Cement

Cement

Cement consists of certain anhydrous double silicates of calcium and aluminum, which are capable of combining chemically with water, to form a hard mass. It lliffcn; from lime mortar in that it hardens while wet, does not require the presence of carbon dioxide for hardening, and is very insoluble in water. It is very well adapted for use in moist places, or c\'cn umler "'ater, amI since its hardening is simultaneous throughout the whole mass, and is quite rapid in most varieties, it finds extensive use in building operations.
There are three general classes of cement: -
1. Those formed from certain volcanic tufas, or from artificial mixtures resembling these. Such cements generally need the addiinsoluble silicates when treated with water. In this group are the natural volcanic tufas, l'ozzuolan, trass, and Santorin earth, together with blast furnace slags and certain coal ashes, which are occasionally used.
2. Those which contain a large proportion of free lime, having been made by burning natural argillaceous limestones at a temperature sufficiently high to drive off all the carbon dioxide, but not to fuse the product. These include "hydraulic limes" (p. 150) and Roman cements.
3. Those prepared by burning an intimate mixture of clay and powdered calcium carbonate, at a very high temperature; so that incipient fusion takes place in the mass. These constitute the Portland cements.
Pozzuolanic cements are chiefly derived from volcanic tufas, which are found in Italy, near Naples (l'ozzuoli), in the islands of the Grecian archipelago, and in Germany near Andernach on the Ithine. These tufas consist of silicates which are easily decomposed by acids. They have resulted from the action of volcanic fires, and need no further treatment than fine grinding and mixing with lime. Such cements are slow in hardening, but have considerable ultimate strength. Pozzuolan has been used since the time of the Romans, who were well acquainted with its prop- CIties.
Blast furnace slag is now used to some extent for the production of cement. But in order to develop the hydraulic property, the melted slag must be cooled suddenly; this is usually done by running it into water. The granulated material thus produced is then very finely ground and mixed with lime. Some slags contain a high percentage of lime-alumina silicate, and hence need less addition of lime. Slag cements lULVenot given entire satisfaction up to the present, although they have been greatly improved within a year or two, by increased care in running the furnace charges. These cements are said to harden more slowly, and to have less resistance to the action of frost than the Portland cements. 'rhey shrink considerably, and are liable to crack, but this is remedied somewhat by coarse grinding. Slag cements, low in silica, but having a high percentage of iron, do not harden well.
Hydraulic limes have already been mentioned (p. 1150). The free lime which they contain is sometimes slaked with just sl~tficient watei' to hydrate the quicklime before the material is sold; but not enough water should be added to set the cement. was first made in England by J'. Parker, who patented a process for preparing it from the septai'ia nodules, consisting of clay and chalk found in the bed and along the banks of the Thames HiveI'. T~ater, the beds of clay limestones were used, but as there was much irregularity in the composition of these rocks, the prouuct did not give satisfaction. But by careful selection of the material and propel' mixing of different kinus of stone, the quality of cement produceu has been improved. These rocks are also found in .France, Holland, and Germany, and in the United States. There are several deposits that are very pure, and vary but little in the different parts of' the bed. Homan cement was first made in this country 'in New York state, from a rock found on the banks of Rondout Creek and near the Hudson River, and is called "Rosendale," from the chief town in the district; it still constitutes a large part of the natuml cement made in this country. Another important region is on the Ohio River, near Louisville, the cements made there being known by the latter name. Pennsylvania, Illinois, Wisconsin, and Colorado also supply much natural cement.
Nearly all these rocks contain a large percentage of magnesia, but this does not appear to injure the cement made from them.
The rock is broken into Inmps about the size of a goose egg, in order to secure evenness in burning. The burning is done in continuous kilns, as a rule, and the temperature must be very carefully regulated, high enough to drive out nearly all of the carbon dioxide, but not to fuse the rock. Then the rock is carefully ground between buhrstones, and sifted. The finer the grinding, the better the product. In order to secure supposed uniformity in the product, it is often customary to mix rock from several beds, in the same kiln, but this is of doubtful benefit. '1'he color of Homan cement varies greatly, from pale yellow to red brown, aJlll is due chiefly to the amount of iron and manganese oxides present. But there should not be great variations in the color of the products made from the same rock, as this indicates inequality in burning.
Roman cement is generally quick setting, and hence is preferred by many engineers for work under water. It weighs from 50 to 56 pounds per cubic foot. Its strength is inferior to Portland cement.

Portland cement is entirely an artificial product, but represents the most important branch of the cement industry. '1'he first patent was taken out in England, in 1824, but the process extended in a few years to France and Germany. In the United States the manufacture of this cement was begun in 1878, at Coplay, Penll., but the industry has not yet developed sufficiently to supply more than one quarter of the annual consumption in this country.
The materials used are very pure calcium carbonate and fine clay, rich in silica. The English manufacturers prefer chalk and clay mud taken from the mouth (estuary) of the l\Iedway River. In Germany, a marl (calcareous clay) is employed. But in any case the proportion of calcium carbonate to clay must be controlled within tolerably narrow limits. A most thorough mixing of the ingredients is very essential. The calcium carbonate and clay are both finely ground, and often levigated, before they are mixed. There are two methods of mixing in general use: the semi-dry and the wet. In the semi-dry method, the materials are mixed in a washmill to a thick" slurry," containing about 40 per cent of water, and which is then ground again in buhrstone mills; then it is run into shallow pits, where much of the water drains away or evaporates. Sometimes it is dried by the waste heat from the kilns. ,when dry enough to ,work properly, the mass is made into balls or bricks (usually by machines) and dried, after which they are ready for burning. In the wet method, the clay and chalk are ground together, and the slurry levigated, so that all coarse particles are eliminated. After settling, the water is drawn off, and the slime dried by waste heat and bricked, as in the semi-dry method. The dry bricks are then burned.
Several varieties of kilns are in use for burning Portland cement. Periodic, dome-shaped, or shaft kilns are still much used in this country, but are uneconomical as to fuel and output. Coke is used as fuel, and is charged into the kiln in alternate layers with the bricks. The coke, amounting to 40 per cent of the weight of the clinker, is burned out, and the kiln cools and is emptied. Continuous kilns are coming into more general use, however. The great difficulty at first ·encountered with them was the tendency of the charge to stick to the furnace walls, owing to the high temperature. But this is largely avoided now by proper attention to the kiln while bU1'l1ing.
The Dietsch two-storied kiln (etageofen) (Fig. 58) is much used for Portland cement. The bricks and fuel (which may be soft coal) are charged continuously at the top (A), and descend into the horizontal chamber (B), from which the charge is raked into the combustion chamber (C) by introducing a tool through the door (D). The burned clinker is withdrawn at the bottom opening (E). Air enters high temperature, where it supports the combustion of the fuel. The hot gases passing off through (B) and (A) serve to heat the charge before it arrives in (C). These kilns now are often worked with forced draught, and produce about 7 tons of clinker per ton of coal.
Hoffmann's ring furnace (Fig. 59) is also much used in cement burning. This consists of an elliptical gallery built around a central chimney (A). The gallery is divided into 15 or 20 compartments (B, B), each having a door (C) opening outside, a flue (D) leading to the chimney (A), and a wide opening (E) into the next compartment. Each flue has a damper, by which connection with the chimney (A) may be opened or closed. The openings (E) between the compartments may be closed with a sheet-iron or heavy paper diaphragm, as will be explained below.
If the door of the compartment on one side of the diaphragm be opened, and the damper of the flue (D) leading from the compartment on the other side of the diaphragm is also opened, while all the other doors and flues are closed, the draught of the chimney (A) will cause air to enter the open door and pass around the entire gallery, through each compartment in succession, and finally out through the open flue (D) to the chimney. In the roof over the gallery are charging holes (G), several being in each compartment, through which fuel is introduced. The furnace is run as follows: Assume that there are 14 compartments, as shown. Twelve compartments contain cement bricks, and their doors and chimney flues are closed. Suppose that No.1 is being emptied, while No. 14 is being filled. The paper diaphragm closes the opening between No. 13 and No. 14, and the flue (D) of No. 13 is open to the chimney. Compartment No.7 is at the height of combustion, while :K os. G, 5, 4, 3, 2 contain bricks which have been burned. In Nos. 8, 9, 10, 11, 12 are bricks to be burned. Cold air is drawn in through the open door of No.1, and passing in order through Nos. 2, 3, ,t, 5, G becomes heated by contact with the hot bricks in these compartments until, after passing through No. G, which is still red hot, it a1'1'i ves in Xo. 7 at a very high temperature. In Ko. 7 the fuel is burning at a white heat, and the hot gases pass on through l{os. 8, 9, 10, 11, 1:!, 13, from which they escape to the chimney. By this passage of the hot gases through the compartments, the unburned bricks are heated, those in No.8 being neady red hot; but as no fuel has been introduced into these chambers, combustion and white heat are confined to No. 7. ~WhenNo. H is filled with green bricks, its doors are closed, and also the chimney damper of No. 13, while that of No. 14 is opened, and th8 diaphragm transferred to the opening between No. 14 and No. 1. Fuel is now introduced into No.8, which becomes the combustion chamber, and the door of No.2 is opened. The burned bricks in No.2, having been cooled by the passage of cold air, are taken out, while No. 1 is being refilled. Thus the cycle of operations goes on, each compartment in turn being charged with fuel and made the combustion chamber. The temperature in that compartment which has just been filled is only high enough to dry the green bricks well. A paper diaphragm is often used between this chamber and the next one, which is to be filled. "\Vhen the temperature becomes sufficiently high to burn away this paper, the next compartment is ready, and is thrown into the circuit.
This furnace is very economical of fuel, one ton of soft coal burning G}tons of clinker, but it requires much labor. The bricks must be accurately piled in order that open channels may be left beneath the charging holes for the fuel, which is thus made to burn in a column extending from the floor to the top of the furnace. The fuel must not contain too much ash. Usually OIlecompartment is emptied each day, and consequently the fire is moved forward each day.
Revolving furnaces (p. 17) are also used to some extent for cement burning. A furnace used in this country is about 75 feet long, slightly inclined, and lined with fire-brick. The lower end is G feet in diameter and the upper end [) feet. Cmde petroleum is used as fuel, and the furnace is continuous acting. The wet slurry runs in at the upper end and, by drying on the rotating surface, forms little balls or gravel-like lumps which are thoroughly burned in the passage through the furnace, requiring two hours 01' more. The annual increase in output from rotary furnaces indicates that the method is a successful one. Producer gas will probably soon take the place of crude petroleum, should the price of the latter advance much.
~Iuch depends on the proper temperature of the burning, during which there is considerable shrinkage; well burned clinker is a semi-vitrified, brown or grayish green mass. If overburned, it may fuse and take on a blue green or black color; such cement will not combine with water. The clinker is then ground very fine, so that nearly the whole of it will pass through a sieve with 2500 meshes to the square inch. Buhrstone or ball-mills are generally used for this purpose. The Griffin mill, which consists of a heavy stcel roll revolving about a vertical shaft, and pressing, by centrifugal force, against a steel ring, is proving very effective. Since only the finest dust is of value in cement, great care is necessary to prevent the coarse material from passing through the mill. After grinding it is generally stored, that any free lime it contains may become air-slaked, before the cement goes to market.
The composition of Portland cements varies somewhat, but the following are the extremes: -
The cause of hardening of cements has been investigated by many chemists and has caused much discussion. 1,e Chatelier's * investigations show that during the burning a tricalcium silicate (Ca;!SiOo)is formed by the action of the clay on the lime. At the same time a certain amount of calcinm aluminate and ferrite are also formed, besides mono- and di-calcium silicates. when treated with water the triealciull1 silicate reacts to form a hydrated monocalcium silicate and calcium hydroxide:-
1) 2Ca3SiOo+ \)Hp = (CaSi03)2' 5H20 +4,Ca(OH)2'
Then a reaction between the calcium hydroxide, water, and calcium aluminate may occur, forming a hydrated basic calcium aluminate:- 2) Ca3AlcOo+ Ca(OHh +11 H20 = Ca4AI20;· 12 H20.
The formation of the hydrated basic aluminate (CaO)., A1203· 12 H20,
probably exerts some influence on the rapidity of the" setting" of the cement, but the hardening is undoubtedly due to the first reaction.
The" set" of cement, in contrast with that of mortar, is not due to a drying out of water, but is the beginning of the true hardening process. The rapidity of this" setting" is influenced by the presence of other salts; alkalies hasten it, while it is retarded by calcium sulphate, magnesium sulphate and chloride, and sodium chloride. It is also retarded proportionally as the amount of sand used is increased. The amount of water used in mixing is about one-third the weight of the cement, and its temperature has much to do with the rate of "setting," warm water hastening it. Portland cement is usually slower in setting than Roman, but when the hardening has begun it progresses more rapidly with the former. There is very little increase of hardness after six months. Portland cement is more durable than Roman under most conditions, and is generally stronger. It forms a denser and heavier powder of a greenish gray color, but when hardened has a drab shade resembling the color of the stone quarried at Portland, England, and used much for building in that country; hence the name. As has been said, variations in color of the same brand of cement may show changes in quality; if underburned, it is generally yellowish. The weight per cubic foot varies from about 70 to 90 pounds; the finer the grinding, the less the weight. But as a rule heavy cements are preferred by builders, as they are supposed to be more thoroughly burned; they are, however, slow in setting.
The testing of cement is generally the work of the engineer, rather than the chemist. Chemical analysis is of but little use in determining its properties, and the usual tests applied are physical. They are for:
(Ct) Fineness.
(b) Expansion (" Blowing ").
(c) Shrinkage.
(d) Rate of hardening.
(e) Resistance,
to tension
to compression. A committee of the American Society of Civil Engineers has recommended certain standard requirements for cements.* In Germany, an official standard and method of testing have been adopted. In testing the fineness, three sieves should be used, - Nos. 50, 74, and 100, the proportion of each sample rejected by each sieve to be determined by weighing. The residue remaining in the coarse sieves is of little more value than the same quantity of sand. Expansion, or "blowing," is shown by the swelling and cracking of a pat of cement, three inches in diameter, one-half inch thick in the middle, and very thin at the edges. After setting, a pat of this kind is put into water and left several weeks, being examined every day. If fine cracks appear in the edges, or disintegration occurs, the cement is of poor quality. Expansion may be caused by the presence of unslaked lime, magnesia, or gypsum, or the cement may be umlerburned. Expansion and shrinkage tests are often made by filling glass bottles or lamp chimneys with the freshly mixed cement. After setting, any expansion will crack the glass. Colored water may be poured on top of the cement. If it runs down between the cement and the glass, shrinkage has probably taken place. The rate of hardening, or the time of setting, is very important in determining the suitability of a cement for a given purpose. Quick-setting cements are usually desirable for work under water. The time is determined by the" normal needle." * Two of these are used. One is a wire about one-twelfth of an inch in diameter, and is loaded with a weight of one quarter of a pound. The other is one twenty-fourth of an inch in diameter, and carries a weight of one pound. After mixing the cement with water, the time is noted until no impression is made upon it by the point of the first wire. This is the beginning of the "setting." when the second wire will no longer penetrate, the "set" is ended. Cement which will set in less than two hours is called "quick setting"; those taking a longer time are "slow setting." The beginning of the set is sometimes noticed within five or ten minutes after mixing, an<l usually within half an hour. It is important that the work be not disturbed after the "set" is once established, otherwise, tight and strong joints cannot be made.
The resistance tests are the most important. In actual use, cements are generally subjected to compression strains; but since compression tests are difficult to make, it is often the practice to make only the tension tests. Both tests should be made if possible from samples taken from the barrel, and from the sifted cement which passes the No. 100 sieve, as well as with these samples when mixed with various weights of standard sand. This sand is made by pulverizing quartz; it is sifted and the portion used which passes a No. 20 sieve, and is retained by a No. 30 metal mould, forming a "briquette," shaped like an hour-glass, the narrow portion having a section exactly one inch square. The briquettes must be made very carefully, or the results will not be uniform. The cement is quickly mixed with water at 600 F., filled into the mould, pressed down well and smoothed off evenly. This is done on a slate or glass plate, to prevent absorption of moisture. As soon as set, the briquette is removed and allowed to stand covered with a wet cloth for 24 hours. It is then placed in water, where it remains until the test is made, when it is fixed in the jaws of a machine, which applies a gradually increasing tension. The number of pounds necessary to fracture the briquette is read on a graduated scale beam. 1'he average of five or ten tests is taken as the breaking strength.
The briquettes are usually tested at one day, one week, and four weeks after making. Sometimes they are kept for still longer periods. vVhen made from cement without sand, they are called " neat." The proportions of sand, water, and cement recommended are:-
For Portland briquettes (neat), 25 per cent of water.
For natural cement (Rosendale) (neat), 30 per cent water.
For 1 part cement, and 1 part sand; the water used is about
15 per cent of the total weight of the sand and cement. Another proportion is: cement, 1part; sand, 3 parts; and water to about 12 per cent of the total weight of sand and cement. Compression tests are made on small cubes of the cement in question. They should show at least ten times their tension resistance. But, as has already been said, this test is difficult to make, and is very frequently omitted.
Chemical analysis of cement is seldom employed, and chiefly to detect the presence of adulterants. Formerly, the most common adulterant was ground blast furnace slag; but clay, ashes, and hydraulic lime have been employed, and are not always easy to detect. The specific gravity of good Portland cement should not be less than 3.1, but adulterants may reduce it. The presence of considerable quantities of manganese would also indicate adulteration. l\Iagnesia in Portland cement is claimed to cause expansion, and 3 per cent is the highest allowable. In Roman cement, however, a large amount is sometimes present without apparent injury. Alkalies in considerable quantities are also questionable ingredients; not more than 2.5 per cent should be present in a good Portland cement.
Plaster of Paris is made by heating the mineral gypsum (CaS04• 2 H20) until about three-fourths of its water of crystallization has been driven off. The process is called burning, and is usually carried on in kilns, muffle furnaces, or retorts. Direct contact with the fuel is not permitted, lest the action of the carbonaceous matter should cause a reduction of some of the calcium sulphate to sulphide. Neither should the flame come in contact with the gypsum, but only the hot gases. The burning is a very delicate operation, and requires much care. Gypsum contains about 21 per cent of water of crystallization, of which good plaster should retain 4 or 5 per cent. The loss of water begins at about 800 C., but the most favorable temperature for burning is about 1250 C. If heated to 2000 C., all the crystal water is expelled, and the product will combine with water but verJ slowly, the property of rapid setting having been destroyed. Thus it will be seen that the limits of heating are very narrow. The gypsum is broken in rather small lumps to secure evenness in burning; for a fine product, it is sometimes powdered and heated on a plate, while constantly stirred. After burning, the lumps are friable and are easily ground. Sometimes wooden rolls, set edge-runner fashion, are used to grind the burned material. \Y~lCn mixed with water, plaster of Paris forms a paste which soon hardens or "sets," owing to a recombination of water with the burned plaster, to form hydrated calcium sulphate. The theory of this setting has been explained by 1,e Chatelier.* The composition of the plaster is essentially (CaSO.I)2· H20, a salt which is soluble, and part of which dissolves in the water used in mixing. But as soon as it dissolves, a combination between it and some of the water takes place, forming CaS04• 2 H20; this, being much less soluble than the monohydrated salt, at once begins to crystallize from the solution, forming a network of crystals. Then more of the plaster dissolves, becomes fully hydrated, and crystallizes out, increasing the solidity of the "set" by the interlacing of new crystals with those already formed. Thus the cycle of reactions goes on until the plaster is fully hydrated.
The theoretical quantity of water necessary to set plaster is about 18 per cent of its weight; but in fact, from 30 to 35 per cent is generally used. Excess of water renclers the mass more plastic of the plaster, if left in contact ,yith it for some time after setting, owing to the solution of some of the crystallized calcium sulphate. l'laster expands slightly while setting, and for this reason is valuable for making casts and reproductions. It is largely used for' interior decorative work and also as a cement for joining glass and metal ware. The surface of plaster after setting is rather soft, and for many purposes it is desirable to increase the hardness. This may be done by mixing alum, borax, or tartaric acid with it, or by adding some alcohol to the water with which the plaster is mixed. However, these substances retard the setting. By painting or dipping plaster casts in meited wax, paraffine, or stearin, or in solutions of these in petroleum ether, the pores of the plaster mass are filled and the surface is made smooth, so that dirt will not adhere and the articles may be washed. When treated with a solution of barium hydroxide, the surface of the plaster is coated with barium sulphate and rendered insoluble. If plaster is mixed with a solution of gllle or size, the material called" 8tueeo" is obtained.