Search Documents

By S Hapugoda, M Peterson, A Haileslassie, A Poliakov, E Donskoi, L Lu

Most metallic iron in the world is produced in blast furnaces from different feed stocks – sinter, lump ore and pellets. Sinter is the main iron-containing feed to the blast furnace and can constitute up to 70–85 per cent of the total ferrous burden. As sinter quality is a key element for stable blast furnace operation, the textural, chemical and physical characteristics of the sinter are very important. Two major factors that affect sinter strength are mineralogy and porosity. This study utilised optical image analysis (OIA) to characterise sinter mineralogy, and the results were compared to those obtained by manual point counting under the optical microscope.Two major mineral characteristics of iron ore sinter are the presence of primary (unreacted) and secondary (derived from melt) hematite. Standard OIA, which is based on RGB thresholding, is not able to distinguish them as they have the same reflectivity; however, primary and secondary hematite in iron ore sinter have a different morphology. The CSIRO-developed OIA software Mineral4/Recognition4 has a unique feature to identify various textures/morphologies of the same mineral. Together with other capabilities of the Mineral4/Recognition4 software, this feature was used to identify different types of hematite and silico-ferrite of calcium and aluminium and other common sinter phases, such as magnetite, larnite, glass and remnant aluminosilicates.A visual comparison of the sinter mineralogy present in optical photomicrographs and that present in the mineral maps created automatically by the Mineral4 module of the CSIRO software confirmed the accuracy of the automated mineral identification.CITATION:Donskoi, E, Hapugoda, S, Lu, L, Poliakov, A, Peterson, M and Haileslassie, A, 2015. Advances in optical image analysis of iron ore sinter, in Proceedings Iron Ore 2015, pp 543–548 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Jul 13, 2015

By K Nakashima, K Aya, T Sumita, N Saito

Electro slag remelting (ESR) is one of the methods used to produce high-cleanliness steel. In the ESR method, the product steel electrode is immersed in flux, and the electrode is melted by the Joule heat of the flux generated by high current. CaF2-based melts are mainly used as fluxes, but the electrical conductivity of CaF2-based melts varies widely among researchers and measurement methods. Therefore, this experiment aims to measure the electrical conductivity of CaF2-based melts based on the van der Pauw-Ohta method, which does not require the cell constant that is a fundamental error factor in electrical conductivity measurement at high temperatures. The measurement temperature was set at 1450–1600°C, and dehydrated Ar flowed through the furnace at 200 mL/min. Electrodes were fabricated by Pt-30mass per cent Rh alloy, and the electrodes were connected to an impedance analyser using a four-terminal probe. The electrodes were immersed in the melt, and the impedance and phase angle were measured by AC frequency sweep (50 Hz to 200 kHz). The electrical conductivity was then calculated by utilising the impedance’s real part values when the impedance’s imaginary component was zero. The electrical conductivity showed a positive temperature dependence. The electrical conductivity of 60(CaO⋅Al2O3)-40CaF2 melt increased with increasing temperature but was found to be roughly one-third of those of the previous studies. In the conductivity measurement of 60(CaO⋅Al2O3)-40CaF2 melt, the AC frequency with a phase angle of 0° was found to be different depending on the temperature and immersion depth. By sweeping the frequency, the impedance at the phase angle of 0° can be obtained, which enables us to measure the electrical conductivity more accurately.

Jun 19, 2024

By N Voller, A Ishmurzin, G Hackl, M Wu, A Kharicha, M Chintinne, C M. G Rodrigues, A Ludwig

Slag fuming (SF) is a critical process for recycling zinc-containing slags, but the corrosive nature of molten slag poses challenges to the reactor durability. The freeze-lining (FL) technique offers a solution by forming a protective layer on the reactor wall. It requires intensive cooling using watercooled jackets, which can stabilise the FL while compromising the energy efficiency of the process. This study presents a computational fluid dynamic (CFD)-based model to optimise the SF process by considering FL formation and its impact on heat transfer and reactor wall temperature. A volumeof- fluid (VOF) model is coupled with a mixture continuum (MC) solidification model to capture the intricate multiphase flow dynamics within the SF furnace. Two FL types are considered: FL solidifying on the reactor wall in the slag bath region; and FL solidifying on the reactor wall in the freeboard region. The FL of the first type forms when the slag temperature drops below liquidus temperature. The FL of the second type only forms when a splash-induced slag droplet collides with the freeboard wall and solidifies. A series of splashing events are necessary to coat the freeboard wall. The simulation was run until a global energy balance was reached. This means that the heat losses from the water-cooled jacket, bottom wall, outlet and fuming balance the heat gains from the hot gas injected through the submerged plasma torches. The increase in FL thickness, due to its low thermal conductivity, reduces the heat losses through the reactor walls. The calculated FL thickness and heat fluxes were in good agreement with industrial data, validating the model’s credibility. The simulation results provided valuable insights into the fuming process, including slag bath temperature evolution, slag splashing dynamics, FL formation patterns, local heat fluxes through the reactor wall and overall energy balance. These findings can inform process optimisation strategies to enhance the energy efficiency and sustainability of SF operations. The authors have built a prior version of the model framework and applied it to simulate FL formation in an electric smelting furnace (ESF). The results from both the ESF and the current SF highlight the applicability of such model framework to a range of industrial processes involving FL formation. This model framework can ultimately contribute to more energy-efficient and sustainable industrial operations.

Aug 21, 2024

By Fejes G

An integral part of the Brisbane Suburban Railway Electrification involved the relining, refurbishment and track lowering in the Roma Street, Brunswick Street and Gloucester Street double track tunnels. The design and the construction procedures had to satisfy the Queensland Railways electrification clearance and track standards and the requirements of maintaining track services through the tunnels during work. Refurbishing of the tunnel linings was carried out to extend the life of the tunnels.

Jan 1, 1981

By F Brenguier, R A. Lynch, G Olivier

The seismic velocities of intact rock are dependent on the applied stress. Unfortunately, the low susceptibility of seismic velocity to stress changes mean that changes in seismic velocity have to be measured extremely accurately to infer stress changes from them. Previously, it was thought that measurements of this accuracy could only be made with active sources or repeating earthquakes. In this paper, we show that the noise generated by mining activity can be used to construct virtual seismic signals with which we can measure changes in seismic velocity accurately enough to see the short- and long-term effects of a production blast on the surrounding medium. The time-dependent velocity variations achieved with this method show that the blast caused weakening of the rock mass for the rock within a radius of 200 m of the blast. The weakening is healed by confining stresses over a period of five days, after which permanent changes in the seismic velocity can be seen. The size and spatial distribution of the seismic velocity changes were compared to modelled changes in static volumetric stress, which supported the notion that seismic velocity changes can be used to measure the redistribution of stress after a blast. The results of this study indicate that this method can be a powerful tool to evaluate the evolution of stress in a mine at a relatively low cost as the data that it uses would normally be discarded.CITATION:Olivier, G, Lynch, R A and Brenguier, F, 2016. Measuring the redistribution of stresses resulting from new excavations with ambient seismic noise correlations, in Proceedings Seventh International Conference and Exhibition on Mass Mining (MassMin 2016), pp 873–878 (The Australasian Institute of Mining and Metallurgy: Melbourne).

May 9, 2016

By G Varley

This paper describes the mining methods and strategy used to extract the 5 Orebody crown pillar at the Isa Lead Mine over the past five years. The bulk of the crown pillar was extracted by the use of sublevel open stope method. An uphole bench stope was mined to scavenge the remaining ore. Geomechanical issues associated with the crown pillar extraction are described. Stope sizes were gradually increased on a stope by stope basis. This has allowed the maximising of ore recovery while relevant risks were managed. The principles in designing ore rib pillars against various types of fill materials are also described.

Jan 1, 2002

By J Mauk, J Rowland

The Karangahake epithermal Au-Ag mine consists of two main lodes, the Maria and the Welcome/Crown hosted in Miocene andesites. Mineralisation is strongly constrained to major structures and is genetically linked to dynamic tectonic processes. Two main periods of faulting are represented in the deposit: 1. NE striking right-lateral/normal oblique linked spatially and temporally with the mineralising system, and 2. ESE striking right-lateral/normal oblique post-mineral faulting that displaces the lodes. Paleostress analyses performed on a limited kinematic data set indicate NW-SE extension during the formation of the vein system. Veins and lodes strike NNNE and accordingly contain a strike-slip component. The kinematics of vein formation at Karangahake is broadly consistent with that of other deposits within a NNE-trending belt of deposits. Oblique shear on deflecting structures will create zones of compression and dilation. Dilatant zones created on steep structures create fluid pathways potentially focussing large volumes of fluid from deep parts of the system into dynamic structures.

Jan 1, 2003

By D K, S Freij, A Rossiter, X Han, B Feng

In mining operations, the removal of overburden to expose the ores is achieved by rock blasting using ammonium nitrate based explosives, eg ammonium nitrate-fuel oil mixtures, commonly known as ANFO. However, ammonium nitrate can react with pyritic shales that are often found in many mines, leading to premature detonation or spontaneous explosion. In order to minimise the occurrence of spontaneous explosion, the reactivity of the shale in a mine must be assessed. This will provide a better understanding of blasting risks and thus will improve the safety and the cost efficiency of blasting through the selection of appropriate explosives. In this study, shale samples were collected from different locations in Western AustraliaÆs Mt Whaleback iron ore mine and geochemically characterised. The reactivity of the shale samples towards ammonium nitrate and ANFO was investigated using a combination of large-scale firing range tests and laboratory isothermal reactor tests. Results of the firing range tests show that the reactive ingredients of the shale are the limiting factor for the reactions between shale and ANFO. Based on the firing range test results, further isothermal reactor tests were conducted to investigate the reactivity of shale samples towards ammonium nitrate under controlled conditions. Analysis of the results from the firing range tests and the isothermal reactor tests shows that there are correlations between the reactivity of shale samples and several geochemical characteristics such as sulfate, sulfur, ferrous and zinc content.

Jan 1, 2005

By D Van Doorselaere

The purpose of blasting during mining is to produce a muck pile that can be efficiently loaded and transported. There is an increasing awareness of the variation in muck pile characteristics and its economic implications. It is well known that good fragmentation by itself is not sufficient to optimise the muck loading efficiency. The swelling and configuration of the muck pile also significantly affect the productivity of the loading equipment. The characteristics of a muck pile depend upon the blast design parameters, explosive properties and rock mass geology. This study has focused on the dependence of muck pile configuration on blast design parameters, explosive properties and size distribution of the muck.The field work was done in two phases. In the first phase 12 blasts were monitored, while in the second, three blasts were monitored. The blast design parameters and explosive properties for each blast were documented and the following muck pile characteristics were measured:muck pile height and its distance from the facedistance of throwwidth of the power troughsize distribution by image analysis technique.The effects of burden, bench height, heave energy of explosive, row-to-row delay, powder factor and the mean particle size of the fragments on muck pile configuration were examined. It was observed that muck pile height and distance of throw are strongly influenced by the burden, bench height, heave energy of the explosive and the size distribution of the muck. The width of the power trough depends upon the row-to-row delay, powder factor and total burden. In addition, the effect of the fragment size distribution on the muck pile parameters has also been discussed.CITATION:Singh, S P and Van Doorselaere, D, 2015. The relationship between blasting parameters and muck pile configuration, in Proceedings 11th International Symposium on Rock Fragmentation by Blasting, pp 369–374 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Aug 24, 2015

By M Scamardella

The CSA Mine is located at Cobar, New South Wales, where it was in continuous operation from 1966 to the time it went into receivership on 29 January 1998. When it closed it was under the ownership of Golden Shamrock. Glencore International, in conjunction with Mt Lyell Mining, undertook a due diligence study to determine the feasibility to re-open the mine. This envisaged operating CSA Mine as a low-cost, copper concentrate producer under the direction of Cobar Management Pty Ltd (CMPL). The new workforce would consist of contractor trades people in operating functions under the co-ordination of CMPL staff, both on the surface and underground. This was the key to establishing the new culture in the workforce. The orebodies at CSA Mine are steeply dipping with copper-lead-zinc mineralisation in zones of strong alterations in a fine-grained siltstone host sequence. The predominant economic minerals are chalcopyrite, sphalerite and galena. The major gauge minerals are pyrrhotite, pyrite, chlorite, sericite and quartz. Ore for the project is being developed and mined from two high grade copper lenses within the QTS North system. Chalcopyrite occurs in two forms. The first consists of coarse chalcopyrite grains in association with pyrrhotite or pyrite. The second form consists of chalcopyrite grains intergrown with pyrrhotite grains. This form of chalcopyrite is quite rare. It allows electrochemical interactions between grains, resulting in rapid oxidation of the chalcopyrite surface. Physically this ore looks like, and has been mistaken for æpeacock oreÆ, ie bornite, but is infact oxidised chalcopyrite.

Jan 1, 2000

By K A. Heasley

The LaModel program was initially developed in 1993, and it is used to model the stresses and displacements in thin tabular deposits such as coal seams. It uses the displacement-discontinuity variation of the boundary-element method with a simplified laminated overburden model and because of this formulation, it is able to quickly and effectively analyse large areas of single- or multiple-seam mines. Over the last couple of decades, LaModel has been used by many mine engineers and researchers for improving pillar designs and mine safety, and it has been continually updated to add functionality and improve its ease of use and accuracy. The present LaModel can calculate seam displacements and overburden, multiple-seam and total seam stresses, pillar safety factors, energy values, roof-bending stresses, fault effects and surface subsidence. It can automatically generate the mine and overburden grids from AutoCAD drawings, and it includes wizards for automatically generating Bieniawski strength coal pillars, easily calibrating the overburden properties to observed abutment stresses and calibrating the gob material to expected gob loading. This paper will present the formulation and features of the modern LaModel program and demonstrate its utility with a multiple-seam case study from the rugged topography in Southern Appalachia.CITATION:Heasley, K A, 2014. LaModel – a program for coalmine modelling, in Proceedings AusRock 2014: Third Australasian Ground Control in Mining Conference , pp 409–414 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Nov 5, 2014

By T R. Justice

The Eastern Distributor tunnel is located approximately 24 m beneath the suburbs of Woolloomooloo and Darlinghurst to the east of the central business district of Sydney. The 1.7 km long tunnel will carry three lanes of northbound traffic over three lanes of southbound traffic in a æpiggybackÆ arrangement, with a maximum cross-section 22 m wide by 12 m high.The tunnel is constructed predominantly within horizontally bedded Hawkesbury Sandstone. In the unweathered to slightly weathered state, the rock is typically medium strength (average UCS about 24 MPa), with an average quartz content of 72 per cent.Major geological features that intercept the tunnel route are the Woolloomooloo Fault Zone and two extremely to moderately weathered dolerite dykes. The fault zone intersects the tunnel at about 25¦ to its alignment and comprises a series of discrete faults over a width of approximately 50 m. The overall movement of the fault zone appears to be reverse. The two dykes are the 1 m wide Oxford St Dyke and the 6 m wide Great Sydney Dyke.This paper outlines the site investigations and the interpreted geotechnical model for the tunnel route prior to the commencement of tunnelling. This model is then compared to the encountered conditions. The very good correlation between the expected and the encountered conditions shows the high quality of geotechnical investigation and interpretation.

Jan 1, 1999

By R Yong, A McEwan, D G. Miljak, D Milinkovic, A Curtain, P J. Coghill, R Stefulj, B Lovric, G Roberts

The CSIRO has pioneered the development and use of certain classes of magnetic resonance (MR) for mining applications. The MR technique is a radio frequency-based spectroscopy that provides highly quantitative measurement of selected ore minerals. MR has many inherent advantages for bulk sensing in mining applications, including deep sensing penetration into most bulk ores with the use of non-contact sensors that are robust and tolerant of demanding mine conditions. The highly discriminating nature of the technique also provides the very desirable attribute of robust and simple calibration, that can often be performed prior to sensor installation. Rapid sensing (within seconds) at very low grade (few hundred ppm Cu) has been demonstrated, as well as sensing at high tonnage rate. Besides sorting applications, MR may also be applied to preconcentration applications involving very large volume sensing including: (i) haul truck sensing using portal arrangements, where truck destination assignment may be dynamically determined based on the truck grade, (ii) in-pit gantries used for loading trucks that may similarly be equipped with sensors for truck assignment and (iii) deep mine bench sensing, recently progressed through a CRC ORE project, where local bench grade is determined at the few-metre lateral scale and at depth prior to blasting, to aid selective blasting and ore extraction. MR sensor prototypes have been developed for each of these applications and have been tested in laboratory or limited field conditions. To inform the value proposition for truck or bench sensing, initial analysis of truck assignment improvement based on an analysis of in-situ mine bench grade has been performed for a porphyry style deposit. The prototype sensors, preliminary results and future technology development and objectives will be described.

Nov 10, 2020

By MattelmSki M

Outokumpu Harjavalta Metals Oy has been producing nickel since 1962, based on conventional flash smelting technology followed by electrowinning of nickel. In the 1980s, a research project was initiated to develop a converterless direct smelting process to produce matte for further refining in the then existing refinery. The process was tested on a pilot plant scale at Outokumpu's research centre in Pori and was found to be viable. The process was named the Direct Outokumpu Nickel Smelting Process or the DON Process. At the turn of the 1990s, Outokumpu studied various options to develop its nickel business and the Harjavalta operations. As a result of this work, the DON Process was further developed to produce two mattes: low iron matte in the flash furnace to be leached in the existing leaching circuit and high iron matte in the electric furnace to be leached in the new leaching circuit. In Harjavalta's case, it was essential and important that the process could use the existing flash furnace and electric furnace without converters.

Jan 1, 1996

By S N SHEARD

The use of geophysics in mineral exploration in Australia, and overseas, has increased dramatically over this century. This paper shows that using simple equipment with clever interpretation that the foundations of modem geophysics were laid, using all the same principles. The plethora and diversity of mineral exploration geophysical techniques does not allow a full description of their evolution to be given in such a paper. To convey this evolution, however, the progress of two different geophysical methods, magnetics - a potential field (passive) method, and electromagnetics - an electrical (active) method, are described. Examples of magnetic survey data from the 1920s and more recent vehicle home and airborne surveys are given to show the increase in data quality through technological improvement. A set of four airborne electromagnetic data sets spanning three decades is used to indicate the improvement in electromagnetic techniques. It is argued, that considerable progress is still required in the electromagnetic field to increase its interpretability.

Jan 1, 1992

By Cooper W. C, Stokes M. A

Proterozoic gold deposits in the Pine Creek Geosyncline are at present receiving renewed attention with the opening of old mines and several new discoveries. This important polymetallic province contains occurrences of gold, base metals, tin, tungsten, tantalum, platinum, palladium and aluminium. These resources have been exploited over the past century with periods of major interest during the turn of the century and during the past 20 years. Economically viable resources of predominantly gold, eg Pine Creek, Woolwonga, Brocks Creek and Union Reefs, and Ag-Pb-Zn, eg Woodcutters are currently in development or being mined. A variety of genetic models, ranging from magmatic through hydrothermal to syngenetic, have been postulated in the past for the formation of gold deposits in the Pine Creek Geosyncline. The first discussion on the regional geological factors that control ore deposition in the Brocks Creek-Howley District was by Sullivan and Iten (1952). They recognised that the most important ore deposits are associated with anticlinal and domal structures and that granites may be a controlling factor in the localisation of the mineralisation. More recently a group of authors, Needham and Roarty (1980), Goulevitch (1980), Nicholson and Eupene (1984), Sanger-Von Oepen, Friedrich and Kater, (1988), Nicholson and Eupene (1990) and Kruse, Whitehead and Mulder (1990) have suggested an exhalitive syngenetic origin, or remobilised syngenetic during dolerite intrusion and/or regional deformation, for gold mineralisation, especially those associated with South Alligator Group lithologies (eg Bridge Creek, Mt Bonnie group of mines and Cosmopolitan Howley). Studies carried out at the Bridge Creek and Western Arm Prospects over the last three years suggest that syngenetic models for gold mineralisation are less relevant to gold mineralisation in the Howley District than first thought (Wall, 1989; Partington, 1990). The Bridge Creek prospect has been cited as an example of a syngenetic-syngenetic remobilised deposit (Oepen, Friedrich and Kater, 1988) while the Western Arm prospect is more clearly a vein hosted epigenetic deposit and a comparison of the styles of mineralisation at both prospects will be made. The nature of the gold mineralisation at both prospects will be described and compared, and possible genetic models for gold mineralisation and their relevance to regional exploration discussed.

Jan 1, 1994

By M Manefield, G Bournival, S Ata

Asteroids contain an abundance of valuable resources such as: nickel, cobalt, gold, platinum and rhodium. Whilst the consumption of mineral resources on Earth continues to deplete, meteoroids have the potential to become a significant source of minerals. Even though the minerals content of asteroids may be relatively high, the valuable minerals are more likely to be only a small portion of the total ore present in the main body. The ores must, therefore, be processed to extract the metals of interest from the waste. Existing terrestrial technologies rely on manipulating the physical and chemical properties of minerals. The application of these techniques to asteroids processing, however, is not straightforward due to the presence of different conditions such as zero gravity and lack of water in space. This paper discusses potential separation and extraction techniques that can be used for processing of asteroids.CITATION:Ata, S, Bournival, G and Manefield, M, 2015. Resource recovery in space, in Proceedings Third International Future Mining Conference, pp 275–280 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Nov 4, 2015

By C Rizos, A Ng

Land surface deformation can be caused by underground mining activities. The maximum subsidence normally occurs directly above the location of the excavation of the coal seam. In the case where the longwall mining technique is applied, the magnitude and shape of the subsidence bowl are constrained by the longwall structure (pillar). This paper describes the use of satellite radar interferometry for monitoring surface subsidence due to underground mining, as a cost-effective geodetic technique and complementary to conventional ground surveying techniques. This paper uses data acquired by the current generation of SAR satellites û the European ENVISAT and the Japanese ALOS û to monitor ground subsidence in the southern highland coalfield in New South Wales, Australia. An accumulated DInSAR subsidence mine map is produced using six successive ALOS acquisitions. The results are post-analysed with the aid of geographic information systems (GIS) and validated using ground survey data.

Jan 1, 2008

By P Kumar, J Safarian

Understanding how ore interacts with flux particles at elevated temperatures to create molten slag is crucial since it governs the dynamics of a chemical reaction. This study explores the smelting behaviour of pre-reduced Nchwaning manganese ore when combined with lime, with the objective of examining the evolving interaction between pre-reduced ore particles and lime over time. The research sheds light on the interaction between solid and liquid and the phases that emerge during this process. To achieve this, a sessile drop furnace was employed to rapidly heat the materials positioned adjacent to each other on an alumina substrate and to observe the smelting process as it unfolded over time. This method allowed for the direct observation of the melting temperatures and the flux-ore reaction progression rate, and the potential disruptive events that might occur. By comparing the molten interfaces of the fluxed materials at various time intervals, this study provides insights into the relative rate of slag formation from the two materials. The results indicate that the main slag formation initiated at approximately 1400°C and continued to advance with time, with complete mixing occurring around 1500°C. The possible phases formed were identified using Scanning Electron Microscopy and modelled using Fact Sage thermodynamic software. In addition, the iron particles in the pre-reduced Mn ore were separated and settled from a rich MnO-containing slag. It was found that the separation of molten iron droplets from the slag depends on the rate of solid MnO particles dissolution into the adjacent slag phase.

Jun 19, 2024

By S K. Michelic, N Preisser

For certain applications, steels have to feature a very distinct and well-defined level of cleanness in terms of the number and composition of non-metallic inclusions. Therefore, one crucial aspect of steelmaking is removing and controlling said non-metallic inclusions during the process. Particle removal during ladle refining comprises three stages, namely: 1. Flotation 2. Transport through the steel/slag interface 3. Dissolution of the particle in the slag. Only if all three stages are completed the particle can be considered as fully removed from the steel melt which, in turn, leads to the slag being one of the most influential parameters for inclusion removal and control. Depending on the composition of the particles and the slag, the dissolution can be controlled by different mechanisms. This study focuses on the diffusion-controlled dissolution of solid oxidic particles in slags. For this, a model of particle dissolution has been developed. For modelling diffusion, either the differential equation following Fick’s second law can be solved, or an analytical solution to this equation valid for stationary interfaces can be used. Both approaches allow for calculations of concentration profiles of the dissolving species in surrounding slag. From this, the mass flux can be derived, which finally leads to calculating the boundary layer velocity due to considerations of mass balance. In this work, both methods as well as results of dissolution experiments using High Temperature Confocal Scanning Laser Microscopy are compared. This research aims to provide a framework for diffusion-based dissolution modelling in further expanding the understanding of particle dissolution in secondary metallurgy.

Aug 21, 2024