Search Documents

By I Grewal, B Klein

The Knelson Continuous Variable Discharge (CVD) is a continuous centrifugal gravity concentrator for high-mass yield recovery applications. It has four operating variables that enable control of mass yield, product grade and recovery namely: fluidisation water, bowl speed, pinch valve open time and pinch valve closed time. Plant testing of a pilot scale Knelson CVD6 was conducted to evaluate the effects of operating variables on mineral separation. The analysis of operating variable effects has provided a framework for the development of operating guidelines for the CVD.

Jan 1, 2001

By P Collins, C Sheriff

The Sand Queen gold mine is a typical Archaean narrow, high-grade gold quartz lode deposit within a shear hosted system. Initial mining during 1904 - 1948 produced 181 659 ounces of gold from 248 564 tonnes of ore milled at an average production grade of 22.7 g/t Au, which equates to a minimum of 750 ounces per vertical metre. The current mining operation is using the old Sand Queen main shaft to access the Sand George deposit, which is a narrow quartz lode system discovered in the same shear structure along strike to the south of the Sand Queen mine workings. Current underground resources contain 360 200 tonnes at a grade of 11.8 g/t Au for 136 500 ounces. Mining is scheduled to produce 20 000 ounces of gold annually. Re-opening of the Sand Queen mine and underground development of the Sand George lodes has revived several grade control issues relevant to a narrow, ænuggettyÆ gold lode mining environment. These include the dependence of assay grade of individual samples upon a number of factors, including how and where a sample is collected, how and where it is analysed, and how grade data is used to estimate an æaverageÆ grade. Ideally, the only way to achieve an acceptable estimate of ætrue gradeÆ is to mine and mill a large parcel of ore (eg >100 tonnes) from each lode or ore type. In a mining operation producing relatively small tonnages from a number of separate ænuggettyÆ gold lodes, availability of resources and operational constraints may preclude intensive sampling for predictive grade control. At the Sand Queen mine, one of the key issues has been to establish a reliable system for estimation of gold grade with which to reconcile milling data. Extensive use has been made of detailed geological data to understand the grade distribution within individual lodes and between lodes, so as not to rely solely on assay results.

Jan 1, 2008

In the UK during the 1990s the energy companies had developed a policy of æDash for GasÆ, building large natural gas power stations to produce more than 30 per cent of the UKÆs electricity demand. The price of imported thermal coal was dropping and the electricity generators were putting pressure on the UK coal producers to produce coal at a reduced cost. Imported coal from Columbia, South Africa, America and Australasia had increased their market share in thermal coal sales to over 50 per cent of a 55 Mtpa market. The pressure on the UK coal industry to reduce costs and improve productivity levels from an ever decreasing resource was high. Companies were benchmarking their operations on the imported coal producers and had to realign their organisations, productivity levels and costs to world best practice, not just technically, but in world class organisational management techniques.

Jan 1, 2005

By M Badawi, Y Foucaud, C Veloso, J Laine, A C. Araujo

Iron is one of the most exchanged commodities worldwide. Its use for steelmaking explains the growing demand in the metal which remains mainly produced by the mining and the beneficiation of iron ores. Flotation is the most used mineral separation technique worldwide, processing several billion tonnes (Bt) of ore annually, including, in some cases, hematite and magnetite ores. Based on the physicalchemical properties of mineral surfaces, flotation aims to selectively adsorb molecules that render mineral surfaces either hydrophobic or hydrophilic, known as collectors and depressants, respectively. Hydrophobic particles are then recovered in the froth by the injection of gas bubbles travelling up in the cell and attaching the hydrophobic particles. However, despite the popularity and the good performances of flotation, mining industry faces new challenges of decreasing grades and increased mineralogical complexness in ores. Many flotation variables, especially when related to reagent adsorption, remain poorly understood and it becomes necessary to gain knowledge to enhance the existing processes. Molecular modelling techniques like the density functional theory or molecular dynamics simulations (ie Classical molecular dynamics or Ab initio Molecular Dynamics) currently represent powerful tools for different applications in material sciences, like catalysis or depollution purposes. They have recently been successfully applied in the flotation area, with an increasing number of publications dealing with this topic every year. These tools provide crucial information on bulk and surface properties as well as on the adsorption mechanisms of reagents on mineral surfaces, key step in the flotation process, and synergetic effect of collector mixtures. A better understanding of the adsorption mechanisms of flotation reagents, provided by molecular simulations, will help to finely choose flotation reagents, and even lead to the development of new reagents formulations. Overall, these tools enable the discovery of practical and industrial solutions to improve the flotation process, which will undoubtedly help the mining industry to face its future challenges.

Nov 8, 2021

By C D. Pocock

Maintaining catch berm capacity in open pit mines through controlled blasting prevents production delays associated with managing risks associated with wall instability. A method has been developed to quantify the drill hole deviation of presplit holes at Century Mine. Different controlled blasting techniques are used to control wall quality and decrease the risk. The performance of presplit drilling, in terms of deviation from design has been investigated. A methodology has been developed to quantify the drill hole deviation in three dimensions. This methodology was converted into a series of algorithms capable of handling an extensive range of data. The methodology was applied to a case study at Century Mine to evaluate the performance of drilling on two presplit blasts. Overall, 67 drill holes were investigated in the case study. The drill hole deviation was quantified in terms of three parameters: skewness, linearity and drill hole deviation parameter. Methods for quantifying each parameter were developed. Skewness, a measure of the amount of curvature and linearity, was defined as the measure of the consistency of the gradient of the drill hole. Drill hole deviation parameter is defined as the average measured deviation of the surveyed drill hole position from the design position. Acceptable drill hole deviation tolerances for each of the defined parameters were developed by analysing drill hole deviation parameters with respect to implications to catch berm capacity. Verification of the methodology was confirmed through the comparing of the completed drill hole deviation results using both visual inspection and cross analysis for each of the three methods. Holes with an R squared value of less than 0.97 for skewed deviation were classified as having skewed deviation, while holes with a difference in gradient of 3.5 were classified as having linear deviation. The results show that 76 per cent of the drill holes had deviation. Of this, 19 per cent were skewed, 57 per cent had linearity and 12 per cent contained both skewed and linearity deviation.

Nov 8, 2011

By C J. Greet

Oil can enter a flotation circuit from a number of sources (for example, from mining equipment, oil leaks from the grinding equipment and coatings on the grinding media, but to name a few). Folklore generally states that any oil that contaminates the ore entering the flotation circuit is considered to have a negative impact on flotation. That is, a loss in concentrate grade and recovery.This represents a dilemma for some grinding media suppliers, who use a variety of quenching medium (air, polymer, oil or water) to achieve the desired physical properties of the ball. In the case of high chrome grinding media it is necessary, in some instances, to quench in oil to give the ball the right degree of toughness, which invariably leads to these balls having a thin film of oil on their surface. Once the grinding media arrives on site and site personnel recognise that the balls are “covered” in oil, they generally insist that the oil will have a deleterious impact on flotation performance.Obviously, any supplier “worth their salt” would take remedial action at the point of manufacture to minimise this problem. However, does quenching oil actually present a significant “threat” to the flotation process? This paper provides both laboratory and plant data describing the effect quenching oil has on the flotation of copper sulphide ore, with surprising results!

Aug 8, 2011

By D Finn, D Kentwell

Experience has shown that the E-Type multiple indicator kriging (MIK) estimation technique often works well for resource estimation in stockwork mineralisation that contains a mixture of high and low-grade mineralisation types that are difficult or impossible to separate spatially. E-Type MIK, like ordinary kriging (OK) is a smoothing estimator and where data is sparse can result in estimates of grade tonnage curves that do not reflect the true block variability at selective mining unit (SMU) scale. The rank and replace (RR) estimation methodology is designed to reproduce both local block accuracy and true block variability at SMU scale and has shown good results when used with highly skewed unimodal diffusive style mineralisation. This paper examines the use of RR with mosaic style mineralisation typically found in stockworks via a case study using real data from Newcrest’s Telfer West Dome deposit. Resource level estimation results for ordinary kriging (OK), E-Type MIK and RR are compared to close spaced grade control models as the reference reality.Results show that stationarity requirements and declustering can severely impact on the usefulness of the rank and replace technique and that it should only be used where trends are not too strong and data are relatively uniformly spaced. This does not rule out its application to stockwork or mosaic style deposits. However, in this case, the presence of trends and different data densities requires the non-stationary domain to be further broken down into sub domains for the RR technique to be effective.CITATION:Kentwell, D and Finn, D, 2017. Testing new estimation techniques – rank and replace: a stockwork case study, in Proceedings Tenth International Mining Geology Conference 2017, pp 365–374 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Sep 20, 2017

By G Barla

In Europe, a major interest in the infrastructure and tunnelling fields is in the new crossings of Alps in the form of base tunnels, which are important components of the high-speed (for passengers)/high-capacity (for freight) Trans European railway corridors. These tunnels cross the mountain belt at the elevation of the main cities and basins located all around the Alps and are longer (up to 57 km) and deeper (up to 2500 m) than all tunnels constructed in the centuries before. The excavation of the first Alpine railway and road traffi c tunnels could in general be carried out successfully by conventional excavation methods. With the great lengths of the new base tunnels, mechanised excavation by using tunnel boring machines (TBMs) is imperative as it contributes significantly to savings in construction time and costs. Questions are however posed on the adoption of such a systematic tunnel construction method given the severe ground conditions which are some time anticipated along the tunnel alignment, with possible major geological hazards such as water inflow in heterogeneous fractured or karstifi ed rock masses, stress-induced tensile failure of hard rocks, and plastic shear failure and squeezing of weak rocks, etc. This lecture is to briefly review the current state-of-the-art in the understanding and modelling of two such critical hazards as spalling/bursting and squeezing ground behaviour. With the present interest in the design of mechanised tunnelling with TBMs in such conditions, the complex interaction between the rock mass, the tunnel machine and its system components, and the tunnel support is studied in detail by numerical modelling. The research and development efforts presently underway with the intention to provide the designer with a novel three-dimensional (3D) simulator of mechanised tunnel excavation are presented. Case examples are discussed with reference to the Lyon-Turin Base Tunnel design.

Mar 8, 2011

By S M. Arndt, D Tyler, F Reusch

Although significant advances have been made in understanding the mechanics of pit slopes, numerical modelling hasnÆt significantly impacted on the reliability of the design process. It is well known that legacy modelling approaches donÆt simulate large slopes well. In recent times, the emergence of very large finite element (FE) models with more than ten million degrees of freedom and the ability to simulate a large number of discrete structures, has allowed a step change in the simulation precision for slopes. The model size allows global and local effects to be realistically simulated in one model for the first time and the process can be better integrated into normal open pit slope engineering processes. Some examples of simulation of very large open pits are shown that demonstrate the challenges and some recent advances in modelling large slope behaviour. Significant limitations in the current measurement and observation regimes of open pits have been identified.

Jan 1, 2007

By S C. Dominy, R C. A Minnitt

Determination of the sampling constant and liberation diameter are of importance when applying the Fundamental Sampling Error (FSE) equation to optimise sampling protocols. The method for the direct determination of the sampling constant is based on the æGy 50-pieceÆ experiment and variants thereof, which is generally known as the heterogeneity test. In the coarse gold environment, the heterogeneity test is prone to poor repeatability and bias due to the very heterogeneity that it aims to quantify. A key issue is the large sample mass that can be required in the coarse gold environment, which may reach 1 t or more. The heterogeneity test often provides an evaluation of the background gold grade heterogeneity, but understates that of the high-grade coarse-gold component. This is because the total mass of fragments is usually too small to represent the mineralisation. Despite this relatively well-known problem, heterogeneity tests are often undertaken on deposits where the presence of coarse gold is ignored, not fully realised or vastly understated. Sampling types and protocol designs are frequently based on poor quality data that do not match the characteristics of the orebody. For this style of mineralisation, an empirical approach to define ore characteristics is recommended.CITATION:Dominy, S C and Minnitt, R C A, 2012. Application of heterogeneity testing to coarse gold ores, in Proceedings Sampling 2012 , pp 143-160 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Aug 21, 2012

By H Kurth

Mining geologists utilise many technologies to map, sample, measure, model and plan before and during the mining process. Predicted ore quality rarely corresponds with feed quality determined by process operators. Objectively measuring conveyed material quality accurately in real time avoids disagreements over mined ore quality and generates opportunities to improve mill feed quality.This paper discusses the well proven PGNAA (prompt gamma neutron activation analysis) technology used in the GEOSCAN-M elemental analyser that is ideally suited to measuring composition of primary crushed conveyed material in real time. Representative, continuous, fully penetrative measurements are unaffected by particle size, belt speed, mineralogy, dust or layering on the conveyor. Material can be directed to processes or stockpiles by utilising measurement increments of typically one to five minutes and then activating diverters to direct increments to the correct location. Waste, low-grade or deleterious material can be rejected to waste dumps or stockpiled for later processing. Material meeting a product specification can bypass unnecessary processing. Paybacks in weeks have been achieved for such applications. Process operators are able to control the blend of ores fed into the mill to improve processing performance and use measurements for ore reconciliation and metal accounting. Utilising the technology to upgrade ore quality after mining and before processing is beneficial in improving plant economics.The technology has been successfully utilised in base metals, ferrous metals and industrial minerals. Mining geologists can determine ore quality being produced in real time and modify the ore quality prior to milling. Ore feed quality measurements are also used by process operators in optimising plant performance. The elemental content can help identify geometallurgical domains as well as other major changes in quality, which may affect processing. This paper outlines proven applications and case studies and proposes some innovative applications.CITATION:Kurth, H, 2017. Measuring ore quality using real-time on-belt elemental analysis – more than grade control, in Proceedings Tenth International Mining Geology Conference 2017, pp 213–220 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Sep 20, 2017

By J-P Franzidis, E V. Manlap

Froth transportation characteristics have a significant impact on the flotation recovery especially in large flotation cells. The probability of an attached particle being recovered into the concentrate depends strongly on the location where it enters the froth phase. A mathematical model for froth transportation in industrial flotation cells was proposed (Zheng et al, 2004). The model has been tested on the experimental data obtained from testwork conducted at the Western Metals Hellyer lead/zinc concentrator. The results show that the model fits the experimental data well, especially when the froth is well developed (ie for a deep froth). The model fitting can be significantly improved by incorporating an air dispersion factor. However, the models encounter some limitation, ie inconsistent prediction at different froth heights, due to change in froth viscosity which cannot be modelled at the present time.

Jan 1, 2005

By E Philippa, B Miller

Material characterisation, ore sequencing and process capability are commonly regarded as the key parameters to consider in designing and operating an effective pit to plant supply chain. Therefore, an understanding of rock mass properties is the starting point for controlling the presentation of problematic material to the ore handling plant and optimising process utilisation. There are different methodologies for characterising problematic material and ore handleability which exist across a spectrum of sophistication. At one end of this spectrum are simple indices built using first principles geology and at the other are complex algorithms developed using advanced analytical methods, machine learning and ore traceability. Regardless of the method, without an integrated process for careful management of problematic ore, material handling delays (ie blocked, bridged, blinded and buried events) will interrupt the stable flow of production. The ability of the ore control geologist to mitigate these delays is central to the ore handleability challenge. This paper describes a quick and easy solution that links the presentation of ore to the operating window at the crusher. This operating window is based on material characterisation, using the physical, textural and mechanical properties of the rock mass, and crusher performance. It also describes how a basic problematic ore index (POI) can be developed using ore tracking to the production face, petrographic studies of ore texture and mineralogy, and existing variables in the geological model. Finally, this paper will demonstrate that by coding the POI into the geological model for mine planning, together with performance monitoring through compliance to plan routines, material handling delays at the crusher can be reduced, unlocking significant resource value.

Mar 22, 2022

By T R. Stacey

South Africa has a long history of rockbursts, some 100 years. During this time much research and many developments have taken place in the country to limit the occurrence of rockbursts, and to mitigate their effects. In this paper I shall present a personal view of what I consider to be important achievements, and contributions made. Some of these are: planning of stope layouts, and the use of strike barrier pillars; recognition of energy as a key factor, the use of the energy release rate ERR as a layout planning criterion, and development of MINSIM for calculation of ERR and stope planning; sequential grid stope layouts; use of backfill for stope support; and, development yielding support elements, and dynamic testing of stope and tunnel supports. The record shows the establishment of a rock engineering research powerhouse over a 40 to 50 year period and, unfortunately, the complete breakdown of this capability in the subsequent 10 years. CITATION:Stacey, T R, 2018. A personal review of South African rockburst research, experience and lessons learnt, in Proceedings The Fourth Australasian Ground Control in Mining Conference (AusRock), pp 5–16 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Nov 30, 2018

By I Canbulat, J Emery, L Yao and C Zhang

This paper documents a process of data analytics and machine learning applied to coalmine roof convergence data. It follows a publication documenting the formation of the database. For each set of analyses Python scripts were utilised to extract data from an SQL database, with subsequent analyses also executed in Python, utilising open-source tools such as scikit-learn. Preliminary analyses with classic machine learning algorithms included multiple logistic regression, K-nearest neighbours, decision tree, random forest, support vector machine, kernel support vector machine, artificial neural network (ANN), naïve Bayes and multiple linear regression. Results indicated that relationships were complex and non-linear, with the artificial neural network most suitable. An ANN model was then refined through a standard process of hyperparameter optimisation and data augmentation to arrive at a final model. Due to the black box nature of the ANN further insight into how the parameters interacted was sought and found with Shapley additive explanations (SHAP values). SHAP values utilise Game Theory, where each input parameter or feature is a ‘player’ while the data set is the ‘team’. The SHAP value is the impact of each player on the target value, essentially establishing the contribution of each input parameter. The results from all analyses are delivered in the paper, together with insights, applications, strengths, and limitations of the methodology.

Nov 29, 2022

By W Zeng, D Walker, D Cai, E Baafi

"The efficiency of the truck-shovel haulage system is significantly influenced by the truck fleet size. A discrete event simulation model with 3D animation has been developed to estimate the productivity of a truck-shovel network system, shovel efficiency, truck cycle time, truck utilisation and the optimal fleet size. The simulation model considers the dynamic variability of traffic conditions and also the truck as an individual entity. The resulting truck cycle time was found to be realistic and thus providing a better tool for strategic mine planning.Citation:Zeng, W, Baafi, E, Walker, D and Cai, D, 2016. A 3D Discrete Event Simulation Model of a Truck-shovel Network System, in Proceedings Ninth AusIMM Open Pit Operators' Conference 2016, pp 265–273 (The Australasian Institute of Mining and Metallurgy: Melbourne)."

Nov 15, 2016

By P Clark

This paper discusses the challenges, solutions and lessons learnt as tunnelling work continues on the Melbourne Main Sewer Replacement (MMSR) project.The MMSR project is a $220 million dollar project for the Melbourne Water Corporation. It involves the construction of approximately 4.2 km of new sewer from the southern fringe of the CBD through the inner metropolitan suburbs of Port Melbourne and South Melbourne.Of the 4.2 km of sewer currently in construction approximately 2.2 km will be constructed using a 2.95 m diameter Lovat EPB tunnel boring machine (TBM) with approximately 1.9 km of small diameter (300 to 600 mm) reticulation sewers constructed using trenchless techniques.These sewers are being constructed in a dense urban environment, which means that the tunnels and shafts are often in close proximity to buildings and structures. The underground constructionof the sewer is further complicated by the variable geology of the ancient Yarra River delta which includes weathered basalts, the infamous Coode island silt formation and fine loose sands.Some of the key challenges and lessons learnt so far during construction include:· steering and alignment control of the TBM through very soft silt with little residual strength,· development and refinement of a monitoring and alarm system for leaving the machine dormant on a day works only project,· control and measurement of tunnel lining defl ection during ring build and grouting,· prediction and monitoring of surface settlements during tunnel excavation,· improved understanding of the geology and hydrogeology at this previously unchartered area of Melbourne,· consolidation settlement due to dewatering at shafts required careful construction controls, and· ground support and excavation in close proximity to the existing 110 year old sewer which requiredvery detailed risk control measures.

Mar 8, 2011

"Recognising that the mining industry is wedded to the break-even cut-off grade rather than a cut-off grade that optimises value, the author examines some perils and pitfalls in the use of break-even grades that lead to what the author calls ‘negatively geared ore reserves’.Negatively geared ore reserves occur when a portion of the ore reserves actually has a negative value and is effectively being subsidised by the more profitable portion of the ore reserves. This situation is generally difficult to detect with the level of information available to the public investor. The ore reserves can still be quite profitable in total, but now possess a level of risk that may not be appreciated.The risk profile (financial risk) of a negatively geared ore reserve is far greater than for a non-negatively geared ore reserve. Like a negatively geared investment property in a falling property market (for readers familiar with the Australian tax minimisation investment strategy), when the price of a commodity falls, a company that possesses a negatively geared ore reserve loses value at a magnified rate relative to a company that has a neutral or positively geared ore reserve. In the extreme case of a major commodity price reduction, as happened with the gold price in mid-2013, the negatively geared ore reserve can easily become negative value in total, not just in part.The negatively geared ore reserve is sometimes intentional – for instance in order to increase published ore reserve tonnages and meet the CEO’s key performance indicators – but is also often unintentional, resulting from various errors in the break-even calculation. Common errors in calculating a break-even grade include using fixed metallurgical recoveries, not including allowances for ideal laboratory versus real plant results, ignoring available operational data and using modal rather than mean parameter values in cost and performance calculations.This paper examines a common value distribution profile in many ore reserves and the effect of common break-even grade errors on this profile.CITATION:Poniewierski, J, 2016. Negatively geared ore reserves – a major peril of the break-even cut-off grade, in Proceedings Project Evaluation 2016, pp 236–247 (The Australasian Institute of Mining and Metallurgy: Melbourne)."

Mar 8, 2016

By Travis G. A

The Mons Cupri Cu-Pb-Zn deposit is situated within a complex Archaean volcanic sequence dominated by rhyolitic fragmental rocks. This sequence is underlain by a 150 m group of rhyolitic ash flow tuffs with minor conglomerate and sandstone, and is overlain by approximately 180 m of dacite and andesite flows. The Mons Cupri sequence itself consists of several elongate or circular rhyolite domes, flanked by rhyolitic fragmental rocks ranging up to 400 m in thickness away from the domes. The lithology of clasts (typically 1 cm to 1.5 m in size) provides strong evidence that the flanking fragmental unit was derived by explosive eruptions associated with the rhyolite domes. An approximately 15 m thick shale horizon, bleached and silicified at its base, overlies both the rhyolite domes and the frag- mental unit. A south-dipping chloritic pipe-like body, with near-surface dimensions of 1,000 by 250 m and narrowing at depth to produce a cone-shaped cross section, lies within the rhyolitic frag- mental rocks about 250 m north of the largest rhyolite'dome in the Mons Cupri vicinity. Its east west elongation parallels that of the rhyolite dome. Ghost fragments indicate that the material in this fissure is a severely chloritized derivative of the rhyolitic frag- mental unit. The chloritic pipe is overlain

Jan 1, 1973

By B Jones, M Chappell

Permanent rock bolts used to support rock tunnels are required to perform for the design life of the structure as a minimum. Quality of installation is critical to the performance over the life of the rock bolt.This paper will present the findings from an in investigation into the integrity of rock bolts installed in an underground hydroelectric power station cavern and access tunnel. The 25?mm diameter rock bolts were installed 50 years ago, tensioned and post cement grouted. Due to a major machine upgrade the whole facility was subject to an asset survey. The owners are responsible for managing the risks associated with the failure of these rock bolts and ensuring the structural integrity of the tunnels and caverns is maintained to continue operations, protect machinery, assure production and more importantly maintain a safe working environment for all employees and visitors.To help manage risk a proactive approach is desired. To this end a test schedule using a non-destructive testing (NDT) system was carried out by Independent Integrity Testing Pty Ltd (i2t). The system used in this application was RokTel™. The system operates by applying a small axial impact to the rock bolt end and measuring the resultant acceleration response. The rock bolt characteristics affect the acceleration response.The data output from the system is analysed principally using Fourier Analysis (FFT) and sinusoidal waves. The energy generated in the bar, the rate of damping and changes in frequency are all assessed. Testing was carried out on 70 rock bolts within the tunnel and cavern. A selection were then over cored, photographed and documented. The data gathered from the testing is then correlated with physical evidence to establish a calibration for the remainder of the rock bolts and assess if this is a viable method for identifying defective elements, and what degree of accuracy can be achieved. The findings will assist with the future assessment of similar hard rock underground?facilities.CITATION:Jones, B and Chappell, M, 2014. The use of non-destructive testing to identify defective rock bolts – a case study, in Proceedings 15th Australasian Tunnelling Conference 2014 , pp 275–280 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Sep 17, 2014