions w h e n concaves were 101 used for long periods, this highly worn region resulted in broken edges on some concave parts. Mantle-wear Although drawing results served to better visualize the effects of mantle-wear on chamber profile, a similar wear-rate average calculation for the mantles did not provide a meaningful result. In contrast to the concave results, the wear-rate o f any specific slice on the mantle shows greater variation for different measurements. This variation may be explained by the fol lowing facts: - different mantle profile types were used during the concave life; - two different mantle materials were used during the analysis period; - the mantle movement in relation to the concave position; - the impact o f crushing operational conditions on mantle-wear is greater due to a lower wear resistance o f the mantle when compared to the concave. 102 6.3 Correlation of Operational Data and Liner Characteristics A s discussed in section 5.3.1, four graphs were used to analyse crushing parameters for both crusher #4 and #5. Each of these graphs corresponds to a period of one concave life and are shown in Figure 6-7 to Figure 6-10, being: - Figure 6-7 Graph # 1 - Concave C002 at crusher #4 (29 June 2001 -7 Feb. 2002); - Figure 6-8 Graph #2 - Concave C004 at crusher #4 (9 Feb. 2002-25 Sep, 2002); - Figure 6-9 Graph #3 - Concave C003 at crusher #5 (5 Jan. 2001-27 Sep, 2001); - Figure 6-10 Graph #4 - Concave C005 at crusher #5 (30 Sep. 2001 -7 June 2002). 103 OTQ S n ON o O ~i to •a it *. I O o a n S3 < n O o o o 1/3 M O O t-i CO O ^1 The graphical analysis allowed the observation of a series of interesting relationships between the parameters plotted and resulted in the identification of key periods for cross reference with the information provided by the wear measurements. The most relevant observations and significant relationships are discussed next. 6.3.1 Issues during the end of concave life As shown in Figure 6-7 to Figure 6-10, on the majority of the occasions that overload conditions appeared, i.e. unstable current draw as well as high amplitude maximum current values (power spikes), the concaves had been running for more than 7 megatonnes on average. The only exception can be visualized in Figure 6-9 for concave G005 during the first 4 megatonnes of operation. However, in this instance the overload condition indicated by the current draw was related to problems with the motor, as explained in section 5.3.1. Although several different types of mantles, containing different profiles, were used during the periods of concave life after 7 megatonnes, none of them achieved a substantial cumulative production. Actually, most of the lowest tonnage per mantle results appeared in these periods. The characteristics related to these issues in the final period of concave life were investigated and the results are discussed next. 108 Non-choking condition of the chambers Assessing the chamber volume characteristics using the graphical method described in section 5.3.2 provided a comparison between the chamber geometry characteristics achieved during different periods. These periods are: the periods when the issues occurred ("bad operation") and other selected periods when not only the current draw remained stable but also product quality and mantle tonnage were the best ("good operation"). From the assessment, it was clear that the chambers that provided the best results were closer to a non-choking type and the ones providing bad results were closer to a straight type (as suggested in section 3.1.4). Figure 6-11 shows a chamber volume graph with the results of three measurements corresponding to periods o f "good operation". In addition, Figure 6-12 shows a chamber volume graph with the results o f three measurements corresponding to periods of "bad operation" (more detail about the determination o f the "non-choking" condition o f chambers is given in section 5.3.2). 109 Increased volume of the chambers It is a common procedure in the mines to increase the size o f the mantles to compensate for worn concaves. The impetus for this is to try to maintain the original C S S . This practice of trying to maintain the C S S dimension by the successive installation of larger diameter mantles affects chamber volumes. The volume modifications may also contribute to rising power requirements. Figure 6-13 serves to illustrate this procedure. In the figure, two hypothetical chambers with the same C S S (152 mm - 6 in.) are shown in section view. Although the mantle on the right is 254 mm (10 in.) larger in radius (idealistic extreme case), both chambers show an equal chamber in cross-section. However, with the use of a three-dimensional representation o f these chambers, as shown in Figure 6-14, the difference between the chamber volumes becomes apparent. | - R40 in |-R50 in. | - R46in. (-R56in. Figure 6-13 Cross-section view of two similar but not identical chambers 111 Y _ Z Figure 6-14 Comparison between chambers with different radius The example chambers in Figure 6-13 and Figure 6-14 do not exactly match the dimensions of the liners used at H V C . However, a rough estimation of two chamber volumes normally utilised at H V C providing similar C S S confirms the difference. One chamber consisting of: a new standard size mantle and a new set of concaves has a volume of approx. 19.57 m 3 (1,194,200 in. 3), and a chamber comprised o f worn concaves and a new over size mantle has a volume of approx. 20.35 m 3 (1,242,100 in. 3); an increase of 4%. Since the lifetime of a concave is much longer than that o f a mantle, the practice o f using several mantles with one concave is worthwhile. However, as the impact o f the chamber volume change is often overlooked, situations can arise where uneconomical mantle concave combinations are used. 112 Increased CSS area Similar to the increase in chamber volume, the discharge area increases by approximately 10% during the final stage of the concave life. Thus, although the C S S can be maintained at a desired value, the amount o f material being discharged increases and the size distribution becomes courser as the area increases. Figure 6-15 shows the difference between the C S S area achieved with the same gap of 122 mm (4.8 in.) for two extremely different, but feasible, situations. The first area, A l , was calculated for a new concave and a new standard-size Esco 3-piece mantle, while the second area A 2 was calculated for a concave as it was measured after approximately 87% of its useful life (measurement # 62 on 22 August 2002 at crusher # 4) and a new over-size 90" Esco-2 piece mantle. Using automatic control on the crusher, product size distribution (percentage of product course) and crusher power (amps value) are part o f the input parameters used to adjust the mantle position and the feed rate. Thus, an increased discharge area may be an additional factor in explaining the occurrences of current "spikes" and intermittent overload conditions o f the machine during the final life o f the concave. In other words, as the product becomes coarser, the control algorithm tends to raise the mantle closer to its maximum limit. Alternatively, as the current draw increases, the control tends to lower the mantle. These contrary trends may result in a greater instability when using the automatic control during the final life o f the concave. 113 —123mm [4.8"] Figure 6-15 - The difference between the discharging areas resulted from the wear of the concaves. 6.3.2 Mantle and crushing performance The analysis of the mantle position adjustment (described in section 5.3.2) revealed the correlation between mantles which lasted for short periods of time and their limited room for adjustment, i.e. when a mantle needs to be set at a high position just after its installation, obviously its life w i l l be short. A n example o f this problem occurred when the mantle M257n was installed in Crusher 5 with concave C005 and needed to be immediately set at the 203-mm (8-in.) position in order to achieve the desired C S S (refer 114 to the 10 mantle in Figure 6-10). Only 595,668 tonnes were produced with this mantle before it had to be replaced. Similar cases occurred with all four concaves analysed and served to identify two facts. First, poor knowledge of concave wear at the time of selection and installation o f the mantle generates short mantle lives. Second, although several different types of mantles were used (8 types) during the period of the analyses, frequently the mantles that were installed could not be used for their full range of mantle positioning, i.e. an initial adjustment close to 0-mm position and a final adjustment close to the 254-mm (10-in.) position. These short life mantle occurrences were mostly observed during two specific phases of the concave life: the initial life period of the concave (0 to app. 2 mega-tonnes) and the second half of its useful life (more than 5.5 mega-tonnes). Based on the fact that all the available under-size and over-sized profiles were used during these two specific periods of concave life, this observation suggests that those profiles are inadequate to provide long mantle life. 6.3.3 Liners that provided optimum performance In general, product quality varied greatly during the period of the analysis. Only a few occasions were observed where reasonable product quality, smooth operation, and normal product rate occurred simultaneously for considerable time periods. These rare events happened for Crusher 4 concave C004 with mantles M410n and M31 In (refer to the 2 n d and 3 r d mantle in Figure 6-8), and Crusher 5 concave C003 with mantle M182u (refer to 115 the 3r mantle in Figure 6-9). Each of these mantles were 3-piece standard 2216 mm (87.25 in.) diameter Esco CZ 18 alloy, and the periods of these occurrences were approximately within 2.0 and 5.5 mega-tonnes of concave life. In addition to performing well, the three mantles each achieved a considerable cumulative tonnage of app. 3.1 megatonnes on average. From this, two things were observed. First, the best matches between mantle and concave occurred between 2.0 and 5.5 mega-tonnes of concave life. Second, the Esco 3-piece standard type mantle has the most suitable profile among the mantles used during this period of concave life. These facts supported an investigation of the characteristics of these chambers and to designate them as "good chamber" characteristics to be targeted during the entire concave life. The results from this investigation are discussed in the following section. 116 6.4 Liner Management An interesting result of this analysis was the observation of a large variation in the number of mantles utilized on each concave as well as the large variation in the tonnage achieved by these mantles as summarized in Figure 6-16. Mantle Production Cr 5 & Cr 4 Mantles • Cr5-6Jan01J27Sep01 • Cr5-30Sep01/7Jun02 • Cr4-29Jun01S7Fev02 • Cr4-9Fev02-2SSep02 Figure 6-16 Number of mantles used per concave and their total tonnage 117 These variations in mantle use plus the issues related to the final life period of the concaves, discussed in section 6.3.1, suggested that the liner replacement policies at HVC should be changed in order to improve crushing performance. The first modification to the usual replacement schedule is the reduction of concave life to app. 7.5 megatonnes. Second, as it is apparent from the analyses, only one of the current mantle designs fulfills the requirements necessary for effective crushing. Therefore, the 3-pieces standard mantle design is the only design considered useable for a new liner replacement policy. In addition, two new mantle types, under-size and over size profiles, should be designed for use during start-up and final concave operational conditions. In the design process, the non-choking characteristic of the chamber and the room for adjustment of the mantle position are two essential design characteristics that must be optimized. A "good" non-choking condition and a CSS-MP relationship that provides a long life for the. mantle are the design targets. Considering that under optimal circumstances mantle lives exceeded 3 million tonnes, and following maintenance department recommendations to maintain the replacement schedule following the current three-week pattern, the expected useful lives for the mantles are designed as follows: - 1st mantle - 9 weeks (app. 2,900,000 tons) - 2n d mantle - 9 weeks (app. 2,900,000 tons) 118 - 3 r d mantle - 6 weeks (app. 1,950,000 tons) Note: Only three mantles are used per concave and the maximum concave tonnage is close to the 7.5 megatonnes targeted. In the design of the new over-size mantle profile, the concave wear rate by slice (described in section 6.2) has been used to estimate the wear o f the concave after 5.8 megatonnes. A new concave profile was used in the design process o f the new under-size mantle. After the design of the new profiles, their chamber conditions and C S S - M P relationships were assessed using the graphical analyses discussed in section 5.3.2. Figure 6-17 compares the expected chamber condition of a new under-size mantle and the best option of mantle profile available (Esco 2 pieces 2184 mm - 86 in.). A s well , the C S S - M P relationships for these two mantles are compared in Figure 6-18. 119 20 40 60 80 100 120 140 Volumes per slice (litres) Figure 6-17 Comparison of chamber condition for under size mantles Figure 6-18 Comparison of CSS-MP relationship for under size mantles 120 Figure 6-19 and Figure 6-20 show a comparison of the chamber conditions and CSS-MP relationship for two over-size mantles, the new design and the best option available (recent Esco 2 pieces 2286 mm - 90 in.). 10 1 , , , , , 1 20 40 60 80 100 120 140 Volumes per slice (litres) Figure 6-19 Comparison of chamber condition for over size mantles. 121 CSS (inches) Figure 6-20 Comparison of CSS-MP relationship for over size mantles. The new designs fol low the 3 pieces arrangement to facilitate the reuse o f the top piece and the middle piece. Two new bottom pieces (one under-size and one over-size) and one new medium piece (over-size) were designed (drawings of the new parts are shown in Appendix D). With the introduction of two new mantle designs and the reduction o f the concave useful life, the use o f three mantles per concave is recommended in a proposed schedule for liner replacement. Figure 6-21 shows how the pieces are used over the life o f the concave. 122 F i r s t s e t ( u s e d 2 X ) ( u s e d I X ) N e w U / S B o t t o m p e e ( u s e d I X ) S e c o n d set | | J | | E s c o 3 p c s t a n d . T h i r d s e t N e w D / S M e d , p e e . N e w D / S B o t t o m p e e Figure 6-21 Suggested replacement policy The application of the new designed mantles and the suggested management policy for the replacement of the liners w i l l not only result in better operational conditions and increased average product quality but also in reduced costs. A cost analysis was performed to compare the current liner costs and the projected costs for the suggested procedure. First, based on the liner information (section 5.2.2) and the historical data for the last six concaves used in Crushers 4 and 5 (summarized in Table 6-3) the current average cost of liner replacement per tonne (in Canadian dollars) was calculated as shown in Table 6-4. In addition, the average downtime was calculated, based on the same period, as shown in Table 6-5. 123 Table 6-3 Summary of liners information for 6 recent concave life periods C o n c a v e Install. R e m o v a l Total C u m . Mant les C o d e Date Date T i m e (h) T o n n a g e Utilised C 0 0 3 1/5/01 9/27/01 6357 11,234,723 8 C 0 0 5 9/30/01 6/7/02 5999 12,172,495 11 C 0 0 2 6/29/01 2/7/02 5355 11,168,894 7 C 0 0 4 2/9/02 9/25/02 5462 10,793,188 6 C 0 0 7 6/8/02 1/22/03 5462 9,733,099 9 C 0 0 6 9/26/02 3/26/03 4334 8,300,496 6 Total 32969 63 ,402,897 47 A v e r a g e Ton/h = 1,923 Table 6-4 Liner costs (parts, rebuilt and installation) and current total cost per ton Liner cos ts Quant . Unit C o s t Tota l C o s t Used mant le 5 17 $1,850 $31,450 New mantle 30 $41,900 $1 ,257,000 Concave 6 $102 ,930 $617 ,580 Total $1 ,906,030 Total tonnage 63 ,402,897 C o s t per ton $0.0301 5 Occasionally in the mine, a mantle that was used in one crusher is utilized in another. In such cases, the mantle-mainshaft assembly is just reinstalled. Thus, part costs and labour/supplies-assemblage costs are not incurred in these procedures. 124 Table 6-5 Specific downtime per liner and current total liner downtime Liner Quant. Unit Total Replaced Downtime (h) Downtime (h) Mantle 41 12 492 Mantle and concave, 6 72 432 Total downtime 924 Total tonnage 63,402,897 Average downtime (hour/Mton) 14.57 Second, assuming that the cost o f the n e w designed mantle parts w o u l d be s i m i l a r to the correspondent standard-size parts, and f o l l o w i n g the same rationale used for the ca lcu la t ion o f the current cost and d o w n t i m e , the projected cost a n d d o w n t i m e were then calculated as s h o w n i n T a b l e 6-6 and T a b l e 6-7, respectively. Table 6-6 Liner costs (parts, rebuilt and installation) for one concave life and projected cost per ton Liners details Quant. Unit Cost Total Cost Mantle Top piece liner 1 $7,650 $7,650 Med. piece liner 2 $10,700 $21,400 Bottom piece liner 3 $12,000 $36,000 Supplies 3 $4,000 $12,000 Labour 3 $5,500 $16,500 Installation 3 $1,850 $5,550 Mantles (total) $99,100 Concave Concave (total) 1 $102,930 $102,930 Liners total cost $202,030 Total tonnage (24 weeks) 7,754,100 Cost per ton $0.0261 1 2 5 Table 6-7 Specific downtime per liner and projected total liner downtime Liner Quant . Unit Tota l R e p l a c e d Downt ime (h) Downt ime (h) Mantle 2 12 24 Mant le a n d c o n c a v e 1 72 72 Total downt ime 96 Total tonnage 7,754,100 A v e r a g e downt ime (hour/Mton) 12.38 A comparison between these two calculated costs indicates that the suggested policy enables a reduction o f app. 13% (from 0.0301 to 0.0261 Canadian dollars) over the total annual liner cost (parts, rebuilds and installations) and the cost associated to the amount o f downtime involved in the replacements is also expected to drop as the average annual downtime would decline by 15% (from 14.57 h/Mton to 12.38 h/Mton). 126 7. Conclusions 7.1 Achievements This thesis has presented the development of a novel approach to assessing wear in gyratory crushers. Through the use of this approach greater understanding into the relationship between, crusher wear, crushing chamber geometry and production capacity and quality has been gained. In addition, the research has had tangible results of direct benefit to the supporting organization. More specific outcomes of the research are as follows: The use o f a new laser profiler device (LPD) to measure the crusher chamber was implemented at H V C and improvements were made to the original measurement procedure. The redesign of L P D ' s support structure resulted in a 45% reduction in the necessary time for measurements. The investigation o f the initial issues presented in the application o f the new device resulted in the correction o f several problems and thus, in the improvement of the accuracy of the generated liner profiles. - Chamber data provided from the measurements and crushing operational data from mine information systems were collected for the two crushers. The data comprised of 40 complete sets of measurements and information of approximately 23,300 hours o f records for crushing operational parameters. Calculations and graphical analysis 127 techniques were developed and integrated in a new software tool to facilitate data analysis. The data analyses conducted provided an understanding of crushing chamber characteristics and their impact in crushing performance. In addition, wear rate as a function of production was determined for the concaves, which enabled wear prediction for the bottom part of the concave. The knowledge gained by the analyses helped in the evaluation of the current liner management policy. The evaluation resulted in a revised maintenance schedule based on the use of two new mantle profiles designed for this application. The proposed liner management policy is expected to reduce overall liner costs as well as to enhance crushing performance. Thus, a real opportunity to increase the profitability of the operations was gained with this work. 128 7.2 Future Work Opportunities Continue with the laser measurement and the improvement of the process. The continued collection of additional measurement data can facilitate the assessment o f mantle wear and enables the determination of mantle wear rates as functions of: chamber characteristic concave life period - mantle position variation - mantle material Expand the work done by combining concave and mantle wear prediction and, therefore develop a C S S prediction tool to be used online with the operations. Re-evaluate L P D performance and application in order to assess the necessity o f building an alternative device based on laser technology or the application o f a new technology such as sacrificial sensors as has been the focus of research for cone crushers. 129 8. Reference List Adams, A . , Personal Communication, Highland Val ley Copper, 2003. A S T M G65-94, "Standard Test Method for Measuring Abrasion Using the Dry Sand Rubber Wheel Apparatus", Annual Book of A S T M Standard, V o l . 03.02, Amer.Soc. Testing Mater., Philadelphia, P A , pp. 232-243, 1996. Bearman, R.A. and Briggs, C .A . , "The Active Use of Crushers to Control Product Requirements", Minerals Engineering, V o l . 11, No . 9, pp. 849-859, 1998. Blickensderfer, R., Tylczak, J .H . , and Madsen, B.W., "Laboratory Wear Testing Capabilities of the Bureau of Mines" , IC 9001, pp. 36, 1985 Burkhardt, E.S., "Primary Crushers Factors that Affect Capacity", Design and Installation of Comminution Circuits, eds. Mular, A . L . and Jergesen, G .V . II, Society o f Min ing Engineering, pp. 387-392,1982. Dance, A . , "Crusher Control Strategy", H V C ' s Crusher Control Help Fi le, H V C , 2000. Dance, A . , "The Importance of Primary Crushing in M i l l Feed Size Optimisation", S A G Conference University of Brit ish Columbia, Vancouver, B C , 2001. 130 Delalande, G . , "Concasseurs Giratoires: Determination de la Duree d'Uti l isation Optimale des Pieces d'Usure", Industrie Minerale, Mines et Carrieres, V o l . 68 May, pp. 325-326,1986. ' ' • Delalande, G . , "Concasseurs Giratoires: Methodes de Releve, en Place, du Profi l des Chambers", Industrie Minerale, Mines et Carrieres, V o l . 68 June, pp. 371-374, 1986. Diesburg, D.E. and Borik, F., "Optimizing Abrasion Resistance and Toughness in Steels and Irons for the Min ing Industry", Cl imax Molybdenum Company of Michigan a Subsidiary o f A M A X , Material for the Min ing Industry Symposium, pp. 15-41, 1974. ESCO®, "A l loys for Crusher Wearparts", v ia the Internet 10 June 2003, Flavel, M.D. , "Selection and Sizing of Crushers", Design and Installation of Comminution Circuits, eds. Mular, A . L . and Jergesen, G .V . II, Society of M in ing Engineering, pp. 343-385, 1982. Flavel, M.D. , "Relevance of Efficient Crushing in Comminution Systems", preprint Society o f Min ing Engineers o f A I M E , Littleton, C O , 1988. Gauldie, K., "The Output of Gyratory Crushers", Engineering, V o l . 178, pp. 557-559, 1954. 131 Gaudin, A . M . , "Crushers", Ch . II in Principles of Mineral Dressing, McGraw-H i l l Book Company, Inc., pp. 25-50,1939. H V C , "Highland Val ley Copper Mi l l i ng Operations 2000", Highland Val ley Copper Report, 2000. MacPhai l , A . D . , "M in ing at Highland Val ley Copper", C I M Bullet in, V o l . 962 July/August, pp. 73-77, 1992. Major, K., "Types and Characteristics of Crushing Equipment and Circuit Flowsheets", Mineral Processing Plant Design, Practice, and Control Proceedings, Vancouver, B C , eds. Mular, A . L . , Halbe, D .N . and Barratt, D. J . , V o l 1, Society o f Min ing Engineering, pp.566-583, 2002. Moshgbar, M . , Parkin, R . M . and Bearman, R.A. , "The Compensation of Liner Wear for Optimum Control of Cone Crushers", Progress in Mineral Processing Technology, eds. Demirel, H . and Ersayin, S., pp. 549-555,1994. Moshgbar, M . , Bearman, R.A. and Parkin, R., "Optimum Control o f Cone Crushers Ut i l iz ing an Adaptive Strategy for Wear Compensation", Minerals Engineering, V o l . 8, No . 4/5, pp. 367-376,1995. 132 Nims, L., Personal Communication, Conveyor Dynamics Inc., 2001. O'Bryan, K. and L i m , K., "Selection and Special Considerations for Pebble Crushers" Mineral Processing Plant Design, Practice, and Control Proceedings, Vancouver, B C , eds. Mular, A . L . , Halbe, D .N. and Barratt, D.J., V o l 1, Society o f Min ing Engineering, pp. 628-635, 2002. Parks, J.L. and Kjos, D .M . , "Martensitic Steel Concaves for Large Gyratory Crushers" S M M E (A IME) : 5 2 n d Annual Min ing Symposium, University o f Minnesota, Duluth, M N , pp. 207-218, 1991. Reddick, S., Personal Communication, Highland Val ley Copper, 2002. Richards, D .M . , "Integrating G P S Technology into the Management and Operation of Highland Val ley Copper " , C I M Bulletin, V o l . 93 No . 1040, May , pp. 76-80,2000. Rosario, P.P., "Laser Drawing Procedure" in Highland Val ley Copper Project Monthly Report No . 4, October 2001. Rosario, P.P., "Laser Measurements for Crusher's Liners II" in Highland Val ley Copper Project Monthly Report No . 8, February 2002. 133 Rosario, P.P., " H V C Presentation - December 2002" Highland Val ley Copper Project Extra Report, December 2002. Sheppard, M . and Witherow, C M . , "Primary Crashing", R.I. 3377, February, U.S. Bureau of Mines, pp. 10,1938. Simkus, R. and Dance, A . , "Tracking Hardness and Size: Measuring and Monitoring R O M Ore Properties at Highland Val ley Copper", Mine to M i l l , Brisbane, Australia, October 11 to 14, 1998. Svensson, A . and Steer, J.F., "New Cone Crusher Technology and Developments in Comminution Circuits", Minerals Engineering, V o l . 3, N o . 1/2, pp. 83-103,1990. Svensson, A . Hedvall , P. and Fjaestad, M . , "New Comminution Applications Using Hydrocone Crushers with Setting regulation in Real T ime", International Journal o f Mineral Processing, No . 44/45, pp. 461-469,1996. Taggart, A . F . , "Coarse and Intermediate Crushing", Handbook of Ore Dressing, Sec. 3, pp. 260-281,1927. Teck Cominco, "Highland Val ley Copper Mine, B .C. , Canada" v ia the Internet, 17 June 2003, 134 Tunstall, A . M . and Bearman, R.A. , "Influence of Fragmentation on Crushing Performance", Min ing Engineering A I M E , pp. 65-70, January 1997. Tylczak, J .H , Hawk, J .A. and Wi lson, R.D. , " A Comparison of Laboratory Abrasion and Field Wear Results", Wear, pp. 1059-1069,1999. Utley, R.W., "Selection and Sizing of Primary Crushers", Mineral Processing Plant Design, Practice, and Control Proceedings, Vancouver, B C , eds. Mular, A . L . , Halbe, D .N . and Barratt, D.J., V o l 1, Society o f Min ing Engineering, pp. 584-605,2002. Valery, W., Morrel , S., Kojovic, T., Kanchibotla, S.S., and Thornton, D . M . , "Model l ing and Simulation Techniques Appl ied for Optimisation of Mine to M i l l Operations and Case Studies", V I Southern Hemisphere Meeting on Mineral Technology, C E T E M / M C T , Rio de Janeiro, Brazi l , pp. 107-116, 2001. Westerfeld, S.C. , "Gyratory Crushers", Chapter 3B , S M E Mineral Processing Handbook, ed. Weiss, N.L. , S M E - A I M E , New York, pp. 27-46,1985. Whiten, W.J . , "The Simulation of Crushing Plants with Models Developed Using Mult iple Spline Regression", J . South Afr . Inst. M i n . & Met., No . 72, pp. 257-264, 1972. Wi l ls , B.A. , Mineral Processing Technology: an Introduction to the Practical Aspects o f Ore treatment and Mineral Recovery, eds. Butterworth-Heinemann, Boston, 1997. 135 Wilson, R.D. and Hawk, J.A. , "Impeller Wear Impact-abrasive Wear Test", Wear, V o l . 225-229, pp. 1059-1069,1997. Wolff, E. , Personal Communication, Highland Val ley Copper, 2002. Yaxley, S.J. and Knight, J .A .G . , "The Development and Integration of Nove l Sacrificial Wear Sensors into the Quarrying Industry", Sensors and Actuators, V o l . 75, pp. 24-34, 1999. Zandee, D.W. and Redfearn, M.A . , "Crushing, Screening and Dust Col lect ion", Operation and maintenance in Mineral processing Plants, ed. Claridge, P.G.,Special V o l . 40, Canadian Mineral Processors Div is ion of C I M , pp. 41-72, 1989. 136 Appendix A Drawings of the new support structure. 137 Appendix B Description of the measurement program 139 To improve the accuracy of the measurement results obtained using the L P D and to enhance the scope of these results, such as with the addition o f C S S and wear rate determination, a new program was developed using Microsoft Excel . The major functions and the calculation procedures applied in this program are described in a sequential format as follows. S t e p l Four files generated from the measurement test are loaded in the program. Each file contains a table o f coordinates generated by the L P D (an example is shown in Table 5-1). The files correspond to the following measurements: 1 s t - shooting the calibration bars with the laser beam perpendicular to the track; 2 n d - shooting the bars after levelling the mirror; 3 r d - shooting the concaves of the crusher; and, 4 t h - shooting the mantle after rotating the mirror 90 degrees. Step 2 The program identifies at which actuator position the calibration bar was first targeted during the first and the second calibration measurements. Then, using the trigonometric relations between these two values, the inclination angle o f the track ( a ) is calculated (Figure 5-4 pg. 58 illustrates the process). 140 Step 3 The program corrects the data tables contained in the other two files (mantle and the concave measurements) uti l izing a model developed from results of field-tests which provided evidence of systematic deviations between distances measured and real distances (the graph in Figure 5-19 pg. 84 shows the deviations). Step 4 Each pair of coordinates, from both liners measurements, is transformed into a pair of coordinates in the Cartesian system with the origin (0,0) being the centre of the shaft located at the top part of the laser track ("track" coordinate system), as shown in Figure B - l . The program performs this calculation using the angle a and the trigonometric relationships between the original coordinates and the ones in the "track" coordinate system, as illustrated in Figure B-2. Figure B-l Location of the origin of the "track" coordinate system 141 0,0 Figure B-2 Schematic of the trigonometric relationships between the original coordinate system and the "track" coordinate system Step 5 The two new sets o f points, corresponding to the mantle and the concave profiles, are transferred and plotted into an A u t o C A D drawing as two polylines. A s shown in Figure B-3, the two profiles are plotted together with the drawings of the original liners and the crusher mainshaft. However, the position of the polylines does not match with the crusher drawing. This is due to the different coordinate system of the crusher drawing 142 which has its origin coinciding with the pivot point o f the crusher mainshaft ("crusher" coordinate system). p Fie £ * . ytew Insert Format look Qraw Dimension Modify Imjge Window Help Figure B-3 Snapshot of the AutoCAD drawing with the measured profiles and the original parts Note that the mainshaft/mantle is drawn in its central position, i.e. with the eccentric in the position where the C S S is equal to OSS. Step 6 The two polylines (measured profiles) are moved together to a new position that better represents the chamber as measured. In order to achieve this objective, the top region o f the liners (low wear region) is used as reference in this aligning process. Once the 143 profiles are relocated, the X and Y offset values from the "crusher" coordinate system's origin are determined. These numbers are input in the Excel program to translate all data to the "crusher" coordinate system. Due to the crusher throw, a rotation of the mantle profile in the X , Y plane may be required and this procedure is discussed in step 8. Step 7 To model the original liner profiles in Exce l , equations are fit to the drawing profile data. A s illustrated in Figure B- 4, five equations are usually necessary to describe an original mantle profile. Figure B- 4 Group of lines and arcs that represents a mantle original profile 144 Step 8 In order to determine the required angle of rotation for the mantle measured profile, two methods are used. The first method utilises Excel to calculate the angle that minimizes the distance between the measured and the original profiles at their topmost regions. In some cases this approach in unsuccessful, in which case the angle is determined in A u t o C A D by manually rotating the measured profile. Step 9 Using the information achieved in steps 7 and 8 and trigonometric relationships, the program calculates new values that describe the measured mantle profile rotated by its maximum and minimum angular displacements (+ and - 0.2415 degrees) which correspond to the crusher eccentric throw dimensions. In addition, fol lowing a similar calculation process, the program develops two new sets of equations that describe the new mantle profile at these two extreme positions. Step 10 The program applies the slicing technique for the bottom part o f the crushing chamber, as described in section 5.3.2. For each slice, the program determines the coordinates o f the original profiles and the measured profiles where they intersect a line through the midpoint of the slice. For the mantle, the process is repeated for the fol lowing three cases: - maximum angular displacement, - minimum angular displacement, 145 - zero degrees of displacement. Using these results, calculations such as what follows are performed: - the distance between the mantle, in its maximum angular displacement, and the concave for each slice, - the minimum distance among the 80 slices (CSS), concave and mantle radial loss of material (wear) per slice, and, - the volume o f each slice formed by the mantle measured profile and concave measured profile. Step 11 Although until step 10 al l the calculations have been described for the mantle located at its bottommost position (0 mm), the program allows the input o f different positions to recalculate all the results. In addition to the ability to determine important characteristics of the chamber for the measurement period, such as the approximate CSS dimension, the program is equipped with a macro that calculates a range o f CSS corresponding to mantle positions varying from 0 to 254 mm (0 to 10 in.) with 12.7 mm (0.5 in.) increments. Step 12 The program can simulate different configurations o f the crushing chamber and calculate results for various operational options at the mine. For example, it is possible to assume the continuation o f the concave and the replacement of the mantle with a new mantle with 146 a different profile by inputting the correspondent mantle information at step 7 and running the program again. 147 Appendix C Example of a measurement drawing result 148 #4 Crusher Liner Profile September 25/02 Eccentric Thro< S n a l l e s t C o n c a v e D i a n e t e r W h e n N e * S r i a l l e s t C o n c a v e D i a n e t e r : ( F e b 2 ! - 91.4") , <Mar 7 -(Mar 21 - 92 .1 ' ) , ( A p r 18 - 93 .1 ' ) , <May 16 - 9 3 . 6 ' ) , (May 30 - 9 4 . 0 ' ) , ( J u n e 2 ! - 9 4 . 9 * ) , ( J u l y It - 9 5 . 0 ' ) , ( J u l y 2 5 - 9 5 . 8 ' ) , ( A u g 21 - 9 5 . 9 ' ) , ( S e p t 2 5 - 9 7 . 0 ' ) L a r g e s t M a n t l e D i a m e t e r V h e n New - 90" ( S e p t 2 5 - 87 .0 ' ) M a n t l e ! I n s t a l l e d S e p t e m b e r 4 , £002 C o n c a v e s : P e n t i c t o n , H igh C i—WI, I n s t a l l e d F e b 7 / 0 2 S e p t 2 5 / 0 2 - T o n n e s C r u s h e d Cn iU ions) - 0.89 ( a p r o x ) Sep- t £ 5 / 0 2 M a n t l e r e m o v e d - r e p l a c e d w i t h E s c o 3 - p i e c e STD M e a s u r e D a t e - Mor 7 / 0 2 - 1.07 n t A p r 18 /02 - 3.32 n t May 16 /02 - 4.79 n t May 3 0 / 0 2 - 5.56 n t J u n e 2 1 / 0 2 - 6.26 n t J u l y 11/02 - 7,00 n t J u l y 2 5 / 0 2 - 7.61 n t A u g 2 1 / 0 2 - 8.7 n t S e p £ 5 / 0 £ - 10.2 n t C o n c a v e s r e m o v e d Appendix D New mantle parts dimensions 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items