Search Documents

By Ashis Bhattacherjee, Rahul Upadhyay, Nearkasen Chau, Aditya Kumar Patra, Dhanjee Kumar Chaudhary

Heavy earth-moving machineries used in open-pit mines intensively expose the drill operators to whole-body vibration and shocks and thus to musculoskeletal disorders. The roles of an individual, machine, and rock-related factors in their exposure remain poorly understood. This cross-sectional study investigated their role among 39 drill operators from Indian iron ore mines. It shows that 70% of the operators were exposed to high levels of vibration, above the limit values recommended by ISO 2631-1. Multiple linear regression models showed that the whole-body vibration and shock exposures were strongly associated with operator’s age, drill machine model, rock’s hardness, uniaxial compressive strength, and density. The role of body mass index was close to significance (p = 0.08). Univariate analyses found that drill’s age, seat pad thickness, and seat backrest height were also strong predictors. These results help in identifying risky operators, materials and occupational situations, and in implementing appropriate prevention and intervention to reduce and monitor the exposures and health risk.

By M. R. Shahan, W. R. Reed, S. Klima, T. W. Beck

"Testing was conducted to determine the ability of foam to maintain roof coverage in a simulated longwall mining environment. Approximately 27% of the respirable coal mine dust can be attributed to longwall shield movement. Developing controls for this dust source has been difficult. The application of foam is a possible dust control method for this source. Laboratory testing of two foam agents was conducted to determine their ability of foam to adhere to a simulated longwall face roof surface. Two different foam generation methods were used; compressed air (CA) and blower air (BA). Using a new imaging technology, image processing and analysis utilizing ImageJ software produced quantifiable results of foam roof coverage. For CA foam in 3.3 m/s (3.3 m/s (650 fpm)) ventilation at 3 minutes after application, 98% of agent A was intact, while 95% of agent B was intact on the roof. At 30 minutes after application 94% of agent A was intact while only 20% of agent B remained. For BA in 3.3 m/s (3.3 m/s (650 fpm)) ventilation, the results were dependent upon nozzle type. Three different nozzles were tested. At 30 minutes after application 74-92% of foam agent A remained, while 3-50% of foam agent B remained. CA foam seems to remain intact for longer durations and is easier to apply than BA foam. However, more water drained from the foam when using CA foam which demonstrates that BA foam retains more water at the roof surface. Agent A seemed to be the better performer as far as roof application is concerned. This testing demonstrates that roof application of foam is feasible and is able withstand a typical face ventilation velocity, establishing this technique’s potential for longwall shield dust control. INTRODUCTION NIOSH personnel have visited many longwall operations at various geographical locations in the U.S., conducting benchmark surveys to characterize current operating practices and dust control measures in use. Sources of respirable dust generated from the longwall face have been identified as: intake (10%), stage loader (20%), shields (27%), and shearer (43%) [1]. Dust generated from the shearer, armored face conveyor, and the stage loader can mainly be controlled by water sprays and increases in ventilation air quantities, however, research has demonstrated limited effective controls for dusts generated by shield movements [2]. One possible approach to mitigating shield dust generation is to apply a layer of water or foam to the shield-roof interface. Once this interface is blanketed with foam or water, dust will be less likely to become airborne and contaminate the airways over the personnel walkways. Though water is an abundant and common dust control measure, roof applications have limited success because water is unable to remain in-place in adequate quantities for the necessary duration between application and shield advance. Past and current testing of foam for material properties has shown that it has longer applicability times, ranging from 10 to 60 minutes [3] and has shown better performance for dust control, ranging from 19-96% improved efficiency over water [4, 5, 6, 7, 8]. The majority of this foam dust control research has focused on dust control efficiency at the longwall shearer or continuous miner cutting drum."

Jan 1, 2017

By John G. Groppo, James C. Hower, Cortland F. Eble, Rick Q. Honaker, Alireza Valian, Stephen F. Greb

Coal and coal combustion byproducts are potential candidates for alternative resources of REEs (rare earth elements). The Illinois Basin is a major coal-producing district in the USA, but little work is available on speciation of REEs in the basin’s coals. In this study, a total number of 382 samples from known locations in the Illinois Basin coal beds were acquired, analyzed for their REE contents, and investigated for any spatial trends or chemical correlations. This study includes Springfield, Herrin, and Baker coal beds—the most heavily mined coals in the basin. An average coal ash from the basin contains 320 ppm REE, with Baker being the highest (407 ppm on average). However, the distribution of REEs in the coal beds was laterally and vertically heterogeneous. A number of trends were found in the REE variations. The most important trend was the relationships of the REE concentrations and fractionations with the distance from the syndepositional sandstones. The total REE concentrations were higher in the proximity of the sandstones, likely due to the increased deposition of detrital material. This increase was the most significant in Springfield coal (90%) and least in Baker (6%). Those coal samples from the Springfield and Herrin coal beds distal from the syndepositional sandstones had a heavy to light REE ratio (H/L ratio) of 0.40 on average, whereas in proximal locations, the ratio was dropped to an average of 0.30. This trend was not significant in the Baker coal bed. In general, higher mineral matter in the coals was associated with higher REEs. In coal preparation plants, refuse rocks can have up to 360 ppmREE. However, the ashes of the cleanest coals had higher REE levels. Such ashes also have a higher ratio of heavy to light REEs. The highest REE concentrations were found in the claystones and the shales associated with the coal beds. Since the naturally occurring calcite, quartz, and pyrite samples were depleted in REEs, an association of REEs with the clay minerals can be suggested.

Jul 20, 2020

By B. R. Benner

Progressive improvements in crusher design, screening efficiency, and operating experience along with the downsizing of many taconite operations has resulted in the abilitv of some crushing plants to produce a substantially finer product. This finer product opens the possibility of replacing the rods in the rod mill with balls as is the current practice at Sydvaranger, Norway. If this replacement could be accomplished substantial savings associated with metal wear and the elimination of rod charging may be realized. The purpose of this study was to compare the performance of a mill charged with rods with the performance of the same mill charged with balls at the pilot scale. The first objective was to establish whether we could achieve the same throughput rate at the same mesh-of-grind in a mill in which the rod media has been replaced by balls. The second point was to determine whether the ball media would create a significant tramp oversize problem.

Jan 1, 1990

By Fika Rofiek Mufakhir, Amru Daulay, Widi Astuti, Slamet Sumardi, Lukman Hakim Nasution, Hendra Prasetia

Rare earth elements (REEs) are utilized in numerous disciplines, including chemical engineering, the nuclear industry, metallurgy, medicine, electronics, and computer technology. Recycling products containing and extracting them from effluent is necessary to satisfy the rising demand for these elements. Some studies investigate the adsorption of rare earth elements from dilute aqueous solutions to remove them from effluent. Based on the results of this study, it can be concluded that lab tests demonstrate high adsorption capacities, which vary widely depending on the adsorption type and conditions. The Langmuir, Freundlich, and Temkin isotherms usually describe adsorption isotherms. In addition, the finest models for describing adsorption kinetics are pseudo-second-order and pseudo-first-order models. The thermodynamic parameters, such as the changes in free energy, enthalpy, and entropy, provide additional information regarding the energy changes. Additional research is required to develop environmentally friendly adsorbents that can be used to remove REEs from actual mine wastewater.

Apr 3, 2024

By A. Bakalarz, M. Duchnowska

"The copper sulfide ore from the Legnica-Glogow Copper Basin in Poland, which is processed by KGHM Polska Miedz S.A., consists of three lithological fractions: dolomitic, sandstone and shale. The copper ore, depending on the mining area, contains all fractions in different ratios. The lithological fractions differ in the particle-size distributions of their useful minerals. In the dolomitic part of the ore, most of the metal sulfide particles are 50-200 microns in size, while those in the sandstone and carbonaceous shale are 30-200 and 5-40 microns in size, respectively. This complex composition results in different mineralization, physicochemical properties and, in particular, upgrading selectivity. Due to the specific nature of the examined copper ore, two of the crucial parameters determining the beneficiation efficiency are size and liberation of sulfide particles. We analyzed the influence of grain size of feed on scavenging selectivity in one of the KGHM flotation plants and performed evaluation and analysis of the upgrading results of industrial beneficiation of the ore, containing mostly dolomite and shale. The upgrading efficiency was estimated using a selectivity indicator obtained by means of the Fuerstenau upgrading curve, which was then correlated with the particle-size distribution of the flotation feed.IntroductionFroth flotation is used extensively in commercial ore treatment flotation plants to separate minerals from gangue and to separate minerals from each other. The floatability of ores and their benefication selectivity are determined by many factors. One of them is particle size. The effect of particle size on flotation performance has been widely studied by many authors (Gaudin et al., 1931; Trahar, 1976; Trahar and Warren, 1976; Lynch et al., 1981; Pease et al., 2004). Particle size also plays a critical role in the probability of particle collision and attachment to bubbles (Solari and Gochin, 1992; Schulze, 1993), kinetic flotation (Taggart, 1956; Ahmed and Jameson, 1985; Debacher, 2002; Brozek and Mlynarczykowska, 2007; Potulska 2008), collector consumption (Sutherland and Wark, 1955; Vianna et al., 2003) and mechanical entrainment (Lynch et al., 1981; George et al., 2004; Konopacka, 2005). Most of the results have shown that there is an optimum particle size range for a given flotation system and that the simultaneous optimization of the recovery and concentrate grade of these particles in conventional flotation equipment is extremely difficult."

Jan 1, 2015

By John W. Straton

Following the completion of the paper presented in February, 1972, data were received from many other mines and some original respondents submitted more complete questionnaires. The computer program was changed to permit the results to reflect more clearly all the data reported by the mines. The original paper was revised to reflect the new information. There were significant changes in some numerical values, particularly where the size of the sample changed appreciably. Some limited samples considered originally in arriving at conclusions were not relied on in this revision. The revised tables and text are of the same design and format as the original paper.

Jan 1, 1972

By Setareh Afrouz, Emily Sarver, Cigdem Keles

The resurgence of severe lung disease among US coal miners since the mid-1990s has led to efforts to better characterize and monitor respirable dust exposures, particularly focusing on mineral constituents from rock strata surrounding the coal. A rapid analysis method for silica (quartz) mass using direct-on-filter Fourier transform infrared (FTIR) spectroscopy has been developed, which can also estimate kaolinite, a primary silicate mineral in many coal mines.

Feb 24, 2022

By Molood Saeedi, Nima Sadeghi, Eskandar Keshavarz Alamdari

This study showed the kinetics of gold chlorination leaching from an anode slime complex containing copper, silver, barium, selenium, and tellurium. AgCl and BaSO4 control the rate of gold leaching, and the interface transfer and diffusion across the product was selected as the desirable kinetic model. The concentration of hydrochloric acid, chlorine gas flow rate, pulp density, and temperature have been precisely studied. The results showed that temperature and hydrochloric acid have a favorable influence on the leaching rate. However, the chlorine gas flow rate is inactive in the reaction rate. Furthermore, the leaching rate decreased with an increasing pulp density. The leaching process was controlled by the interface transfer and diffusion across the product, and the activation energy was determined to be 13.66 kJ mol–1.

Oct 19, 2021

By Mahamuda Abu, John Kalimenze, Benatus Norbert Mvile

The central Dodoma region of Tanzania is one of the regions defined by the Tanzania Craton (TC). Although the Craton gold mineralization has been well explored and hosts most of the gold deposits in the country, the understanding of the mineralization potential of the Dodoma region is still an enigma. This study using trace elements geochemistry of soil samples aims basically at providing information for future exploration works in the region by delineating the possible mineralized zones, the pathfinder elements, and the lithological facies to target. Generally, there is an anomalous spatial distribution of a probable Cu-Au deposit in the N, NE, S, and SE parts of the region and are prospective areas worth exploring for gold from the Gaussian variogram models. The multivariate statistics shows that Cu, Pb, Zn, and Fe are the main pathfinder elements while searching for Au mineralization in the region. The gold deposit is most probably a Cu-Au porphyry type, although VMS type of Au deposit cannot be precluded in the area. The mineralization in the area is largely associated with tonalitic orthogenesis, tonalities, granodiorite, and granitic rocks with Cr-Ni-Co-V forming elemental association. This elemental association is similar to other gold bearing greenstone belts. The style of mineralization in the Dodoma region is similar to other gold deposits within the TC; hence, successful exploration methods in these ancient gold deposits like Geita and Kahama gold deposits can be adopted in the event of any further exploration exercise in the central Dodoma region.

Jul 12, 2021

By W. G. Pariseau

"The Crandall Canyon Mine (CCM) is located in the Wasatch coal field in central Utah. Extensive longwall mining at CCM preceded a decision to recover coal left in barrier pillars on each side of a set of main entries extending far into the mine. During mining in March 2007, a major bump occurred that led to cessation of mining in the barrier pillar on the north side of the mains and to the start of mining in the south barrier pillar. Subsequent loss of ground control in August 2007, through coal pillar bursts had tragic consequences with loss of six miners and shortly afterward three rescue team members. A U.S. Mine Safety and Health Administration (MSHA)report describes the events in detail (MSHA, 2008).An important lesson to be learned from CCM is the need to select a computer program suitable for pillar design. Computer programs are engineering tools and selecting the right tool for the job is essential. In this regard, minimum program requirements include capability for computing stress distribution within a pillar and adjacent roof and floor strata. The reason is simple: strength of rock is stress dependent; both vary from point to point in pillars and adjacent strata. Preferably, one should follow the evolution of stress as the mining plan is executed from start to finish to assure safety or to warn of a potentially hazardous situation developing as mining progresses.The approach to the problem posed by barrier pillar mining at CCM is conventional in the sense of applying well known finite element modeling techniques to analysis of a proposed mining plan. The engineering purpose is to evaluate the mining plan with respect to ground control questions including roof, floor and pillar safety. Results of previous two-dimensional finite element analysis indicated pillar safety was by no means assured (MSHA, 2008; Pariseau, 2008; Pariseau, 2011). Geometry of barrier pillar mining is limited in two dimensions for example, crosscuts are not seen in vertical cross sections. While three-dimensional analyses are preferred, even moderately sized regions often lead to very large problem sizes that exceed computer capabilities. However, a recent development in finite element modeling of tabular deposits allows for whole mine analysis in three dimensions. This technique involves a dual node - dual mesh concept. Both have been used separately in much earlier work, but are only now used in combination. It is possible to have an estimation of subsidence at a kilometer scale and details of stress, strain and displacement about barrier pillars, main entries and pillars, and so on at a meter scale (Pariseau, 2014). No compromise involving empirical formulas or similar guess work is required. Indeed, all calculations are based on first principles: physical laws, kinematics, and material laws. Effects of joints and in situ variability or uncertainty in strata properties, elastic moduli and strengths, may be included for greater realism. An important objective of the present study is to assess the role of three-dimensional pillar geometry in arriving at conclusions regarding barrier pillar mining versus two-dimensional results."

Jan 1, 2014

By R. C. Moe

Pelletizing studies were conducted to find ways of improving the metallurgical properties of acid and fluxed pellets. In acid pellets, melting temperature was controlled by the formation of low melting temperature fayalite during reduction. Formation of a permeable metallic shell during reduction allows diffusion of reducing gas to the pellet core, decreasing fayalite formation. With fluxed pellets, slag chemistry had the main influence on metallurgical properties. Mineralogical, chemical and structural transformations that occur during reduction were studied at various intervals for a number of Mesabi Range pellet types. Six pellet types were selected for further investigation based on significant differences in softening/melting characteristics. Retention of porosity in the metallic shell was found to be the overriding factor in determining softening/melting behavior.

Jan 1, 1990

By L. Sandbak

The Turquoise Ridge Joint Venture (TRJV) is located in northern central Nevada, and is situated within the Basin and Range province, near the northeast end of the Osgood Mountains (see Figure 1). The TRJV mine consists of typical Carlin type- gold mineralization characterized by low strength rock of altered limestones, siltstones, and mudstones. The weak nature of the ground facilitates the underhand cut and fill mining methods. TRJV has been working to optimize ground support and mechanize top cut drift sizes and eliminate operating hazards associated with jackleg drilling. Safety Factor analysis based on bolt strengths has allowed us to pursue larger sized top cut drifts and to evaluate bolt length sizes and spacing. Switching from jackleg sized drifts of 10?W x 10?H to a mechanized 12?H x12?W top cut drift design has the potential to decrease overbreak, provide more uniform drift design sizes and bolt patterns, and increase ore production by at least 38%.

Feb 23, 2014

By M. Shriwas, F. Calizaya

"Recirculation is prohibited in many coal mining countries because of the fear that the re-use of return air would allow the build-up of air contaminants at the workings. The incorrect design and location of a booster fan in any ventilation network can create unsafe condition due to recirculation. The current approach to investigating recirculation using simulation software requires manual effort which becomes tedious in a complex and a large network. An algorithm-based C++ program was designed to detect the recirculation in a booster fan ventilation networks. This program needed an input file prepared from output file generated by any ventilation simulator. This program created an output file for recirculation. This program demonstrated the strong capability to detect the recirculation in a sample network and a coal mine ventilation network. The outcomes of this program were documented in this paper. INTRODUCTION The proper design and location of booster fans eliminate the safety hazards associated with high main fan pressure. Conversely, the incorrect design of booster fans in any ventilation network can create unsafe conditions due to uncontrolled recirculation. The proper utilization of a booster fan derives benefits to reduce the main fan pressure, leakage, and consequently ventilation cost especially with high resistance network due to deep and extensive working area in a mine. However installation of booster fan may create uncontrolled recirculation, if it is not sized and located properly with respect to main fan. Recirculation is prohibited in many coal mining countries because of the fear that the reuse of return air would allow the build-up of air contaminants at the workings. Therefore, it is imperative to check the recirculation while designing and operating booster fan for a ventilation network. This paper deals with an algorithm based C++ program to detect the recirculation in a booster fan ventilation networks. This program was implemented to detect the recirculation in a sample network and a coal mine network. STATEMENT OF PROBLEM The currently available ventilation simulation programs do not give much information about recirculation for the generated solution (fan pressures). The current approach to investigating recirculation using simulation software requires manual effort, i.e., traversing from node to node in the output network by the simulator while keeping tracks of all the nodes that form recirculation paths. This is an easy task in a simple ventilation network, but it becomes tedious and sometimes very confusing in a complex and large network. Moreover, the presence of multiple recirculation paths makes the manual effort more complex and difficult, especially in separating multiple loops that are connected together. Thus, there is a need for a more efficient program that will detect and quantify the recirculation efficiently, for the generated fan pressures. A program is developed to detect the recirculation(s) in a network."

Jan 1, 2015

Operation of a pilot-scale Floatex density separator was studied using a pyrite concentrate slurry feed. Experimental measurements were further utilized in parameterization of a hindered settling simulation model. The model incorporates particle slip velocity calculation and also a separation efficiency function, adjusted to describe the feed material characteristics. The model was used to compute the overflow and underflow compositions when both fluidization bed pressure and teeter water flow rates were varied. In addition, prediction accuracy and the model validity were evaluated. The simulations gave good results when the teeter interstitial velocities were moderate enough to avoid excessive turbulence in the fluidization. Prediction of the cumulative overflow and underflow size distributions in most experiments yielded essentially less than 10% relative error between the sample data and the estimated value of the 50% passing sizes.

Jan 1, 2011

By Roberto Carranza

The planning of experiments and the importance of statistical interpretation of the results have been recognized for years by scientists and engineers. The Factorial Design of experiments is among the various methods used for this purpose and is being widely applied in all branches of research and production. The basic principles of this tool have been described in detail (1, 2, 3, 4) and several examples of its application in various fields have been published (5). In recent years the mineral industry has also used Factorial Design to improve the efficiency of research, testwork and plant operation (6, 7, 8, 9). This method permits for a qualitative determination of the influence of the variables studied in a process and also a qualitative measure of the interaction of the variables. By graphic means or derivative methods, one may also establish the optimum conditions of the process (ibid 7, 8, 9). It is desirable, however, to develop mathematical methods to evaluate quantitatively these effects and restrict the optimization to the practical limitations of the variables. It is also convenient to adopt a notation system to generalize the Factorial Design.

Jan 1, 1971

By Qiang Wang, Jiaqiang Yu, Chunhui Fan, Yong Wang, Changyu Jin, Jun Bai, Yu Lu

Accurate ore blending necessitates an in-depth understanding of movement trajectories of fragments and the spatial evolution of the displacement field in the bench during blasting, Accurate ore blending is currently one of the most complex challenges faced by mining engineers. In light of this problem, a systematic study was conducted in this paper using neural network and numerical simulation based on field monitoring data. A nonlinear mapping relationship was established between the mechanical parameters of rock mass and the displacements of the monitoring points. The measured displacements of monitoring points under blasting were input into the neural network, which used back-calculation to derive the mechanical properties of the rock mass. These parameters were then applied in a forward calculation to analyze the movement trajectory of the monitoring points. The results showed that the calculated displacements were in good agreement with the monitoring results during the. The average relative errors in the horizontal and vertical displacements were 4.12% and 15.2%, respectively. Furthermore, the muckpile profile was in an inverted recumbent “S” shape after bench blasting, which was consistent with the actual shape of the muckpile. This verified the rationality and accuracy of the method proposed in this study. Finally, the movement trajectories of particles in the bench and the spatial evolution law of the displacement field during bench blasting were determined based on the calculation results, offering theoretical support and fundamental information for accurate ore blending.

Dec 16, 2023

By P. D. Rao

Although phosphorus is a very important biogenic element, its concentration in coal is generally low. Phosphorus concentrations typically range from 0.001 % to 0.229% in raw coals of the contiguous 48 states. Some Alaskan coals in the Cook Inlet region contain horizons unusually high in phosphorus. A foot by foot analysis of deep drill core, containing 70 ft. of sub bituminous 'C' coal from the Beluga coal field, Alaska, showed certain increments contained as high as 17.03% P2O5 in the coal ash; 1.01 % P on a moisture-free coal basis. Crandallite has been identified by electron microprobe analyses as a phosphorus numeral in these coals. The authors could quantify calcium, barium, strontium and phosphorus concentrations from the coal attributable to crandallite. These high phosphorus coals provided a unique opportunity to study the association of phosphorus in coals using chemical fractionation for exchangeable cations and acid extractable elements. Since the source of high phosphorous horizons in Cook Inlet coals is attributed to volcanic ash, it is suggested that the profiles of acid extractable phosphorous concentration may be useful for coal seam correlations.

Jan 1, 1992

By Jeffrey A. Yonkey, Timothy J. Lutz, Jacob L. Carr, Christopher C. Jobes

"Since 1984, there have been 39 fatalities involving striking and pinning of the operator and other workers by the continuous mining machine (CMM). In January 2015, the U.S. Mine Safety and Health Administration (MSHA) finalized a regulation to require proximity detection systems on all continuous mining machines except full-face machines (MSHA, 2015). Proximity detection systems are designed to stop machine motion to protect miners from striking/pinning hazards. Several MSHAapproved proximity detection systems are commercially available.NIOSH researchers have been involved in the development and testing of proximity detection technology since creating the HASARD system (Schiffbauer, 2002) and have recently conducted research intended to develop more advanced proximity detection technology for CMMs (DuCarme et al. 2013; Jobes et al. 2012; Carr, DuCarme et al. 2013 and Carr, Li et al. 2013). In addition, NIOSH has also conducted research investigating how the signals of a proximity detection system propagate (Li, Carr et al. 2013 and Li, Jobes et al. 2012) and have developed a novel method for determining the location of a personal wearable device (PWD) relative to the magnetic field generators (Carr et al. 2010 and Jobes, et al. 2012).To evaluate the performance of proximity detection systems implemented in industry, NIOSH researchers conducted tests in 2013 on three systems installed on CMMs at three underground mines in Illinois and West Virginia. The results of these tests indicated that, while generally consistent performance was observed, there were opportunities for improving performance. These results were presented at the 2014 SME Annual Meeting (Carr, Lutz, et al. 2014). Since that time, second-generation systems have become available to the industry and are becoming more commonly used.Beginning in early 2014, NIOSH researchers conducted underground testing of these secondgeneration proximity detection systems installed on CMMs. Five tests were conducted on four installations of second-generation proximity detection systems, with one installation being tested twice. The tests were conducted at mines located in Illinois, Virginia and Kentucky with seam heights ranging from 137 cm (54 in.) to 213 cm (84 in.). These tests were conducted using the same test protocol used during first-generation testing, which is summarized in the following section."

Jan 1, 2014

The Cutoff grade optimization is very critical to the open pit production scheduling. In case of single mineral deposits, algorithms have been developed for determining cutoff grade policy that maximizes the net present value (NPV) of yearly cash flows. For deter-mining the cutoff grade policy for projects with two minerals including stockpiles the problem becomes much more complex. The purpose of this paper is to present fundamentals of a generalized cutoff grade optimization algorithm that will maximize the NPV of a project subject to mining, milling, and marketing constraints. Benefits of the methodology developed are demonstrated on three different case studies.

Jan 1, 1997