Search Documents

By C. W. Owings

When explosions have occurred in mines where rock dust has been used to some extent, generally to a very limited extent, the press has immediately characterized those mines as "rock-dusted." To persons familiar with the method of rock-dusting, it is apparent that if rock-dusting is done properly, if all mine surfaces are rock-dusted to within at least 40 feet of the faces of rooms and entries, and if the incombustible content of the dust is more than 65 per cent, there is little likelihood that an explosion of coal-dust will occur. The report that an explosion has occurred in a "rock-dusted" mine immediately conveys to the mind of the alert safety advocate the information that the mine is inadequately rock-dusted or that either dust sampling of the rock-dusted surfaces has not been done or, if done, the mine has not been redusted with sufficient promptness. Rules for Collection of Samples Collection of dust samples is by no means intricate, but a few simple rules must be kept in mind if the results are to be dependable. A sample must represent the true average conditions of the dust in the section or zone in which it is taken, and therefore 'the point at which the sample is to be collected must be selected with care. During an explosion much if not all of the dust on the ribs, roof, timber, and road in the affected region is blown into the air; the sample, therefore, should contain all loose dust and material on: the surfaces to a depth of about 1 inch, if that amount is present. If there is considerable sand locally along the track, care must be taken not to include enough of this material to vitiate the sale and distort the results; however, if the sand is distributed throughout the region being sampled it should be included. Dust should not be collected from recesses in the ribs, such as hitches cut for timber, as the character of the dust in such places is often different from that on the ribs proper. Samples should not be collected at points where the dusting machine has been loaded or at a point where the machine has stopped and discharged for a longer period than at other points on the entry or in the rooms. In short, the sample must represent AVERAGE CONDITIONS, and if there is any doubt as to the character of the dust it is far better to have too much combustible matter than too much incombustible matter in the sample, as it is best to err on the side of safety if it is necessary to err.

Jan 1, 1929

By K. K. Kelley

"In an earlier consideration of the vapor pressure of iron 3/ it was found necessary to estimate the specific heat of gaseous iron to write satisfactory free-energy-of-vaporization equations for the several forms of iron. That the specific heat of gaseous iron will exceed greatly the classical value of Cp = 4.967 calories per gram-mole per degree is evident from even a cursory Lamination of its spectrum. The two lowest energy states are separated by only 416 cm-1, which is small enough so that the second state is effective, theoretically, at below room temperature. Such phenomena are by no means uncommon, but, while they have been recognized with substances that ordinarily are encountered as gases, classical gaseous specific--heat values still are used, often without question, for the vapors of metals and other elements that are solids under ordinary conditions. The following calculations for iron and copper gases are presented to illustrate the sort of deviations from classical values that are encountered with metal gases.The methods employed in calculating from spectroscopic data of thermodynamic quantities such as specific heat, entropy, and free energy have been described, in detail by Giauque 4/ and others. These will be briefly reviewed here. It is customary to separate the translational energy of a molecule from the rotational, vibrational, and electronic energies. In the case of monatomic gases, only translational and electronic Energies are involved. The classical expression for the specific heat of a mona¬tomic gas, Cp = 4.967, deals only with energy stored in the molecule as translational energy and consequently is valid only in temperature ranges in which no appreciable number of molecules undergoes a change in electronic state."

May 1, 1937

Aluminum alloys are used in two broad classifications: wrought and cast. Wrought alloy products, which are formed into final shape by rolling, forging, extruding. etc., contain low percentages of alloying elements; whereas cast aluminum products, which generally are not worked before use, may contain higher percentages and a somewhat different selection of alloying elements. For example, wrought aluminum alloys usually contain less than one percent silicon and cast alloys contain as much as 18 per-cent silicon. While there are good techniques available for separating aluminum alloys from mixed scrap, there are no efficient ways to separate wrought aluminum from cast aluminum. The high levels of silicon, iron) copper, and zinc from the cast alloys make wrought-cast mixtures unacceptable for wrought alloy production. Consequently, separated wrought aluminum scrap commands a premium price on the market. Since increasing amounts of mixed aluminum alloys are being recovered from automobile scrap, this separation is important for efficient recycling of these alloys. Approach The difference in thermo-mechanical properties between the two classes of aluminum alloys is used as the basis for the separation process. Casting alloys lose their ductility within a certain temperature range while wrought alloys remain ductile. As temperature increases, aluminum alloys gradually become weaker, and at a certain temperature, each alloy loses strength and ductility due to incipient melting in localized areas. For casting alloys that temperature generally lies between 520° to 560° C, but it is above 600° C for most wrought alloys. Mild crushing, therefore, in the proper temperature range will result in complete fragmentation of the casting alloys but only plastic deformation of the wrought alloys. A simple screening step then is used to separate the oversize wrought pieces from the fragmented cast alloys.

Jan 1, 1984

By Donald W. Mitchell, John Nagy, Edwin M. Murphy

Information is presented on remote sealing of a mine passageway through a borehole by five methods : PPnneeuummaattiicc injection of mineral wool and sand, ureaformaldehyde foam , balloons , urethane foam, and caving by blasting . An adaptation of the method using pneumatically conveyed mineral wool was tried in an operating mine . Further research is needed to simplify and to improve the techniques involved , as well as to explore the possibilities of other methods and combinations .

Jan 1, 1964

By David R. Hanson

Experimental work performed by the Bureau of Mines in a laboratory environment has shown that the formation of failure zones within certain rock types is accompanied by the emission of significant amounts of radiofrequency (RF) electromagnetic (EM) energy. This radiation was detected using nonresonant, "electrically short," broad-band antennas. Rock types for which emission was observed include granite, quartzite, and taconite; sandstone and marble did not emit measurable radiation. Amplitude spectra of the radiation showed the energy to be concentrated between 10 and 40 kilohertz. These data indicate that emission is directional, but this has not yet been proven. In addition, amplitude of RF emission increases with increasing crack size. Since emission was observed only for brittle quartz-bearing rocks, it appears the formation of piezoelectric fields is a necessary condition for RF radiation. Plausible mechanisms for emission include rapid decay of piezoelectric fields accompanying the sudden stress release at failure and/or the acceleration of an "exoelectron plasma" through the intense local piezoelectric fields. Since emission appears to increase with the scale of failure, and since antennas do not need to be coupled directly to rock surfaces as with conventional geophones, the possibility of developing a portable system to monitor seismically active areas exists.

Jan 1, 1981

By R. V. Ramani, R. Stefanko, G. W. Luxbacher

Comprehensive ventilation surveys are made to review the adequacy of the mine system with regard to both statutory requirements and/or company policy, to evaluate improvements to the current ventilation system and to plan for £stun expansion. If computer simulation is to be used for these purposes, accurate quantity and pressure measurements become an absolute necessity since the validity of the computer generated output is heavily dependent on the accuracy of the input data. Consequently ventilation surveys that are run to generate input data for computer simu¬lation must use the best data collection procedures to guarantee as accurate results as are possible. These requirements have led to modifications in procedures that are used for routine ventilation surveying. McElroy and Kingery (1957) have recommended that pressures be measured independent of quantities since minor changes in mine conditions have a greater effect on absolute pressure measurements than on quantity measure¬ments. They estimate that, for large mines, a pressure survey may take only 1 to 3 days while the corresponding quantity survey would require 1 to 3 weeks. Although this procedure would undoubtedly lead to more ac¬curate pressure measurements, its use for the development of computer input data is not recommended for the following reasons. Atkinson's equation, H = RQ2, is used in computer oriented analysis to develop the airway resistance, R, based on quantity and pressure loss measure¬ments. It can easily be seen that since R a 1/Q2 and R a H, the effect of errors in quantity measurement is, in fact, more significant than those in pressure measurement. If the quantity and pressure measure¬ments are made at different times another problem results. Minor changes in quantity that may occur in the time between the pressure and quantity

Jan 1, 1977

By S. H. Katz

"At the annual meeting of the International Association of Fire Engineers at Kansas City, Mo., in June, 1919, the Bureau of Mines offered to cooperate with municipal fire departments for the purpose of obtaining information regarding composition of the air in burning buildings. (Katz, S. H., Statement made at Convention of International Assoc. of Fire Engineers: Fire and Water Eng., vol. 66, July 16, 1919, pp. 129-30, also Fireman's Herald, vol. 78, July 12, 1919, pp. 48.) The gas-mask laboratory of the Bureau's experiment station at Pittsburgh sent vacuum bottles for taking samples to fire chiefs in twenty-five cities, who had signified their intention to aid in this work. It was hoped to obtain one hundred or more samples, because statistics in quantity are necessary for drawing reliable conclusions concern-in the likelihoods of encountering irrespirable as in burning buildings. This far, eleven samples have been received by the laboratory. Results of analyses and pertinent data regarding these samples are given in the accompanyingEight of the eleven samples contained, no carbon monoxide; seven of the eight were taken above the ground level and one below ground. One sample taken above ground contained 0.03 per cent carbon monoxide, and this quantity is not dangerous to breathe for a limited time. Of the three samples taken below ground, one contained no carbon monoxide; the other two contained 0.10 and, 0.35 per cent, respectively. This indicates that in a burning building, the hazard from carbon monoxide is much greater below the ground level. The cause undoubtedly lies in the poorer ventilation that is usually found in cellars or basements. Abundance of ventilation provides the necessary air for complete combustion, or if incomplete combustion does occur, the carbon monoxide generated by the fire is greatly diluted with fresh air and swept away by the current. It may be noted that the highest proportion of carbon monoxide was found in a cellar where a fire as burning among piled cases. The difficult access of air to the burning parts facilitated the formation and lib ration of carbon monoxide. The smoke here was very light and no difficulty was experienced in breathing. Some of the air containing dense smoke showed little change from normal air by analysis. Thus the density of smoke or the irritation and difficulty it causes in breathing are no indication of the presence of poisonous gas. None of the samples shows important oxygen."

Nov 1, 1920

By Louis A. Panek

This report deals with the combined friction and suspension effects in the reinforcement of horizontally bedded mine roof by bolting. Other reports have treated these two effects separately. Friction effect refers to the reduction of bending in a laminated roof due to the clamping action of tensioned bolts. If the beds tend to have unequal deflections, the bolts oppose this tendency, thus creating a suspension effect, which consists of a transfer of part of the weight of weaker or thinner beds to the flexurally more rigid beds within the bolted unit. The combined effects of friction and suspension on the bending of the beds, whether an increase or a decrease, are difficult to generalize because they depend on the bedding sequence and the bolt pre-tension. Some analysis of the bedding sequence usually is required to determine the critical bed within a set of bolted beds. However, if the suspension effect is substantial and if the bolts are pre-tensioned, then the greatest bending stress is likely to occur in the thickest bed. Findings are based partly on theory and partly on tests of mine-roof models in a centrifuge, which were found to verify the theory. Mathematical formulas are obtained for the change in maximum roof bending stress and for the change of roof deflection due to the combined friction and suspension effects. The maximum bending stress is needed for designing a roof-bolting system and the deflection for evaluating its effectiveness.

Jan 1, 1962

By T. J. Driscoll

In this Bureau of Mines investigation, the electrochemical polarization behavior of Fe and Fe-1BCr has been characterized in rigorously controlled 1[N] H2S04 solutions stirred with oxygen or an inert gas. Samples that under-went various levels of corrosion or passivation were examined using Auger electron spectroscopy (AES) and ion sputtering. Passivation of iron occurred at a potential of approximately 0.6 volt (1) in poorly stirred, oxygen-saturated solutions for forward polarization (scanning from cathodic to anodic potentials); (2) in pH-adjusted (increased), well-stirred, oxygen-saturated solutions for forward polarization; and (3) in well~ and poorly stirred, oxygen-and helium-saturated solutions for reverse polarization. Passivation of Fe-18Cr occurred at a potential of approximately -0.1 volt in helium-, nitrogen-, and oxygen-saturated solutions for all cases of forward polarization and most cases of reverse polarization. Elemental depth profiles obtained from AES sputtering measurements indicated that films or adsorbed species were present on iron samples in solution subjected to open-circuit, active, and transpassive corrosion, cathodic reduction, and passivation. This conclusion was reached by considering the distribution of the solution species, primarily sulfur, in the films, which were composed primarily of iron and oxygen. Sulfur was present throughout the films on actively corroding samples, in the outer regions of films of passive and transpassive samples, and on the film surfaces of open-circuit and cathodically reduced samples. For Fe-18Cr, sulfur was found on the surfaces of films on passive samples only, indicating that only the passive samples had films on their surfaces in solution. The films on electropolished and passive Fe-1BCr samples had the largest chromium concentrations, estimated to be approximately 45 at. pct, with less chromium in films of cathodically reduced, active, and open-circuit samples, in that order.

Jan 1, 1979

By J. S. Tosh

An equilibrium study of the system potassium carbonate, potassium bicarbonate, carbon dioxide, and water has been conducted with solutions of 20-, 30-, and 40-percent equivalent potassium carboaate concentrations. These equilibrium data cover the operating range of the hot carbonate scrubbing system for removing carbon dioxide from gas mixtures. Such data are necessary to establish the limits of purification and are used to design absorption and regeneration towers. Values of K in the relationship K = (KHCO3)2/(K2CO3)PCO2 were determined for higher temperatures of 70" to 140° C. and greater concentrations, I 20 to 40 percent, than employed by previous investigators. For the 20- and 30-percent concentrations, K is constant at a given temperature; thus, once K has been determined the equilibrium pressure of carbon dioxide can be calculated for any conversion. Variation of & values with changing conversion was observed with the 40-percent solutions, so that equilibrium calculated pressures are less reliable from average K values.

Jan 1, 1959

By K. Kar, M. Gautam, J. T. Jurewicz

The 10-mm Dorr-Oliver cyclone has been widely used as a personal sampler for size selective sampling of particles within the respirable range (Dp < 10 gm). Because of their inertia, the particles are unable to follow the swirling flow (outer vortex) within the cyclone. The particles migrate to the bottom of the cyclone, depending on their size, and get trapped into the inner vortex which leads to an eventual blow-off from the top of the cyclone. By measuring the size distribution of the particles in the free stream and in the collected sample, the aspiration efficiency of the cyclone can be calculated. Velocity difference between the free stream and sampling velocity. Inlet orientation, particle impaction and deposition, particle inertia and free stream turbulence result in the departure of the aspiration efficiency from unity and this, in turn, leads to inlet bias. Although the effect of inlet bias on sampling efficiency has been investigated in past, the results lack detailed information and show only some general trend. In this effort, a systematic, in-depth experimental work on the effect of inlet bias on the sampling efficiency has been undertaken using non-intrusive Laser Doppler Anemometry (LDA), Aerodynamic Particle Sizer (APS) and impactors.

Mar 1, 1992

By Harry G. Dosh

Sillimanite, kyanite, and andalusite are aluminum silicates having the same chemical composition (Al2SiO5) and are used in the ceramics industry for the manufacture of high-grade refractories. Natural sillimanite with added corundum or other alumina to just short of mullite (3Al2O3 -2SiO2) proportions is unequalled for jobs wherein strength at high temperatures, resistance to spalling, and resistance to slagging by acid slags are desired. High-grade porcelain ware also requires these minerals. At the same cost, sillimanite would be preferred for most uses, but the lack of any large source of sillimanite bas given economic preference to kyanite or andalusite for special refractories and high-grade porcelain wares. Sillimanite-bearing schists were observed in Spartanburg County, S. C.; as early as 1937 by D. Hoye Eargle, geologist with the U. S. Soil Conservation Service. The mineral was not identified until 1942, when Miss Jewell Glass, of the U. S. Geological Survey, examined several of the deposits. This identification was confirmed in 1943 by Dr. Laurence L. Smith, of the University of South Carolina, when he examined the Spartanburg County deposits and found numerous additional occurrences of sillimanite schist in Greenville County. Subsequent work by the Bureau of Mines and the South Carolina Geological Survey has been responsible for the location of additional sillimanite-rich schist deposits in Spartanburg County. Between October 1, 1948, and November 21, 1948, the Bureau of Mines conducted an exploratory diamond-drilling project on the John Gideon property in the southeastern section of Spartanburg County (fig. 1) and drilled a total of 681.7 feet in six holes. Drilling indicated that the deposit is relatively shallow, as no ore was cut below the zone of weathering.

Jan 1, 1950

By R. V. Ramani, V. T. Burgos, J. A. McClay

Introduction The Cost program was developed to provide the planning engineer with a tool to evaluate the cost implications of the technical mine design decisions. Direct costing methods are employed, where feasible, to provide greater detail and accuracy in the primary cost centers. Options have been provided in the cost computation, economic analysis, and output data subsystems to increase the flexibility of the system. Computer Program The program is written in Fortran IV for use in the IBM System 360/370 computers. i. Main - The main routine reads and stores input data from three sources, user input, Cost Data Library, and data transfer from the Master Design Simulator. Data is trans¬ferred to the appropriate subroutines, as they are called, by means of labeled common blocks. ii. Subroutine CSTCOM - is a data modification routine which computes the peripheral charges related to labor and supervisory costs. iii. Subroutine MANPWR - interrogates section data to determine the total labor and supervisory costs for each cost center. iv. Subroutine CAPCST - computes the total capital investment for each cost center, including the economic life and salvage value of each element. Capital costs are determined by user selection of quantities and models of equipment. Equipment costs are stored in the Cost Data Library for retriev-

Jan 1, 1975

By Bruce I. Thomas

In 1943, the Bureau of Mines investigated the occurrence of cassiterite in the gravels of Morelock Creek which flows into the Yukon River 32 miles east of the confluence of the Yukon and Tanana Rivers in central Alaska. Samples of the gravels were obtained by churn drilling test pitting, shaft sinking, and open-cut mining. The gravels along Morelock Creek are about 6 feet in depth and covered with about 2-1/2 feet of overburden. The overburden is frozen, but the underlying gravels are not. Cassiterite, with gold and other heavy minerals, is usually concentrated in the first 2 feet of&apos; gravel above bedrock and in some places penetrates into joint planes in the bedrock. The various methods of obtaining samples are discussed, and drill logs and analyses are tabulated in this report.

Jan 1, 1948

By Isaac A. Zlochower, Kenneth L. Cashdollar

The Pittsburgh Research Laboratory of the National Institute for Occupational Safety and Health (NIOSH) conducted a study of the explosibility of various metals and other elemental dusts, with a focus on the experimental explosion temperatures. The data are useful for understanding the basics of dust cloud combustion, as well as for evaluating explosion hazards in the minerals and metals processing industries. The dusts studied included boron, carbon, magnesium, aluminum, silicon, sulfur, titanium, chromium, iron, nickel, copper, zinc, niobium, molybdenum, tin, hafnium, tantalum, tungsten, and lead. The dusts were chosen to cover a wide range of physical properties-from the more volatile materials such as magnesium, aluminum, sulfur, and zinc to the highly "refractory" elements such as carbon, niobium, molybdenum, tantalum, and tungsten. These flammability studies were conducted in a 20-L chamber, using strong pyrotechnic ignitors. A unique multiwavelength infrared pyrometer was used to measure the temperatures. For the elemental dusts studied, all ignited and burned as air-dispersed dust clouds except for nickel, copper, molybdenum, and lead. The measured maximum explosion temperatures ranged from "-1550K for tin and tungsten powders to "-2800K for aluminum, magnesium, and titanium powders. The measured temperatures are compared to the calculated, adiabatic flame temperatures.

By James R. Coldwell

Mining and processing cost analyses were conducted by the U.S. Bureau of Mines on stratiform zinc-lead-silver, coal, and stratiform barite deposit types that are found in the Colville Mining District. Reserves and gross revenues which would allow these deposits to be minable were modeled. A 15% Discounted cash-flow rate-of-return economic threshold was selected as the minimum acceptable return on investment for the three deposit types. The economic modeling indicated combinations of grades and tonnages for stratiform sulfide deposits that could be economically viable for orebodies ranging in size from 25 to 130 million metric tons. The gross revenues for a surface mine range from $200/mt for a 5,000 mtpd operation using the proposed South corridor to $140/mt for a 17,000 mtpd mine using the proposed North corridor. The economic modeling indicated underground coal mining was subeconomic for deposits ranging in size from 42 to 250 million metric tons. The required gross revenues ranged from $300/mt for a 2,200 mtpd mine to $21 D/mt for a 13,000 mtpd mine. The economic modeling indicated supplying projected in-state barite demand was subeconomic for the three seasonal surface mines modeled. The required gross revenues ranged from $600/mt for a 90,000 mtpy mine to $840/mt for a 6,000 mtpy mine.

Jan 1, 1995

By Spencer P. Hwell, Charles E. Munroe

"Among the surplus explosives left on band with the War Department was a quantity of that styled Trojan Grenade Powder, which was especially designed during the war for use as a bursting charge in grenades and for which naturally there would be very little military use in time of peace. With a view to its disposal through use in civil projects a quantity was sent to the Bureau of Minas explosives experiment station, where analysis showed that its characteristic explosive ingredient was nitrostarcn and that the powder bad the following composition:Trojan Grenade Powder appears as a dark gray granular substance, part of which is slightly pulverulent. The granules fairly hard. The explosive is not at all dusty, and has the consistency of granulated sugar.Nitrostarch is a substance produced by treating starch with nitric and sulphuric acids and in appearance it looks very little different from powdered starch. It resembles nitrocellulose and nitroglycerin in its constitution and explosive properties and has been longer known than either, it having been prepared by Braconnet in 1833. Nitrostarch explosives have in recent years been developed for industrial use in this country, they having been used to an extent in blasting in the Canal Zone, yet as this grenade powder was a special composition it seemed desirable that a complete series of tests should be run upon it. This was done, and the results are reported in the following table in comparison with those previously obtained in testing ammonia dynamite."

Jun 1, 1920

By Manuel Gomez

Areal distributions for ash, total sulfur, and sulfur forms were developed for the top 12 inches, the middle section, the bottom 8 inches, and the total seam of Pittsburgh coal in southwestern Pennsylvania. The sulfur and ash exhibit an apparent high local variability. As a general rule, both the sulfur and ash show high concentrations in the top 12 inches and in the bottom 8 inches of the seam relative to the concentrations observed in the middle section of the seam. The high-low sulfur and ash concentrations in the top 12 inches of the seam are not coincident with the high-low sulfur and ash concentrations in either the middle section or the bottom 8 inches of the seam. With respect to sulfur and ash, it is suggested that the top and bottom sections of the Pittsburgh coal may reflect both a syngenetic and epigenetic origin. Present data support the belief that paleotopography controlled Pittsburgh seam deposition. Evidence is presented to show that channels were cut at various times into the semilithified sediments that buried the Pittsburgh seam. In this channeling process, portions of the Pittsburgh coal may have been eroded away. Following channel development, the lower part of the Pittsburgh sandstone filled the complex system of braided channels to produce the "want areas" encountered in present mining operations. In localized areas of the Pittsburgh seam, impingement of Pittsburgh sandstone may be responsible for the lateral displacement of the coal. Information of this type becomes important in the assessment of exploration data, in the development of a quality control system for mine output, and in mine life planning.

Jan 1, 1974

By John M. Devine

The oxygen bomb-is used extensively for determining sulphur in petroleum, although it has been suspected that the results obtained with this method are somewhat lower than the true varies. In this paper are compared the results given by the oxygen-bomb and Carius methods, the latter being generally accepted as the standard procedure for the determination of sulphur in organic compounds. The Carius determinations given hero were made in connection with a special study of this method.4 Only petroleum oils of the less-volatile type are covered in the present work. Summary and Conclusions. When used with oils containing 2.3 to 0.15 per cent of sulphur the oxygen-bomb method gives results that are consistenly if slightly lower than those obtained with the Carius method. This variation, for the oils used, amounts to 3.1 per cent (on tae basis of sulphur present) for the sample containing 2.47 per cent of sulphur, and gradually increases (as the percentage of sulphur in the oils decreases) to 7.3 per cent for the oil containing 0.15 per cent of sulphur. With the oils studied the presence of ammonium nitrate in the bomb does not influence the results materially.

Jan 1, 1928

By Frank E. Cash

The iron-ore ranges in the Lake Superior district are the liarquette in Michigan, Menominee and Gogebic in Michigan end Wisconsin, and the Vermilion, Mesabi, and Cuvuna in Minnesota. These ranges produce about 85 percent of the iron ore mined in the United States. Accident experience in connection with this production is better than that of the regions producing the remaining 15 percent. Production and shipments vary slightly from year to year but closely approximate each other over a period of years. During the 4-year period 1940 to 1943, inclusive, 323,667,599 long tons of iron ore was shipped from the Lake Superior district.4/ Michigan and Wisconsin shipped 65,386,608 tons, or 20.2 percent of the total, and employed approximately 15,000 men; Minnesota shipped 258,280,991 tons, or 79.8 percent of the total, and employed approximately 10,000 men. The tons produced per man employed is much, greater in Minnesota than in Michigan and Wisconsin owing to the large open-cut production in Minnesota.

Jan 1, 1945