Search Documents

In Australia we usually define "resource" in purely material terms, and rarely relate it to intellect and human capacity. We look at the sheer abundance of our natural resources and are satisfied with the feeling that we don't really need-to rely on using our brains. Certainly our country is fortunate to have the mineral resources it does, but in many ways Australia's worst enemy is its "lucky country" syndrome because we tend to underutilise the great resource we have in the capacities of our people. Raw material exports will undoubtedly continue to contribute a major share of our export revenue, but to harness the riches our country has to offer most effectively we need the vision and planning necessary to bring together our material and human resources. There is a need not only to find, exploit and process our minerals economically, but to turn as much of our mineral wealth as possible into high value products before it leaves the country. In particular we must use our brainpower if we want to maximise economic benefit from and political influence over the strategic minerals which leave our shores. In the past we have been successful in finding and exploiting minerals because our mining industry has been in the forefront of private sector research and development. But if Australia is to have an advantage over other mineral rich countries in insulating itself from the effects of cyclical swings in commodity prices, then the future of the industry must lie even more in technological sophistication rather than depending. on such considerations as size and location of mineral deposits. To survive, our mineral industry must have the foresight and resourcefulness to make its own luck. There are two keys to increased efficiency in exploration and production and in further processing methods: enhancement of the skills of the workforce and continual improvement of R&D.

Jan 1, 1986

By A Nigam, K Biswas, R Sarkar

Because of the problems with 1st and 2nd generations of Advanced High-Strength Steels (AHSS), a 3rd generation of AHSS steels has become prominent and these steels have properties in between the 1st and 2nd generations of AHSS. However, although the 3rd generation of AHSS is a promising candidate as a replacement for its predecessors, there remain some challenges in processing these steels which are essentially medium Mn (Mn content ~ 5–7 wt per cent) and high Al (Al content ~ 1– 3 wt per cent) steels. The use of conventional casting powders based on the CaO-SiO2 system is unsuitable for high and medium Mn/Al steels. This work investigates the development of CaO-Al2O3- based mold fluxes for casting third-generation AHSS steel. Mold fluxes, with otherwise similar compositions but different C/A ratios, are tested and their crystallisation behaviour is examined using differential scanning calorimetry (DSC). A calcium aluminate phase having a composition Ca12Al14O33 was found to be the main crystalline phase in the mold fluxes. A decrease in crystallisation temperature was observed as the CaO/Al2O3 ratio increased from 1.00 to 1.33. The effective crystallisation rate constant exhibited an increase with decreasing crystallisation temperature, indicating a potential influence of nucleation rate on the overall crystallisation rate and suggesting an anti-Arrhenius behaviour in the crystallisation process of these mold fluxes.

Aug 21, 2024

Coal washery wastes comprise tailings, finer than 0.5 to 0.1 mm depending on the sophistication of the washery, and coarse reject, typically finer than 50 mm. Subaerial disposal of coal tailings slurry at about 30 per cent solids by mass results in the formation of a delta with an average slope of about 1 in 100, representing one extreme. The ravelling of coarse reject on loose-dumping, results in an angle of repose slope with an angle of about 37o (1 in 1.3), representing the other extreme. The pumped co-disposal of coal tailings and coarse reject results in the formation of an upper delta with an average slope of about to one in ten, between the two extremes. The depositional behaviour of these three waste streams is discussed and compared in this paper. The important depositional and material parameters, and a theoretical description of the processes involved, are also presented. Comparisons are made between the depositional behaviour of the various coal washery waste streams, and natural depositional systems, such as rivers, alluvial fans, and scree slopes.

Jan 1, 1998

By K B. Quast

During the prefeasibility stage of the mining cycle for a gold deposit, an evaluation of the ore leaching characteristics from various defined areas is essential prior to plant design. There are a number of simple laboratory tests which have proven successful in establishing gold and silver leaching kinetics, carbon adsorption and fouling, as well as elution characteristics. A number of case studies of carbon-in-pulp (CIP) simulation are presented which show excellent agreement between calculated and assayed gold and silver grades in feed and adsorbed on carbon which validate the methods used.

Jan 1, 1998

By J J. Eksteen

"In the platinum group metals (PGM) industry, smelters are often perceived as ‘black boxes’ that have to process concentrates of variable mineralogy and originating from multiple mines operating along strike or downdip from each other in the same geographical region, as is the case in the Western Bushveld of South Africa. However, PGM smelter complexes are much more sensitive to concentrate blend variation than most metallurgists or geologists would anticipate.This paper will review the impact of concentrate blends deriving from a number of concentrators processing ores from the Bushveld Igneous Complex (BIC) on smelting and converting behaviour and the constraints they impose upon smelter operation. Significant variation in behaviour is not only noted between the main mineralisation types, namely the Merensky and Upper Group 2 (UG2) reefs, but also within a given reef type. Due to the way PGMs are mineralised and finely disseminated, concentrators have to grind concentrates ultra-fine to liberate the PGMs and provide a sufficiently high-grade concentrate for smelting.The chemistry and mineralogy of the concentrates vary with the type of mine (open pit versus underground), the reef type and varies along strike; this variation impacts numerous key smelter parameters such as the base metals content, the matte fall, the converter aisle capacity, sulfur dioxide gas production, furnace matte and slag temperatures, and the accumulation of high smelting chromite-spinel minerals that has led to significant smelter upset conditions such as catastrophic furnace hearth and sidewall failures.Due to the production intensity of smelting, ie its high rate of processing per unit volume, smelters have been found to be far more sensitive to rapid variation in concentrate properties, highlighting the need for adequate blend control.CITATION:Eksteen, J J, 2016. Impact of Mineralogical Variation on the Smelting of Cu-Ni-PGM Concentrates, in Proceedings The Third AusIMM International Geometallurgy Conference (GeoMet) 2016, pp 321–326 (The Australasian Institute of Mining and Metallurgy: Melbourne)."

Jun 15, 2016

MIM Holdings Limited (MIM) committed to a long-term future in copper in the Carpentaria Mineral Province in 1996 with the development of the Enterprise and Ernest Henry mines, and an upgrade of MIMsÆ copper smelter in Mount Isa, and copper refinery in Townsville. The upgrade of the copper smelter will enable the smelter to produce 250 000 tpa of copper with a single primary smelting unit, the copper Isasmelt furnace installed in 1992. The copper Isasmelt furnace currently smelts concentrate containing 145 000 tpa copper. To reduce the risk associated with the copper smelter upgrade, and shorten the ramp up time of the upgraded smelter, two high-rate trials were undertaken in 1997 to demonstrate that the copper Isasmelt furnace could, with an additional 525 tpd of tonnage oxygen, treat concentrates containing 250 000 tpa copper. During the trials the concentrate feed rate to the Isasmelt furnace was increased from its average rate of 100 tph of concentrate to 150 tph of concentrate. The increase in the capacity of the furnace was achieved by: supplementing the installed tonnage oxygen capacity of 595 tpd with an additional 300 tpd supply; reducing variability in the furnace feed rate from ¦25 per cent to less than ¦5 per cent; and reducing tramp air. The Isasmelt furnace smelted concentrate at a rate of 140 to 150 tph for four hours 25 minutes during Trial 1, and at a rate of 150 tph of concentrate for five hours and 30 minutes during Trial 2. The oxygen utilisation was 100 per cent during the trials with the process remaining under control. As a result of the development work for the trials the average treatment rate of the furnace under normal conditions increased by ten per cent.

Jan 1, 1998

By M D. Urbani

HiTec Energy Ltd has been conducting research since 1996 into novel process routes for the treatment of low-grade and waste manganese sources for the production of high value and high quality electrolytic manganese dioxide (EMD) for the alkaline battery market. HiTecÆs approach is to treat previously unsaleable manganese deposits and mine site tailings using a novel sulfur dioxide leach route, followed by purification and electrowinning to generate the sought-after EMD product by an environmentally superior processing route to those presently practiced throughout the world today.HiTec Energy Ltd expects to be in commercial production with its Stage 1 facility in Kalgoorlie in the last quarter of 2004 for the production of 12 500 tpa EMD. With this production facility being adjacent to a laterite nickel producer with available waste manganese units, the opportunity arose for HiTec to combine its existing technologies with processes (that will be discussed in this paper) to turn two waste by-products into products that are used for high-tech purposes.HiTec Energy Ltd and the Parker Centre have had a collaborative research relationship for a number of years in areas involving EMD and battery related technologies. Meanwhile the solvent extraction (SX) group of the Parker Centre (CSIRO Minerals) has developed SX processes to separate manganese from nickel and cobalt. A project was initiated by HiTec Energy to recover manganese from waste laterite leach solutions using SX technology. Batch tests were conducted to select a suitable organic system for the extraction of manganese in terms of selectivity and extraction and stripping kinetics.A total of eight organic systems were tested by determining pH isotherms of metals using a laterite nickel process waste stream. Two organic systems, the Versatic 10 and Versatic 10/synergist systems, were chosen for further test work with Shellsol D70 chosen as the diluent for both systems.The Versatic 10/synergist system would be advantageous over the pure Versatic 10 system because fewer SX stages are required, resulting in a smaller organic inventory and lower capital costs. The risk of gypsum formation is also significantly lower with the synergistic system.

Jan 1, 2004

By Kwitko G

Exploration for coal has been on-going in South Australia for the past 100 years. During this period substantial resources of low rank coal have been delineated throughout the State. These deposits are of variable quality often with high moisture, sodium, chlorine, sulphur and ash content and although generally regarded as suitable for use in conventional pulverised fuel boilers, many are difficult to utilise. The only operating coal mine in South Australia is at Leigh Creek where coal bearing shale was discovered by chance in 1888 during excavation of a railway dam near the township of Copley. Leigh Creek has had a very diverse mining history from the initial small scale underground mining through to open-cut mining utilising draglines and more recently large shovels and trucks.

Jan 1, 1992

By L Keeney

Geometallurgy is not a mechanism designed to replace current geological and metallurgical deposit characterisation but an added component required to understand orebody variability. Current geological characterisation approaches focus on understanding deposit genesis aimed at identifying new mineralised regions and metallurgical characterisation determines parameters required for engineering design of processing plants. Geometallurgical integration has the task of drawing on aspects of each stream to enable spatial mapping of performance indices across a deposit. It is a complimentary activity to current practice and adds a new dimension to orebody knowledge. Although the notion of conducting an integrated geological and metallurgical analysis is not a new concept, limited publications exist in the field proposing integrated methods, with several technical issues existing in current geometallurgical integration approaches. An iterative method drawing on existing geological and metallurgical characterisation and new measurement tools and inputs is proposed. The integrated method comprises of discrete components dynamically linked together through iterative feedback loops with each component having a specific purpose achieved by applying a range of data acquisition technologies and statistical tests. The key outcomes of the method are:overcoming the lack of optimised geometallurgical domains relevant to processing performance,development of a rigorous approach to identifying parameters controlling processability,improved protocols for relevant metallurgical test data and sample selection, andability to generate greater data support to enable rigorous geostatistical modelling of geometallurgical attributes. This paper outlines the integrated mapping and modelling method developed and draws on multiple data acquisition technologies applied within the AMIRA P843 GeMIII Project.

Sep 5, 2011

The Sishen iron-ore-mine produces 3 pro= ducts with size gradings of 25 to 8 mm, 11 to 5 mm and 5 to 0,2 mm and has a product capacity of 20 Mtpa. Ore is produced at an average Fe content of 66,2 % Fe for the two coarser pro= ducts and an average of 65,3 % Fe for the fine ore product. Three stage crushing is practised to re= duce the run-of-mine ore to -90 mm after which it is screened into five different size frac= tions of which four are heavy media beneficiated in static baths and cyclones and the fifth frac= tion discarded to slimes dams. A final three stage crushing and screening process is employed to produce the final three Products which are homogenized by means of blen=ding systems nrinr to despatch.

Jan 1, 1987

By Pratt V

The economic, social and political environment of the mining and minerals industries is, as we all know, volatile and affected by constantly changing international forces. In addition, the outcome of negotiations with overseas customers is often uncertain for long periods, or even unfavourable. The industries and companies represented at this Conference are in an arena of high competition, risk taking and enormous capital outlay. They are industries which our overseas markets (rightly or wrongly) have criticised for industrial relations problems, unreliable production and uncertain deliveries. They are industries which have an urgent need to be cost effective and competitive at all levels of activity.A critical factor in the mining and metallurgy industries is, therefore, its human resources and its human resource management. It requires the full use of available talent, innovation, skills and qualifications within its workforce, and the appropriate management of these resources.

Jan 1, 1985

By A Jyoti, J R. Black, R R. Haese, E O. Ansah

Copper (Cu) is a critical metal in the transition to a low-carbon power supply. Presently, low-grade copper sulfide ores account for almost 80 per cent of the global reserves. Large amounts of water are required to process these low-grade copper ores. The mining industry utilises lixiviants from recycled water from different sources to conserve water, whilst enhancing Cu recovery. Some utilise intermediate to pregnant leach solutions, sea water, and processed water from electrowinning circuits with highly variable concentrations of ions. Previous studies have highlighted the formation of precipitates that passivate the surfaces of minerals in these low-grade copper ores and/or block flow paths for lixiviant during heap leaching. In this study, we utilised concepts of phase replacement to devise an innovative means of controlling precipitation and increasing copper recovery using industrial brine as opposed to freshwater. We propose the formation of Al-rich phases (such as AlSO4 + or alunite) to replace iron sulfate and iron hydroxysulfates (such as jarosite) during the chalcopyrite dissolution-precipitation mechanism. The rationale is to have available Fe3+ throughout the chalcopyrite leaching process. We tested this method by carrying out various batch experiments using CaCl2, AlCl3, and NaCl industrial brine types compared to freshwater and explored the capacity to leach Cu from chalcopyrite. There is higher Cu recovery during oxidative leaching of chalcopyrite in industrial brine with 91 per cent recovery using AlCl3-rich brine, 84 per cent with NaCl-rich brine, and 56 per cent with CaCl2-rich brine as compared to 35 per cent recovery using freshwater. In the absence of Al3+, various forms of iron sulfates, and jarosite are formed, reducing the availability of Fe3+ for chalcopyrite dissolution. The introduction of Al3+ leads to the formation of AlSO4 + species, thus lowering sulfate activity and the tendency for K-jarosite formation.

Aug 24, 2022

By R G. Reddy

The complexity in treating environmentally harmful impurities such as As, Sb, Bi from the base metal matte or metal in the smelting stage are responsible for the high cost of the refining process. The impurity capacities (such as arsenic, antimony and bismuth) of slags were calculated a priori using Reddy-Blander (RB) model. The capacity predictions were for a wide range of matte and slag compositions in copper smelting conditions. The calculated impurities capacities and impurity distribution ratios results are in good agreement with the available experimental and industrial slags data. The a priori knowledge of impurities is useful for reduction of energy consumption and enhanced environmental control in the current and future non-ferrous metal processes.

Aug 21, 2024

By K L. Avarmaa, M A. Rhamdhani, A D. P Putera, H T. B M Petrus, G A. Brooks

Multistage slag treatment was carried out to investigate boron removal from metallurgical grade silicon. Boron is known to be one of the most difficult elements to remove from silicon. Depending on boron initial concentration, boron can be removed from silicon into certain level; most previous studies have attained final 50–100 ppmw (parts per million weight) boron in the silicon using slag treatment. The term ‘multistage’ refers to a process where the resulting silicon from the initial slag treatment is further refined multiple times using a new slag. In this study, an initial slag composition of 45CaO-45SiO2-10Al2O3 (wt per cent) was utilised. Using this technique, boron level in the silicon was expected to be gradually lowered because the new slag will always have more capacity to reach the equilibrium distribution coefficient (LB = (ppmwB)slag/(ppmwB)Si). The current study was carried out through combined thermodynamic calculations (using FactSage™ ver 8.2, by Thermfact Ltd, thermochemical package) and high-temperature experimentation. The thermodynamic calculations were carried out for reactions at 1500, 1550, and 1600°C for three different scenarios: 1. Single stage refining with varied slag to Si ratio. 2. Multistage refining with slag to Si ratio of 1:1. 3. Multistage refining with slag to Si ratio of 2:1. The multistage experimental work was carried out at 1550°C under inert gas atmosphere each time for 2 hrs, with slag to Si ratio of 2:1. The simulation results indicate that the boron distribution coefficient stay essentially constant with varying slag to Si ratio and temperature, around 1.8 to 2, and the experimental results measured LB in the same range, between 1.5 and 2.5. By harnessing this feature, the simulation showed that boron can be removed by approximately 78 per cent in each stage. The experimental result confirmed that the boron concentration in the silicon was reduced by a total of 97.8 per cent from its initial value using four stages of slag treatment, from initial 270 ppmw down to 6 ppmw.

Aug 21, 2024

By D Gagnon, P Bouchard, F Vanier, E Soares de Lima Filho, J El-Haddad, M Sabsabi, G J. Wilkie, A Beauchesne, A Hamel, A Blouin, T Vaillancourt, F Boismenu, A Harhira, A Plugatyr, C Padioleau

Quantitative mineralogy is an established discipline in the geosciences and is aimed at providing mineral grade and texture information for geological and mineral processing applications. The two core technologies underpinning Quantitative Mineralogical Analysis (QMA) are Scanning Electron Microscopy (SEM) in combination with Energy-dispersive X-ray Spectroscopy (EDS). Technologies such as QEMSCAN and MLA are now routinely used to optimise the performance of large-scale mineral processing plants in the base and precious metal sectors. One of the major disadvantages of the QMA methods used today is extensive sample preparation requirements which make application of this technique to real-time mineralogical characterization not feasible. A breakthrough has been achieved by the National Research Council Canada in collaboration with CRC ORE by developing a novel Laser Induced Breakdown Spectroscopy (LIBS) based technology capable of real-time mineralogical characterization of process streams without sample preparation. The potential applications of this technology include, but not limited to, in-pit muck piles, underground draw points, cross-belt analysis as well as slurries. This paper describes the development of the LIBS-based technology from a proof-of-concept (TRL2) to the construction of a prototype sensor and its validation in a simulated environment (TRL5). Future work is being planned to test and further validate the LIBS sensor on a mine-site which will progress the technology to its next readiness level (TRL6).

Nov 10, 2020

By Z Fu, E Topal, O Erten

"Mine waste rock management involving waste rock movement and placement, is a critical aspect in the planning of open pit mining operations, as the waste rock haulage cost can be the largest portion of the mine operating cost. The concept of sustainable development in the mining industry also encourages environmentally friendly waste dumps that have zero hazards to people and the surrounding environment. Therefore, a detailed waste management schedule must satisfy economic and environmental considerations. However, the available tools in the market cannot provide optimal waste dumping schedules to meet all these constraints. This paper introduces a user-friendly and automatic tool to optimise the waste dumping schedule, which minimises the haulage cost, and covers the encapsulation of reactive materials to minimise acid mine drainage problems. This tool has been developed by using a new mixed integer programming (MIP) mathematical model that considers both single-lift and multilift dumping scenarios. The application of the proposed tool on a data set collected from a gold mine shows that the multilift dumping scenario has provided a more flexible and economical waste dumping schedule than the single-lift scenario, and both scenarios satisfy environmental requirements.Citation:Fu, Z, Topal, E and Erten, O, 2016. A Modified Software Tool for Mine Waste Rock Dump Planning, in Proceedings Ninth AusIMM Open Pit Operators' Conference 2016, pp 229–242 (The Australasian Institute of Mining and Metallurgy: Melbourne)."

Nov 15, 2016

By K Myers

Since the mid-1990s, the Chino Mine Facility (Chino) located near Silver City, New Mexico has been developing a closure plan with financial assurance to protect water quality, as required by State regulations. In February 2002, Chino and the New Mexico Environment Department (NMED) reached agreement on a closure plan; a closure cost amount of approximately $US 394 million ($A 591 million) was finalised in April 2003. Both the New Mexico Water Quality Act Regulations (1977) and the New Mexico Mining Act Regulations (1993) guided development of the closure plan and financial assurance amount. The process incorporated many years of correspondence, studies, negotiations and administrative hearings. The major technical issues focused on regrading slopes of the leach ore stockpiles and waste rock piles; cover systems; water treatment systems; and abatement of ground water contamination. The proposed financial assurance amount covers all closure costs and provides for perpetual water treatment.

Jan 1, 2003

By M Gwynne

McArthur River Mine is located approximately 50 km south-east of the NT township of Borroloola and approximately 120 km from the Gulf of Carpentaria. The mine hosts a series of layered shallow dipping tabular orebodies, with current reserves of approximately 22 Mt, grading 13.6 per cent zinc, 6.3 per cent lead and 65 g/t silver. Owing to economic considerations, the lower most orebody (No 2 orebody) has been selected for initial mine production. Following the completion of the No 2 orebody, mining will transfer to the lower grade and metallurgically more complex No 4 orebody. The mine employs a total of 235 people throughout the underground and surface areas. The majority of the workforce work a fly-in/fly-out, seven-day- on/seven- day- off roster with 12-hour shifts. Presently, a further 18 contractors are involved in trial mining areas. An apparent dip room and pillar technique has been selected as the primary mining method. Officially opened in September 1995, the current production rate of 1.2 Mt per annum will eventually escalate to 1.5 Mt per annum. Though most production is achieved through the use of conventional loaders, trucks and jumbos, the recent development of methods utilising scrapers, hand held machines and modified jumbos has produced encouraging results in steeper dipping areas of the mine. Initial production has been hindered by several factors including extensive and largely unpredictable faulting, worse than expected ground conditions and difficulties with some mining equipment. Compounded by the relatively thin 3 m ore seam, initial mining resulted in high dilution and poor production. This paper describes the development of unique planning and production systems that were evolved in order to overcome these problems and achieve the targeted production rates and head grade.

Jan 1, 1998

By M Mohammad Khani

An economic block cave is dependent to the characteristics of the rock mass at stages of caving. These are cave initiation, propagation and ore flow. In this research, the rock engineering system (RES), first established by Hudson (1992) to solve complex rock engineering problems, is introduced and verified against the well-established caving parameters established by Laubscher. There are two types of factors influencing caving and these are natural (cannot be controlled) and induced (operator controlled). A RES based interaction matrix is developed by determining the rating parameters required to reproduce the caving database using Laubscher’s caving parameters. This provides the relative importance of the parameters in the interaction matrix. After discussing the rock mass caving process and identifying all the effective parameters, an interaction matrix is introduced to study the cave initiation of rock masses. The expert-semi-quantitative (ESQ) coding method is utilised to assign the numerical value to each parameter’s interaction. Thereafter, the interaction intensity and dominance of each factor in the system is determined to classify these factors. Finally, all the parameters affecting caving operation and their weights on the RES system are identified. This new application of RES in block caving is very efficient and useful for any stage of caving studies and helps experts to make decisions more accurately.CITATION:Mohammad Khani, M, 2016. The application of rock engineering systems to block caving to identify the effective parameters, in Proceedings Seventh International Conference and Exhibition on Mass Mining (MassMin 2016), pp 853–860 (The Australasian Institute of Mining and Metallurgy: Melbourne).

May 9, 2016

By L W. Palmer

"Evaluating the geological risk associated with ore deposits is becoming common practice in the mining industry. While geostatistical simulation is a useful tool for evaluating this risk, the effect of applying different types of simulation is less understood. For simulation of categorical variables like rock types, available methods perform simulation in fundamentally different ways. In this study, the geological risk relevant for informed mine planning is studied with two categorical simulation methods, PluriGaussian Simulation (PGS) and Sequential Indicator Simulation (SIS). Risk is expressed as a probability of repeated occurrence of lithological units in the same location across simulations. Significant differences are observed between the methods in terms of spatial distribution of risk. To identify the most plausible risk model, the input parameters at every step of both techniques are scrutinised. PGS and SIS inputs are classified as four parameters that are: fundamental to implementation of the methodfixed to enable comparison of PGS and SISdetermined by user interpretation, hence variablenormally optimised during application of PGS and SIS. The effect of parameter selection from categories 1 and 4 on the risk model is investigated using drill core data from a leucogranite deposit. A corresponding block model is created for assigning outputs from PGS and SIS. Risk models produced by either method are illustrated in a cut-out from the block model to enable bench-by-bench comparison. For PGS, risk associated with relatively rare lithologies is relatively sensitive to variation of parameters. Independent of the abundance of lithology, a location-dependent effect of variogram and vertical proportion curve parameter selection is observed, with risk increasing towards the deposit edge. Although SIS requires fewer inputs, single lithologies must be modelled in turn. Using small numbers of samples in the search neighbourhood, all risk models differ substantially. For larger numbers of samples, the risk models are similar. Overall it appears that the risk model produced using SIS is less sensitive to changes than PGS but further work is required to confirm this.CITATION:Palmer, L W and Glass, H J, 2014. Risk associated with rock type prediction using simulation techniques, in Proceedings Ninth International Mining Geology Conference 2014 , pp 217–228 (The Australasian Institute of Mining and Metallurgy: Melbourne)."

Aug 18, 2014