Document vB9LvrDNbxNYx2o532ZO7bJJR

FILE NAME: Zenith (ZEN) DATE: 1942 Mar DOC#: ZEN004 DOCUMENT DESCRIPTION: Journal Article - Journal of Scientific Instruments l i f f i l l wms A publication dealing with their principles, construction and use PHYSICS IN INDUSTRY jf e !.' t e -' [;4' <V > f f s ?>$';.? e of Physics with the co-operation of the N0V/29 1#I1 lal Physical Laboratory \,,,c. -^v UNIVERSITY OF CALIFORNIA Berkeley CONTENTS \ifK ; L. pOoDAtL, M.Sc., F.Inst.P. Note on an Approximate^ Methc |f e f 'yv tnove Spherical Aberration Laboratory and W orkshop N otes: %*% J. Di Graggs, M.Sc., A.Inst.P. erical Refracting Surfaces to reJ. L. H oughton, B.Sc., A.Ihst-P. G. J. T homas lew Instrum ents, M aterials and Tools; ' N.P-L. Type Micrometer ^ - ;v Mercury Diffusion Pumps. v: . i'l'lf Beat-Frequency Oscillator . A .: Fused Alumina Cements -.-s V Lastometer for Leather Testing P & T te in jsd ; , V- t New Books ' . '{ 1 .; Jf .'*" ` 'fify < . J j.O, >< Notes and Comments 'i C. EeHoMER, B.Sc., Ph.D. and H, c ! WATEKnts vv::' . :: 'O- Ir'- 1 ' .-` V V ' :1i-V.-K ? V**:., x- C wvr- a a ra s JO U R N AifeoFi^IEN TIFIC INSTRUMENTS In response to many requests fo ra 'slower plite than th e `Kodak' B.20 or B.10, t h e `IyOdak' Slow Spectrographic plate has been evolved to meet the following ?pecifi.catloni -l- 1 ` '>{ , (1) HIGH CONTRAST AND RESOLVING POWER. The `Kodak' Slow Speetrographic plate has J ' higher contrast and somewhat higher resolving power than the `Kodak' B.TO-k-hitherto ^ the speetrographic plate with maximum resolving power. / ' x ' (2) ' ROBUST EMULSION. The ` K odak' Slow Speetrographic^plate has a-toughened emulsion layer and is less susceptible to damage in handling than the *Kodak *B.10 plate. (3) RAPID PROCESSING CHARACTERISTICS. The `Kodak' Slow Speetrographic plate has pro cessing pharacteristies appioaqhing "in rapidity those which have helped to make the ()j /Kodak*5B.10 plate .sopoptilar with spectrographers. 1\ # ' -v (4) HIGH ULTRA-VIOLET SENSITIVITY RELATIVE TO BLUE. The ultra-violet^sensitivity is remarkably ' ^Tiigh' thpugh, as with all normal emissions, the sensitivity falls off at h-bout 2300 A. C ' The `Kodak' Slow Speetrographic plate is not^colour-s&siiised. * , (5) LOWER SPEED THANEITHER | KODAK ' B .20 OR B.40 PLATES. T h e `Kodak' SlowSpectrogrhphic gag plate is about l/8th" the speed o f B.20, the ^standard speetrographic plate, i or one*' $ quarter the speed of B.10, the plate with exceptionally high speed towards ultra-violet. (6) ACCURATE SPECTROPHOTOMETRY, ' If these plates are processed uniformly (see Kodak Data WEALDjTOHI' ' -HARROW 'klDblESEX Physics and Protection against Industrial Dust By Kenneth L. G oodall, M.Sc., F.Inst.P., H.M. Engineering Inspector of Factories Introduction have available a knowledge of the concentration (in In d ustrial dust presents two main hazards to. the workers in factory and mine: (1) risks to health, (2) risks to life and limb through fire and explosion from certain types of dust. Space will only permit reference to but a few aspects of the subject and therefore only factory dust problems will be dealt with, and, in particular, the physical aspects of those problems. The corresponding problems in mines, whilst covering a somewhat nar rower field, are essentially similar though modified by the special environmental conditions of working under ground in confined spaces. M any factory operations give rise to dust, the chief being those o f crushing, breaking, disintegrating, grinding, abrading, drilling, sawing, turning, mixing, Nieving, rumbling, sand- or shot-blasting, linishing, spraying, polishing, carding, weaving, spinning, and almost any operation involving material which is itself n dust or contains a percentage of dust. T h e risks to health from industrial dusts arise from the damage they can cause to the respiratory system, Hnd from the action of certain dusts as general systemic poisons or skin irritants. Thus the inhalation of silica or asbestos dusts, which id cumulative in effect, can cause the incurable lung diseases of silicosis(1) and asbestosis (2) respectively, in which the lung tissue is gradually destroyed, with the result that blood aeration diminishes; the inhalation of animal or vegetable dusts such as cotton, jute, flax, hemp, wood, hair, silk, and of some mineral dusts such ns limestone, glass, cement, marble, and emery may (feet the respiratory passages and give rise to bron chitis, asthma, or catarrh. Again, the breathing of certain metallic oxide fumes such as zinc oxide, magnesium oxide, manganese dioxide, and copper numbers of particles per unit volume), the size distribu tion, mass per unit volume, and mineralogical nature and chemical composition of the air-borne dusts which are breathed by industrial workers with known duration of exposure to the dusts, so that correlations can be made with the workers' clinical history of silicosis and other pulmonary diseases, and with the medical data to be obtained from X -ray examination of workers' lungs in vivo, and from examination of lung sections of workers known to have died from such lung diseases. In the examination of lung sections themselves it is important to ascertain by physical methods the nature and size of the dust particles to be found there. In ex perimental work on dust inhalation by animals of toxic or fibrosis-producing dusts it is equally important to have an accurate knowledge of the amount, size and type of dust being inhaled to correlate with the medical observations on the effect of the dusts. It is now established (3) that the majority of the particles found in the lungs are less than 5 microns in diameter (1 micron = I /a--io -4 cm.), very few exceed ing io/x. Within the range of sizes below 10 j. it has still to be established conclusively which size or range of sizes of dusts such as silica and asbestos is the most dangerous, and how much of such dangerous dusts can be regularly breathed by workers without appreciable shortening of life. Large numbers of comparable dust samples from various industries and processes need to be accumulated and analysed to enable conclusions to be drawn on these and related matters. For the measurement and control of actually existing dust conditions it is equally important to have accurate data on which to base estimates of the protection re quired, the type and arrangement of the devices most oxide, can give rise to metal fume fevers, while the in halation or swallowing of certain toxic dusts such as lend, arsenic, cadmium, or manganese, and their com- Enunds, leads to poisoning. The dust produced when nndling certain kinds of infected raw wool, hair, and hides, may carry spores, which, by skin contact or in halation, may cause the disease known as anthrax, while Hlinlly almost any dust appears to be capable of causing (he inflammations and affections of the skin, known as dermatitis, to workers who happen to be allergic, or emotive, to them. ()f these dangers to health the most serious are un doubtedly those due to silica and asbestos dusts and tc (lie poisonous dusts such as lead. Physical principles are involved at many stages in the intimation and control of dusty conditions in the work ing atmosphere, and it is these rather than the health rinks with which this article is concerned. E stim ation of the dust h ealth hazard Protective measures against health risks arising from the inhalation of industrial dust must take many forms, which must be based on a sound knowledge of the way lit which dust can enter and affect the lungs and re spiratory system, and through them the rest of the body. This is largelya medical problem, but one to the elucida tion of which physics has contributed substantially. Thus it is of primary importance for the physiologist to Viii.. 19, m arch 1942 Reproduced by the courtesy o f Messrs C . F . Casclla and Co., L td Fig. 1. Thermal precipitator $ $ * -v m '* .n -V;/ <>4 likely to be effective, and to assess their efficiency and value when provided. A large number of scientific instruments is available for these purposes, all based on well-established physical principles. For the measurement of air-borne dust concentrations and size distributions, the chief instru ments available for field use are : the particles of dust and in their being charged by the corona electric discharge, the effect of which is to drive them in a direction normal to the motion of the air along the tube. The dust particles are precipitated on the inside surface of the glass tube or upon a thin celluloid lining placed inside the tube, from which they (a) The Thermal Precipitator (4) (Fig. 1). This instru ment, which is rapidly becoming the standard instru ment in this country, employs the principle of the dust free space surrounding a heated body(s). Th e dust laden air sample is drawn between two thin microscope cover-glasses, part of whose surfaces lies in the dust free space of a heated wire, and the dust is completely deposited on the cover-glasses. Reproducedfrom the U .S .A . Bureau o f M ines Report o f Investigations, 3360 Fig. 3. Modified Greenburg-Smith midget impingers. Right: with hose and inlet guard, ready for use. Left: sealed for carrying Reproduced by the courtesy o f M essrs C . F . Casella and Co., L td . Fig. 2. Owens jet dust counter. (6) The OwensJe t Dust Counter (fi) (Fig. 2) and similar type instruments in which a hand-operated suction pump fitted with a non-return valve draws a measured volume of the dusty air via a damping tube through a slit. T h e rapid expansion and consequent drop in temperature of the air passing through the slit results in the condensation of moisture on each dust particle and so assists the particles to adhere to a chemically clean cover glass, which, placed just behind the slit, acts as the impinging surface. (c) The Greenburg-Smith Impingertrt, 8) (Fig. 3), based on the high-velocity impingement of the dust aspirated through a jet against a glass disk immersed in a known volume of water, which captures the dust. A sample of this water is then transferred to a shallow sedimentation cell for evaluation of the dust content. (d) Konimeters of various kinds (9,10), one of which is illustrated in Fig. 4, in which the dust is abstracted from the air sample by impingement at high velocity through a jet on to a slide coated with a sticky substance. (e) Electric Precipitators (u . 12) or Electrostatic Dust Samplers (Fig. 5), in which a measured volume of dust laden air is passed through a cylindrical glass collecting tube fitted with a central wire electrode maintained at a potential difference of up to 15,000 V with respect to a gauze electrode wrapped round the tube. Th e electric field produced results in an electrostatic attraction of Reproduced by the courtesy o f Messrs A . Gallenkam p a n d C o ., L td . Fig. 4. Konimeter can be washed into water or other liquid for sedimenta tion and evaluation, as with the Impinger samples. For counting and sizing the dust records obtained, some of these instruments incorporate their own micro scope equipment, but this is usually only capable of giving a low overall magnification of about x 200 so that the smaller particles under o-8g are missed. Such in- struments are chiefly used for routine dust checks and for preliminary surveys of dust conditions. For research work or for accurate dust surveys the dust samples need to be examined by a microscope of high resolving power incorporating a a mm. oil-immersion apochromatic objective and giving an overall magnification of about x 1500. High power microscopy is essential because industrial dust clouds contain considerable numbers of very sm all particles down to 0 * 2 /t or even less, and with low resolving power these are not seen. The tendency now is to substitute microprojection of the dust samples for direct counting with the microscope, to obviate or reduce the eyestrain inseparable from the direct method and to save time and work. A n example of such equipment as used in the dust estima tion work of the U .S .A . Bureau of Mines in Washington is shown in Fig. 6. particles sampled and so lead to erroneous figures for concentrations and size distributions, It may be emphasized that there is a real need for international agreement on the choice and specification of dust-counting instruments and standards of counting suited to fundamental research work and industrial dust surveys in the various industries, in order that the results obtained by workers in this field may be com parable, The dust samples obtained from industrial atmo- spheres with the thermal precipitator and similar instruments, which are intended to be counted and sized under high resolution microscopes, are heat treated to remove atmospheric pollution particles which are necessarily sampled along with the mineral dust which Reproduced by ike courtesy o f the U .S .A . M ine Safety Appliances Co. Fig. 5. Electrostatic dust sampler For fundamental laboratory work on dust cloud con centrations and sizes the special ultramicroscope (13) and sedimentation cell (14) instruments are available, and the development of the electron microscope will un doubtedly extend knowledge of the nature of dust particles in the regions of size beyond the capability of ordinary high power microscopy to resolve. Thus Preston(15) has shown, using a x 33,000 electron micro scope, that zinc oxide fumes contain particles in the form of needles of remarkably uniform length (about 0-4ft) and less than 0-04/4 thick; while Prockat and Windel (16) with a similar apparatus have demon strated that the fibrous structure of Cape asbestos is maintained down to particles of only 0-02 ft in thickness. For dust samples not air-borne but in powder form, methods of sedimentation or lutriation analysis as described by Heywood (17) are available. It is not proposed to discuss here the relative efficiencies of the various instruments, as an excellent critical survey (18) is already available which shows that they differ widely in their sampling efficiencies and that some instruments, which depend on impingement of the dust for collection, shatter a proportion of the VOL. 19, MARCH 1942 Reproduced from the U .S .A .B u r e a u o f M ines R eport o f Investigations. 3373 Fig. 6. Microprojector for dust samples, utilizing high power microscopy. Black cloth screening of counting cubicle and projection cone removed, to show details of construction and method of remote control of microscope adjustments it is desired to evaluate, since the size ranges of both overlap. W ith the lower power microscope equipment incineration of records is unnecessary as atmospheric pollution particles are almost wholly under o-8 ft in size and so are not seen. Th e problem of removing all atmo spheric pollution particles from a dust slide without removing mineral particles as well or cockling the thin cover glasses on which the dust is deposited is a very real difficulty worthy of further research work. The amount of atmospheric pollution can be gauged from a recent series of dust counts taken in a field in the centre of the Potteries at Stoke-on-Trent. Th e heaviest concentration measured was 27,000 particles per c.c. of pollution, of median size 0-25/1 (as counted down to the limits of visibility with a 2 mm. oil-immersion apochromatic objective and overall magnification X 2000) on a January afternoon; the lightest, on a day in the middle of Wakes week (the holiday week when all potteries were closed down), of 760 particles per c.c. 3-2 The dust records obtained with the instruments mentioned above are sized by direct comparison of each particle in a representative strip of the deposit with a series of calibrated circles in a special graticule fitted in the eyepiece of the microscope equipment. Some typical size distributions of air-borne industrial dusts are as follows: Flint milling. Median particle diameter o-8 -i -3 l. 2 - 3 % < '5f*. 70-80% < 2/x , 9 4 -9 9 % < 5 Potters' shop general atmosphere (Plate making and towing in progress). Median particle diameter i-o/x. I 9 % < 0 -2 fx, 5 0 % < I J a, 7 3 % < 2 /x, 9 2 % < 5 /i. Sandstone dust produced from mason's mallet and chisel work. Median particle diameter o-6 -i -i i. 2 5 -4 5 % < o -5 fq 70-90 % < 2 /x, 9 2 - 9 9 % < 5 H-Asbestos dust (in a spinning room). Median fibre length 2-6ja. 15 % < 1 /x long, 75 % < 5 M long, 9 2 % < xo /x long. Median fibre diameter 0-2 fi. 5 1 % < 0-2 /x diameter, 9 2 % < 0 -5 /x diameter. Median size of mineral particles present in asbestos dust clouds, 0 -3/x. 4 0 % < 0 '2fx, 73 % <o-5/x, 9 7 % < 2 ( x . The results of numerous determinations of size distribution of particles in industrial dust clouds have led to the conclusion (19) that particles under 0-2/x in diameter are present to the extent of only a few per cent. That so much of the dust in typical factory atmo spheres where mineral dusts are disseminated is under S /x is largely accounted for by the rate at which dusts settle. Thus particles of quartz of specific gravity 2-65 fall in free air at the rates given in the following table. Diam. of particle P 10 5 1 Rate of fall cm./sec. 0-784 0-196 0-008 Time to fall 1 ft. 39 sec. 2 min. 36 sec. 56 min. Actually these theoretical times to fall x ft. will be increased appreciably by convection currents and air eddies, but dust over ro/x very rapidly falls below a worker's breathing level under ordinary conditions. Dust particles of 1 /x and less in diameter are probably kept indefinitely in suspension in the atmosphere through the agencies of convection currents and Brownian movement. For the measurement of the mass of dust per unit volume in air-borne dust clouds there are available, amongst others, the following methods: (a) The Sugar Tube (20) or Potassium Nitrate Tube (21) in which a bed of sugar or potassium nitrate acts as a filter for a known volume of air-borne dust. T h e sugar etc. is subsequently removed from the collected dust by solution in water and filtration through a filter paper which is ashed and the residue weighed. (b) The Tyndallometer (22) which measures the in tensity of the light scattered by the dust in a measured volume of air, the intensity being a measure o f the mass of dust present. (c) The' Volatile Filter (23), e.g. a bed of naphthalene or anthracene through which a measured volume of dusty air is drawn. Th e filtering material is subse quently sublimed away at 70 "- 7 5 C and the dust residue weighed on a microbalance, (d) The Salycylic A cid Filter (24) in which a bed of this material traps the dust from a measured volume of dusty air. Th e dust is recovered for weighing by ex tracting the salycylic acid with alcohol and ether, and reclaiming the dust by centrifuging. Such determinations of mass concentration are not of great significance in practice as a few large size particles present in the sample outweigh completely the very much larger number of particles under, say, io/x. This results in a misleading idea of the amount of potentially injurious dust being obtained. For micro-chemical analysis of air-borne dust the salycylic acid filter will abstract enough dust, some 50 mg., in a reasonable time (3 hr.) from a dust cloud of moderate concentration. Sufficient dust for micro chemical analysis can also be obtained from the sugar tube apparatus or the naphthalene filter. For complete chemical analysis, and for solubility determinations, much larger quantities of dust, of the order of 1 - 5 g., need to be extracted from the dust cloud, and can be conveniently obtained by the use of the Labyrinth instrument designed by Briscoe and his co workers (zs). T h e instrument consists of a tube, con taining a series of baffles, through which the dusty air is pulled, the dust particles being deposited on the baffle plates by impingement. The mineralogical nature of air-borne dust particles as small as 3 /x in diameter can in favourable cases be determined by direct examination with the petrological microscope. T o obtain similar data for particles under 3 jx and thus cover the range o f particle sizes of particular interest in connexion with industrial health risks from air-borne dust clouds, recourse must be had to the X -ray and electron diffraction methods, which are now attracting the close attention of physicists. Clark and Reynolds (26) have, for example, developed the ordinary powder X -ray diffraction method to enable accurate estimations to be made of the percentage of free silica in the form of quartz in a few milligrams of mine dust, while von Hevesyto) has used secondary X-rays for the direct qualitative and quantitative micro-analysis of finely powdered mineral and rock samples. Again, X -ray diffraction is being increasingly utilized b y medical workers in pneumoconiosis research (28) on mineral dusts, down to colloidal sizes, such as are used for animal inhalation or injection experiments or are present in chemically extracted mineral residues of lung tissue or in pulverized and dried lung tissue. Protection against industrial dust B y a suitable combination of the physical and physical-chemical methods outlined above, a fair assess ment of the amount and nature of the dust clouds being breathed by workers in any particular case can be made. Methods of protection to suit can then be devised. This is the ideal approach to the problem which can only be adopted at present in a limited number of cases on account of the time involved in such surveys. In prac tice a selective survey of dusty conditions associated with typical processes in an industry (29) offers valuable guidance for more general application of protective measures to generally similar conditions in a whole industry. In the normal individual the human nose, throat, and lungs in themselves provide a defensive mechanism for arresting fine dust and preventing its reaching the ultimate air sacs in the lungs. Th e nose traps large particles in the nostrils and finer particles in the maze of passages above the nostrils in which dust is deposited from eddy current action, particle aggregation, and from impingement on the sticky secretion of mucus which lines all the air passages from the nose down wards. Th e windpipe and air tubes of the lungs are also lined with minute whip-like processes known as cilia, whose function is constantly to sweep upwards to the windpipe for expectoration the mucus and the dust particles it has trapped, the process being aided by the rhythmic contraction of the smaller muscular air tubes. Fine particles which succeed in reaching the air sacs of the lung are attacked and ingested by phagocytic cells, some of which manage to carry the dust back into the air tubes where they can be expelled. T h e importance of nose breathing in dusty atmospheres, in order that all the natural defences of the respiratory system may be brought into full play, will be appreciated, though individuals differ widely in the power of these defences. Th e main methods of protection against industrial dust take a wide variety of forms. Application of localized exhaust ventilation This is the commonest and most usually effective way of protecting workers against local dust concentrations.. Th e basic idea is to capture the dust or fume as near as possible to its point of origin by means of concentrated air currents flowing into suitable exhaust hoods, which need to be designed on sound physical principles to suit each individual case. Th e collected material is sucked away through ducting by a fan and passes to filtration plant. It is found that most industrial dusts can be cap tured by exhaust hoods designed to produce a linear air speed of from 100 ft.-50o ft./min. at the point of origin of the dust. T h e design of efficient exhaust hoods, of which Fig. 7 is a good example, is an important matter and is dealt with adequately in another publication (3) as is also the design of the associated ducting system, fan and Reproduced by the courtesy o f The Chloride Electrical Storage Co., L td . Fig. 7. A well-designed localized exhaust ventilation system protecting conveyor workers assembling lead accumulator plates from the risk of lead poisoning. Fumes from lead burning, and lead dust created in the assembly process are sucked away by the exhaust draught. A glass screen forming part of the hood is interposed between the work and the workers' noses Application o f general mechanical ventilation The purpose of general ventilation is to dilute with fresh air the dust laden air liable to be breathed, so that the resulting dust concentration is below the toxic or otherwise harmful limit. Such ventilation can take two forms, the extraction system with wall fans removing the contaminated workroom air, which is replaced by warmed fresh air, entering through suitably disposed openings in other parts of the room, or the plenum system in which clean, warmed and possibly humidified air is blown into the workroom under slight positive pressure through a ducting system, the air finding its way out via doors, windows or other openings. These methods of dealing with dust in workrooms, or a com bination of them both, are of strictly limited application and are only successfully applied where the dust con centrations are generally disseminated in the work rooms. Even then, the use of general mechanical ventilation is often a confession of failure to provide what is more effective, namely, adequate localized exhaust ventilation. dust collecting apparatus, all matters calling for the careful consideration of appropriate physical principles. Ducting resistance is an important item contributing to running costs, and proper application of the laws governing the flow of dust laden air in pipes will enable resistance to be reduced to a minimum by securing a symmetrical layout of adequate sectioned piping, fed by branch ducts at small angles in the direction of air flow and provided with gradual bends, tapered connections to fans, cleaning panels suitably sited, smooth joints and the like. Th e type of fan used for dust exhaust systems is the paddle type, receiving the dusty air approximately radially and discharging it tangentially against re sistances which may reach as high as 14 in. w.G. In many cases the fans aire placed downstream from the air cleaning plant, which is thereby kept under suction, an advantage where filters are, against the best practice, sometimes located in workrooms. Th e fan itself also thereby handles practically clean air. T h e main types of dust collecting apparatus are all VOL. 19, MARCH 1942 based on physical principles and are summarized below in the. approximate order of their ability to filter out the finest dust. (a) The Gravitational Settling Chamber, in which the dusty air is introduced from ducting into a com paratively large chamber, with a correspondingly large fall in velocity. T h e dust particles settle out in the chamber under gravity if their time of passage through it is long enough. On account of the large size of the chamber necessary for efficiency, practicable size chambers will not separate out appreciable quantities of dust under 50 x. (b) The Centrifugal Settling Chamber, or Cyclone, which utilises centrifugal force to throw the dust out of suspension in the air stream, and is useful for dusts above 10 ju. in size. (c) Inertial Separators, in which dust is caught by being given a sudden change of direction by impinge ment on baffles in a chamber. T h ey can be designed to catch dust efficiently down to 10/x. (d) The Viscous Filter type, in which the dusty air is sucked at high speed through a maze of oiled springs or steel wool and the like, the dust being caught through sticking to the oily surfaces. These filters will trap dust efficiently, down to about 3 j., only if carefully main required statutorily in a number of dusty industries is of great value, whilst much may also be achieved by the education of workers in the risks they run, and in their personal responsibility for preventing and suppressing dust clouds. A respirator of good design and with high filtering efficiency for low breathing resistance is useful as a secondary line of defence against dusty conditions, but is no substitute for other preventative measures such as have been outlined. Such a respirator, the Government approved Mark IV Dust Respirator (32), is now available. F ire and explosion risks from INDUSTRIAL DUSTS T h e extent to which protection is needed against the risk of fire and explosion from many industrial dusts under certain conditions is not generally realized, yet such dust fires and explosions frequently involve serious loss of life and injury to workers, and extensive damage to factory premises and plant. The accompany ing photograph (Fig. 8) speaks for itself. tained. (e) The Wet Scrubber Type Filter, such as the wet coke bed filter, water and steam spray towers, and various patent filters utilizing fine water mists with or without wetting agents added to reduce the surface tension. Good filtration of particles down to 2 in diameter, and lower in the case of easily wetted particles, can be achieved. (/) The Bag Filter, an excellent all round type of filter which will filter dust down to 0-5 ja with good efficiency. (g) The Electrostatic Precipitator, which will filter out dust and fumes of the very finest sizes with high efficiency. Other methods Space does not allow of more than a brief reference to other methods. Considerable use, for instance, is made of the wetting of materials to prevent dust creation, but Watson (31) has pointed out that in certain cases this practice may be more dangerous than dry working. Th e substitution of enclosed mechanical methods for dusty hand work generally, as in the con veying, elevating, machine packing and weighing of dangerous dusty materials, is already a widely adopted form of protection, though there is considerable scope for its extension. In some cases safe material may be substituted for dangerous dusty material, for example in the classic cases of the substitution of steel shot for sand in the operation of sandblasting iron castings, and of alumina for powdered flint in the process of china bedding in the pottery industry. A ll dusty processes should, of course, be separated from other processes so as to limit the number of workers exposed to any particular dust risk. B y what is called ` good housekeeping' a great deal can be done to protect workers from dust; for example by widely spacing dust producing machines, by storing dusty materials not in immediate use outside workrooms, and by keeping plant and workrooms as clean as possible, preferably with the aid of vacuum methods of cleaning. Th e periodical medical examination of workers now Reproduced by the courtesy o f S . H . Wilkes Fig. 8. Wreck of a factory following on the ignition of a starch dust cloud. Eighteen workers were killed and many injured Examples of dust liable to bum or explode under appropriate conditions are pulverized coal, sugar, dextrine, cocoa, various cereal dusts, cork, shoddy and rags, sulphur, wood flour, various resins and dyes and plastics, malt, tea, magnesium, and aluminium, to name but a few. Oxygen readily combines with many substances, in cluding carbonaceous materials and most metals, the reactions being usually exothermic. If the reaction is fast enough, incandescence results, leading to fire. If the reactions are exceptionally fast, the volumes of the [ 38] products of reaction may, due to the heat generated, increase very rapidly with the correspondingly rapid pressure rise which we refer to as explosion. Oxidation of combustible material is a surface phe nomenon and so is most rapid and capable of causing explosion with material in the finely divided state, i.e. with dusts, and particularly when these are in suspen sion in air as dust clouds. Th e increase of surface area with decrease in constituent particle size, for a given weight of dust, is striking. Thus Berger (33), on the basis of certain assumptions, shows that the surface of a cube of aluminium weighing 1 grm is approximately 3 sq. cm., whereas if the cube is comminuted until the thick ness of the resulting lamellae is 20 i (coarse stamped powder) the surface increases to 429 sq. cm. If the process is continued until the thickness of the lamellae is reduced to o-6/x (fine stamped powder), the surface of the i gram of metal becomes 14,290 sq. cm., while for the finest powder (prepared industrially in large quantities) with lamellar thickness 0-3 /r, the surface of a gram becomes 28,570 sq. cm. The increase in in tensity of explosion of dust clouds with decrease in constituent particle size is also of interest in this con nexion. Gliwitzky(34) obtained the following figures from experimental explosions of three types of com mercial aluminium pigment powder in a bomb of 53,000 c,c. capacity, for two powder concentrations. Particle size 1*2 0-6 0-3 Duration of explosion (sec.) for 700 mg./l. 0-06 0*03 0-012 Explosion pressure (atm.) 700 mg./l. 7 .7 8-6 10-6 580 mg./l. 7'7 9-2 ii*6 It will be seen how the velocity and pressure of the explosion increase with increase in the powder fineness until the maximum pressure of all was recorded, n -6 atmospheres for a powder concentration of 580 mg./l. This pressure is higher than the maximum pressures recorded for the explosion of gas/air mixtures, including the hydrogen/oxygen mixture with which 9-7 atmo spheres has been obtained. Pressures up to 18 atmo spheres have been recorded (35) for coal dust and flour explosions at certain concentrations and finenesses. Much greater values still have been measured for starch dust. T h e dangerous potentialities of the finely divided carbonaceous and metallic dusts for fire and explosion are manifest, since factory buildings, unlike mines, are not capable of withstanding such high pressures and rates of pressure rise. For each combustible .material there is an upper and a lower limit of concentration and beyond these limits the material will not burn. Within these limits there is also a minimum temperature or ` ignition temperature', below which the material will not ignite. Safety measures against dust cloud fires and explosions therefore aim fundamentally at keeping con centrations outside (and in practice usually below) the inflammable concentration limits, and avoiding sources of ignition; and if this should not be successfully achieved and a fire or explosion should result, one should aim at so designing the plant concerned and arranging the manner of working the process that there is a minimum of damage to workers and plant. T o describe how these aims can be achieved in con nexion with grinding mills, disintegrators, grinding wheels, mixing machines, elevators, conveyors, cyclones and other dust collecting apparatus, storage hoppers, and other plant in which dust clouds may occur would need a separate article, and those interested are referred to other publications(33,35,36,37.38), which deal with the design, location and construction of grinding and associated plant and buildings, methods of grinding dust in inert gases, the design and provision of bursting disks (39) and other explosion reliefs on appropriate parts of dust plant, the provision of rotary valves and other flame arresting devices, fire precautions, the use of magnetic separators, and methods of avoiding sources of ignition, including, electrical sources. It is of interest, in connexion with sources of ignition, to mention that electrostatic charges are known to be formed when powder issues from nozzles, or is blown into a dust cloud, or rubs frictionally against machine surfaces (40). Belts running over pulleys also accumulate charges. Discharges of such accumulations may ignite dust clouds, and although very little is known about this danger, it needs to be taken into account. Safety mea sures take the form of keeping the air as humid as possible, or in special cases ionized, using electrically conducting belting or suitably earthed collectors, and of bonding all plant to earth to prevent accumulation of charges. It has also been claimed by W olf (41) that where aluminium foil or ppwder is produced it can be ignited not only by fire and sparks but also by the effect of light. In the case of grinding particularly dangerous dusts, such as magnesium, the only safe practice is to prevent any worker being in the mill room whilst grinding is in progress, to which end the grinding mill door should be so interlocked with the electrical control to the grinding mill motor, that unless the door is shut the grinding plant cannot be set in motion, and the door cannot be opened until the current has been switched off and the mill has come to rest. Above all there needs to be emphasized the vital need for so enclosing, working, and constructing dust grinding and associated plant that large dust accumula tions outside such plant on walls, floors, ledges and beams are avoided. Otherwise a primary explosion in the plant (itself perhaps of small intensity) may raise a dust cloud outside the plant, ignite it, and so propagate a disastrous secondary explosion throughout a large area of a factory. M any of the above methods of protection against in flammable or explosive dusts are statutory requirements under special sections of the Factories A ct or under the Electricity Regulations (42), and the physical principles underlying most of them will be self-evident. There is as yet no approach to finality in the design of safeguards and protective measures, since much has yet to be learned concerning dust explosions. Dusts hitherto unsuspected occasionally prove to be explosive under certain conditions. T h e explosive limits and ignition temperatures of many dusts are not yet known or are not known with certainty; little is known about the best methods of removing ` static' from dust clouds, about the danger of ignition of dust clouds by sparks caused in various ways, about flame arrestors, or about the best methods of safely relieving explosion pressures in long elevator casings or in cyclones. It would be very useful to know the pressure and rate of rise of pressure which develop when the dusts in question explode under various conditions and in varying concentrations and grades of fineness. Th e physical and chemical mechanism by which dust clouds of various types inflame and ex VOL. 19, MARCH 1942 [39] W ! 1 i plode is also obscure, and methods of inhibiting fires and explosions might well follow from greater know ledge in this sphere. In the investigation of such prob lems as these and in the furthering of preventive and control measures necessary for the elimination of the health hazards, physicists must play a large and ever increasing part. REFERENCES (1) M iddleton. The Lancet, p. i, 4th July 1936; p. 59 n th July 1936. (2) M erewether and Price. Report on Effects of Asbestos Dust on the Lungs. Home Office Report (London: H.M. Stationery Office, 1930). (3) M iddleton. See (1). (4) G reen and Watson. Physical Methods for the Estima tion of the Dust Hazard in Industry. Medical Re search Council, Special Report Series, No. 199 (London: H.M. Stationery Office, 1935). (5) Watson. Trans. Faraday Soc. 3 2 , p. 1073 (1936). (6) Owens. Proc. Roy. Soc. A, 101, p. 18 (1922); J . Industr. Hyg. 4 , p. 532 (1923). (7) G reenburg and Bloomfield. Publ. Health Rep. (Wash,), 4 7 , p. 654 (1932). (8) Hatch, Warren and D rinker. J . Industr. Hyg. 14, p. 301 (1932). (9) Final Report of the Miners' Phthisis Prevention Com mittee (Johannesburg, 1919). (10) M avrogordato. S'. African Inst. Med. Res. Publ. No. 17 (Johannesburg, 1923). (n ) D rinker, T homson and F itchet. J . Industr. Hyg. 5 , p. 162 (1923). (12) D rinker. J . Industr. Hyg. 14, p. 364 (1932). (13) Whytlaw-G ray and Patterson. Smoke, Ch. 4 (London: Edward Arnold and Co., 1932). (14) G reen. J . Industr. Hyg. 16, p. 29 (1934). (15) Preston. Nature, p. 298 (8th March 1941). (16) Prockat and Windel. Staub. 10, p. 264 (1939). (17) Heywood. Trans. Inst. Mech. Eng. p. 257 (Dec. 1938). (18) G reen and Watson. See (4). (19) G reen. Trans. Faraday Soc. 32, p. 1091 (1936). (20) M avrogordato. See (10). (21) G raham and Lawrence. Trans. Inst. Min. Eng. 92, P. 1 (1936). (22) J . Sci. Instr. 14, p. 252 (1937)(23) M atthews and Briscoe. Bull. Inst. Min. and Met. (Nov. 1934). (24) M atthews, H olt, S anderson and B riscoe. Bull. Inst. Min. and Met. (Nov. 1936 and April 1937). (25) B riscoe, M atthews, H olt and S anderson. Trans. Inst. Min. and Met. p. 269 (June 1937). (26) C lark and R eynolds. Ind. and Eng. Chem. 8, p. 63 (1936). (27) von Hevesy. Chemical Analysis by X-Rays (New York: McGraw-Hill Book Co., 1932). (28) B erkelhamer. J . Industr. Hyg. 23, p. 163 (1941). (29) B loor, G oodall and Webb. Trans. Ceram. Soc. 38, p. 1 (1939). (30) D rinker and Hatch. Industrial Dust (New York: McGraw-Hill Book Co., 1936). (31) Watson. J . Industr. Hyg. 20, p. 155 (1938). (32) S add, G reen, D avies, N icholson, H il l and Watson. Chem. and Ind. 5 7 , p. 781 (1938). (33) B erger. Light Metals, 4, p. 44 (1941). (34) G liw itzky. V.D .I. Zeit. 80, p. 687 (1936). (35) Memorandum on Dust Explosions in Factories, Factory Dept. Home Office, Form 829 (London: H.M. Stationery Office, 1930). (36) Report on Experiments into the Means of Preventing the Spread of Explosions of Carbonaceous Dust. Home Office (London: H.M. Stationery Office, 1935). (37) Precautionsfor the Construction, Working, and Repair of Coal Drying and Pulverising Plants. Home Office, Form 831 (London: H.M. Stationery Office, 1937). (38) B rown. U .S. Department of the Interior. Bureau of Mines Information Circular 714 8 (19 4 1). [(Reprinted in Metallurgia, 24, p. 49 (June 1941).] (39) L ake and Inglis. Engineering, 148, pp. 58 and 88 (1939)(40) R udge. Proc. Camb. Phil. Soc. (19 12 -14 ). (41) Wolf. Chem. Ztg, 54, p. 796 (1930). (42) Regulationsfor the Generation, Transformation, Distribu tion and Use of Electrical Energy in Premises under the Factory and Workshops Acts 1901 to 1911. S.R. and O. 1312 (London: H.M. Stationery Office, 1908). Note on the Preparation and Properties o f Metal-coated Mica Con densers. By J. D. C raggs, M .S c., A .Inst.P., King's College, London* [M S. first received 20ih February 1941 and in finalform 28th October 1941] A B S T R A C T . In this note, the production of stable thin metallic films of low resistance is described, and alternative techniques of cathode sputtering and evaporation considered. Using such films as electrodes, mica dielectric condensers have been prepared with capacitance stabilities of the order one part in one thousand over a period of weeks. This estimate is conservative, as the circuit used for measurement had a stability of the same order. The experimental con densers had power factors between 2 and 5 x io-4 at 500 kcyc./sec. Differences between condensers prepared by sputtering or evaporation were negligible. Introduction T h e need for stable low loss condensers has, in recent years, become more urgent on account of the improved tuned circuits required' for high-frequency electrical apparatus. In many cases it is important to reduce the size of condensers, for a given capacitance, as far as possible, and mica is often used as a dielectric in pre ference to air. * J<ow at Metropolitan Vickers Electrical Co., Ltd. There are several possible methods of depositing metallic films on mica, viz. (i) by cathode sputtering, (ii) by condensation in vacuo, (iii) by spraying, (iv) by chemical precipitation, and (v) by deposition from col loidal solutions. In this note only the first two methods will be described. Condensers so made have been known for some years, but, so far as the author knows, no details of preparation have been published. F ilm s made b y sputtering T h e theory of this process is still obscuref and only the experimental procedure will therefore be described. As stable films with high electrical conductivity are required for condenser electrodes, silver and possibly copper appear to be the most suitable metals to be used. Preliminary experiments with a bell jar (25 cm. x 17 cm. internal diameter) and a 5 cm. square silver t Langmuir and Kingdon, Phys. Rev. 22, pp. 148, 357 (1923); von Hippel, Ann. Phys., Lpz. 8 0 , p. 672 (1926); ibid. 86, p. 1006 (1928); Cartwright, Rev. Sci. Instr. 1 , p. 758 (1930). [ 40] xlii JOURNAL OF SCIENTIFIC INSTRUMENTS CONSTANT VOLTAGE A new " A D V A N C E " in transformer design * * * Line voltage variations of / o " f * l / cu Stabilize your Electrical Instruments by using an " A D V A N C E " Constant Voltage Transformer. ADVANCE COMPONENTS LTD., Back Rd., Shernhall St., Walthamstow, London, E 17 Telephone : La rk iw o o d 4366-7 PRICE ON APPLICATION Write for details TYPICAL SPECIFICATION: In p u tV o lta g e :........... 190-260 v. 50c.p.s. Output Voltage: ... 230 v . 1 % M ax, lo a d :................... 150 watts Input power factor over 90 % REGISTERED ~ TRADE-MARK RESISTANCES FO R LABORATORY AND TEST ROOM You cannot use better Type D Double Slate Unit Catalogue post free The ZENITH ELECTRIC C. Ltd. SOLE MAKERS ALSO OF THE WELL-KNOWN " ZENITH " TRANSFORMERS ZEN ITH W ORKS, VILLIERS ROAD W ILLESD EN GREEN , LO N D O N , N.W .2 Phone: WILIesden 4087-8-9 Cram s: Voltaohm, Phone, London 46 RANGES -- ONE METER A self-contain ed com bination A .C ./D .C . meter providing for 46 ranges o f direct readings o f A .C . and D .C . volts.; A .C . and D .C. amperes; resistance; capacity.audio-frequency power output and decibels. A ny range instantly selected by means o f two rotary switches. N o external shunts or multipliers. Gives direct readings on 5 in* hand-calibrated scale with anti-parallax m irror; dead beat in action. Current consumption is negligible. Conforms w ith B .S. 1st Grade accuracy requirements. Autom atic cut-out disconnects meter from supply in event o f severe over load. Automatic compensation for temperature variations. Robustly built to ensure lasting accuracy. W rite for fully descriptive Pamphlet and current prices, Delay in delivery of Trade Orders is in evitable, but we shall continue to do our best to fulfil your re q u ire m e n ts as promptly as possible. J lu 46-RANGE UNIVERSAL BRITISH MADE Electrical Meaturing lntrumnt 7 Sole Proprietors and Manufacturers: THE AUTOMATIC COIL WINDER & ELECTRICAL EQUIPMENT CO. LTD. Winder House, Douglas Street, London, S.W.1 Phone: Victoria 3404-7 JOURNAL OF SCIENTIFIC INSTRUMENTS xlix IND EX TO ADVERTISERS : Accurate Recording Instrument Co., T h e ............................xxx Advance Components, L td .. xlii Aimer, G. C., St Co. . . .3 cover American Institute of Physics, T h e .................................. Hi Automatic Coil Winder St Electrical Equipment Co., Ltd., T h e ..................................xlii Baldwin Instrument Co., Ltd. xlvi Barr St Stroud, Ltd. . . . xliv Beck, R. St J., Ltd. . . 3 cover Becker, F. B-, St Co. . . . xv BirminghamSound Reproducers, L t d . ....................................xlviii Black, B., St Son, Ltd.. . 3 cover British Precision Springs, Ltd. xxxii British Rototherm Co., Ltd., T h e ...................... ..... . xxxv British Thermostat Co., Ltd., T h e ................................. iv Cambridge Inst. Co., Ltd. . xxviii Casella, C. F., St Co., Ltd. . xl Chance Brothers St Co., Ltd. back cover, e e b . Edwards, W ., St Co. (London), Ltd........................................ xxxix PAGE Electro Methods, Ltd. . . xxxi Everett,EdgcumbeStCo.,Ltd. xliv Flowerdale; Ltd..................... li Foster Instrument Co., Ltd. . xliv Foyle, W . St G., Ltd. . . . x General ElectricCo., Ltd.,The xxvii Gowllands, Ltd..........................xxxi Griffin St Tatloqk, Ltd. . . xxxi I.A.C., Ltd......................back cover Ilford, Ltd...................................xliii Institute of Physics . back cover Jobling, James A . St Co., Ltd. . . . . . .- . liii Kodak, Ltd....................... 2 cover Leland Instruments, Ltd. . lii Lewis, H. K., St Co., Ltd. . . 1 Londex, Ltd...............................xxiv Lyons, Claude, Ltd. . . . xxxvii Macmillan St Co., Ltd. . . xxxvi Mallory Metallurgical Products, Ltd........................................... xxv Marconi Instruments, Ltd. . xli Masteradio, Ltd..........................xlvi Metropolitan Vickers Electrical Co., Ltd.............................. liv PAGE Mulrhead St Co., Ltd. . . xlv Negretti St Zambra . . back cover Partridge, N.,B.Sc., A.M.I.E.E. lii Pulsometer Engineering Co., Ltd..................... back cover, j a n . Record Electrical Co., Ltd., T h e ............................ xlviii Salford Electrical Instruments, Ltd..................................... li Short St Mason, Ltd.. . . xlii " Solus" Electrical Co., Ltd. . xxx Sullivan, H. W., Ltd. . . xxxviii Sunvic Controls, Ltd. . . xl Taylor Electrical Instruments, Ltd. . .................................. liii Thermal Syndicate Ltd., The xlvi Tinsley, H., St Co., Ltd-. . . xxxix Tintometer Ltd., T h e. . 3 cover Townson St Mercer, Ltd. . li United Insulator Co., Ltd. . xlix Vactite Wire Co. ( 1919), Ltd; 1 Watson, W., St Sons, Ltd. 3 cover Watts, E. R., St Son, Ltd. . xlvii Wiggins, F., St Sons' . . 3 cover Zenith Electric Co. Ltd., The xxii Kindly mention the " Journal of Scientific Instruments" when corresponding with Advertisers