Moulded and Ornamental Bricks and Blocks, Including Copings and Lintels, Cutters and Rubbers,
Fireplace Bricks, Etc.
Ornamental bricks and bricks of special shape are generally made by hand moulding, but where the nature of the ornamentation permits them to be made by the wire cut process or the stiff plastic process these are cheaper and applicable to most shapes such as mullions and squints, of which the profile can be cut in a mouthpiece or die (sec Cutter and Rubber Bricks and Wire cut Ornamental Bricks).
Where a very small number of special shapes are required, ordinary bricks may be made by the wire cut process or by the stiff plastic process and then cut by a taut wire preferably in a frame or guillotine.
Where a sufficient number of bricks of the same pattern and size are required, a metal die may be used and where only a small number of such bricks are desired they should be moulded by hand in wooden or metal lined moulds, but for more ornate work plaster moulds sometimes made in several pieces must be used. A brick of the required design is first curved in plastic clay a little larger than the size of the finished brick, so as to allow for contraction in diying and firing. This model must he very carefully and accurately made, as any defects in it will be reproduced in future bricks. As soon as the modeller has completed his work, the mould maker places it on a board and brushes it over with a solution of soft soap in water to which a little sallow has been added, all the boards being similarly treated. He next places several boards or a piece of linoleum around the model, wins them tightly tngcthet to form a strong casing, and carefully stopping up any holes with clay paste, so that a case is formed into which the liquid piaster can be poured without any leaking awiy. Plenty of clay paste should be used, as a leak is very troublesome.
The model and the inside of the case are brushed over with the soap solution, and the mould maker next mixes a quantity of superfine plaster of Paris with water in a bucket, so as to obtain a thick slip, and stirs this well with his hands, so as to mix it thoroughly. The amount of plaster needed must be judged by experience the beginner will not go far wrong if he half fills a bucket with water and sprinkles the plaster rapidly into it until it no longer sinks into the water, but the proper proportions can only be ascertained by trial. They usually lie between 3 and 5! b. of plaster to each quart of water.
The plaster slurry must be worked with the hands until it is free from lumps and is of a smooth, creamy consistent it is then poured slowly and steadily into the case by an assistant, whilst the mould maker uses one or both hands to stir it slightly and prevent air bubbles forming between the model and the plaster. Sufficient plaster must be poured in to cover the model to a depth of about 2 or 3 in. The whole is now left until the plaster has set, after which the casing is removed, the plaster mould turned upside down, and the clay cut out with a knife or torn out with the fingers, great care being taken not to damage the mould. Sometimes the mode! will drop out whilst the mould is being turned, but if it does not do so it must be cut out. The mould is then set aside to dry and harden before it is used.
If sufficient care is taken not to spoil the moulds by overheating them, they may, with advantage, be dried by heating them in a warm stove, or even by placing them on the boiler.
When complex designs are required, it may be necessary to make the mould in several pieces, especially if some part of the work is undercut, i.e. with part of the surface projecting beyond an adjacent (lower) part, and so gripping the mould that the article cannot be withdrawn. By making the mould in several pieces this difficulty may be overcome, but it is often cheaper to make a single piece mould for a modified model and afterwards to undercut portions of the brick where required.
To reproduce bricks in such a mould, it is laid on a bench and a piece of clay paste thrown into it with considerable force and pressed well into the crevices of the mould. More paste is thrown in and pressed in until the mould is full. Any excess of clay is removed by drawing a strike or a stretched wire across the face of the mould, the clay being then smoothed (if necessary) with a large, flexible bladed pallet knife. The mould with its contents is then set aside until the clay is sufficiently dry for it to be turned out of the mould. If the mould is properly made and filled, the bricks should not require any further finishing, but it will often be found necessary to touch them up slightly with a modelling tool before drying them. The burning may be carried out in any ordinary kiln, but as the colour of ornamental bricks is usually important, they should be so placed in the kiln as to be discoloured by dust or not flame.
Ornamental moulded bricks, especially those which are not symmetrical, require special care in drying, as the ornamental portion tends to dry more rapidly than the remainder of the bricks and to crack or flake. Such bricks are best dried on hot floors or on shelves in intermittent steam heated chambers, and not in tunnel or other continuous dryers. The drying should be slow and without draughts.
There is no end to the shapes of bricks that can be made by hand moulding, although the cost of some of the moulds used in producing the beautiful bricks used in some Tudor architecture would be prohibitive today. Some of the old Tudor chimneys required fourteen separate moulds for a total of 144 pieces.
There is still a large demand for hand made bricks, and although it is very difficult to copy exactly the beautiful colours of the old bricks which were arrived at more by accident than design, owing to the manner in which the bricks were burned in those days. Innumerable different colours and mouldings can still be produced using the various sands and stains available, which give different effects, both in the texture of the faces of the bricks and also in the colour.
Carved Brickwork may be produced by carving and finished bricks in situ in the wall, but as this removes the skin and renders the bricks less resistant to weather, it should be avoided. Another method consists in making a large slab of plastic clay, modeling the desired design on it, and then cutting it with wires into bricks, which are afterwards burned in the usual way. In doing the carving, the artist must remember to allow for the effect of the joints when the bricks are laid in mortar or a ludicrous effect may be produced in the finished brickwork.
The design may also be modeled on separate plastic bricks, which are laid on thin boards instead of mortar, the modeling being afterwards cut so that the bricks can be separated and burned. When this method is used, the burned bricks should be assembled again before they leave the works, so as to ensure the design being accurately produced.
When the burned bricks are to be cut or carved, those made of very sandy loams and known as cutter bricks are generally used.
The carving of brickwork is still practiced, but not to so large an extent as formerly.
Cutter and Rubber Bricks are made of very sandy loams, and are so soft that, when burned, the former can easily be cut with a hammer and chisel, whilst the latter can be rubbed to the required shape on other bricks or on a stone. They are made by hand moulding or in a box mould.
Both Sir Christopher Wren and Inigo Jones did much to popularize rubbed and gauged work, which was very fashionable about the middle of the seventeenth century, and large surfaces of wall were built with rubbed and gauged bricks, and in addition to fiat surfaces, bricks were cut and rubbed to form mouldings and quite elaborate cornices and projecting string courses with several courses of axed or axed and rubbed bricks. Some of these cornices had projections of more than 19 in.
For some purposes, ordinary bricks can be cut to a special shape or size (e.g. key bricks, wedges, tapers, semi tapers, and closers) by means of a masonry saw such as those supplied by the Clipper Manufacturing Co., Ltd., Leicester.
For cutting over size or distorted bricks to the correct size an abrasive (carborundum) wheel is usually employed. It is dangerous to press the brick against the circular face of the wheel either the edge must be used or a cup wheel suitable for surface grinding must be employed.
Fire bricks and Other Refractory Bricks
The subject of refractory bricks is now so large that it cannot be dealt with fully in the present volume. Readers requiring more information than is contained in this one should see
A. B. Searle, Refractory Materials their Manufacture and Uses (London Charles Griffin & Co., Ltd.).
J. H. Chesters, Steel Plant Refractories (Sheffield the United Steel Companies, Ltd.).
J. R. Rait, Basic Refractories (London Iliffe & Sons, Ltd.) and the current Literature.
A. T. Green and G. H. Stewart, Ceramics a Symposium (Stoke on Trent British Ceramic Society).
The manufacture of fire bricks and blocks was carried on for many years in a somewhat rudimentary manner, and it is only during the last forty years or so that the more important firms attempted to improve their product and bring it up to date. In earlier times fire bricks and blocks were only required to withstand relatively low temperatures, but, with the increasingly stringent requirements of modern metallurgists and other users of furnaces, it is necessary at the present time to make use of every available assistance which science can render to the fire brick maker.
With this development has come an increasing use of the term refractory bricks primarily to distinguish those with a greater resistance to heat (refractoriness) from bricks unsuitable for use at temperatures above about 1500ºC. The term refractory is used rather loosely and many ungraded fire bricks are sold as refractories.
Investigations have shown that various users require widely different characteristics in fire bricks and blocks, and a material which suits one customer well may be entirely unsuitable for another. It is, therefore, necessary to know what characteristics are required before the value of a fire clay can be stated.
The materials from which fire bricks and blocks are made are of four main classes (1) fire clay (2) rocks consisting of almost pure silica (3) rocks composed chiefly of silica, but containing about 10 per cent of clay and known as ganister (artificial imitations of ganister are also used) (4) neutral and basic materials such as chromite, alumina, and magnesia.
The treatment of the materials depends on their nature, and the three chief processes used must therefore be described
Fire clay Bricks are made from seams of fire clay found in several parts of the country, the most noted deposits being in West Scotland, Northumberland, Yorkshire, the Midlands (including Burton on Tren! and Ashby de la Zouche), Cheshire, Stourbridge, Shropshire, Devonshire, and Wales. The materials from these various sources differ widely in composition and character.
The West Scotland fire clays (including those of Glenboig) are noted for their unusual heat resisting power. They require to be fired at a very high temperature, as otherwise they are soft and weak.
The North Umbrian fire days are chiefly found near the Tyne, and are richer in alumina than most of those of Scotland, though this advantage is more than neutralised in several cases by the presence of an excessive proportion of fluxing material (alkalies and lime), which greatly reduces the heat resisting power of the bricks. Several seams in Northumberland and Durham are, however, of excellent quality.
The Yorkshire fire clays are found chiefly near Leeds and Halifax, but the material crops up unexpectedly in several other parts of the county. In South Yorkshire it is associated with ganister (silica). The fire clays in Yorkshire are peculiarly variable in composition, the alumina varying from 15 to 39 per cent. The clays richest in alumina are found nearer the surface, but are much more tender than the stronger ones found at greater depths. Taken as a whole, the Yorkshire fire clays are amongst the most refractory, but they have not hitherto been worked so as to develop this property to the fullest extent, as they are almost invariably under fired, and so shrink abnormally when in use at high temperatures.
The Midland fire clays are more readily vitrified than must others of equal quality, and are, therefore, in great demand for the manufacture of close grained bricks and sanitary pipes. They are not usually so resistant to heat as some others, but where other factors (such as the cutting or corrosive action of dust and fire gases) have to be considered, they are very valuable, and under some conditions prove more durable than more infusible bricks from other districts.
The Stourbridge fire clays have a world wide reputation for refractoriness. The composition is remarkably constant, though unexpected variations occur at times. The average proportion of alumina is about 22 per cent thus corresponding to the Scotch and some Leeds clays but portions of clay with over 36 per cent of alumina have been found.
The Devonshire fire clays, like those of the Ashby district, are relatively easily vitrified, but considerable variations in quality exist. The most noted fire clays in this country are found in the Teign valley, and often contain considerable proportions of undecomposed granite. They are, therefore, used for the manufacture of vitrified bricks where the greatest resistance to heat is not required, but where a brick which will stand what is ordinarily considered to be a high temperature is needed.
The Welsh fire clays in some ways resemble those of Stourbridge, but are seldom so pure, and must, therefore, be worked with caution. Even the best deposits in this district are not of first class quality for refractory work, yet are excellent with respect to resistance to abrasion.
The fire clays are chiefly found associated with the Coal Measures and the Millstone Grit, and are usually obtained by mining. Some brickmakers work up the rubbish heaps of collieries, but the best fire clays are obtained by direct mining.
The seams vary in thickness, just as do those of coal, but are less uniform than the latter, and it has generally been considered that the only seams which can be worked at a profit are thick ones near the surface or those mined along with coal. Curiously enough, the best fire clay is often raised from pits containing little or no coal.
The chief constituent of fire clays is a mineral resembling kaolinite and also halloysite but not identical with cither, as shown by Roberts in 1947 and named livesite. In some fire clays the other two minerals are also present in small proportions. This mineral, when heated, behaves like kaolinite and is decomposed into free alumina, free silica and water vapour. On further heating a liquid glass formed by the reaction ofalkalis and other fluxes on some of the silica this glass contains most of the impurities in the fire clay and gradually dissolves the alumina and silica. At and above 1200ºC mullite is formed by the catalytic action of the alkalis liron oxide and other fluxes. At a still higher temperature, mullite crystallizes from the molten glass, the size of the crystals depending on the temperature and duration of the healing. The amount of mullite formed can be increased by the use of more flux, but this is commercially unsatisfactory and a much better way is to increase the temperature of the kiln and to prolong the heating at that temperature.
The essential constituents of fire clay bricks as well as of other aluminosilicate refractory bricks consist of crystals of mullite and silica (tridymite or cristobalite) and a glassy matrix. Minor constituents may include unaltered quartz, free alumina (corundum) calcium and other silicates.
The proportion of mullite crystals is always small in proportion to the whole mass, but for the best, bricks it should be as large as possible as it forms the skeleton or core of the bricks.
There appear to be several modifications of mullite with slight differences in properties and composition. Natural and fusion cast mullite appears to contain more alumina in solution than the mullite obtained in the firing of fire clay bricks.
Formerly, the manufacturer of fire bricks had chiefly to see that his material was right and that the men worked well. A few degrees more or less in the kiln made but little difference, and so long as his goods were saleable little else mattered. Within the last forty years, however, a great change has come over the fire clay industry. This is due to a variety of causes, the chief of which is the demand for better bricks and blocks from various users. This demand is increasing as progress with high temperature work continues, and the fire clay worker of the future must use his best endeavours to meet the demand. Fortunately, the cost of building and rebuilding is so high compared with the cost of fire bricks that a good price can be obtained for a really satisfactory article.
The Stiff plastic Process of Brickmaking
The stiff plastic process owes its name to the fact that the bricks appear to have been made of plastic material, though they are stiffer and stronger than most bricks made by a plastic process. The stiff plastic process is specially suitable for certain shales, which are becoming increasingly popular for the manufacture of hard burned, slightly vitrified building bricks.
The chief advantages lie in the saving of (i) the capital cost of a dryer, (ii) the cost of placing the bricks in the dryer, (iii) the cost of labour in working the dryer, and (iv) part of the cost of the fuel. The total of these amounts is sometimes quite large. These advantages can best be gained with bricks which can be safely and rapidly dried in a continuous kiln, but the stiff plastic process can also be made suitable for (i) materials whose plasticity or excess of water can be reduced sufficiently by drying by artificial heat or by adding a suitable non plastic material and (ii) materials which only require the addition of water to enable them to be pressed satisfactorily.
Hence, the stiff plastic process can be used for almost all clays and shales if they are first subjected to a suitable preliminary treatment. If no such treatment (other than grinding and screening) is desired, the process is confined to clays and shales which are dry enough to be ground and screened.
THE SIMPLE STIFF PLASTIC PROCESS
When no such preliminary treatment is required, the clay or shale is taken from the pit in wagons and fed into a grinding mill, generally of the edge runner type, with a revolving perforated pan, though a preliminary breakage of the large lumps is desirable. The clay is ground dry or in a slightly moist state, and is then taken by an elevator to the screens, of which there is generally one to each mill. The clay which passes through the screens goes down a chute into a mixer, where a little water is usually added and the whole is then thoroughly mixed. It next goes into the making machines and is pressed into rough blocks or clots about the size of a brick. These are then re pressed, this latter operation giving the brick its proper shape, making the well or frog and printing the name of the firm. The bricks are then dried, if necessary, and taken to the kilns. Drying is avoided when possible, this being the great advantage claimed by the stiff plastic process, though even where it cannot be entirely avoided its cost is greatly reduced. The kilns are the same as those used for bricks made by the plastic process but it may be noted here that as the stiff plastic process is generally used for large outputs some form of continuous kiln is usually employed.
The material must be sufficiently ground, and for the best bricks must be able to pass through a sieve with twenty holes per linear inch without leaving any residue though for common bricks a coarser sieve may be used, one with eight holes per linear inch being popular.
The ground material may require the addition of a little water, hut in any case it should be mixed so as to form a granular material of uniform composition and of constant stiffness, and the machinery used must be kept in first class order.
Economical grinding and pressing by this system requires the provision of a comparatively dry clay, or one in which a wet clay can be mixed with a large amount of dry material so as to make a relatively dry mixture. This is necessary, because in this process the clay is ground and sifted, and this cannot be done if the clay is very moist. If these matters are attended to and the material is suitable, no serious difficulties should occur in the manufacture of stiff plastic bricks.
Fine grinding and accurate screening are essential, and avoid many difficulties which otherwise arise. Saleable brick can be made with imperfectly ground material, but the process is costly and the results are always uncertain.
A convenient arrangement of the plant for the stiff plastic process in its simplest form is shown in Fig. 1, in which I represents the grinding pan, 2 the elevators, and 3 the brickmaking machine in this instance a Fawcett brick making machine being included.
Various modifications of this simple process are usually desirable.
CRUSHING
Although many firms using the Stiff plastic process send the clay or shale direct to the grinding mills, it is usually more economical to subject it to a preliminary crushing.
Stone breakers or Jaw Crushers can only be used for dry hard shales. If damp clay is passed through such a machine it soon clogs it and may cause serious damage! Many attempts made to use them for damp clays have failed.
Jaw crushers are very satisfactory for the preliminary crushing of burned clay (waste bricks used for grog) and for soft sandstone used for reducing the shrinkage of clay during drying, but not for wet (slightly sticky) shales.
No attempt should be made to crush the material very small and there is no need for the jaws to be set closely, and consequently they can be arranged to give a large output. The jaws should be examined occasionally and any wear and tear made good, as the machine will waste power if it is unduly worn.
Crushing Rolls are often very satisfactory for a preliminary crushing of lumps of material prior to their entering an Edge runner Mill.
Such a machine with two rolls, each 18 in. in diameter and 16 in. long, using 25 to 30 h.p. will crush sufficient shale or hard clay in an hour to make 6000 bricks and will greally reduce the wear and tear on the Edge runner Mill.
Prior to the material entering the crusher it is often advantageous to pass it over a live grizzly or other form of screen to separate pieces which are too large for the crusher and to break these by hand. A second screen in take all pieces smaller than the outlet of the crusher will also save power, though when the proportion of such small pieces is insignificant the whole of the material can be passed through the Fig. 3. Section of Light type of Breaker crusher. A jaw crusher is generally the best machine for reducing large lumps of dry clay or shale, though a gyratory crusher has a greater range of reduction.
Hammer Mills (Disintegrators) In the United States the term disintegrator is applied to crushing rolls, but in Great Britain it is used for an entirely different type of machine, consisting of a series of hammers hunt loosely on a shaft which rotates at the rate of about 1000 revolutions per minute, and so rapidly reduces any moderately dry clay or shale to a coarse powder. Another type of disintegrator consists of two cylindrical cages, one inside the other, which revolve in opposite directions, and so break up lumps of clay, shale, etc., and reduce them to a coarse powder.
The Lightning Crusher shown in Fig. 4 consists essentially of a casing enclosing a rotating shaft, bearing two discs or flanges which carry two or more shaped hinged hammer bars, which are carried round by the revolving shaft. As these bars strike a lump of material they deliver a violent hammer like blow, which splits the lump rather than crushes it, so that the product is more cubic than that from edge runner mills and crushing rolls. A grid is provided when a liner product is desired.
In the disintegrator made by British Jeffrey Diamond Ltd., the rotor consists of a series of discs mounted on a strong shaft, with a number of hammers loosely mounted on pins on each disc. The end discs are flanged to form a seal.
In all hammer mills, the number of hammers and their position on the rotor should be made to depend on the nature of the material to be ground and on the desired fineness of the product.
Most machines of this type are not suitable for plastic clay, but are excellent for shale, stone, or grog which is to be reduced lo pieces 1/16 in. or less in diameter.
These machines are not generally suitable for fine grinding, but a disintegrator, which can both dry and grind shales and hard clays (if they are not too sticky) is the Atritor, made by Alfred Herbert, Ltd., Coventry. It requires the material to be reduced to pieces not more than ¼ in. by a preliminary crusher an edge runner mill being usually the most convenient and effects the drying by a current of hot air which also conveys the ground material to a cyclone separator from which it is delivered to a storage bin or conveyor. The fineness of the product of this machine often improves the appearance and texture of the bricks or hollow blocks.
Whilst disintegrators are not usually regarded as fine grinders, an Atritor will grind hard shale and clay sufficiently fine for more than half of it to pass through a 100 mesh sieve. One pattern of the Atritor has been specially designed for grinding shale or clay to specified degrees of fineness from 5 to 200 mesh. In this machine, the air is passed through it by means of an independently driven fan, so that the particle size of the product can be controlled by varying the speed of the mill without interfering with the air supply. By pre heating the air, damp clays and shales can be dried sufficiently to be ground satisfactorily.
Disintegrators usually require the material supplied to them to be in pieces not more than 3 in. diameter, and they do not work economically when the articles of the product are less than 1/8 in. moreover, they are chiefly useful for clays employed for common bricks in which the minute particles of metal from the hammers do not spoil the colour. With a suitable material they require less power than an edge runner mill, but the wear and tear is greater. To obviate damage by stray pieces of metal entering the machine, the beaters should be hinged so as to stand straight out by centrifugal action in the ordinary course of grinding, but to fall back when a mass of metal is encountered. One section of the casing is hinged and held in place by an easily opened catch, and upon the attendant hearing the noise caused by a stray article he at once opens the catch with a long pole, standing well aside and out of the way of the material which is ejected from the machine.
When using a hammer mill or disintegrator it is important to adjust the machine so as to give a product of the desired particle size with one passage through the machine, as such mills depend on hammeraction (shattering) and not on direct crushing pressure and the larger the pieces of material fed into the mill the greater is the shatter effect. For the same reason it is seldom advisable to pass any material a second or third time through the mill unless it has previously been mixed with a large proportion of fresh material. If the product is too coarse, increasing the speed of the mill will grind it finer but care must be taken not to exceed the safety limit. A considerable increase of speed may also require the substitution of superior bearings and may involve the use of smaller hammers arranged in a staggered position instead of a few larger hammers.
It is unwise to use bars with too small a space between them, as the function of these bars is not that of a screen and insufficient space between them merely results in clogging the mill, i.e. the material is carried round and round inside the mill, without being ground. It is usually possible to adjust the particle size by moving the striking plate closer to or further away from the hammers.
Gyratory or Conical Crushers are seldom used for clays and shales in this country, but they are extensively used in some of the much larger works in the United States.
Hand Moulding Processes
Hand Making is chiefly practiced in an area to the south and east of a line drawn from Kings Lyrin to Portland Bill for ordinary building bricks and in the Midlands and North for the manufacture of fire bricks, specially moulded bricks, and terca cotta. As almost any clay with sufficient plasticity can be moulded into bricks by hand, the number of clays of widely differing characteristics described as brick earth is very large, and the prospective brick maker must be careful in his choice of material, for some clays are impossible to use commercially, even when, apart from the cost of manufacture, it is quite possible to make good bricks from them.
Only about 5 per cent of the total output of building bricks made in the British Isles is made by hand moulding, as such a process is much more costly than others which are available.
Bricks were made by hand moulding during more than 4000 years and in fairly large numbers in the British Isles during the Roman occupation but after the Romans left in the fifth century it was a long time before bricks were again made in appreciable quantities, viz., in the thirteenth century.
They were very largely used in the fifteenth and sixteenth centuries when some of the most beautiful Tudor buildings were erected and ever since they took an increasingly important part as a constructional material suitable for almost every kind of building and for many other purposes until the use of machinery greatly reduced their production relative to the total output of bricks.
Notwithstanding its great age, hand moulding is by no means an obsolete method, for by it bricks of a beautiful appearance can be made better than by any other method. It is costly and slow, and skilled moulders are scarce, so that the use of machinery has many advantages but where a beautiful appearance is an important consideration, hand made bricks still occupy the first place.
Materials. Hand moulded building bricks are usually made from surface clays of a mild character highly plastic or tough clays must be mixed with a suitable non plastic material before they can be used satisfactorily. The materials are selected according to the kind of bricks to be made thus, for an artistic facing brick a suitable loam or sand may be added to the clay, and the moulding process may also be arranged so as to produce bricks coated with sand. The well known stock bricks of Kent and Essex, on the contrary, are made of clay rich in calcium carbonate, or this is added purposely to the clay in the form of chalk.
Clays used for hand made bricks should usually be of such a nature as not to require very powerful machinery to convert them into a suitable paste. Hence, they should be free from gravel, nodules, and stones, or these materials must be removed by washing. Some clays which contain an excess of sand can be used satisfactorily after being washed.
The most popular clays for hand brickmaking are the Kent, Reading, Bagshot, and Gault beds in the South and East, and some Midland beds, but many surface clays in different parts of the country are locally considered to be of great value for this purpose.
Many clays which are too strong or tough can be made suitable by the addition of 20 to 30 percent of sand, which must be thoroughly mixed with the day, either by repeatedly turning over with a spade or by using an open trough mixer and a pug mill or brick machine.
HAND MADE FACING BRICKS
Hand made bricks vary greatly in appearance Kent Stocks are usually smoother than Sand faced Facing bricks and some antique Hand made Facing bricks have (purposely) a very coarse texture. Many imitation handmade bricks now made are described as antiques, rustics and by other names.
For red facing bricks, the clay must be as free as possible from lime, as this would affect the colour. If necessary, the clay may be washed to free it from objectionable impurities, but this is not usually necessary. When the bricks are sand faced their colour is largely due to the sand used in moulding, and by selecting this carefully, using different sands for different coloured bricks, many distinct and many beautiful shades of colour ranging from red to purple may be produced. Whatever sand be used, the colour of the clay body will show wherever the brick is chipped, and predominate eventually on the large face of the brickwork. The inherent colour of the burned clay produces the main effect of colour, but it can be varied to a limited extent by the use of sand.
The stock bricks made in Essex, Kent, and Middlesex from natural or artificial marls are made by washing the brick earth, with or without the addition of chalk, and then running the liquid into wash backs where it remains until sufficiently solid to walk on. The material is then covered with a layer of fine ashes (Soil) and when required the mixture is dug vertically and sent 10 a pus mill to be tempered. The resultant paste is moulded by hand, the bricks are then dried on hacks or in dryers and afterwards burned in clamps or kilns.
Many of the resultant bricks are by no means pleasing when viewed singly they may have a bad arris, be much chipped, and are often irregularly burned and sorted, and when they arrive on the job they do not appear to be in such good condition as machine made bricks. On the other hand, they form a very strong mass when built up, on account of their adhesion to the mortar they are excellent for plastering, and are the most durable of any bricks when exposed to the London atmosphere. If well made and properly burned, they are as strong as many other building bricks, though samples picked at random vary greatly in this respect.
Rubbers are relatively soft bricks which must be fine grained and of uniform colour throughout. Some of the best of these are box moulded and so uniformly burned that even when the outer skin is removed by carving or rubbing, the new surface exposed will weather perfectly.
HAND MADE FIRE BRICKS
The materials used in the manufacture of fire bricks are too hard to be sent direct from the mine to the pug mill. They are crushed or ground before being made into a paste. It is possible to use crushing rolls, but fire clays are usually best crushed in an edge runner mill and, after sifting, are mixed with water in a pug mill until a uniform paste is obtained of a consistency suitable for hand moulding. It is not advisable to mould it immediately, but to keep it for several days or even for a month in a moist state, and then to pass it a second time through the pug mill. The reason for this is the souring, or putrefaction, which most clays undergo when kept in a moist state, ensures the water being move uniformly distributed, and a more homogeneous paste is the result. The soured or matured and re pugged material is then slop moulded and then the bricks are dried on a hot floor or in another dryer, and are afterwards burned in single or continuous kilns. As fire bricks are required to be highly resistant to high temperatures, they should be burned under such conditions that they will not shrink seriously when in use. For this reason, the temperature attained in burning fire bricks should not be less than the bending point of Seger cone 5a (1180ºC.), and may be as high as that of Seger cone 18 (1500ºC.).
Glazed Bricks
Glazed bricks are used for three distinct purposes (a) to provide a smooth and readily washable surface which is impervious to dirt, (b) to increase the resistance of the body of the bricks to acids, e.g. by salt glazing acid proof bricks, and (c) to provide a pleasing and ornamental facing to the building.
For the last named, the surface may be glossy or matt and in the United States very attractive effects are produced by spraying one colour irregularly over another, so that walls built of a dark shade of brick at their base tone gradually to a light shade in the upper courses of the buildings, or contrasting shades of light and dark are used in columns running to the height of the building, thus emphasizing the vertical construction a striking example is the American Radiator Building with its manganese and gold tower.
In conjunction with larger glazed blocks (known as glazed terra cotta and by various trade names) very effective facings are provided.
There is a general impression amongst brick makers that any kind of brick can be glazed, provided that the composition of the glaze is known. This half truth has been the cause of much trouble and loss of money, because few people have realized that unless the brick to which the glaze is to be applied is practically perfect the glazed brick will be a failure. Trifling defects in a facing brick are often overlooked, but even smaller defects in a brick which is afterwards glazed will render attempts to sell it entirely abortive. Thus, a few tiny specks of lime in a facing brick may be passed unnoticed by the purchaser, but, if such a brick be glazed, the glaze will shell off above each lime speck and the brick will be of no value. Again, small defects in the arris of an unglazed brick are not obvious, but in a glazed brick they are at once noticeable.
Speaking generally, red burning clays are very liable to defects which are trifling in themselves, but which render successful glazing impossible, and, whilst a few firms have succeeded in building up a good trade in glazed bricks made of red burning clay, the majority of those who have attempted to use this material on a large scale have failed to show any profit. Glazed bricks are therefore, chiefly made of fire clay, the second grade clays with a refractoriness corresponding to cone 26 to 30 being used.
A brick to be suitable for glazing must be regular in shape, exact in size, with clean arrises, and a fine face free from small irregularities or discoloured spots. It must be sufficiently porous to absorb the water in the glaze slip, and must be refractory enough to keep its shape whilst heated at a temperature which will suit the glaze.
Such bricks are usually made by the plastic process and are re pressed before being fired, so as to obtain a good shape and face and to make them accurate in size. Any of the re presses may be used that by Pullan and Mann has a special measuring mechanism which automatically makes all pressed bricks of the same thickness, as any excess of clay is absorbed by making a somewhat shallower frog than usual.
When made of fire clay, bricks to be glazed are often hand moulded, as are fire bricks and are re pressed. Stiff plastic and semi dry prcssed bricks are slowly coming into use for glaring purposes, but they have not proved popular so far, owing to their liability to develop tiny surface cracks, which are of little or no importance in unglazed bricks, but prevent the glaze from adhering properly.
Much difference of opinion has been expressed from time to time on the desirability or otherwise of burning bricks before glazing them. It is considered that the cost of burning the bricks is so much wasted money, as they have to be reburned when glazed. Experience shows, however, that if the glaze is applied to unfired (green) bricks, the damage suffered in handling makes a large proportion of the bricks useless when they come from the kiln. These spoiled glazed bricks cannot be sold except as rubbish, as it is obvious that they are damaged. If on the contrary, the bricks are first burned without glaze, any defective ones sorted out may be sold as building bricks of good quality, or even as fire bricks at a higher price. The bricks selected to be glazed are stronger and less liable to damage, the amount of glaze wasted is reduced, and the number of unsaleable glazed bricks is brought to a minimum. These various savings often combine to make it cheaper to burn bricks twice instead of once.
At the same time, it is often possible with extraordinarily careful handling to glaze the unfired bricks and put them into the kilns in a remarkably perfect condition, and if work people will give sufficient care to the matter, it is quite possible (though seldom realized) to obtain a large proportion of excellent glazed bricks with a single tiring.
A mistake made by several purchasers of glaze recipes is to consider that they can buy all the bricks they require from a neighbouring yard. Such people forget that bricks intended for glazing need most careful handling, and when chipped at the edges they are useless. As few bricks which have been carted from one yard to another are not slightly chipped, it is practically impossible to buy bricks suitable for glazing unless the glazer is allowed to work on the same premises as the brick maker.
The glazed brick manufacturer cannot be too stringent or careful in the selection of his bricks.
The clay being suitable for the purpose of making a clean, well shaped brick, free from any impurities which could affect the body or the glaze, the most important part of the manufacture is the pressing. The presses should be placed conveniently near to the drying floor, or to the dipping sheds, according as the bricks are glazed in the burned or green state, as a little roughness in handling the impressed bricks will do no damage, but the pressed bricks must be handled as little as possible and carried to the dryer and after wards to the kiln with all necessary care to prevent them from being damaged especially at the arrises and corners.
Two serious errors, which sometimes arise in pressing, must be prevented at all costs. The first is due to the use of worn moulds or dies, whereby the bricks are formed with an arris or rough edge on them, which does not leave a clean edge unless the arris is very skillfully removed. The second is where the pressman fails to clean out the die completely, with the result that succeeding bricks have small pieces of clay forced into their faces, and these rise during the dipping and later cause the glaze to peel.
Pressing bricks for glazing is necessarily a slow operation (about four bricks per minute being the maximum), and any attempt to hurry the pressman may result in the loss of several hundred bricks, because these are spoiled by loose arris getting on to the faces of the bricks, or in other ways.
Glazed bricks must be laid with the thinnest possible joints, and, for this reason, must be pressed accurately. Any good power press may be used for this purpose, but it is sometimes a convenience lo use one in which the die can be drawn out on slides to the front of the press in order to discharge the brick, and enable the die to be cleaned before pressing another brick. When the die is movable in this way, it is much easier for the workman to see that it is properly cleaned and oiled than when a die fixed permanently beneath the plunger is used. It is, however, essential that the slides on which the die moves are kept perfectly clean, or the male part of the die will not fit accurately into the female portion and the die will be damaged.
Bricks which are glazed previous to burning require to be set in the kilns with the greatest care to prevent chipping, and the temperature throughout the kiln must be as uniform as possible, or the bricks will be unevenly glazed. Bricks to be glazed in the green state are often first beaten with a flat wooden blade to close up the face, but with a good press in charge of a careful man this operation is not necessary.
The burned bricks to be dipped are conveniently sorted at the kiln, then placed on a large off bearing barrow fitted with ample springs to prevent undue vibration, and are taken to the dipper, who has a small wagon carrying his tub of slip.
If the bricks are to be dipped before firing, they are placed directly they come from the press on to the barrow already mentioned, a sufficient number of these barrows being provided to allow the bricks to dry somewhat after they have been pressed. This is better than placing the bricks on the floor, as the double handling thus necessary is certain to damage them, and ihecost of a few additional barrows is not usually prohibitive.
The barrows with the bricks on them may be run into a warm shed so as to allow the bricks to stiffen and dry sufficiently within two or three hours, or they may be left overnight, bricks pressed one day being dipped on the next. The bricks must not be so dry as to show a lighter colour at the edges. Some firms dip the bricks after they have been dried white hard, but this is seldom satisfactory, as the sudden soaking of the dried face often cracks it.
GLAZES AND BODIES
Glazes are seldom applied directly to the bricks as the result would not be pleasing. Anengobe, body or dip, consisting of white burning clays, such as chinaand ball clays together with a flux such as Cornwall stone or felspar and some free silica (usually flint) is therefore made into a slip or dip and applied to the surface of the bricks. This engobe is later covered with a coating of glaze.
Engobes are intermediate in composition between the clays of which the bricks are made and the glazes used and so forms a suitable buffer between them it covers small defects in the surface of the brick, prevents any staining impurities in the brick from coming into contact with the glaze and enables the glaze to adhere more tightly to the engobe than it wouid to the brick.
The compositions of some engobes and a glaze which are largely used for glazed bricks are shown below they are merely guides as they will probably have to be adjusted to any bricks which the reader desires to giaze.
The glazes used for bricks must be sufficiently durable to withstand ordinary climatic changes without crazing1 or forming hair like cracks. They must be sufficiently hard to withstand accidental blows, and must adhere to the bricks so completely that they will not chip or peel off. Glazes which melt at low temperatures (below 1000ºC.) do not usually possess these necessary characteristics when fired on a porous body, but tend to craze or peel. Glazes fired at a higher temperature are, therefore, employed for glazed bricks, as the higher temperature enables a mixture of material to be used which produces a mass more nearly resembling the brick itself. Low temperature glazes are frequently termed soft fired or soft, and high temperature ones are spoken of as hard fired or hard the terms hard and soft when applied to glazes have no necessary connection with the softness or hardness of the glaze.
The materials used in the preparation of glazed bricks are very numerous, and would require a large volume to describe them fully. For temperatures near 1000°C. they are similar to those used by potters, but for the higher temperatures less fusible glazes are employed, and these are usually composed of felspar, Cornwall stone, flint and whiting, the corresponding bodies being composed of china clay, ball clay, Cornwall stone, and flint, a little of the finely ground or finely screened brick clay often being used in the first dip. Other materials, such a barytcs, zinc oxide, soda, and plaster of Paris, may be added at the discretion of the glaze maker, and the materials must, in some cases, he fritted into a kind of glass and ground before use.
Lead compounds are seldom necessary in hard fired glazes, and their use should be avoided whenever possible, for several reasons.
Enamelled Bricks are, strictly, those with a coating of enamel (i.e. opaque glaze), but the term is applied to all glazed bricks.
Production of cement clinker
Introduction
The objective in making cement clinker, namely the combination of the four principal oxides to make a material high in di and tri calcium silicates but low in free lime, was discussed in terms of the composition and reactivity of the raw materials. In this chapter, commercial methods used will be outlined and some of the chemical and physical processes that occur at high temperature considered. The manufacturing process is primarily concerned with the selection of the most efficient engineering methods for crushing, grinding, blending and conveying of solids on a large scale as well as with their heat treatment (pyroprocessing). Energy usage is considerable and is constantly monitored so that improvements can be made.
Two distinct processes are employed in the production of clinker. In the wet process a slurry of the finely divided raw materials is made and pumped into a long rotary kiln. In the dry process the raw materials are prepared for pyroprocessing as a blend of finely ground powders and initial heating is usually carried out in a preheater using the hot gases from a relatively short kiln. However, in those parts of the world where the raw materials are relatively dry, and fuel costs are not prohibitive, drying, calcining and clinkering may be carried out in a long dry kiln. The wet process was at one time predominant but a rapid increase in fuel costs in the 1970s accelerated the final stages of its replacement, since it involves the evaporation of a substantial quantity of water, typically 30 35% of the mass of the kiln feed. However, in areas where the primary raw material is a porous, high moisture content chalk containing flint, the wet process has survived in the preparation of raw materials.
Preparation of kiln feed
After extraction from the quarry, raw materials must be crushed, ground and blended to make the raw feed (meal) for the kiln. The choice of equipment is dependent on the physical properties of the materials. The importance of adequate grinding and blending of raw materials cannot be overemphasised, as there is a limit to the size of the regions of different chemical composition which can be eliminated by the rate determining processes of dissolution and ionic diffusion in clinkering.
Wet and semi wet processes
Soft materials are converted to a slurry with water in a wash mill. This involves vigorous agitation with harrows and tynes hanging into a cylindrical tank from a centrally pivoted rotating arm. Fine material in suspension passes through a vertical screen at the side of the tank, against which it is thrown by the harrows. The slurry produced, controlled by measurement of its density and the addition rates of the materials should contain the highest concentration of solids at which it is pumpable. Reduction of the water content necessary for pumping can often be achieved by adding a deflocculant, such as sodium carbonate or silicate, at a cost which is less than that of the kiln fuel saved. However, if significant amounts of a smectite (swelling clay) are present in the clay, the addition required will be excessive.
Either clay or chalk may be slurried first and the second component blended with it in a second wash mill. When chalk contains flints it may be added to the clay slurry in a form of tube mill, called a wash drum, in which the flints act as grinding media and are then scalped off. Remaining coarse material is removed from the slurry by fine screens or by hydrocyclones. The suspension enters a cyclone tangentially in an upper cylindrical section that produces a downward spiral motion in it (a vortex), which carries it into the conical section. The quantity of material leaving the bottom is limited so that an upward vortex is produced in the centre. Heavy material is thrown outwards and downwards and light material is carried up by the fluid flow to the second exit.
Further adjustment, using minor components such as ground sand, pulverised fuel ash or iron oxide, may be necessary to optimise chemical composition. The refined slurry is monitored by determination of residues when a sample is washed through standard sieves. Acceptable residues depend on the reactivity of the coarse material but 0.5% greater than 300 mm and 12% greater than 90 mm (in would be typical. Chemical composition is checked by sampling at several stages. An automated X ray fluoresence spectrometer is used to determine silicon, aluminium, iron and calcium, and an on line instrument can control equipment feeding the blending mill. The slurry is then held in tanks in which it is agitated both mechanically and by compressed air to prevent segregation.
Water has a high enthalpy of evaporation, and so the curve related kiln fuel consumption to the water content of a slurry is a steep one [(Fig. 1 (a)]. Consequently, a number of the remaining wet process plants have been converted to the semi wet process by installing filter presses (Fig. 2) which produce a cake with a water content of about 18 20% as feed for a kiln or preheater. In some variants of this process part of the kiln feed is introduced to the preheater in a dry state.
Dry and semi dry processes
The sequences of operations used in preparing the raw meal for the dry and semi dry processes are shown schematically in Fig. 3. Raw materials are crushed and put into stockpiles, usually under cover. A considerable degree of homogenisation is obtained by laying them down in strips or layers and systematically reclaiming from the stockpiles produced. They are then continuously proportioned into the milling and drying system by weight.
Gases from the kiln or the cooler are used for drying, although supplementary firing may be necessary. Grinding is carried out either in a ball mill or a vertical spindle mill in which rollers grind the material in a pan (Fig. 4). The latter is now usually selected for grinding raw materials in a new plant (unless they are exceptionally abras ive) as it uses significantly less energy for a given fineness than a Fig. 4. Vertical roller (spindle) mill used in drying and grinding raw materials A feed B roller on grinding table hot air flow (arrows) carries material to classifier-C-from which the finer particles are carried to exit-D. A peripheral dam ring determines the depth of the bed on the table which rotates. Hydrostatic pressure is applied, to the rollers ball mill. Early problems of wear and maintenance have been reduced by using replaceable wear resistant alloy surfaces, and steadier running is achieved by recycling some of the ground product to establish a denser bed in the pan.
Separation of coarse from fine material leaving the mill may be effected by entraining the powder in exhaust gas from the kiln, followed by separation of coarse material in a separator and/or cyclone system. Alternatively, the mill product may be transferred to the separator by a bucket elevator. The separator or classifier makes use of the balance between centrifugal and air drag forces, produced by a rotating plate which strikes the particles and by a fan. The milled raw meal is transferred to blending silos, which may be more than 30m high, and then lo final stage silos. To reduce the capital investment required in the blending and storage silo system, some new plants now employ limited storage capacity with direct on line chemical analysis and continuous adjustment of the kiln feed composition, using corrective materials such as ground limestone, sand and iron oxide.
Pyroprocessing principal manufacturing processes
The processes that occur when a raw meal is heated to clinkering temperatures, and where they take place in the principal types of plant employed, are summarised in Table 1. In wet process and some semi wet plants they all take place in the rotary kiln, which can be divided into the zones in which they occur (Fig. 5). In the most modern dry process plants almost all but the clinkering is carried out by suspending the powder feed in hot combustion gases before it enters a short kiln.
Material moves down a rotary kiln by sliding and rolling induced by the rotation and inclination of the kiln. The powder begins to form nodules as melting begins in the approach to the clinkering zone, the driving force for this being the high surface tension and low viscosity of the melts formed. Timashev reported a linear dependence of nodule size on melt surface tension. Gas velocities in a kiln are high. In the wet kiln, where a high fuel input is needed for drying the slurry, velocities may exceed 5 m/s so that dust entrainment occurs. Any dust not captured by the slurry in the drying zone and leaving the kiln constitutes a significant energy loss since its sensible heat (heat content) is lost. However, in the dry process most the dust/heat is retained in the preheater. Dust leaving the preheater may be captured where exhaust gases are used in a raw mill/drying system. Before going to atmosphere, gases are passed through a cooler and bag filter or a multi chamber electrostatic dust precipitator after conditioning to a suitable humidity and temperature. Collected dust which is not too rich in alkali may be returned to the raw feed blending system.
The choice of process for a plant is determined by balancing fuel economy against the capital investment required and its depreciation, as well as the amount and quality of the cement demanded by the (mainly) local market. Where there is limited demand and a relatively unsophisticated product suffices, a simple vertical shaft kiln may be used. A new dry process plant is unlikely to be constructed to produce less than about 800 000 t per annum. Kilns are fired by pulverised coal, heavy fuel oil or natural gas, according to local availability and cost. Coal has usually been employed in the UK but with a substantial supplement of petroleum coke. Other low cost materials employed as fuel supplements include old tyres and solvent wastes, use of the latter being subject to the constraints imposed by any deleterious effect on cement quality, refractory life or potential emissions to atmosphere from the system.
Wet and semi wet processes
A long kiln (length/diameter (L/D)~30) is necessary to carry out the endothermic processes of drying the slurry, decomposing the clay minerals and the calcite, and then to raise the temperature of the feed to a level at which clinkering takes place. Some 30 minutes residence in the burning zone is usually required. The strongly endothermic nature of both water evaporation and calcination is reflected in the plateaux in the material temperature profile in Fig. 5 Their length is an indication of the time taken for completion of these processes, involving minutes rather than the seconds needed in a fluidised powder suspension. In the drying zone, heat transfer from gas to slurry is optimised by heavy chains fixed to the shell of the kiln. These transfer heat from the combustion gases to the slurry and lift it to increase the surface at which evaporation can occur. In the semi wet process, filter cake is either introduced 10 a shorter kiln chain section or it enters a kiln after being dried on a Lepol grate or in a crusher dryer.
Grinding and fineness of cement
After it leaves the cooler, clinker is conveyed to a covered store in which some blending may be possible. Cement is produced by grinding clinker and gypsum, usually in a tube mill. This is divided into two or three chambers by means of slotted partition walls (diaphragms) which permit the forward movement of cement but retain the size graded grinding media. Milling is continuous and the residence time of the material in the mill, and therefore the fineness of the cement, depends on the rate at which clinker and gypsum are introduced. A large mill drawing 4500 kW, 4.6m in diameter and 14m long, would contain about 2801 of steel balls with diameters from 90 mm in the first chamber down to 15 mm in ihc last chamber. A mill may operate either on open circuit that is with the product going direct to a storage silo, or on closed circuit with the product being conveyed by air or a mechanical elevator to a separator (classifier) from which coarse material is returned for further grinding.
The wall of a ball mill lifts the media as it rotates and at a certain height they fall to grind the cement the mill must not rotate above the critical speed producing a centrifuging action. The shell of the mil! is protected by liner plates which may have a rippled profile to optimise lifting, since slippage of the media results in energy loss. Efficiency is rated in terms of the surface area produced per unit of electrical energy consumed and a standard energy requirement of 1.15 x 104m2/kWh may be used as a basis of comparison. Energy consumption is approximately linear up to about 300m2/kg, above which it increases progressively per unit increase in surface area as cushioning becomes more serious. In everyday running, the residues in the mill product on 90 mm and 45 mm sieves are used to monitor mill performance, which may decline as a result of media wear, for example. Increase in such residues at a given surface area will result in a change in the compressive strength/curing time relationship of the cement. A comparison of the effects of open and closed circuit milling on these residues can be seen in Fig. 1.
Fig. l. Effect of classifier efficiency on the particle size grading of ordinary (42.5N) Portland cement Power consumption in ball milling Portland cement is of the order of 45kWh/t for a surface area of 360m2/kg. This may be reduced by employing a closed circuit system, the saving of 2 5 kWh/t depending on the efficiency (fan power requirement) of the separator. The principal variables to be considered in optimising energy consumption in a ball mill include the speed of rotation of the mill, its ball size grading and loading, and the design of its lining. The use of a grinding aid reduces energy consumption, especially with higher surface area products. A mill is designed on the basis of the throughput required, using data for the grindability of the clinker determined in the laboratory.
The ideal way to grind a material would be to break each crystal or aggregate of crystals separately by simple cleavage. The energy consumed is then the surface energy created plus that lost to the fragments and media as heat. However, those milling systems which most efficiently keep the particles separated for grinding, such as the roller (vertical spindle) mill or the roll press are the most susceptible to wear with a hard, abrasive material like clinker. This resulted in serious maintenance problems in early versions of these mills before replaceable, wear resistant alloy surfaces were developed. Consequently, although a ball mill wastes energy in multiple impacts (cushioning), it has remained predominant in cement grinding. Capital costs usually preclude the complete replacement of a grinding plant in an established works.
Initial grinding of large clinker nodules in a ball mill is particularly inefficient. In existing plants the introduction of a roll press for preliminary or semi finish grinding, with finish grinding in a ball mill, has proved a cost effective way of significantly increasing both energy utilisation and throughput, the latter making it possible to maximise use of low tariff (off peak) electricity. Improvements in both are even greater in raw materials grinding. This combination of grinding techniques has the advantage of avoiding two adverse effects when a roller mill or roll press is used for finish grinding, namely an increase in the water demand of the cement and the possibility of an unacceptable reduction in initial setting time. The former is ascribed to the narrowing of the particle size distribution resulting in an increased voidage in the cement (decreased bulk density) and the latter to the production of coarser gypsum particles and a lower degree of dehydration resulting from a lower grinding temperature.
The relatively recently introduced Horomill (horizontal roller mill), which is suitable for the finish grinding of cement clinker and raw materials, is essentially a tube mill in which a cylindrical roller constitutes the grinding component. Cordonnier describes the performance of the first industrial (25t/h) installation of this mill in Italy. For a similar capacity to a ball mill, it has a slightly smaller diameter and is only one third of the length. Energy saving was quoted as 30 50% with wearing surfaces having a satisfactory life. For a cement with a surface area of 360m2/kg, an energy consumption below 30kWh/t can be expected. Cement produced by the Horomill had a similar particle size distribution and similar physical properties to one produced in a ball mill.
Factors influencing the grindability of clinker The grindability of clinker depends on its chemistry and on the conditions it experiences in burning and cooling. Hard burning and high melt content resulting from a low silica ratio increase initial grindability since they result in a clinker with a low porosity. (Grindability increases with increasing difficulty of grinding. Maki et al. observed that grinding was impaired in clinker containing clusters of belite crystals. After most of the larger aggregates of crystals have been broken, the fracture properties of the individual phases assume greater importance, although it must be remembered that a majority of the final cement is made up of multiphase particles. Hardness of a crystal is less important than its brittleness in comminution and since alite cracks much more readily than belite in a microhardness measurement, clinkers with a high lime saturation (and substantially complete chemical combination) can be ground more readily than those with a low lime saturation.
Hornain and Regour found that the grindability of a clinker sample sintered to a density of 3000 kg/m3 was determined by its fracture energy and the size of the microcracks present. The number and size of the latter could be related to the cooling regime experienced by the clinker. They measured fracture energies in the range 12 20J/m2, using notched, sintered prisms of clinker. From measurements of the impression made in each clinker phase by a Vickers micro indenter and the size of the cracks radiating from the indentation, they calculated values for a brittleness index C3S 4.7 C3A 2.9 C2S and C4AF 2.0.
Scrivener examined cement particles by BSE imaging and X ray microanalysis and found that fracture of alite crystals predominated in polymineralic clinker particles rather than fracture along phase boundaries. Many smaller particles had surfaces rich in interstitial phases. Bonen and Diamond used the same techniques to examine two chemically very similar cements, one produced in a ball mill the other in a roller mill. The latter contained more nearly isodimensional particles and a significantly smaller proportion of the finest particles present in the sample from the ball mill. The particles in the ball milled sample also exhibited much greater surface
roughness. X ray microanalysis revealed that the surfaces of the particles in the two cements differed in composition, that from the roller mill having a higher content of belite and interstitial material, apparently reflected in a significant decrease in heat release in the first day of hydration.
Minor additional constituents
The current British Standard follows the European prestandard to be ground with clinker as part of the nucleus of a Portland cement that is excluding the gypsum. Unlike the European specification, it limits the materials which can be used to one or more of a natural pozzolana, blastfurnace slag, or pulverised fuel ash but not if a blended cement is being manufactured with one of these as a main constituent or if a filler such as ground limestone is being added. The term filler covers any inorganic natural or artificial material which, owing to its particle size grading when ground, enhances the physical properties of cement without any detrimental effect on concrete durability. Moir concluded that ground limestone is likely to be a preferred mac for practical and economic reasons. It aids the control of cement workability and strength development and inhibits bleeding.
Tests of cement quality
Introduction
The assessment of cement quality relies primarily on direct performance tests because of the complexity of the factors influencing its rate of hydration and its hydraulicity. It was seen in the last chapter that the value obtained for the specific surface area of cement is particularly method dependent, so that a prescribed procedure must be followed and named when referring to the result obtained. A similar constraint applies to the determination of the hydraulicity of cement by incorporating it in a concrete or mortar.
Committees of specialists, representing all interested parties (manufacturers, consumers, government and academic institutions), have in some countries produced national specifications and test methods for the assurance of cement quality. In other countries, British Standards or those published by the American Society for Testing and Materials are used. Standard test procedures have also been published by the International Standards Organisation and by the Comite Europeen de Normalisation (CEN). Those developed by the latter, Methods of Testing Cement EN 196, have been published with useful National Annexes by the British Standards Institution as pans of BS EN 196. CEN is currently finalising a specification for cements commonly available in Europe. A pre standard is voluntary (ENV 197) but drafts have formed the basis of the present British Standard, BS 12 1996. It should be noted that standards and specified test procedures are regularly reviewed and revised when deemed necessary. BS 12 was first published in 1904. ASTM standards are published annually in book form cement and concrete specifications and testing are included in Volume 04.01.
A standard specification lays down the chemical, physical and performance characteristics required of a cement for it to be sold as conforming to the standard. In the current European approach, a programme of product sampling (defining minimum frequency and method) is indicated for the manufacturer who must sample at the point the product leaves a works and Employ the test methods laid down in EN 196. Two statistical procedures, and values for the necessary statistical parameters, are provided for the assessment of test results obtained against the specification (manufacturers auto control). Criteria for conformity take the form of specified characteristic values for properties, which can only be breached by a defined number of test results in a given set (test period). They are derived from probability theory using a defined, low risk of acceptance of a batch not meeting the required characteristic value. These procedures are described in detail by Brook banks with helpful worked examples. In addition, for some properties, limit values are also specified and no individual auto control test result must fall outside these.
The British Standard BS 12 1996 specifies chemical physical and performance requirements for Portland cement following ENV 197. In addition to characteristic values, the British Standard specifies acceptance limit values for certain properties which are somewhat more stringent than the limit values in ENV 197. They can be used with results obtained for single samples by a customer or independent test laboratory. Maximum permitted deviations above or below the stated acceptance limits for individual results are also specified.
Chemical composition
The compositional requirements specified for Portland cement cover both clinker and cement. The test methods to be employed are those described in BS EN 196. A detailed discussion of these is given by Taylor. The compositional requirements for clinker are C3S + C2S > 66.7% C/S > 2.0 MgO < 5.0%. They are comfortably met in the UK. The requirements and acceptance limit values specified for cement in BS 12 1996 are given in Table 1. The limit for chloride ion content is necessary to reduce the risk of corrosion of steel in reinforced and prestressed concrete. Limits for loss on ignition and insoluble residue protect the consumer from a product which has suffered either excessive exposure to the atmosphere during storage or contamination.
Setting times
These are the times after completion of mixing at which a neat cement paste presents specified resistances to the penetration of a needle. The principle variables influencing penetration are the water content of the paste, the temperature, the load on and dimension of the needle and, of course, the reactivity of the cement. The needle employed (diameter 1.13 mm, total load 300 g) is named after Vicat. It is released at the surface of the hydrating paste at intervals until it penetrates only to a point 4 ± 1 mm from the bottom of the standard mould. When the paste has attained this degree of stiffness it is said to have reached initial set, for which a minimum value is specified in BS 12. A second similar needle with a concentric ring attached can then be used to determine final setting time, although a maximum value for this is no longer specified in the British Standard. Final set is reached when the needle makes an impression on the surface of the paste but does not penetrate the 0.5 mm necessary for the ring to mark the surface.
The higher the water content of the paste, the longer it will take for the cement hydration products to form a structure with the chosen resistance to penetration. BS EN 196 does not, however, specify a fixed water/ cement ratio. Instead, pastes are examined at a range of ratios toestablish that needed to produce a paste into which a 10mm dia. plunger, which is held in the apparatus used for the Vicat needle, penetrates to 6 ± 1 mm from the bottom of the same mould. This paste is described as having standard consistence and since the result is sensitive to shear history, the mixing procedure is specified.
Compressive strength
The most important test of cement quality involves the determination of the compressive strength it produces in a mortar or concrete. In the past, a specified concrete mix was usually tested in the UK using British Standard 4550 although this also gave a procedure for a mortar. The USA and many European countries favoured mortar testing and widespread use of the ISO RILEM R679 mortar strength test is encountered in the literature. In the spirit of membership of the European Union, the mortar prism test method of EN 196 was adopted in BS 12.
The mortar mix specified is 3 1 0.5 by weight of sand, cement and water, respectively. It is cast into 40 x 40 x 160mm moulds. Flatness of the resulting surfaces of the mortar prisms is important because surface irregularities would concentrate stresses during compressive strength measurement and affect the result. Compressive strength may be related to the volumes of cement (C), water (W) and air (A) in a mix by Ferets empirical law
Where k is a constant for the aggregates, cement and curing employed.
The volume of air present depends on the degree of compaction achieved and the object is to achieve full compaction using, in the EN 196 procedure, a jolting table or a vibrating table giving equivalent results. Each prism then contains approximately 600 g of mortar. Excessive compaction must be avoided, however, as it causes particle segregation. Some air may be entrained during mixing of the mortar and checks against a reference sand are important because some sands cause more entrapment than others, possibly because of an abnormal, although small, amount of clay and/or organic matter adhering to the grains.
The effect of the amount of water used is marked but easily controlled. Compressive strength is usually measured after 2 and 28 days. The latter gives what is referred to as the standard strength and BS 12 classifies cements on the basis of the level attained for low strength classes the 2 day test is replaced by one at 7 days. After curing, a prism is superficially dried and tested immediately. It is first broken in flexure in a specified manner and then the separate halves broken in compression across the 40mm thickness. Prisms are cured in batches of three yielding six results for compressive strength at each age. If any one result deviates by more than 10% from the mean it is rejected and if any one of the remainder deviates by more than 10% from the new mean, then all the results must be rejected and the test repeated.
Where p is the mortar prism compressive strength (N/mm2), c the concrete cube compressive strength (N/mm2) and d the curing period in days at test. At 28 days this is equivalent to a strength ratio p/c of 1.30.
Some results for concretes prepared using the sand, granite and the procedure specified in BS 4550 Part 3 1978 are given in Fig. 2. They illustrate the effect of cement surface area and that of water/cement ratio. In construction contracts, test cubes are prepared from samples of production concrete taken as it is placed. Since 28 day strength is regarded as an important indication of concrete quality for structural engineering purposes, various accelerated curing procedures involving elevated temperatures have been proposed to reduce the time needed to get an indication of its potential value. Unfortunately, correlations with strength developed under normal curing conditions are poor, presumably because at elevated temperatures there is a coarsening of the pore size distribution in the on the development of strength in BS 4550 concrete mixes (the w/c specified in the standard is 0.60) hydration products of the cement. A prediction of 28 day strength from the 7 day strength, using a knowledge of the form of the growth curve, is preferred. This assumes that curves such as those in Fig. 2 may be displaced parallel to the strength axis but retain their shape, an assumption which is acceptable as long as the chemistry of the cement employed does not change significantly.
Admixtures and special cements
The development of new cements has been a major part of research activity for most of the twentieth century, driven by the need to obtain cement compositions which improve on Portland cement by providing particular properties such as rapid setting and hardening, improved workability, or increased durability in severe environments. In the past 25 years, the need to reduce energy consumption and, where possible, emissions of the greenhouse gas CO2, have added to the incentives to introduce new cementitious compositions as well as improved production processes. Since long term satisfactory performance in use is a major requirement of any new cement, testing to a point where it is widely accepted can involve a prolonged examination of its durability in the environments in which it is to be used.
In this chapter, some special cements are described briefly and, since the properties of Portland cement mortars or concretes can be given special properties by the use of admixtures, a short account of their nature and applications is included.
Admixtures
The properties of a concrete or mortar containing Portland cement can often be beneficially modified for a particular use by the addition of small amounts of certain chemicals. When the addition is made as the concrete mix is being prepared, the material is described as an admixture. A classification of the commonest admixtures groups them as accelerators or retarders of set and hardening and water reducers, although a particular substance may combine one of the first two characteristics while also reducing the water needed to produce a mix with a given workability. The addition is usually made with the substance in solution to maximise the uniformity of its dispersion, although its introduction may be slightly after the addition of the main bulk of the water where experience has shown that this increases effectiveness. Sensitivity to dose of the admixture is first examined with a sample of the particular cement to be used since the addition required may be influenced by cement fineness and chemistry, in particular the contents of C3A, free lime, and soluble alkali sulfates. Dose is usually expressed as mass percentage on cement of the active ingredient of an admixture.
Accelerators
These may be employed in precast concrete production or cold weather concreting. They act by increasing the rate of hydration in the acceleratory period, leaving the dormant period largely unaffected. Calcium chloride at 2% is particularly effective in increasing early strength development but the corrosive effect of the chloride ion means that it cannot be used in reinforced or prestressed concrete. Calcium formate, nitrate and nitrite are less effective alternatives but no single accelerator is widely accepted. It is often more practicable to employ a water reducing admixture to enhance early strength development.
Retarders
These are valuable in extending the working time of a concrete or mortar in warm conditions since their effect is primarily confined to the dormant period. Hydroxycarboxylic acids (citric acid and those, such as gluconic acid, derived from sugars) and sugars themselves are examples, the latter having drastic effects if an overdose are used. Additions of about 0.25% are usual, but sucrose can extend setting time by as much as 10 h at levels as low as 0.05%. The low dosage of retarders needed is considered to indicate that they function by adsorption on the surfaces of cement grains or, more probably, the hydrates formed initially on them. The poisoning of portlandite (CH) nuclei has also been suggested as a mechanism of retarding set. With cements having a high C3A content, a greater retardation with a given dose may be obtained by a delay of just two minutes in adding the retarder after the bulk of the water. The initial interaction of C3A with gypsum is believed to reduce its interaction with the admixture.
Fluorides, phosphates, zinc and lead salts are all retarders which are precipitated from solution (the first two as calcium salts, the second two as hydroxides) as coatings on the surface of cement grains. Phosphate based admixtures show superior retention of their effectiveness at elevated temperatures. Proprietary blends of retarders and plasticisers are employed in ready mixed mortars with a life of 36 48 h.
Water reducing (plasticising) admixtures
These enable a reduction of up to 15% in water content to be made while retaining a chosen workability. Usually, a sodium or calcium ligno sulfonate (by products of wood pulp manufacture) or a hydroxycarboxylic acid is employed. Sugars in unrefined samples of the former are said to give it an additional function as a retarder. These materials inhibit segregation and can be used in pumped concrete.
Sodium salts of sulfonated naphthalene formaldehyde co polymers (PNS) or sulfonated melamine formaldehyde co polymers (PMS) are described as superplasticisers because they can be used at levels of addition (ca. 0.5%) which make possible water reductions of around 30% without introducing either air entrainment or retardation. The production of very high strength concrete becomes practicable and Neville cites 28 day compressive strengths of
150N/mm2 for concrete made at a water/ cement ratio of 0.2. Alternatively, the water content may be maintained with no loss of cohesion even with slumps as high as 200mm. These materials are therefore important in pumpable concrete and self levelling screeds for flooring. A mix will progressively stiffen (exhibit slump loss) but an additional dose of superplasticiser may be used to prolong working time.
The enhancement of flow by superplasticisers is attributed to their adsorption on initial cement hydration products resulting in an increase in the zeta (z) potential at the shear plane of the electrical double layer at the interface of particle and aqueous phase.Bonen and Sarkar found that the adsorption capacity of a cement depended on the molecular weight of the PNS, the fineness of the cement and its C3A content, while slump loss was strongly dependent on the ionic strength of the aqueous phase. Cement grains in a paste usually possess a low z potential, presumably because of the high ionic strength, but the strongly acidic (dissociated) sulfonate groups (SO3H) maintain a negative surface charge and C potentials in the region of 30 mV have been quoted. This would significantly reduce the extent to which particles form agglomerates trapping water which cannot contribute to flow, a factor which is particularly important in concentrated suspensions.
Uchikawa et al. found lower z potentials (ca. 10 mV) in a cement paste to which a PNS co polymer had been added, although it exhibited satisfactory flow characteristics. They used the relatively recently introduced electrokinetic sonic method, which does not require dilution of a cement paste, and the lower potential could be ascribed to the higher ionic strength of the undiluted suspension. However, the most interesting observation was that a co polymer of polyacrylic acid and polyacrylic ester was especially effective in increasing paste fluidity, although it did not produce a significant change in the z potential of the paste (ca. -1 mV). They concluded that this was evidence of steric stabilization. Everett has provided a useful introduction to the electrical double layer, as well as the possible roles that polymers can play in either coagulating or stabilizing colloidal dispersions, depending on their molecular size and structure.
Calcium aluminate cement (CAC)
Formerly referred to as high alumina cement (HAC), calcium aluminate cement was developed in France in the early years of the twentieth century to meet several needs rapid hardening, concreting in cold conditions, sulfate (sea water) resistance and refractoriness. The major product, Ciment Fondu, is manufactured by total melting of a mix of a ferruginous bauxite and limestone at about 1700° in a reverbatory, open hearth furnace. A typical major oxide analysis of the product is SiO2 4.5% Al2O3 38% CaO 38% Fe2O3 10% FeO 6%. The principal phases present are CA (40 50%) and a ferrite phase (20 40%), the A/F ratio of which varies from grain to grain but is generally less than 1. Solid solution effects are extensive and minor phases include C2S and C2AS. Some C12A7 and a glass may also be present. The minor iron containing phases present, which include FeO (wusite), depend on the ratio of Fe2+ to Fe3+ and they give the clinker its black colour. The cooled, solidified product is ground without addition to a fineness of 300 400 m2/kg.
Concrete made with Ciment Fondu, unless very vigorously mixed, usually sets more slowly than that made with Portland cement but it can develop the same strength in 1 day as that found in Portland cement concrete in 28 days and the heat generated aids hydration in cold conditions. Strength is mainly derived from the rapid hydration of CA which occurs by a process of dissolution followed by nucleation and precipitation of a crystalline hydrate. The hydrate precipitated depends on the temperature of hydration. If C12A7 is present, C2AH8 is formed even at the lower temperatures. The hydration of calcium aluminate cements involves the conversion of Al from the 4 coordination in the anhydrous phases to the 6 coordination present in the crystalline hydrates and AH3, so that it can be followed by 27Al NMR.
Where AH3 is an aluminium hydroxide gel which subsequently crystallises to form gibbsite. However, following the conversionreaction by synchrotron radiation energy dispersive diffraction, Rashid el al. found that in the temperature range 70 90°, C2AH8 was an initial, transitory intermediate. In a later investigation of the reaction at 50°, they found that both crystal forms of this phase were formed sequentially. The direct conversion to C3AH6 was only observed after its nuclcation by the indirect route.
The reaction in Equation (1) is accompanied by a significant increase in the density of the solid phases and, consequently, a marked increase in the porosity of the concrete and reduction in its strength result, although some of the water formed may react with remaining anhydrous cement to give a modest subsequent increase. The design of a structure can be based on the converted strength and the value for this can be optimised by using a water/cement ratio of less than 0.4. This produces a concrete in which the reduction in strength is not too severe but ensuring adequate strength by use of such low water contents has been considered too difficult for normal working conditions. In a humid environment serious problems have also arisen where CAC concrete structural members have been in contact with Portland cement concrete. Loss of strength in the former is the result of alumina leaching from the CAC by soluble alkali from the Portland cement (alkaline hydrolysis) and loss of integrity in the concrete is accelerated by simultaneous carbonation.
Calcium aluminate cements are now primarily used in the UK for non load bearing and refractory applications. The more refractory cream and white grades are produced by a clinkering process using low iron content bauxites, or gibbsite from the Bayer process, and either a pure limestone or quick lime. They may contain CA2, CA6, and alumina itself, as the proportion of the latter is increased in the raw mix to raise the refractoriness of the product.
Characterisation of Portland Cement Clinker
Introduction
Provided that the kiln feed has been properly prepared with a regularly checked composition and fineness, then it is only necessary to check the adequacy of the burning process and this is done by frequent determination of the free lime in the clinker. However, as an additional check, elemental chemical composition is determined regularly by X ray, fluorescence analysis. The sulfate content of the clinker is also frequently determined so that the level of gypsum to be added in cement grinding can be calculated. In this chapter the application of a number of methods of phase characterization will be briefly described.
In routine checks on the kiln product, the nodule size grading is watched for variation from the norm. Well nodulised clinker with a minimum of dust is needed for efficient air flow in the cooler. Nodule formation and densification also indicate a high degree of chemical combination, although if the burning temperature is too high the relationship breaks down and dust is formed. For a given burning time and temperature the clinker size range and density also depend on the silica and alumina ratios of the kiln feed. Portland cement clinker is normally black and the appearance of a paler than usual product has in the past been taken to suggest excessive chemical reduction of Fe3+ in the kiln. However, Scrivener and Taylor found significant reduction in some black samples of clinker and, perhaps more surprisingly, they also found that a lighter colour did not necessarily mean that reduction had occurred. Very pale clinker indicates serious under burning.
When problems arise, indicated either by the checks mentioned above or by test results obtained with cement produced from a clinker then a more fundamental investigation is necessary and clinker phase composition and microstructure are examined, usually by optical microscopy. Scrivener has provided a useful introductory account of both optical and electron microscopy and their application to cement and concrete.
Chemical analysis by selective dissolution
Free lime can be extracted from ground clinker (or cement) by treatment with a hot glycerol ethanol mixture or by hot ethanediol (ethylene glycol). The latter is more usual and the extraction produces a solution of calcium glycollate which, as the salt of a very weak acid, can be titrated after filtration with a standard solution of HCl using a methyl red bromocresol green indicator. Calcium hydroxide present is also extracted.
The total silicates plus free lime in a clinker can be determined by dissolving them from a ground sample using either 20% salicylic acid or 20% maleic acid in dry methanol and weighing the washed and dried residue which consists of C3A, ferrite, MgO and sulfates. From this assemblage it has been suggested that C3A can be determined by dissolving it in 3% aqueous saccharose solution, although the ferrite phase is also slowly attacked and the soluble sulfates would be extracted. Selective dissolution is a considerable aid to qualitative phase analysis and characterisation by X ray diffraction. Gutteridge used a solution of sucrose in aqueous potassium hydroxide to dissolve the aluminate and ferrite phases in order to identify the forms of alite present in cement samples.
Optical microscopy
Optical microscopy has proved very effective in characterising cement clinker because much of the microstructural detail occurs from 1 mm upwards. Sampling of a clinker requires great care since only a very small proportion of any batch can actually be examined under the microscope and clinker from each of the range of sieved size fractions should be included. Pieces of lightly crushed clinker and fines are stabilised by vacuum impregnation with an epoxy or polyester resin, which is hardened in situ.
Although a thin section of a specimen is sometimes examined in transmitted light, which makes the determination of optical properties easier, polished surfaces are more frequently examined. Nitric acid in ethanol and hydrofluoric acid vapour are common eichants. Both attack the silicates at rates dependent on their silica content, forming surface films which in reflected light produce interference colours. These provide good contrast between the di and tri calcium silicates, the colours depending on the thickness of the reaction layer and therefore varying somewhat with expo sure to the etchant and the reactivity of the individual crystal. The use of HF also allows the (more difficult) distinction of free lime, periclase, C3A and the sulfatc phases with care. Differences in reflectivity and morphology are also useful aids to identification. Reflectivity is greatest for the ferrite phase.
Microscopy is used to examine clinkers qualitatively and, less frequently, quantitatively. Even without point counting of individual phases, an experienced microscopist can make a very useful assessment of the proportions of the phases present. However, the principal value of microscopy is in detecting non equilibrium effects such as the heterogeneity of phase distribution porosity, crystal form and size. The size of the crystals of the interstitial phases, for example, may indicate the cooling rate experienced by the clinker. Maki and his co workers related the crystal size and fine structure of alite to the conditions it experienced in the burning zone of a kiln and the microstructure of belite to the cooling regime.
Microscopy is also valuable in a number of other ways in the cement industry, in particular in the identification of the mineralogy of build ups and deposits which form in kiln and preheater systems and in the examination of clinker refractory interactions in kiln coatings.
Characteristics of the principal clinker phases
Alite (C3S density 3150 kg/m3). Crystals of alite are prismatic, sometimes pseudo hexagonal, frequently having clearly defined faces. However, less regular shapes with rounded corners and faces with re entrant angles are found. Internal structure observed includes lamellae and other manifestations of twinning and inclusions of belite are common. Thin sections can be used to distinguish trigonal, monoclinic and triclinic forms but to distinguish between variants with the same symmetry may require X ray diffraction.
Sizes of crystal sections within a specimen usually range up lo 100/mm with an average around 30 mm. In general, the size distri bution in a clinker is not random. Smaller alite crystals form in the lime rich regions while larger crystals form in the silica rich regions of a nodule by conversion of belite. Crystal size is also influenced by burning time and temperature, melt quantity and composition, and, if present, mineralisers such as fluoride ions.
Twin planes or fault lines may be marked by material exsolved from solid solution during cooling. For example, Fe3+ may be formed on cooling as striae of C2F by oxidation of Fe2+ present in solid solution at clinkering temperatures. If clinker cooling is too slow, dissolved iron can significantly accelerate the decomposition of alile below 1250°. Scrivener and Taylor described the complex sequence of reactions which can occur between 1200° and 1100º. When severely reducing conditions exist in the burning zone (ca. 1400°) alite absorbs Fe2+ and S2- from the melt, the latter being derived from the sulfur present. In the absence of oxidation during cooling, some of the alite decomposes with the formation of belite, free lime and CaO FeO solid solution. Subsequently, an aluminium rich ferrite and metallic iron are formed. Reduction can result in a serious fall in cement quality but is minimised by proper flame control. This problem is usually suspected if yellow brown clinker nodules, or nodule cores, are produced and free lime rises significantly.
Belite (C2S density 3280 kg/m3). Belite usually occurs as rounded crystals, frequently made up of lamellae or marked by striations indicating twinning and/or exsolution from solid solution, both effects resulting from polymorphic transformations. Belite crystals may occur in clusters which, if large, indicate an unacceptable level of heterogeneity in the kiln raw feed. If a belite cluster is centred on a pore, a site formerly occupied by a silica particle is indicated. In a clinker from a coal fired kiln, inefficient dispersion and incorporation of the ash can be detected as a localised concentration of belite, occasionally in the outer part of a nodule.
In a slowly cooled clinker, small belite crystals separate from the melt. When the A/F ratio is high they are also formed by interaction between alite and the melt, typically fringing alite crystals. Together with the crystal size of the aluminate and ferrite phases, these features of the clinker microstructure are indicators of cooling rate.
Interstitial phases. Tricalcium aluminate (density 3030 kg/m3) and the ferrite phase crystallise from the melt with a degree of separation which increases with decreasing rate of cooling. The crucial cooling is thai occurring in the kiln itself, that is above about 1300°. If clinker is air quenched from the burning temperature then these individual phases may not be resolved by optical microscopy. Normally C3A is cubic but, in the presence of alkali not combined with sulfate, alkali containing solid solutions may form. The alkali is retained by the C3A during cooling and part or all of the resulting solid solution will be orthorhombic. This phase grows as elongated, lath like crystals and exhibits birefringence.
The ferrite phase (density in the region of 3700 3900kg/m3 variable as a result of variation in its composition) is distinguished by its colour in thin section and its high reflectivity in polished section. The black colour of clinker is believed to result from elements such as magnesium, titanium and silicon in solid solution in the ferrite phase and oxidation as it cools in the kiln cooler system.
Minor phases. Magnesia When the magnesia content of a clinker is greater than that which can be taken into solid solution (ca. 1.5%) periclase crystals may be observed, sometimes well formed and in clusters. More often it is observed as relicts of a dolomitic material in the kiln feed or present in the coal. Alkali sulfates and double sulfates calcium langbeinite (K2SO4.2CaSO4) the sulfates are the last phases to crystallise during cooling of the clinker.
They may also condense directly from the vapour phase. A solid state phase inversion of potassium sulfate leads to characteristic cracking. If an aqueous etchant is used, these phases are removed leaving pores. Free lime free lime is often globular as a result of attack by the melt but where interaction is limited it exists as relicts of the original calcite particles, usually surrounded by alite crystals.
Quantitative determination of phase composition
Using a suitably prepared polished section, point counting with a crosswire and steps of a fixed size in a grid pattern is employed with each field examined in the optical microscope. The phase observed at each step is noted and some 2000 4000 counts are made per clinker specimen. Automatic, instrumental counting may be used as in metallurgical analysis, but a number of different etches may be needed to obtain the range of contrast required for the instrument to operate satisfactorily. Counting gives volume percentages of the phases present which are converted to mass percentages using individual phase densities. The principal sources of error in quantitative microscopy derive from the difficulty in obtaining a representative sample and incorrect phase identification, especially where the crystals of the interstitial phases are very small.
The Mineralogy of Asbestos
INTRODUCTION
The asbestos minerals have an almost unique combination of physical and chemical properties. They take the form of extremely thin and flexible fibres, which at the same time have great thermal stability and tensile strength. In appearance the fibres are often silky, and their flexibility is such as to allow them to be spun into yarn and subsequently made into woven fabric. The most widespread modern uses of asbestos are in fireproof textiles, papers and boards and in brake and clutch linings for many kinds of vehicle and machinery. The low thermal conductivity of asbestos products can be as important as their incombustibility for fire prevention. Composites of asbestos with cement, resins, and plastics are commonly manufactured.
The term asbestos derives from the Greek meaning inextinguishable and was perhaps used by the ancient Greek writers in the sense of indestructible to describe material which was unscathed by flames. The first use of the term in the context of mineralogy came in the middle of the 19th Century as applied to a fibrous amphibole mineral discovered in the Italian Alps.
The three main kinds of asbestos which have had wide commercial exploitation are chrysolite, amosite, and crocidolite. Of these, chrysolite is by far the most abundant and most used. Three others, in decreasing order of importance, are anthophyllite, tremolite and actinolite. The last material, although not uncommon as a mineral, is rare in asbestiform habit.
Of the above six varieties of asbestos, chrysotile alone is a member of the serpentine group of minerals, all of the others belonging to the amphibole group. Chrysotile has yellowish or greenish white fibres which are usually silky in nature, crocidolite is blue and less silky, and amosite has white, grey, pale yellow, or pale brown fibres which are more brittle than those of the former two varieties. Typical specimens are shown in the frontispiece.
Amosite is not strictly a mineral name but is a commercial name derived from the initial letters of Asbestos Mines of South Africa. Amosite usually consists mainly of the amphibole mineral grunerite, but appreciable amounts of other amphibole fibres are present in some specimens.
Silicates other than serpentines and amphiboles, and also some non silicates, can occur in asbestiform habit, but they are not exploited because of their limited occurrence or inappropriate physicochemical properties.
DEFINITIONS
Some would confine the use of the term asbestos to material which, if present in sufficient amount, would be commercially exploitable because of the special properties mentioned above, but it is probably difficult to exclude specimens which fall slightly short of this because of greater coarseness, brittleness, or weakness of fibre.
The most abundant minerals which could be termed asbestos are chemically hydrous silicates. Those hydrous silicates that are asbestos are composed of particles which are extremely thin and so generally have a very high length to breadth ratio. We are concerned, therefore, with particle morphology and a property known in mineralogy as crystal habit. Figure 1 shows a range of crystal habits and their descriptive terms, namely tabular, equant, prismatic, acicular, and asbestiform.
The limiting lengths to breadth ratios that differentiate between these terms have not been precisely defined, and it has not in general been thought necessary to do so. Thus, it is a matter of subjective judgment as to whether a crystal is described as acicular (meaning needle like) or prismatic, but it is doubtful if anyone would use the term acicular for material with a length to breadth ratio much less than 10 1.
Since the fundamental fibrils of asbestos have typical widths in the range 20 200 nm, a fibril of any appreciable length will have a very high length to breadth ratio. Thus, for example, assuming an average width of 100 nm, a fibril of length 1 mm will have a length to breadth ratio of 10 000 1. Even a fibril as short as 10 mm has a length to breadth ratio of 100 1, and only a 1 mm length will give, a ratio as low as 10 1. The characteristic feature of high length to breadth ratio is not, however, applicable on a single fibre basis, but it is more useful if applied statistically. In natural untreated asbestos, although some very short fibres will be present, a high proportion of fibrils with a length to breadth ratio of the order of 100 1 or greater can be expected. In samples of milled asbestos most particles were found to have a length to breadth ratio between 5 1 and 20 1, whereas for a non asbestos amphibole the majority had a ratio of less than 3 1. For milled chrysotile asbestos most particles had a ratio of greater than 50 1.
Related to, but not entirely dependent upon, the length to breadth ratio is the property of flexibility. Needle shaped crystals of some substances will be very brittle, others may flex slightly without breaking and spring back elastically, while others can deform inelastically. The term acicular, meaning needle shaped would not normally be applied to crystals that are curved.
A bundle of asbestos fibrils will tend to bend and recover elastically. but when the diameter of the fibre bundle is small so that only few fibrils are involved, then inelastic deformation occurs, giving the curved fibrils common for natural asbestos.
Variations in minerals can be produced not only by the different habits of their component crystals, as described above, but also by the way in which these crystals are aggregated. An important feature of asbestos is that it is an aggregate of hair like fibrils. These have their lengths approximately parallel to one another but lie in more then one azimuthal orientation. More extensive discussion of definitions pertaining to asbestos is given by Campbell et al.
Serpentine minerals
The serpentine minerals, of which chrysotile is one variety, have a layered silicate structure. The layers can be regarded as made up of Si atoms, each surrounded by four oxygen atoms at the corners of an almost regular tetrahedron, and Mg atoms each surrounded by six oxygen atoms at the corners of an approximately regular octahedron. Si O tetrahedra lying on their triangular bases and with apices all pointing in the same direction are linked by sharing all basal oxygen atoms to form a continuous layer. This has approximately 3 fold symmetry but is more conveniently described in terms of a rectangular unit cell which has repeat distances a 5.3 Å, b 9.2 Å. Gaps in the plane formed by apical oxygen atoms are filled with (OH) ions to form a regular and approximately close packed array of (O, OH) at this level. Lying above this plane is an array of Mg atoms and above these a plane of (OH) ions such that the Mg atoms are surrounded by octahedra of (O, OH) ions. Each Mg has a triangle of 3 (OH) above it and a triangle below of 2(O) and one (OH), rotated by 60° with respect to the first. Plan and elevation views of this structure are presented in Figure 2. The chemical content of a unit cell can be seen to be a multiple of Mg3Si2O5(OH)4.
In building up the three dimensional serpentine structure, composite layers of the type described are superimposed, and the inter layer distance is about 7.3 Å. However, one layer may be placed directly above another, or there may be some displacement or rotation between layers. This leads to the possibility of various stacking sequences, different symmetries and multi layer the symmetry may be trigonal, hexagonal, or monoclinic. Disordered stacking can also occur.
The comparatively simple structure described above is that of the serpentine mineral called lizardite. Lizardites are mostly fine grained and have platy morphology, but some have lath like particles with elongation parallel to a.
The asbestiform morphology of chrysotile is not obviously reconcilable with a layered crystal structure, and this paradox has been the subject of much painstaking research. The ultimate and complete solution came with the direct evidence from high resolution electron microscopy that in chrysotile the structural layers are curved about the direction to form either scrolls or concentric cylindrical tubes. The diameters of such tubes are of the order of 200 Å thus; a sliver of chrysotile asbestos with cross section 0.1 mm square contains about 20 x 106 tubular fibrils all in approximately parallel orientation. It is possible, therefore, to strip from an asbestos fibre bundle very fine threads each of which still contains many thousands of fibrils.
A theoretical reason for the curvature of serpentine layers can be found in the composite and polar character of the fundamental sheet. The Si component tends to have a smaller repeat dimension than the Mg component, and the mis match between the two can be overcome by bending, with the Mg octahedra on the outside of the curve. In another variety of serpentine, antigorite, the layers invert at regular intervals thus producing a regularly repeating corrugation. Well formed platy crystals, not tubes, are the result. In lizardites the platy crystals tend to be buckled and disordered, and they are usually limited to very small (sub microscopic) dimensions.
The growth of very thin fibrils is more readily understood for chrysotile than for amphibole asbestos (discussed later). When chrysotile tubes are formed, a particular radius of curvature may be the most stable, and this may impose an upper limit on the diameters of the tubes.
Amphibole minerals
The fundamental unit of the amphibole structure is a chain of SiO4 etrahedra linked by sharing corner oxygen atoms in the manner shown in Figure 4. The characteristic chain formula is Si4O11 and the repeat distance along its length is approximately 5.3Å. The chains are four tetrahedra wide and of very great length fibres of asbestiform amphiboles run parallel to the chain length.
In the amphibole structure the Si4O11 chains are linked laterally by cations as shown in Figure 5. In tremolite the cations are Mg and Ca, the Mg ions linking chains by means of strips of Mg (O, OH) octahedra. The oxygen atoms in these strips are the apices of tetrahedra, and the (OH) ions occur as in Figure 4. The Ca ions link neighbouring chains across the bases of the tetrahedra and they occur in distorted polyhedra of oxygen atoms. The alternative occupation of Mg (Y) and Ca (X) sites in different am phiboles has been discussed above.
Most amphiboles are monoclinic in crystal symmetry, as a result of the way in which successive chain units are stacked with respect to each other. Anthophyllite, however, is orthorhombic, and the relation of its unit cell to that of a monoclinic amphibole is shown in Figure 6. The cell parameters of some amphiboles are given in Table 2.
The chain like structure in amphiboles leads to their possession of good cleavages on (110) planes. If the chains themselves and their linkage across Y sites are regarded as the strongest elements of the structure, paths of weakness can be traced as in Figure 7 and an average path can be taken as defining a likely cleavage plane.
It is tempting to associate the physical nature of asbestos with the chain like structure and with cleavages, but while these features may be indirectly pertinent they are not fully responsible. Thus, many amphibole specimens are not asbestiform although they possess the chain structure and good cleavage. A distinction should be recognized between the process of cleaving fragments from a single crystal of an amphibole, and that of stripping a fibril or bunch of fibrils from the aggregate that constitutes a specimen of asbestos.
The processes that lead to growth in asbestiform habit are not clearly understood. To form asbestos there presumably has to be multiple nucleation followed by relatively rapid growth along the fibre direction and very limited growth at right angles to it. Such a process might be influenced by chemical factors, and it is noteworthy that the class of amphibole called hornblendes, in which Al is an important substituent for Si in the structural chains, are not often found with fibrous and still less with asbestiform morphology. Major element chemistry cannot, however, be the only factor, since we find among tremolites some which are and some which are not fibrous or asbestiform. One can only conjecture that other factors such as the temperature regime during crystallization, trace element chemistry, speed of growth, or a combination of these, exert an influence on crystal habit.
The question of variations in nature and properties from one amphibole to another is worthy of further discussion. Although the crystal structures of all amphiboles are broadly the same, these structures are derived (by X ray diffraction methods) from specimens which, although small, contain about 1015 unit cells. They are therefore average structures. Real crystals do not have the ideal exact repetition of unit cells, but contain defects of various kinds, and the abundance and distribution of defects are known to influence mechanical properties, and may also influence morphology and surface properties. Structural detail at this level is not easily detected by X ray diffraction, but may be investigated more readily by electron microscopy.
An important kind of structural imperfection in amphiboles is the stacking fault. In a perfect crystal of a monoclinic amphibole, slabs of structure parallel to (100) are stacked alongside one another with regular and identical displacements. In a faulted structure, errors in the direction and magnitude of these displacements occur and the frequency of such faults can vary from one specimen to another. When the faults are relatively infrequent, the result can sometimes be described as a twinned crystal. When the faults are frequent and regularly repeating, they are no longer really faults but are the regular displacements of a structure with larger unit cell and perhaps different symmetry. This latter condition describes approximately the relationship between monoclinic and ortho rhombic amphiboles.
Monitoring and identification of airborne asbestos
INTRODUCTION
Airborne asbestos dust is usually monitored for one of three reasons. Firstly, large numbers of samples are taken to check compliance with legislation. As part of the standards issued by most controlling authorities recommended methods are described by which this monitoring should be done. Secondly within the asbestos industry regular sampling is carried out to determine the efficiency of dust suppression equipment. Here it is frequently necessary to know only the relative amount of dust present, and direct reading dust monitoring instruments play a key role. Finally, an increasing number of samples are taken for epidemiological purposes. For this it is essential that standard methods be used which can be related to one another, and which remain constant over many years. This work includes monitoring the exposure of people outside the asbestos industry and may involve measuring extremely small amounts of asbestos.
In practice, it is not possible to obtain an absolute measure of the dust inhaled and retained in the lung. With any sampling method that is adopted, therefore, there will be inherent errors that must be understood and allowed for in the interpretation of the sample results. Care must be taken to ensure that the samples obtained are representative of the airborne dust at the sampling point, and are sufficient in number so that variations of concentration with time and space can be allowed for. Initially, dust sampling instruments were either too heavy to be easily portable (e.g. the thermal precipitator) or only took very short duration samples (e.g. konimeter, Draeger pump). The long period sampling instruments are normally used to monitor the environmental air, and take what are commonly called static samples. The short period or snap samplers operate most satisfactorily in high dust concentrations, where a measurable amount of asbestos can be captured. Neither static nor snap samples may be representative of the airborne dust breathed by exposed personnel, however. Hygiene standards normally refer to Threshold Limit Values (TLVs). These are the levels at which it is believed that nearly all workers may be repeatedly exposed throughout a 40 h week without adverse effect. Many asbestos workers do a series of different jobs throughout a shift and may be exposed to dust for only relatively short periods. In such instances, and indeed in the majority of situations, a sample taken with the instrument attached to the person being monitored has distinct advantages. This provides what is known as a personal sample. In this case the instrument should be completely portable and the air sampled in the breathing zone of the person concerned. Normally this is taken to be anywhere within 30cm of the ori nasal region.
In order for the results of the monitoring to have any meaning, it is necessary that they provide a consistent measure of the dust of pathogenic significance. When evaluating a potentially hazardous dust either its mass or number concentration is normally measured. The mass can be either that of all the airborne particles (total dust), or that of the proportion thought to be capable of reaching the lung alveoli, commonly referred to as respirable dust. For coal dust the work of Jacobsen et al. has shown that the mass of respirable dust present in the air corresponds much more closely with radiological change than does particle number. This work was based on a long term study in British coalmines, but unfortunately similar epidemiological data are not available for asbestos exposure, and both fibre mass and number concentration measurements are therefore used. The total mass of airborne dust is monitored in the U.S.S.R. provided that the asbestos content is greater than 10%. This, however, creates a difficulty when interpreting the results, because whereas in an asbestos textile factory most of the dust will be asbestos, in mining or the asbestos cement industry the majority of the dust may be other material. A more satisfactory method of evaluation is to determine the weight of asbestos in this total dust sample using chemical or physical methods of analysis. This, however, will monitor all asbestos fibres, including those too large to penetrate into the alveoli. One of these fibres could have a mass of many times that of the potentially harmful dust in the sample, and so this method in fact only provides an upper limit to the amount of asbestos present. Alternatively, instruments containing particle size selectors, known as elutriators can be used for sampling the dust. These were designed, however, using the aerodynamic properties of spherical particles and rely on the larger particles separating out by virtue of their higher falling speed. At present, however, the sampling characteristics of these instruments with fibrous dusts are not fully understood and an additional degree of uncertainty is introduced when they are used. One of the alternative German Federal Republic standards for asbestos, however, requires the use of one of these instruments. Once again physical or chemical analysis can be used to determine the amount of asbestos present.
The measurement of the airborne fibre number concentration is normally preferred in most countries. This avoids some of the difficulties and uncertainties of mass monotoring by using a microscopical method to separate out only those particles thought to be potentially harmful from a sample of all the dust from the air. The results obtained, however, depend upon the microscopical magnification and technique used and also upon any restriction placed upon the type of particle counted. Some early attempts to monitor the exposure of asbestos workers recorded all the particles thought to be respirable, but most present methods involve counting only those fibres within a limited size and shape range. One advantage of this method is that, unlike the mass standard, the 2 fibres/cm3 British Standard for chrysotile asbestos (which has been adopted by several other countries), is based upon an epidemiological study. This type of measurement is also likely to provide the basis of any revised standard within the foreseeable future.
Outline of technique
Measurements are taken by drawing a known volume of air through a membrane filter. This filter is then made transparent, and the number of fibres fitting a standard definition of size and shape which are in the deposit are counted using a phase contrast microscope. The mean fibre concentration during the sampling period can then be calculated. Where fibre identification is needed, different types of sampling filters and analytical techniques may be required.
It is also possible to use an entirely different technique, with a green or black membrane, which is fixed with Perspex but not cleared. This is then mounted without a coverglass and examined by incident light using a 4 mm objective and a suitable vertical illuminator (such as the Cooke Universal Illuminator). Tests seemed to indicate that this method gives similar results to the usual phase contrast method, but in general it has not found wide scale usage for asbestos monitoring, probably because of the increased difficulty in setting up and operating the microscope.
Definition of the Fibres which are Evaluated
The membrane filter method was developed by the British Asbestosis Research Council in order to try to monitor only those fibres thought to be capable of causing lung damage. At the time of its inception, asbestos or ferruginous bodies were thought to play a major role in the development of asbestosis. These bodies are fibres surrounded by protein and iron frequently found in the sputum of asbestos workers. These fibres are normally longer than 10 mm and it was therefore concluded that it was necessary to evaluate only the longer fibres. Initially it was decided to monotor those fibres between 5 and 100 mm in length, although this upper limit is no longer applied in most countries, except South Africa. As only a relatively small proportion of airborne fibres are longer than 100 M.m, however, the difference is insignificant. In addition, at the time of inception of the membrane filter method, it was thought that the number of longer fibres was a constant proportion of the total number, but it has since been found that some processes, notably carding, preferentially produce longer fibres. More recent experimental evidence, however, has shown that the longer fibres (>10 mm) are in fact the most potentially dangerous, in particular in the development of cancer, snowing that the limitation of monitoring to the larger fibres was probably justified.
An added advantage of monitoring the longer fibres was that this can be done by optical microscopy, thereby avoiding the sophisticated electron optical equipment that is required in order to detect the smaller fibres. It was also found that by restricting the type of microscope and range of magnification used, good agreement could be obtained between different laboratories evaluating the same samples. A magnification of approximately 500x was chosen in the U.K. With increasing magnification more fibres are seen, but the level of inter laboratory agreement depends more and more upon the quality of the microscope and the skill of the observer. The lack of ability to see all fibres does not invalidate the method, however, as it is only required to produce an index of the hazard, which may in fact not be proportional to all the fibres present. In order to distinguish a fibre from the other dust in the sample, the definition of a fibre having a length to diameter ratio (i.e. aspect ratio) of at least 3 1 was chosen. This is now used internationally by occupational hygienists when monitoring fibrous dusts, whereas 10 1 is more commonly used by engineers and fibre technologists.
In order to cause disease, the airborne particles must be capable of penetrating into the lung. Many, however, are caught in the nose and larger airways by sedimentation or impaction. For spherical particles a generalization of experimental measurements of the proportion of airborne particles of different sizes able to reach the alveoli produced a plot known as the Johannesburg curve. Here the size of a particle is expressed in terms of its aerodynamic diameter, that is the diameter of a sphere of density 1 g/cm3 with the same falling speed in air as the particle. According to this convention, all particles larger than 7.1 mm aerodynamic diameter and about 50% of the 5 mm aerodynamic diameter particles are removed before reaching the lungs. Fibres, however, behave very differently aerodynamically, and it has been shown that those with high aspect ratios fall through the air at a rate which is proportional to their diameters, but independent of their lengths, and that asbestos fibres have aerodynamic diameters approximately three times their actual diameters. Asbestos fibres with high aspect ratios and with diameters greater than approximately 2.5 mm are therefore unable to reach the alveoli. When monitoring asbestos, the diameter limit is made greater than this so as to ensure that all of the respirable fibres are evaluated. In Australia, Belgium, France, Sweden and the U.K., only fibres with diameters less than 3 mm are evaluated, whereas a 5 mm limit is used in South Africa and Finland. Canada, Denmark and the U.S.A., however, do not place any restriction on the diameter of the fibres counted, provided that it is less than one third of its length. A summary of the definitions of a countable fibre used in different countries, together with the standards to which they apply, is given in Table l.
A major advantage of using the microscope to select a definite size range of fibres is that this permits the operator to sort by size and exclude fibres not considered to be respirable. All fibres can therefore be collected on the filter and there is no necessity for a pre selector such as a cyclone or horizontal elutriator, as used for spherical dust. This is important as a satisfactory air elutriation method for asbestos fibres has not yet been reported.
A further constraint placed by some codes of practice for evaluating asbestos is that the fibres fitting the above classification shall be asbestos. This can lead to difficulties as some non asbestos fibres have the same morphology as asbestos fibres, for example organic fibres and chrysotile, and gypsum fibres and amphibole asbestos. In a comparison of the counting procedures in nine countries, it was noted that many laboratories counted all fibres fitting the size definition, whereas others limited them selves to those fibres whose morphology appeared to be that of asbestos.
The membrane filter
Cellulose ester filters are normally used for asbestos fibre monitoring. The Asbestosis Research Council in the U.K. recommends the use of 0.8 5.0 mm pore size filters, whereas the Australian code of practice suggests that only a pore size of 0.8 mm should be used. Although many asbestos fibres have diameters much less than this, they are in fact captured by the filter and the optical fibre count is not affected by penetration. Smaller pore sizes (e.g. 0.2 mm) may be used when it is essential for all fibres to be retained on the filter surface, for example when coating for transmission electron microscopy. These samples are normally restricted to special situations, as a large pump is required to overcome the pressure drop across the filter. When it is necessary to evaluate a sample by scanning electron microscopy, Nuclepore filters are frequently used. These are manufactured by etching the tracks of high energy particles through polycarbonates, and have a very smooth surface, giving a uniform background on which the fibres can be easily identified. These filters, however, have different filtration characteristics from the cellulose ester membranes, and very significant fibre penetration occurs when sampling at pore sizes of 5 mm and above.
Filters of diameter 25 and 37 mm are most commonly used, and sampling filter holders are available commercially. Smaller 13mm filters, however, are being developed for some uses in particular where there is very little dust or where the sampling volume is limited. These, however, present difficulties at high sampling rates where larger pore size filters may be required to overcome the increased pressure drop. Some laboratories prefer cellulose ester membrane filters with a gridded pattern printed on the surface. This enables the microscopist to find more easily the plane of the dust deposit, which may be difficult for sparse samples.
Sampling
The filter is placed in a holder, where it is supported by a gauze or thick pad, which helps in controlling the distribution of air through the filter. The 25 mm Gelman holder is normally used with the filter surface completely exposed. The 37 mm Millipore holder can also be used in this way, or alternatively the face cap may be left in place and the small plug removed. The latter reduces the risk of damage to the filter by large high velocity particles, but has the disadvantage that if any large particles are present, a small portion of the centre of the filter may be obscured by them. This should not alter the count, but the effect can be overcome by drilling six additional 4 mm holes in some face caps and fitting these to the holders during sampling. The addition of a 1¾ in. plenum between the air inlet and the filter has been reported to give a more uniform deposit with this holder. Breslin and Stein measured the collection efficiency of the Millipore holder in still air and found that it was good for respirable sizes of spherical particles. In wind velocities of 2 m/s, however, its sampling characterstics were less satisfactory.
When monitoring asbestos in moving air streams, for example in air ducts of filtration or exhaust systems, isokinetic sampling is required. This involves matching the velocity of the air entering the sampling equipment to that flowing in the duct. A sampling head must therefore be designed so that when it is inserted in the air stream it does not alter the flow pattern.
Personal samples require the filter to be placed in a holder within the operators breathing zone. Various methods of wearing the holder have been evaluated, e.g. using a head harness, plastic jacket, shoulder harness, or lapel filter holder, but no significant difference was found. Normally a shoulder harness or a lapel filter holder is used with the sampling surface facing downwards or vertically. Upwards facing filters are avoided because of the high risk of contamination. In addition, the filter may be protected by a cap. or by leaving the end on the Millipore holder, in order to prevent the contamination from extraneous sources such as dusty overalls. The volume of air sampled is determined by measuring the flow rate of the air through the filter. This can be carried out connecting a variable gap flow merer or bubble meter on the sampling side of the filter holder. Flow meters are usually calibrated at atmospheric pressure and, where they are incorporated between the pump and filter head, allowance must be made for the reduced pressure in which they are operating. Flow meters incorporated in the sampling pumps are liable to error in that they record all of the air drawn through the pump, including any from leaks or bleed systems. The flow rate should be checked at regular intervals and at the end of sampling, so that any changes due to filter blockage or pump malfunctions can be noted. Some personal sampling pumps are, in fact, able to compensate for small increases in pressure drop across the filter. A comparative study of the pumps at present available on the U.K. market has recently been completed.
Alternatives to Asbestos in Industrial Application
Introduction
The numerous applications of asbestos are a consequence of its desirable physical and chemical properties, combined with a low material cost. It is this unique combination that makes the replacement of asbestos very difficult in many applications.
Some of the properties of asbestos are summarized in Table 2, together with the comparable properties of some of the synthetic fibre materials that have been suggested as replacements for asbestos in some applications. Some of these properties require further comment.
Thermal properties
The most widely known property of asbestos is its heat and fire resistance, although this resistance is not as great as is popularly believed. Asbestos cannot be classed as refractory, although normally its properties are sufficient to withstand super heated steam and other high temperature industrial environments. Degradation of the crystal structure of asbestos and major loss of strength occur at temperatures in the range 300 500 °C. However, a useful performance can be obtained at higher temperatures than this specified working temperatures for some asbestos products may be as high as 600ºC. The reasons for this are unclear, but some points of significance are apparent.
Chrysotile contains 14% by weight of hydroxyl groups, which are lost from its structure as water vapour (2OH- > H2O + O2-) at temperatures greater than 450°C. The latent heat of vaporization of this water content is thought to be a potent heat sink, protecting the remaining undegraded fibre. Further, the solid decomposition products are inert and of low thermal conductivity, providing additional protection to the remaining fibres, and maintaining structural integrity. It has been shown that, in some cases, asbestos can maintain its integrity at temperatures up to 1700°C.
Mechanical properties
It can be seen from Table 1 that the values quoted for the strength of asbestos fibres are very high. However, even the average values quoted in the table may not tell the whole story. The measurements of strength are inevitably derived from testing in controlled laboratory conditions, and the values obtained may not be representative. Discussion with suppliers of asbestos has suggested that the reliable strength of chrysotile fibres, as produced and used commercially, is no higher than approxi mately 700MN/m2.
Other properties
Various other properties make asbestos a valuable material. For instance, its resistance to chemical and biological attack is valuable in applications involving hostile environments, and in achieving a useful service life.
The friction and wear characteristics of chrysotile and its thermal decomposition product forsterite, a non fibrous silicate, make chrysotile a widely used material in such applications as friction clutches, brake linings, and bearings.
The high aspect ratio of asbestos fibres makes them useful as a mechanical reinforcement in both polymer and cement based products.
Price and availability
This favourable combination of properties in one material, which is obtainable at a price significantly lower than that of its competitors in specific applications, makes asbestos an extremely attractive material. However, as noted in Table l, prices of asbestos are likely to rise. Further, since asbestos is a limited natural resource, with an estimated resource life of about 25 years, prices are likely to continue to increase, and alternative materials will have to be found on availability and cost as well as on health grounds.
Applications
The breakdown of U.K. usage of asbestos is given in Table 1. The function of asbestos in the different indust rial applications and the alternatives that are available are discussed.
INDUSTRIAL APPLICATIONS OF ASBESTOS PRODUCTS
Asbestos textiles
Chrysotile fibre forms the basic raw material for almost all of the activities of the asbestos textile industry. The length and flexibility of the longer grades of chrysotile are such that spinning into yarn and cloth weaving are possible. Two basic types of yarn are produced plain, possibly braced with an organic fibre and reinforced, which incorporate either wire or another yarn such as nylon, cotton, or polyester. The wire reinforced yarns and textiles can retain their mechanical properties at temperatures up to 600 °C. Recently developed textiles combined with resins and ceramic binders have successfully withstood short term exposure to temperatures up to 2200ºC. The main applications of asbestos textiles are represented in Figure 1. Some of the applications and their alternatives are considered in other sections as indicated.
Fire and heat protection clothing
These garments are manufactured from asbestos cloth which is aluminized to give a heat reflecting surface. The metallic layer is bonded to the cloth by a thermosetting resin.
As an alternative, clothing made from temperature resistant nylon fibre has found application in fire fighting and foundry work and as protective underclothing for racing car drivers. The materials in suitable form can provide short term protection from exposure to temperatures up to 1370°C and also for protection against molten metal impingement. Gloves made from this material are suitable for use with contact temperatures up to 300°C. The product is marketed and manufactured by Du Pont under the name Nomex.
Nomex is suitable for protection against most chemical hazards, with the exception of some strong acids, and can be laundered with little deterioration in properties. Nomex is about 3 times more expensive than asbestos. However, the ability to launder Nomex clothing means that its useful life is much longer than asbestos clothing.
Clothing for heat protection is also made from special wool blends, such as Multitect, manufactured by Multifabs Ltd., Derby. These materials are suitable for direct heat and metal splash protection and have been tested for protection against splashes of molten steel at 1500°C. Aluminized grades are available for greater protection from radiant heat. The materials are resistant to chemical attack and can be laundered with no deterioration in properties. The wool fibres are surprisingly resistant to ignition and flame spread and the clothing is competitive in price with asbestos based products.
Bleached Teflon fluorocarbon fibres, also manufactured by Du Pont, are advertized as being the most fire resistant organic fibre in oxygen rich and high pressure atmospheres. Garments woven from this material have been worn by Apollo astronauts and for missile fuel handling. These suits protect the wearer from extremely high flash temperatures and corrosive missile fuels.
Fire blankets, curtains, and aprons
In general, the materials detailed above are also suitable for these applications. Additionally, a recently developed product, a sandwich of a layer of ceramic fibres between two woven glass cloths has become available from Marglass Ltd. This has been shown to be effective against flame and molten metal hazards. In the latter case, the surface glass layer melts and the ceramic fibres provide the protection. In certain circumstances blankets or rolls of mineral wool or ceramic fibres may be used, although these may tend to disintegrate more readily than the woven products.
Electrical insulation
Most cases in which asbestos textiles are employed for electrical insulation also demand a degree of thermal and/or chemical protection.
Filters
Asbestos cloths are widely used for filtration of bulk liquids such as beer, which has latterly been the cause of some concern in relation to possible cancers of the stomach and gut. Expanded Perlite has been used successfully as a substitute for asbestos in some applications. Vermiculite products are similarly employed.
Filter bags of woven fabrics and needled felts of 100% Teflon fluorocarbon fibres are used in many filtration applications where temperatures up to 300 °C coupled with corrosive chemical environments are encountered.
Ropes, yarns, tapes, etc.
In general, satisfactory substitution of asbestos may be made with glass for many of the applications of these materials, provided that the softening of the glass at about 300 °C is not significant. For higher temperature applications, textile forms of the continuous ceramic and silica fibres may be suitable replacements if price permits.
Other applications
Other textile applications for which Nomex is a suitable substitute include press covers and pads for laundry dry cleaning and the textile industry, racing car seats and iron rests on ironing boards. Teflon is employed for wicking felts and fuel cell membranes. The small number of minor textile applications for which no satisfactory alternative exists at present include lamp and stove wicks, wipes for molten metal, diaphragms for some of the electrolytic cells currently employing asbestos, and some filter cloths.
Getting, Cleaning, and Delivering the Clay
Clays and shales may be obtained by (a) open working a pit, hole or quarry, if the material is sufficiently near the surface, or by (b) mining, if the material is deep seated. The mining of clay is usually subsidiary to that of coal, and may be regarded as outside the scope of the brick manufacturers work, who usually purchases the mined clay when it has been delivered to his works or erects the latter near a heap of mined clay, shale, or blaes. For this reason, the mining of clay is not described in this volume the methods used are practically the same as those for coal and described in text books on coal mining.
In open working, the clay or shale is obtained from the quarry by digging by hand, or by a mechanical excavator, with or without the aid of explosives to loosen the harder portions of the material. As in most clay deposits the composition of the bed varies at different parts, it is necessary to exercise much care in choosing portions of the bed from which the clay or shale has to be taken. It is, therefore, usual to work horizontally in a series of terraces or steps, each step being the height of the particular strata worked, but conditions vary so in different deposits that each brick manufacturer must, to a large extent, be left to use his own judgment in the matter. Care and attention are required if the clay hole is to be worked economically, as otherwise a large amount of useless material may be shifted. Water in the clay hole is often a source of trouble, as its removal entails considerable expense. This can be minimized if the clay or shale is excavated in such a way that the bottom of the working slopes to a well or sump, into which all the water can drain rapidly and be pumped out as fast as it accumulates. A little foresight in arranging this sump and the necessary drains leading to it will save much trouble, expense, and annoyance at a later stage. As the water will contain clay in suspension, a form of pump suitable for dirty water should be used.
When steam can be carried to the clay hole, a Pulsometer pump is the most suitable, as it can deal with very dirty water and has no wearing parts otherwise some form of diaphragm or centrifugal pump should be substituted. The ordinary types of pump with valves, whilst excellent for clean water, are not desirable for use in clay holes,
REMOVAL OF OVERBURDEN
The first step in opening a new quarry or extending an existing one is to remove the overburden, fen, or callow, i.e. the soil and other material lying above the brick earth. Some manufacturers of common bricks do not take this precaution, and so reduce the quality of the bricks.
The overburden need not consist merely of the soil in some localities it is desirable to remove material 3 ft. or more in thickness in order to expose the brick earth. This is particularly the case where the overburden is of a different nature from the brick earth. For instance, if a clayey Glacial Drift, 4 ft. thick, overlies a compact clay rock 50 ft. or more in thickness, it will be profitable to treat the whole of the Drift as overburden and discard it otherwise it will spoil the bricks made of the underlying material, as the Drift requires a different type of machinery from that most suitable to the compact clay rock beneath. Failure to recognize this has ruined several brickworks.
The overburden should be removed so far back from the face that there is no risk of any overburden falling down the face, should a slip or landslide occur. The minimum distance from the face will depend on the nature of the brick earth it is usually wise to have a strip not less than 30 ft. in width cleared of overburden, and to keep this width continuously clear.
Whilst ample surface should be cleared of obstruction it is not satisfactory to clear too large a space, because coarse vegetation may grow on the cleared surface and may cause more trouble and spoil more bricks than if the overburden had not been removed.
The overburden was formerly removed by digging by hand and loading into small wagons, which run on rails and tip the material where it can do no harm. This method is too costly nowadays and mechanical means for removing the overburden are more common.
Any trees of sufficiently large size should be felled, sorted, and sold, or stored for timber. When this has been done any of the following machines may be used according to local conditions
A Plough may be used to loosen the soil if there is no serious quantity of brushwood. It is followed by a bulldozer or scraper to form the loosened material into heaps.
A Tractor rooter may be used when the surface is exceptionally hard or contains trees, roots or scrub vegetation or high boulders, or if the brick earth is hard and dry. This machine will break up the overburden and loosen it, as well as uproot trees and bushes. A gyro tiller can also be used for this purpose. Both these machines require considerable skill and it is usually advisable to hire a man with them for this purpose.
A Scoop or Scraper drawn by a tractor and provided with teeth or blades to loosen the ground and enable a suitable quantity to enter the scoop. A popular type of scraper will hold 9 cub. yd. of clay the tractor hauling it must develop 80 90 h.p. This is too large a machine for constant use in many brickworks and it is therefore, preferable to hire a machine (with driver) when required. On distances of more than 300 ft. rubber tyred machines are often cheaper to use and maintain than caterpillar tractors.
A scraper may be towed by a tractor or mounted on the tractor chassis or it may be connected to the boom of a single bucket excavator (the bucket having been removed) by ropes.
A Bullgrader or Bulldozer consists of a tractor with a powerful blade in front which pushes the loosened material forward so as to form a series of shallow heaps which can then be loaded mechanically into wagons and removed. It is not economical to use a bullgrader for distances of more than 200 ft. Where possible the machine should work down hill.
A Skimmer, which may be regarded as a single bucket excavator arranged to dig horizontally. A Skimmer does not remove the whole of the overburdencleanly and where one is used it is generally advisable to regard the top 3 in. of the clay, or shale as part of the overburden and to remove it. This procedure will often avoid the use of a second machine to follow the skimmer.
A Dragline Scraper, consisting of a scoop, with teeth in fioni. This is attached to two ropes one to pull it to where the material is to lie dut and to cause the scoop to travel forward and fill itself the other rope returns the scoop to where it can be discharged.
A Single bucket Excavator or Navvy may be used where the overburden is sufficiently thick to justify its use.
A small Multi bucket Excavator may be used for a friable, fairly thick overburden not less than 3 ft. or more than 8 ft. thick. Under suitable conditions the overburden may be removed the same time as the clay, yet each be kept separate, by attaching a separating device to a multi bucket excavator.
When a Ladder Excavator is used it can sometimes be provided with a fitment which enables the overburden to be removed separately without a second machine being needed.
Where the overburden is very thick and the output of bricks is very large, as in some parts of the Oxford Clay Areas, it is advantageous to use a conveyor bell, tipping frame or other suitable device to carry the excavated overburden over the face of the pit and to deposit it mechanically some distance away. In smaller works the overburden is usually loaded into wagons and tipped in part of the worked out portion of the pit.
Much greater care than is usual should be exercised in deciding where the overburden is to be tipped. In some works it has been deposited on some of the best clay!
As roots and stones in a brick earth are very objectionable, it is more profitable to remove a small excess of material as overburden than, by attempting an undue saving, to permit roots and stones to be mixed with the brick earth.
Special oversight is needed to prevent the wrong strata becoming mixed with those containing suitable material, particularly at certain stages in the quarrying, but with capable men no special difficulty in this direction need be experienced.
It is very important to see that the overburden is completely removed. Carelessness in this respect may result in many defective bricks being made.
The cost of removing overburden varies very greatly because of the differences in its thickness in various localities. It bears no necessary relation to the number of bricks made, though it tends to be less (in proportion to the output) in the larger than in the smaller works.
When machines of the types described are used, continuous operation by men of sufficient skill is of great importance. With unskilled men the cost can easily be trebled.
As the removal of overburden is an intermittent process sufficient being removed at once to liberate a year or mores supply of clay or shale it is often cheaper to hire an independent firm to undertake this work. In this case, great care should be taken to insure its being properly done or the damage may be serious.
DIGGING AND EXCAVATING
The main bulk of usable clay or shale is obtained by digging or excavating, but fireclays and some colliery shales are mined.
Hand digging. In works with an output of less than three million bricks a year, hand digging is usual, and notwithstanding the fact that hand labour is more expensive than mechanical excavation, a very large proportion of the bricks in this country is made of hand dug clay. This is due to several reasons of which the most important are (i) the output is so small that it is thought it would not justify a mechanical excavator this may be incorrect and is worth further investigation (ii) several kinds of clay have to be dug separately and afterwards mixed in suitable proportions (some may have to be rejected as harmful) (iii) even when the bulk of the clay is excavated mechanically some portions must be dug by hand and separated or they would damage the bricks and (iv) hand digging is often necessary in clearing up part of the pit or in preventing clay lying in a dangerous position from doing harm to workmen or to useful portions of the clay.
Mechanical Excavating. Where practicable, mechanical excavation is much cheaper and is used in most of the larger brickworks.
Excavation in Benches. When the bed of clay or shale is too thick for an excavator to dig the whole thickness at a single cut, the machine may be made to form a series of terraces, or benches, each of suitable height. In a few very large works two or more excavators may work simultaneously at different levels and the products of their total working mixed on arrival at the machine house.
Thin, Shallow Beds. Shallow beds near the surface are best excavated in the same manner as the overburden (p. 41), and thin beds of harmful material lying between thicker beds of usable clay or shale may be similarly treated.
Plastic Moulding by Machinery
Plastic bricks may be made by machinery by two entirely different groups of processes, known respectively as the Machine moulding and the Wire cut Processes, they produce quite different kinds of bricks.
The machines used for shaping bricks in both these groups of processes require to be supplied with a rather soft plastic paste, though in some cases several machines are united in one structure so as to prepare the paste as well as shape it. It is in the use of a somewhat soft plastic paste that the processes described in this chapter differ from those in the three succeeding ones.
THE MACHINE MOULDING PROCESS
The chief purpose of the machine moulding process is to produce bricks as similar as possible to those made by hand, but using machinery instead of skilled moulders. In so far as they are successful they overcome the difficulty of obtaining skilled moulders a difficulty which has been serious for about forty years.
Materials The loams and other mild clays generally used for hand moulded bricks are the most suitable for machine moulding, but many other clays, shales, etc., which can be made into a soft plastic paste can be used satisfactorily.
In the large American works on the River Hudson, the sandy alluvial clay is so wet and soft that it can best be shaped in box moulds. It receives no preliminary treatment prior to entering the large square pug mills which force it into the moulds. Preparation of the Plastic Mass Some clays merely require to be passed through a pug mill with sufficient water, but others require to be weathered or soaked before pugging as these processes add considerably to the cost of manufacture, they are much less used than in former years when labour was cheaper and brickyards were usually small and apparently primitive.
Some tougher materials require to be passed through a pan mill, crushing rolls, and a trough mixer, and then pugged.
Impure clays may need to be washed and if machine moulded London Stocks are required, the clay must be washed and mixed with chalk and soil. Stones may be removed by means of a clay cleaner or by eliminating rolls. Rock clays and shales may be ground to powder, and then passed through a trough mixer and pug mill, in order to produce a suitable paste, though such elaborate plant is unusual in this country. When much mechanical treatment is required in the preparation of the clay, it is usual to shape the bricks by the wire cut, stiff plastic, or semi dry process, and not in mechanically operated moulds.
Whichever method is used, the plastic paste must be as homogeneous as possible, or it will shrink irregularly, and in other ways produce defective bricks.
To Increase the Plasticity of a Clay or Brick Earth, one or more of the following methods may be used
The clay may be wetted and exposed to the weather (see Weathering) Hot water, followed by immediate tempering, is often more effective than cold water, but is seldom used in brickmaking.
When it is difficult to get sufficient water into the clay and the addition of soda or mineral acid is undesirable, the use of a wetting agent is sometimes beneficial. The most useful ones are various alkyl aryl sulphonates, alkyl naphthalene, sodium sulphonate. dioctyl sodium sulphosuccinate and the sodium salts of sulphonated hydrocarbons. These wetting agents are usually sold under registered names. They act by reducing the surface tension between the clay or shale and the water. Very small proportions suffice, sometimes as little as 0.3 per cent. Cationic wetting agents are less useful.
The prepared clay paste (pug) may be kept in a cool place for a suitable time which may be several days or several months according to the nature of the material.
By the addition of a very small proportion of soda or sodium silicate (or both) and mixing very thoroughly with the clay. It is generally better to have the clay in the form of a slip but some clays can he effectively treated in the paste form. An excess of the alkali will do no good and may do harm. A few clays are made more plastic by adding a little acid it depends on the nature of the Base Exchange portion of the clay.
Some clays respond best to the addition of quicklime (only about 10 lb. per ton) followed (after thorough mixing) by a solution of aluminium sulphate.
By adding gum, humus, starch or other suitable colloidal matter. The addition of bentonite is being increasingly made for this purpose in other branches of clayworking.
The addition of a very small proportion of barium chloride, previously dissolved in water, to some clay will increase the plasticity by making any soluble sulphates insoluble. At the same time this substance reduces the tendency to efflorescence or scum. Barium hydrate is even more effective but is less soluble in water.
MOULDING MACHINES
The machines for moulding bricks are always supplied with box moulds, each capable of holding four, six, or eight bricks. These moulds are filled by forcing the plastic paste into them by means of either a pug mill or specially shaped rotating knives. The mode of filling is crude, there is always a tendency for it to be incomplete and for the inclusion of air within the bricks, but in the machines described in the following pages these disadvantages have been overcome sufficiently to make the manufacture of bricks by this method satisfactory and simple, at any rate as far as certain mild clays are concerned.
It is possible to prepare a plastic paste from almost every kind of clay but for economic reasons the use of moulding machines is largely restricted to mild clays which need little or no preparation.
The machine moulding process is not as cheap as the wire cut process, but it has two advantages over the latter (i) it can use a much softer paste and is, therefore, available for some river muds for which the wire cut process is unsuitable and (ii) it produces bricks which more closely resemble hand made ones and, therefore, fetch better prices. The very low costs often mentioned in connection with the soft mud process in America are not due to superior machines, but to the large extent to which mechanism replaces manual handling. This reduces the cost enormously so far as actual running expenses are concerned, but the capital charges are so great that mechanical handling would not pay in this country, except on much larger outputs than are customary here.
In many districts the wire cut process of brickmaking is displacing moulding machines, though where a facing of sand on the bricks is demanded the latter machines, or hand moulding, must be used.
All machines used for moulding sand faced bricks must be provided with a safety release which comes into operation when stones or other causes of excessive pressure occur. Otherwise, the machine will be damaged, and however desirable a machine may appear to be in other respects, the absence of some form of effective relief escapement should be regarded as sufficient to condemn it.
One of the earliest machines to imitate hand moulding was patented by R. A, Morris in 1899. The first machine was in full work for more than 35 years after its installation, but many changes and improvements were made in later machines of this type. It was made for many years by the Brightside Foundry & Engineering Co., Ltd., but this firm no longer makes such a machine.
In the Morris Machine, the clay is given a light pugging and is forced below a descending plunger which fills a mould or series of moulds at a stroke and compresses the paste. On the plunger rising the mould is pushed forward, the surplus paste is struck off, and the mould is bumped to loosen its contents which are then turned on to pallet boards.
Each mould makes three bricks at a time, the patentee claiming that this is better, with his machine, than producing a .larger number simultaneously. Ample time is allowed for the operation of cleaning, sanding, and replacing the moulds and effectual means are adopted for preventing the clay from displacing the sand as the former enters the mould. The Morris machine requires about 3 h.p. to drive it and can make 8000 bricks per day under normal conditions. It is also suitable for making lightly compressed fire clay bricks. It can be worked by either a horse or an engine.
For many years a popular machine was the Monarch sand faced brick making machine, made by the Monarch Company, but not now obtainable. The upper part of the machine consists of a double pug mill, from which the clay is passed down to a series of knives (known as wiperx or presses), and delivered to the moulds immediately beneath it. The action of the presses is somewhat similar to that of the mans fingers and thumbs in hand moulding, and is reciprocating, not rotary. A lad takes the moulds out of a sanding tank, places them at the back of the machine, and after the clay has been mechanically pressed into the moulds in the front of the machine, the mechanism at the back brings another set into position under the die. A man standing in frontof the machine takes the mould, scrapes of the surplus material with a strike and hands it to another man, who inverts it on to a pallet board which has previously been placed ready and lifts the mould from the bricks. The man then
turns round, puts the mould ready to be re sanded, and seizes another full mould. The pallet and bricks are taken by another man to a hack or dryer, two to six bricks being made at a time. The whole operation is very simple and requires no skilled labour.
The amount of pressure exerted on the clay in the moulds can be instantly regulated by moving a small lever in the front of the machine. This regulation is essential in order to prevent difficulties due to variations in the stiffness of the clay.
Another machine for making bricks of the same type was invented by Berry & Son and is now supplied by John Hart & Co., Ltd.. London, N. 14 this machine consists of a horizontal cylinder with the driving gear at one end and the moulding table at the other. The clay is fed into a hopper near the gear and carried by large pugging knives to the moulds, into which it is forced by eccentrically shaped cams or wipers. These blades avoid a defect in the shape of the bricks, which often occurs when the moulds are filled by a plunger. When one set of bricks has been made a freshly sanded mould is fed into the back of the machine and is automatically pushed into the correct position whilst the full mould is forced out at the front of the machine and any surplus clay is removed by an automatic strike. The bricks are then turned out of the mould on to pallets on the turntable and the empty mould is returned to the sanding box for re sanding. Only two men are needed to operate the machine. The normal output is 1200 bricks per hour.
The machine made by Ernest Hole &. Son, of Burgess Hill, Sussex, forms one brick at a time, and this makes the subsequent bumping of the mould much less laborious than when three or more bricks are made simultaneously. Holes machine consists of a horizontal pug mill, which delivers the clay paste into a chamber provided with a plunger. The mould is carried on a carriage on which is a dummy block, forming the bottom of the mould. The plunger remains stationary whilst the mould is drawn forward. The single machine is capable of producing five bricks per minute or 15000 bricks per week. A duplex machine produces double this quantity.