Search Documents

By Kenshi Kawabata, Mitsuji Ueda, Junko Umeda, Katsuyoshi Kondoh, Kantarou Kaneko

Solid-state recycling process of in-house scraps of machined magnesium (Mg) alloy wastes has been established to directly produce wrought materials via the repetitively plastic deformation by Roll-compaction (RCP) and the conventional hot extrusion. The coarse chips, wasted in machining AMX602 (Mg-6Al-0.2Mn-2Ca) cast billets used for the gate bar used in Electronic Toll Collection (ETC) of the high-way system, were employed as the input materials in this study. RCP process enabled to serve coarse Mg alloy powder consisting of refined grains and intermetallics with no re-melting. They were easily consolidated by hot extrusion, and their wrought alloys showed both homogeneously refined grains less than 1~2 µm by dynamic recrystallization and randomized texture by repetitive plastic deformation on the input wastes. These materials revealed extremely advanced mechanical properties of high tensile strength of 356MPa and high ductility of 14.8%, compared to the originally virgin materials with 270MPa TS and 5% elongation. When applying the same Mg alloy wastes via RCP process to the mass production in the company, the large-scale extruded materials also indicated the superior tensile properties. It means that this direct solid-state process is held as “Up-grade recycle of metal wastes”.

Jan 1, 2008

Designing rigging systems for complex castings has been more of an art than a science. Foundrymen have used thumb rules based on past experience to develop working designs. Of late, attempts are being made to formalize the procedure through the use of rule based expert systems, process modeling and other tools. Designing risers is one of the challenges of the rigging system design. Once the riser location have been determined, an optimization technique coupled with a casting simulation code can be used to determine the optimum riser dimensions. The main drawback of using a finite element or a finite difference based process simulation code along with the optimization routine is that the computation time involved is too large. Such an approach, thus, is unrealistic for practical situations where different options have to be explored and conclusions drawn in a short period of time. This difficulty has been avoided in the present work. The optimization scheme has been coupled with a modulus based solution routine to optimize the riser dimensions. The modulus calculations provide a good and quick estimate of the freezing time at different points in the casting. A functional, obtained from these freezing time values, is minimized by altering the dimensions of the riser until an optimum shape is obtained. Some examples have been included to illustrate the utility of this technique.

Jan 1, 1994

By Atinuke Oladoye, David Esezobor

"Particulate Copper Silicon carbide (Cu-SiCp) composites find application as wear and heat resistant materials in electrical sliding contacts such as in homopolar machines, railway overhead current collector systems where high electrical/thermal conductivity combined with good wear properties is required. However, challenges occur during machining due to the presence of hard reinforcement in the matrix. This may lead to high turnover of tool wear and poor surface finish. The adoption of near-net shape technology to produce such hard-to-machine metal matrix composites and subsequent finish machining with its attendant cost has reported limited success. This paper critically appraises the challenges and opportunities in the improvement of tribological characteristics of particulate Cu-SiC composites and identifies low cost reinforcement material that could be used to improve its tribological characteristics.IntroductionTribology consideration is increasingly becoming a decisive factor in the design and selection of materials for applications involving interacting surfaces in relative motion. The consideration is mainly on the combined effects of friction and lubrication on material's wear integrity and performance amongst other. Annual losses due to tribology in the US industries are estimated at between 1-2% of the Gross National Income [1], in Russia it is about 1-4% of the GDP [2]. Thus, the attributes of tribology has become imperative in modem machinery especially those applications involving sliding and rolling surfaces, and electrical contacts [3- 8].These are readily found in homopolar machines, internal combustion engines, aircraft engines, gears, cams, seals and railway overhead current collector systems [9-12]. In these systems, the dissipation of frictional heat arising from the interacting surfaces as quickly as possible imposes constraints on the choice of materials for such application to prevent softening, plastic and shear flow and the transfer of metal as wear debris onto the interacting surfaces. If these possibilities occur, system performance is compromised and premature failure may result. Materials for these systems are expected to possess high electrical and thermal conductivities, low coefficient of thermal expansion, good corrosion resistance and high melting point [3-8, 13]."

Jan 1, 2011

By Dian Adisty, Leiv Kolbeinsen, Eli Ringdalen

In Si and FeSi production, the main Si source is SiO2, in the form of quartz. Reactions with SiO2 generate SiO-gas that further reacts with SiC to Si. During heating, quartz will transform to other SiO2 modifications with cristobalite as the stable high temperature phase. Transformation to cristobalite has been investigated and shown to be a slow process where the rate varies between the investigated quartz types. The effect of quartz versus cristobalite as SiO2 source has been studied for reaction (1) 2SiO2 + SiC = 3SiO + CO, by heating of pellets with SiO2 + SiC in Ar-gas. The reaction was faster with cristobalite than with quartz. Initial experiments indicates that this is the case only in Ar and not in CO-atmosphere The higher specific surface area of cristobalite is believed to contribute to the higher rate. The industrial consequences of the observed difference between quartz types are discussed.

Jan 1, 2014

By Bao Xiancheng

Jinlong's copper flash smelting process started up in April 1997. In the early days of production, lots of equipment and process problems were encountered. But, through continuous improvements in equipment and operation, the designed productivity was reached two years later and the smelting operation has been greatly upgraded since then. After more than 4 yew of production, we have made great achievements, such as increasing the flash furnace availability, significant improvements in operation control, optimization of concentrate burner's structure and operation, and smooth operation of the whole smelter. This paper outlines Jinlong's operational achievements made after commissioning, and discusses problems that were encountered and their countermeasures. Operational experience is also summarized

Jan 1, 2002

By NRMR Bhargava

With the increasing use of lead-free solder alloys, low melting point alloy is often required due to the heat- sensitive parts or substrates in modern electronic devices. For such demand of low melting point, the nanostructure lead- free solder alloy, having excellent metallurgical characteristics and a melting point of <215oC, were developed by mechanical attrition of conventional lead free solder alloys (Sn-Ag-Cu, Sn-Ag and Sn-Cu). The decrease in melting point due to nanosize effect was calculated theoretically as function of particle size and the theoretical results were compared with experimental values measured by differential scanning calorimetry. The newly developed alloys have eutectic composition of Sn-3.5Ag-0.7Cu, Sn-3.5Ag and Sn-0.7Cu and the melting point depression found to be 5.10C, 4.10C and 6.20C when the bulk alloys reduced to 80 nm, 92nm and 96nm sizes respectively.

Jan 1, 2009

By Guiming Yang, Xiaoxian Huang, Zhiyun Ji, Min Gan, Lishun Yuan, Xiaohui Fan, Zhiyuan Yu

"The roasting characteristics of hematite pellets containing MgO were studied where magnesia powder, serpentine and magnesite were used as magnesium-additives respectively. The mechanical strength of both preheated and roasted pellets decreased when MgO content increased, which implied the addition of MgO had an adverse effect on pellet consolidation. By adding flux containing CaO, the mechanical strength of both preheated and roasted pellets were improved. The addition of calcium flux facilitated the formation of calcium ferrite and provided a certain amount of liquid phase, resulting in faster diffusion ofMg2+ and Fe3+. The result showed that diffusing speed of Mg2+ was enhanced by 1.48 times, and rate constant of Fe203 crystalline growth was enhanced by 62% when basicity was around 0.4. Improvement of mineralization degree and recrystallization of Fe203 assisted in consolidation of hematite pellets containing MgO.IntroductionOxidized pellets are ideal iron-bearing burdens in blast furnace. At present, magnetite is the main raw material for oxidized pellet production in China. However, as the rapid development of iron and steel industry, resources of magnetite have been depleted and cannot satisfy the requirement of pellet production. Using hematite concentrates to produce oxidized pellets has become a trend [1, 2]. Many pellet enterprises abroad nowadays have adopted high proportions of hematite concentrates when producing oxidized pellets, e.g. the percentage is 100% in a lot of Brazilian factories [3, 4]. But high proportion of hematite concentrates will lead to low mechanical strength and high reduction swelling index of pellets. Some studies show that Mg-bearing pellets have excellent metallurgical properties, and the reduction swelling phenomenon can be solved by adding fluxes containing MgO [5, 7]. However, there are problems such as low pellet strength and bad roasting characteristics for MgO-bearing pellets, which remains an obstacle in pellet industry. Hence, it's vital to undertake the study on strengthening measures for indurations of Mg-bearing hematite pellets."

Jan 1, 2013

By Zhucheng Huang, Ronghai ZHONG, Tao Jiang, Jun ZOU

The method of reduction roasting and magnetic separation of low-grade, micro- fine disseminated refractory iron ore following granulation with coal produced unexpected success. Pellets of iron ore with diameter between 3 and 8mm were granulated with coal, bentonite and water. The raw pellets and reduction products were analyzed by scanning electron microscopy and x-ray diffraction. Results show uniformly distributed coal powders inside the pellets improved the kinetics of the reduction process and accelerated the reduction rate. Reduction time was higher at temperatures below 890℃, which resulted in the formation of substantial amounts of fayalite. When temperature was increased, micro- fine hematite was then quickly reduced to iron metal and the amount of fayalite decreased sharply. When pellets were reduced at 950℃ for 25min, reduction products with 56.73% metallization rate were obtained. After grinding and magnetic separation, a concentrate with 72.18% Fetot was obtained.

Jan 1, 2015

By J. S. Becker

Technology developed and first implemented in the United Kingdom enables very low grade aluminum scraps containing oils, organics, plastics and ferrous metals to be directly charged, without pretreatment, into a unique melting furnace. The system has been in full scale production for four years. The oxy-fuel-based furnace design incorporates internal afterburning in the furnace&apos;s hot zone, producing excellent thermal efficiency and very low emissions. The furnace is inherently low in NOx. This paper will provide both theoretical background and extensive operating results for this new melting concept. Applications describing both purpose built new furnaces and retrofits of existing reverberatory furnaces will be presented. Extensive operating emissions data will be presented.

Jan 1, 1997

By Kotaro Ogura

"The electrolysis process in which CO2 is selectively reduced to C2H4 at a three-phase(gas/solution/metal) interface on a copper-mesh electrode has been developed in a concentrated solution of potassium halide. At open-circuit potential, CO2 is captured on the electrode by being inserted into the double-layer consisting of halide anions specifically adsorbed and potassium cations which are located in the Helmholtz inner and outer layers, respectively. CO2 thus activated is subjected to the electrochemical reduction. In the early stage of reduction, CO2 is converted to formate radicals that are adsorbed on the electrode by taking the place of halide anions. The reductive coupling of these radicals results in the formation of intermediates such as -CH2C(O)- and CH2=CH-, and C2H4 is finally produced. The catalytic activity of the electrode process involving the insertion of CO2 into the double layer is not deteriorated at all during the electrolysis, and the continuous reduction of CO2 to C2H4 is possible with the addition of inorganic acid under the assistance of natural energy.Carbon dioxide is a serious pollutant released from fossil fuel power station, internal combustion and many industrial operations. The easy release of CO2 to the air has caused the global warming, and the technique which intends to curtail CO2 by pushing it back into the ground has been suggested. However, this problem should be greatly mitigated if CO2 can be converted to innocuous substances. Especially, the reuse of CO2 as fuels and chemicals is extremely effective not only for the mitigation of the global warming but also against the depletion of the fossil fuel reserves. Several processes for converting CO2 to valuable compounds have been proposed including hydrogenation over heterogeneous catalyst at high temperature and electrochemical and photochemical reduction. In the conversion and cutback processes of CO2, however, it is of importance that the process is feasible under mild conditions, because the secondary generation of CO2 is feared if high energy is required and supplied with a fossil fuel. In this meaning, the electrochemical reduction of CO2 taking place at room temperature, which is similar to photosynthesis, is promising. From the practical point of view, it is further necessary that the conversion rate is high enough and the product is satisfactorily valuable."

Jan 1, 2008

By Joo Hyun Park, Kyu Yeol Ko

The viscosity of the CaO-SiO2-xMnO-yCaF2 slags (C/S=1.0; x=10, 40%; y=0 to 15%) was measured to clarify the effect of CaF2 on the viscous flow of molten slags at high temperatures. Furthermore, the Raman spectra of the quenched glass samples were quantitatively analyzed to investigate the structural role of CaF2 in a depolymerization of silicate networks. The critical temperature of the slags abruptly increased at 15%CaF2, which was confirmed to originate from a crystallization of cuspidine using XRD analysis. The viscosity of the slags continuously decreased by CaF2 addition in the 10%MnO system, whereas the viscosity of the 40%MnO system was not significantly affected by CaF2 addition. The effect of CaF2 on the viscosity of the slags was quantitatively analyzed using micro-Raman spectra of quenched glass samples accompanying with a concept of silicate polymerization index, Q3/Q2 ratio. A polymerization index continuously decreased with increasing content of CaF2 in the 10%MnO system, whereas it was not affected by CaF2 in the 40%MnO system. Consequently, the bulk thermophysical property of the CaO-SiO2-MnO-CaF2 slags was quantitatively correlated to the structural information.

Jan 1, 2014

By A. Tsubouchi, F. Noguchi, T. Takasu

"A high current density operation is desired to improve the productivity in copper electrorefining. However, a lot of problems still remain such as passivation at anode, dendritic deposition at cathode and impurity behavior during the electro refining.Effects of the current densitY on the morphology of copper deposition on the cathode, and passivation at the anode were investigated at various fluid velocity of electrolyte in the present study.Although a granular appearance of the copper deposition was obtained with increasing high current density, it was depressed by an increase of the fluid velocity of the electrolyte. A uniform deposition on the cathode was also observed when the fluid velocity of the electrolyte become higher. It was ascertained that passivation at the anode was suppressed by circulating electrolyte during electrolysis. INTRODUCTIONIt is desirable to operate at high current density in copper electrorefining in order to improve plant productivity without enlarging the size of electrodes(1). However, there are many problems in copper electrorefining at high current density with the usual commercial practice, for example, anode passivation, dendritic deposition and an increase of impurities in the cathode.In general, it is effective to agitate the electrolyte in order to suppress these problems(2). Smith et a1.(3) have performed electro refining experiments at various current densities in the range of 8 to 32 Aldm2 and at electrolyte fluid velocities in the range of 40 to 120 mlmin, and they obtained well consolidated, void free copper even at high current density in a high electrolyte velocity. However, they did not discuss the recovery ratio and the passivation behaviour at the anode. Walsh et a1.(4)(5)(6) experimented with the rotating cylinder electrode, and ascertained that electrolysis can be carried out at high current densities under the condition of pure electrolyte.With regards to anode passivation, there are several papers by previous workers(7)(8). Our group ascertained the process of producing anode slime during electrolysis(9). However, these experiments were carried out without forced circulation of electrolyte. Therefore, in the present study, to solve these problems, we electrolyzed at current densities in the range of 4 to 50 Aldm2 with circulating electrolyte in the range of 0 to 48 mlrnin. The effect of electrolyte fluid velocity on the surface morphology of deposited copper at cathode and passivation at anode were investigated when electrolysis were carried out with circulating electrolyte at high fluid velocity and at high current density."

Jan 1, 1997

By Garth James

Bacterial consortia capable of degrading cyanide and metal-cyanide complexes were obtained from gold mine process waters using enrichment cultures. The cultures contained either cyanide, tetracyanonickelate, or ferrocyanide as a sole nitrogen source. Several bacteria were isolated from the consortia, but none of these organisms were capable of degrading cyanide or cyanide-metal complexes at rates equivalent to those of the intact consortia. The consortium from the cyanide enrichment was capable of completely degrading cyanide from an initial concentration of over 100 ppm in less than one day, when cultured using a fed-batch method. Tetracyanonickelate and ferrocyanide were degraded more slowly. Overall, these results demonstrate that bacteria capable of degrading cyanide and cyanide-metal complexes can be isolated from mine process solutions and are capable of significantly reducing cyanide concentrations under appropriate conditions.

Jan 1, 1998

By Hanlu Xie, Liang Yu, Ping Wang, Yuanchi Dong

"The leaching process is the first step when recycling BF slag as raw material for some other products. While studying the kinetics of the BF slag leaching process, a fractal phenomenon was found, which indicated some interesting mechanism in the process. The result suggested a new understanding of the leaching process in detail which useful for the industrial applicationIntroductionTypical composition (wt. %) of blast furnace slag varied in a pretty wide range: SiO2 30–40, Al2O3 10–20, CaO 30–40, MgO 5–10.The ground granulated slag is commonly used as base of roadway (mixed with natural inert for paving, soil stabilization, etc.) or one of the resources of Portland cements because of its good hydraulic activity [1]. To recycle the slag economically, the selective recovery of some components of the slag may be considered. As main components exist in the slag, silica [2, 3 and 4] and alumina [4] recovery was first considered.While recovering of silica from the slag, silica gels are recovered selectively from the leaching solution deriving from the treatment of the slag with waste sulfuric acid. After separation of the exhausted solid of leaching, the silica gel is precipitated from the solution at pH 3.2. Although the silica gels obtained by this procedure contain many impurities and need some more work afterward, it is the first and important step for the process.In the prior work [5], silica extraction using blast furnace slag had been investigated, focused on the pH value of selectively silica extraction. In the present work, kinetics of the leaching process was focused. It is important to design an economical process.ExperimentalThe composition of the slag and the calcined gels has been determined by a LINK860 (Oxford Instruments, Oak Ridge, TN) energy dispersive spectrometer (EDS) in conjunction with a CAMBRIDGE S250-M2 (Cambridge Instrument Ltd., Cambridge, U.K.) scanning electron microscope(SEM). The leaching of slag and the precipitation of silica gels have been conducted at atmospheric pressure in a stirred reactor placed at temperature controlled by a heated water bath."

Jan 1, 2008

By W. J. Guay

Scrubbing of simulated copper reverberatory furnace gases containing 1.0 percent sulfur dioxide with the minus 74 ~m fraction of the San Manuel tailings has shown sulfur dioxide removals of almost 100 percent over an effluent pH range of 3.1 to 4.6 and a wide range of liquid to gas ratios with absorptions of 11 to 17 Kg of sulfur dioxide per tonne of minus 74 um slimes. The high reactivity of the tailings slimes with S02 indicates that the reaction rate is not controlled by the slow dissolution of calcium carbonate in sulfurous acid which characterizes scrubbers using ground limestone. The interpretation of these results has been considerably enhanced by wet chemical analysis and emission spectrographic analysis which show that calcium and iron account for about 90 percent of the SO2 absorption with the rest being accounted for by aluminum, magnesium, potassium, and sodium. Mineralogic examination confirms the, chemical and spectrographic analyses showing that the major constituent removed from the slimes is calcium carbonate, with lesser amounts of chlorite and montmorillonite (silicates containing iron, magnesium, aluminum, and calcium) being removed.

Jan 1, 1981

By James C. Sever

Two green field magnesium reduction plants are under consideration using dolomite as the raw material. The Calix calciner was independently selected for both plants. The unique design of this calciner is detailed and the advantages defined for the specific application. The low carbon foot print of this unit combined with the ability to capture food grade carbon dioxide makes this an environmentally proactive approach to carboniferous ore calcination.

Jan 1, 2015

By J. D. Zhu, T. Ohmichi, H. Yasuda, S. Kashiwai, I. Ohnaka

"Since the casting process affects the properties of castings, its computer simulation has been greatly developed in the last decades. This paper presents our recent works, especially on mold filling simulation, because most casting defects such as porosities and inclusions that greatly affect the properties originate during mold filling. The first topic is the mold filling simulation with consideration of cavity pressure change during mold filling, including direct observation of flow phenomena with X-ray and pressure measurement. The second .one is the mold filling simulation in the lost foam process, in which gasification of polystyrene pattern and gas outflow through the coating and mold have been considered. The simulation can predict not only the mold filling sequences but also periodical vibration of the gas layer between the melt and pattern. The third one is the comparison of simulation of Al melt flow in a metallic runner and direct observation through a transparent mold, showing that the entrapped gas or inclusions is expelled ahead through the melt and its simulation needs very fine meshes. These examples show that the comparison of simulation and observation is very important to develop the numerical simulation further.IntroductionThe mechanical properties of castings are greatly affected by casting defects such as porosities and inclusions, and solidification structure. Although recently the structure simulation has greatly progressed [1-4 ], the application in plants is very limited, because the structure is mostly determined by chemical compositions and cooling rate, enabling us to estimate it from the structure of past similar castings without simulation.On the other hand, the casting defects often vary with each casting or time. Among the defects, porosities and inclusions are main causes to deteriorate the properties and mostly originate during mold filling. Therefore, a lot of efforts to simulate the mold filling and estimate those defects have been made in the last decades [l,5,6]. These efforts have been mostly to develop numerical techniques to solve the Navier-Stokes equation. However, the flow phenomena in molds have not been well understood, because the direct observation of the flow phenomena in mold is not easy. This means the accuracy of the simulation is still not clear."

Jan 1, 2000

By A. F. Kaczmarek

Cerro Negro is an undeveloped ore body located at the Cerro Verde mine near Arequipa, Peru. The mineable reserve contains 70 Mt of oxide ore grading 0.532% copper. Mineralization is hosted in three distinct rock types: tourmaline breccia, crackle breccia and granodiorite. Rock type effects copper recovery and acid consumption. The paper describes Cyprus&apos; testwork philosophy and the resulting four-phase metallurgical program used to define the commercial leaching parameters. Copper recovery and acid consumption were monitored as functions of crush size, heap height, irrigation rate and acidulation conditions. Single lift, on-off and permanent multi-lift heaps were considered. Leach cycle times were based on incremental copper extraction and acid consumption rates, with added constraints imposed by operational requirements. The optimum leach system involves an on-off pad using tertiary crushed ore acidulated with 20 kg/t acid and stacked to a height of 6.5 m. The active leach cycle is 82 days at an irrigation rate of 0.26 l/min/m2. Under these conditions copper recovery is expected to average 81% over the life of the mine.

Jan 1, 1999

By Lin Li, Wen Shi, Hua Wang

"Specimens of the ultra-low carbon bake hardening steel were soaked for 1 min to 5 min, and then cooled by water quenching. The internal friction experiments were carried out to detect the spectra of the specimens. Results show that Snoek peaks are present in all the specimens, however, an obvious Snoek-Koster peak is found only in the 5 min-annealed specimen. Besides, tensile results indicate that the bake hardening values increase with the annealing time prolonging from 1 min to 3 min, but decrease as the annealing time prolongs further to 5 min. The solute distributions in the specimens were characterized by three dimensional atom probe. Some segregated C and V are found in the 1 min-annealed specimen, and the C and V contents decrease as annealing time prolongs to 5 min.1. IntroductionAs an excellent member of automotive body panel materials, various types of bake hardening (BH) steels have been developed since 1980s. Bake hardening property, as one of the most important mechanical properties, has been studied by many researchers [ 1-5]. As is known, BH phenomenon is generally controlled by the interaction between solute C and dislocations. The Snoek peak of internal friction (IF) is very popular in the investigation of solute C [6-9]. However, result of internal friction cannot suggest the distribution pattern of solute C as well as its influence on BH phenomenon.Three dimensional atom probe (3DAP) can diagnose solute element atoms as well as nano-sized precipitates, and it is the only instrument that currently can identify the identities of individual atoms in a material and measure their spatial coordinates with the resolution close to 0.3 nm [10]. There have been a number of studies about using 3DAP to detect solute elements as well as nanometer sized precipitates [11-19]. For example, Hono et al. [16] used 3DAP to determine the solute elements in an age-hardened Mg-Ca-Zn alloy, and they also obtained the distribution patterns of the observed elements."

Jan 1, 2011

By Jr. Langhans, J. S. McPartland

"The presence of zinc in cyanide solutions is generally thought to not have an adverse effect on precious metal recovery or recovery rate by heap leach processing. However, analytical interferences of zinc in solution can cause problems in accurately determining and controlling cyanide concentration in the leach solution, leading to increased cyanide consumption. Analytical methods and the chemistry of cyanide in the presence of zinc are reviewed. The effects of solution pH and zinc concentration on the free cyanide determination were evaluated in detail. Use of the standard silver nitrate titration and a modified titration method for controlling cyanide concentration during heap leaching were investigated experimentally. The practical implications of cyanide heap leaching in the presence of cyanide soluble zinc minerals are discussed.IntroductionThe presence of zinc in precious metal cyanidation circuits is not uncommon. Zinc can be introduced into cyanide leach solutions through dissolution of various zinc minerals1>, or through the use of a Merrill Crowe (zinc precipitation) circuit. Zinc minerals typically occur infrequently and in low concentration in gold ores, but are more abundant in silver ores. Merrill Crowe systems are most commonly employed for recovery of precious metals from solution for relatively high silver ores. It is desirable to control the amount of dissolved zinc in a cyanidation circuit for several reasons. Excessive zinc build-up in cyanide leach solution can, under certain circumstances, inhibit precious metal dissolution and adversely affect solution recovery systems. Additionally, high zinc concentrations can lead to excessive in-process cyanide inventories and potentially, to higher cyanide consumption"

Jan 1, 1999