m m -■>?r->: \m w}^ . i-^,^ %i^ EtiJ Ml -TSI^ l« Kj yfPSvi ss fi^KE^**^^^ . 4|J^V< j F«'.j f^* 9^^^- '•^J ^3 P^ i s^- ■**-r; ^^ #vs^^\ ^'^■S, ii-vx •*( ■• »■><' f >--?*^' >^ H %^i;. >^* •4 c^^m ' W r i? \'.^ r< ^■-■v :^ .A**. . • • 1 UNIVERSITY OF ILLINOIS S LIBRARY 1 • Class Book Volume ! . ^ ■ r * ;^ -^^. .^ii:v"^ \> . '^l 1- V, ^"*^ ">N. >^ { \ «0 a Cl \ \ \ \ JO \ K14- PAV1N6 BRICK SHALE RATTLER LOSS El a* \ \ IX ! X > )< c 10 8 6 C 4 a ? ' > II TEMPERATURES EXPRESSED IN COMES 10 HARDNESS OF BRICK XnD THEIB RESISTANCE TO FROST. a series of test pieces at different stages of burning, made by Wegeman. His observations sliowed that witli tlie fusion of tlie material, the pore space apparently increased, and took the form of blebs or bubbles in the glass. Reasoning from known facts, the change in pore space that takes place during the burning of the brick is believed to be as follows : Period of dehydration and oxidation. In dried brick, the pores are mostly very minute on account of the fineness of the grains. The salts, left from the escaped water, are lodged most abundantly in the pores near the surface. As a result, these pores are more or less clogged. During the period of dehydration and oxidation, several of the consti- tuents of the brick give off gases. Among these are: the combined water, COo from the carbonates, and carbona- ceous matter, SO2 from the sulphides and sulphates, etc. To a certain extent the escaping gases act in a manner analogous to that of yeast in bread, and open up and ex- pand the pores in forcing their way out of the brick. In addition, many of the precipitated salts are volatile and are removed, thus freeing the pores. The final result of those co-operating factors is to increase the pore space, by opening the pores and expanding them. The brick, as a result, take up water much more rapidly and abundantly than before. Period of fusion. Fusion begins with the amorphous matter between the grains, causing them to run together so that they lose their individuality. The pores, composed of the spaces between the grains, begin to lose their triangu- lar form and grow smaller. Eventually the walls come in contact at two or more points and portions of the pores are sealed. The pores are filled with gases, and when they are sealed, these are confined, preventing further collapse of the pores. As tlie glass becomes more fluid, surface tension causes the gaseous inclusions to take on a spherical form. These minute bubbles, as they come in contact with each other, merge and form larger bubbles. The writer believes that these last are the blebs, described by Wege- mann. - HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. 11 Wlieu the pores are not completely sealed by glass, they are nevertheless obstructed, and their etfeetive diame- ter decreased. As will be shown later, this has an import- ant bearing upon the rapidity with which they absorb water, and therefore upon the durability of the brick. It is probable that the only decrease in actual pore space comes about through shrinkage of the brick, and collapse of the unsealed pores. Much of the apparent decrease is due to the sealing of portions of pores by glass, and consequent isolation of these parts from the all-pervading system of pores that existed in the unfused brick. The pore space is commonly measured by the amount of water a brick can absorb, and only that portion of the total pore space is included which offers a free passage to water. This leaves out of consideration the minute and sealed pores, and the apparent decrease in a large measure represents these. The changes, then, that take place in pore space dur- ing burning are: First, the enlargement and clearing of the pores during the period of dehydration and oxidation resulting from the volatilization of the obstructing salts and the mechanical effect of the escaping gases; second, the obstruction of the pores by glass formed during the period of fusion, and the partial or complete isolation of the pores included in the fused portions of the brick. CHANGES IN STRENGTH AND RIGIDITY DURING BURNING. As brick are burned, they gain in strength, as has been shown by numerous crushing tests. This is due to the better consolidation of the grains of clay, and the more perfect contact that is produced by their partial fusion with increased heat treatment. The brick gains in tenac- ity, therefore, and develops greater resistance to disinte- grating forces as burning progresses. Along with this increase in strength, goes increased rigidity and consequent brittleness. While it takes a greater initial force to start disruption in the harder burned brick, the distance through which the force must 12 HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. act in order to cause complete failure is lessened. This is caused by the fact that liard brittle substances cannot be strained as far without breaking, as the softer and tougher materials. This has an important bearing upon the resist- ance a brick offers to disintegration by frost. As an illustration, consider the conditions and action of a brick in the wall of a burning building. The sudden change of temperature expands the surface of the brick more rapidly than the body, on account of the slowness with which the heat is conducted to the interior. When water is thrown upon the wall the surface is cooled very much more rapidly than the interior. The surface of the brick becomes relatively smaller than the body, and a stress results between them. If this- stress is great enough to overcome the resisting strength, and the amount of con- traction is enough to exceed the elastic limit, — or the dis- tance which a brick may be strained without injury — the brick gives way and spalls off. If the amount of contrac- tion does not exceed the elastic limit, the brick is not injured, since the strained parts return to their former positions as soon as the stress is removed. A little brick has a small elastic limit, and will often fly to pieces under conditions where a softer brick will stand. The harder burned brick, therefore, have the greater initial strength to resist strains, but give way more rapidly after movement is once started. Tensile strength versus compressive strength. The distance through which the walls of the pores are forced by the expansion of freezing water depends upon whether the tensile strength of the brick is greater than the compressive resistance, or vice-versa. As an illustration, consider a pore near the surface of the brick. If the tensile strength is the weaker, the material at the surface will give way and spall off. As a result the pore will expand in but one direction, i. e., towards the surface. If, on the other hand, the tensile strength is the stronger and the material holds, the pore will have equal pressure on all sides and expand in all directions. HARDNESS OF BRICK AND THEIR RESISTANCE TO FKOST. 18 If tlie poros be considered as circular tubes, this fact may be stated quantitatively. Kepresentin<>- the radius of the pore as r and its length as l^ the volume of the pore will be ''Ur-," Representino- the expansion of water as "'^a' the total expansion of a tilled pore will be "aUr~.'^ If the expansion takes place in one direction only, as in the first case, tlie distance that part of the wall must move is equal to the total expansion of the confined water divided by the length of the tube, or ^^anr^/' That is, the distance the wall of the pore is strained is in this case proportional to the square of the radius. If, as in the second case, the expansion takes place in all directions, tlie distance any part of the Avails move is the total expansion divided by the number of directions of movement, or alTrr- ar .'iTrr That is, tlie distance the walls of the pore are strained is jtroportional to the radius. This indicates that the dis- tance any part of the wall of a pore moves in consequence of the expansion of the water, is much greater in the first case than in the second. The chances are much greater, therefore, that the elastic limit of a brick will be exceeded when it possesses greater rigidity than tensile strength. The actual expansion in a brick probably lies between the two values given, and depends not only upon the size of the pores, but the rigidity of the brick also. In the softer l)rick, which are less rigid, the expansion of the walls will be more nearly proportional to tlie radii of the pores. In the harder brick, on the other hand, in which rigidity is greater, the ex])ansion will be more nearly proportional to the square of tlie radii. The advantage gained b}^ the smal- ler pores of harder brick is in a measure offset by their increased. rigidity and smaller elastic limit. Special case of the salmon hrick. Purdy and Moore 14 HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. have called attention to the fact^ that some clays disin- tegrate and slake when placed in water, becoming plastic again even after having been burned to the temperature of 1100 degrees F for many hours. They explain this phenom- enon as due to the action of adsorbed salts, and it may be possible that this is more common than is at present sup- l>osed. If this is true, the rapid weathering of some salmon brick is easily explained. In all harder burned brick, lioweA^er, this factor does not enter, as the adsorbed salts are changed to a harmless form at the temperature at which dehydration and oxidation is finished. Consequently, it may be omitted in the discussion of the brick burned be- yond this stage. The changes that take place in the strength of brick dur- ing burning are, therefore, increase in tenacity, due to consolidation and amalgamation of the clay grains; in- crease in rigidity and brittleness; increase in durability due to change of adsorbed salts to a harmless form. The increase in strength increases the initial strain the brick can withstand, and the increase in rigidity decreases the elastic limit and, therefore, the distance the parts of the brick can be strained without injury, DISINTEGRATING FORCES DUE TO CHANGE IN TEMPERATURE. Factors inherent in hrick. Brick, which are merely artificial stones, have many characteristics in common with rocks and many known facts concerning building stones may be applied to bricks. Every geologist is familiar with the immense amount of disintegration that rocks undergo from simple changes in temperature. This is especially true in situations, as on mountain slopes, where diurnal changes are rapid. MerrilP cites several observations of the effect of rapidly changing temperature on rocks. He tells of finding numerous fresh chips and flakes in the valleys and on the slopes of a mountain in Montana, that 'See page 213, this volume "Rocks, Rock Weathering, and Soils, p. 181. HARDNESS OF BRICK AND THKIR RESISTANCE TO FROST. 15 c-oul(l only be accounted for by the action of rapid cliange in temperature during day and night. Another observation lie quotes is tliat of Livingston, who reported that rocks in Africa were frequently heated to a temperature of 13 T degrees Fahrenheit during the day, and that rapid cooling during the night split off fragments weighing as much as 200 pounds. The fundamental cause of this disruption is the poor heat conductivity of the rocks. Kocks and clay products are poor conductors of heat. A difference in temperature of 100 degrees may arise in a depth of one inch when a rock is simply heated by the rays of the sun. The coefficient of expansion of rocks is approx- imately .000005. In a rock 100 feet long when the above conditions exist, the difference between the length of its heated surface and that of a zoiu^ one inch lower would be nearly one-half inch. This places the rock under a tremen- dous strain, and since rocks are very rigid, the strain is concentrated at the weak(^st point. Eventually the strain becoiiH^s greater than the rock can bear and it gives way. This same principle operates on brick, and is apparent in the chipped surfaces of a brick wall that has ])assed through a severe fir(\ Tender ordinary circumstances, the greatest difference in tem]»(Mature exists between the faces of a wall heated to room temperature on the inside and cooled to the temperature of the air on the outside. This difference is at a maximum during winter and probably amounts to 100 (legi*ees V. Since, however, the bricks are separated from each other by much weaker mortar joints, the wall does not act as a single unit, as a rock mass does, but as a multitude of units. Therefju-e the differential ex- pansion oi* contraction cannot be concentrated u])on a weak point as it is in rocks, lnit is confined to each brick. The difference in length of the outer and inner surfaces of a brick under these circumstances is only .0025 of an inch. This certainly does not exceed the elastic limit of the brick and need not be taken into account. The same process that is so effective in disrupting the rocks is inef- fectual in l)rick under ordinary conditions, simply because 16 HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. their small size and elastic connections render it impossible for cumulative strain to concentrate at any one point. A factor operative in the crystalline rocks, especially of the coarse grained granite type, is the internal strains set up by unequal expansion of the different minerals. Here( as above, the amount of differential strain depends directly upon the size of the grains, increasing as the size of the grains increases. As most of the material that forms^ brick is very tine grained and becomes more homogeneous with burning, this differential strain is too small to have any effect. Factors foreuin to the hriel\ Since it has been shown that there is nothing inherent in brick that will cause their weakening with ordinary temperature changes, the disrupting factor must be some external substance that may find entrance into them and, by its different rate of expansion, set up strains. Of necessity this must be a sub- stance that is fluid at least at the time of its entrance into the brick. The conditions require that it be mobile and able to flow through the pores. Further, it must have a different rate of expansion from that of the brick, in order that a differential expansion and consequent strain may exist with change of temperature. In order that this may be effective as a disrupting force, it is further necessary that the substance be confined to a considerable degree, so that the strains will not be relieved by reverse flow of the substance. Liquids and gases fulfill the first two conditions per- fectly, but as the same properties that permit their en- trance into the brick also allow their escape, they cannot cause strain that will mechanically harm the brick under ordinary conditions. All three conditions are only fulfilled by some substance that, fluid at the time of entrance, be- comes rigid with change of temperature, or other ordinary conditions, and thus renders it impossible for the brick to confine it. Of the three states of matter, solid, liquid, and gas- eous, only the first is rigid, and the last two fluid. The dis- HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. 17 I'uptiug- substance must, therefore, enter as a liquid or a gas, and then throusjli ordinary change in conditions, be- come solid within the bi'ick. As at ordinary temperatures, common g^ases do not reach the solid state, they need not be considered. A liquid may be solidified by lowering its temperature, by changing the pressure upon it, or by evaporation of one or more of its constituents. When a solid is dissolved, it becomes to all intents and purposes a liquid. Conversely, when it is crystallized from solution it becomes solid again and, if absorbed when in solution, may, upon the evapora- tion of the solvent, be confined within the brick. Seger''^ states that the concentration of the salts in the surface layer of brick by the evaporation of water often causes the destruction of the surface. Other observers have made the same statement. Tt would seem at first sight that it would be impossible for a crystallizing salt to exert any pressure or cause any strain, for as the crystal grows it shuts off automatically the supply of solution. It would thus completely close the pores, preventing the further growth necessary to cause strain. Recent experiment^ has shown, however, that growing crystals can exert, consider- able pressure and Avill continue to grow even under con- siderable resistance. It is possible, therefore, that brick may be injured by the crystallization of salts in the surface layers. It is believed, however, that this is a minor factor in the weathering of brick, and that the greater destruction results from another source. Some substances, such as water, expand as tliey pass from the liquid to the solid state. Increase in pressure lowers their freezing point. Consequently, Avhen such a liquid is confined while freezing, the increase in volume will cause a pressure upon the walls of the confining vessel. In order that this may be effective in the brick, it is necessary Coll. Writings, p. 372. G. F. Becker and A. L. Day, Trans. Washington Acad, of Sci. Vol. 7. 18 HARDNESS OF BRIOK AND THEIR RESISTANCE TO FROST. that enough of the liquid become solid to seal the surface pores, and thus make it possible for the liquid to be con- fined while freezing. In so doing, a strain is set up within the brick, which if great enough, may burst its bonds and cause damage. It is, then, to liquids that expand upon solidification that we may look for the principal cause of the observed weakening of brick by changes of temperature. Of this class of liquids, there is only one that is universally abund- ant and solidifies at ordinary temperatures. This is water. PHYSICAL PROPERTIES OF WATER. Change of volume with change of temperature. The great majority of substances contract as their temperature is lowered. Water confonm? to this rule until the temperature of about 39 degrees F is reached. At this point water ceases to contract and as the temperature is lowered further, expands until it solidifies. Generally solidification takes place at 32 degrees F but may be de- layed — as will be explained later — until a lower tempera- ture is reached. This expansion has been measured down to 18 degrees F, and is at this temperature .00186 units.^ That is to say, one cubic inch of water measured at 39 de- grees F becomes 1.00186 cubic inches when cooled to 18 degrees F without freezing. Since water is nearly incom- pressible, this small expansion of the water, if it were rigidly confined, would cause a pressure of over 500 pounds to the square inch. Change in volume upon freezing. When water changes to its solid form, ice, its volume increases approximately one-tenth. That is, ten cubic inches of water at 32 degrees F will become eleven cubic inches when frozen. This increase in volume, whenever the freezing water is confined, may give rise to a pressure upon the walls of the containing vessel that may be great eiiouah to burst it. Chwolson, Lehrbiick der Physik, vol. 2, p. 134. , HARDNESS OF BRICK AND THKIK RESISTANCE TO FROST. 19 Near the close of the eighteenth ceutury, Major Wil- liams, while stationed at Quebec, filled two 13-in(h bomb shells with water and closed the fuse holes b}^ driving in as tightly as possible iron plugs weighing three pounds each. The shells were then exposed to the cold of the winter night about twenty degrees below zero F. The next morning one of the shells was found to have burst and a thin fringe of ice projected through and beyond the crack. The plug of the other shells was found at a distance of 415 feet and a column of ice eight inches long protruded from the fuse hole.^*^. Evidently a powerful pressure had resulted from lowering of the temperature of the confined water. Assuming that all of the water froze, and the shell re- mained intact, which is possible if the shell was only partly filled, leaving enough s])ace for the ice to form, both shell and ice would contract as they cooled. Since the coeffi- cients of expansion of ice and cast iron are practically the same, whatever pressure may have been witliiu the shell could not change materially. Consequently, if the shell did not give way at the moment the last of the water froze, further cooling could not cause its bursting. If the water began to freeze and the pressure resulting from the increase in volume of the forming ice became great enough to lower the freezing point of the remaining water progressively witli its falling temperature, an in- creasing i)ressure would accumulate within the shell as the water cooled. An inciease in pressure of one atmosphere — 15 pounds to the square inch — upon water will lower the freezing point .01388 of a degree F. The pressure neces- sary to prevent freezing at 31 degrees F is 1080 ])ounds, or half a ton, to the square inch; at 22 degrees, 10800 pounds; at zero, 34500 ])()un(ls; and at minus 20 degrees — the tem- perature of the air in Major AYilliams' ex])erini<'nt — 00480 pounds, or over thirty tons, to the square inch. This cer- tainly would be pressure great enough to account for the bursting of the shells. It s(^ems probable, therefore, thai Carnot's Physics, 13th Ed. Trans., p. 320. 20 HARDNESS OK BRICK AND THEIR RESISTANCE TO KROSfl'. part of the water was still liquid at the time the shells burst, aud that they were ruptured bv the cumulative pres- sure that prevented this water from freezing. Moussan" performed an experiment that throws some light upon this question. He had a strong cylinder made that was closed at the lower end by a small cone held in by a screw-nut, and at the upper end by a piston moved by a screw-nut. Removing the bottom nut and cone he filled the cylinder with water, placed a bit of copper rod in it as an index, and allowed the water to freeze by exposing the apparatus to the winter air. He then carefully cleaned away enough of the ice to allow him to put the bottom cone and nut in place, screwing them down as tightly as pos- sible. Then he inverted the cylinder and placed it in a salt and snow mixture at a temperature of about zero F. By slowly turning the top screw he compressed the ice to about 0.87 of its original volume. This required about four hours. Then keeping the cylinder in the cold he opened its bottom. The index at the beginning of the experiment was in the upper part of the ice. If the ice remained solid during the compression, the index should remain in this position and when the cylinder was opened appear last. If the ice was melted by the pressure, the index would sink through the water formed and should be at the bottom of the cylin- der, appearing first when it was opened. When ^[oussan opened the lower screw, and loosened the cone, it came out rather .suddenly and ice formed instantly upon its sides. Immediately behind it followed the index and then, for the first time, came a thick cylinder of ice which must have formed at the instant of opening. ^^ 'iPogg. Annalen, vol. 105, 1858, p. 170. '-" ALs man nocli dieseni Verfahneu die iiutere Sclduss-soliraube immer in voller Kalte ofifuete. iind den klienen Konns losfe, trat der selbe sofort etwas heraus und an seiner seite bildete sich augenblick- lich Eis. Gleich hinter den Koniis folgte der Index und erst nach diesem ein dichter Eiscylinder, der sich im augenblick des Oeffnens gebildet haben musste." HARDNESS OF BRICK AND THKIK RESISTAXCE TO FROST. 21 Cousequently he had proved that it is passible to melt ice bA' pressure aloiie. It seems probable iu the light of his experiment that when the enclosing vessel is strong enough to resist the pressure resulting when ice first begins to form, that the pressure lowers the freezing point of the remaining water pi'ogressively with its decreasing temper- ature, and therefoi'e the pressure increases as the tempera- ture is lowered. It lias been proven by Tammann and others that if the pressure be great enough and tempei'ature low enough, the ice will change its form, but these condi- tions are beyond the tenii)eratures and pressures ordinarily prevalent. Recognizing, therefore, that there is a limit, the statement nevertheless holds good that under ordinary circumstances, the pressure resulting from freezing- water, when it is confined, increases as the temperature is lowered. . i DISINTEGRATING EFFECTS OF FREEZING WATER. W'licn water is confined, therefore, and cooh'd below its freezing point, it nun- cause damage either by its sudden expansion when freezing or from the pressure resulting from its attempt to freeze. The pressure is the same in either case, and since it depends directly upon the temi)ei-- ature at which freezing takes place, may be calculated from the known relations of freezing point and pressure. Cojiditions prcvcuiiufj (J T 11 iO 1 K HKSISTA NCE TO KKOS'l'. 23 shock and vibration from the traffic in the streets and buildinjj;. The water on the outer face certainly freezes and the water in the pores is in contact with ice as a con- sequence. On the other hand, most of the pores are very small or at least the numerous constrictions in them pro- duce the same effect. As there is no method known at present by which tlie actual freezing of water in the pores can be proven beyond question, we are compelled to depend upon indirect reason- ing. It is known that brick suffers a loss in strength by having been subjected to freezing. The amount of this loss increases with the fineness of the pores — as shown by the results of the present freezing tests — and not with their coareness, as it should if overcooling was the dominant fac- tor. It has been shown that there is nothing inherent in brick that will cause this loss, and that the only important cause is the pressure resulting from freezing water. Loss of strength in the brick has been reported^in all properly conducted freezing tests, although in these a wide range in the temperatures has been used. It seems extremely prob able, therefore, that at least some of the water freezes and causes danmge. The amount of damage done, {\ssuming the pores to be full and unable to drain, undoubtedly de- pends on the temperature and the relative size of the pores, since the finest pores freeze last and at the lowest temper- atures. It is possible that a portion of this loss of strength may be due to incipient fractures caused by the rapid and repeated heating and cooling to which the brick were sub- jected. fiiiprrfed scaJinti of tlje Hurfufc pores. Obviously the power of resistance of the walls of the containing vessel is no greater than that of its weakest point. AVhen, as in the case of the bombshell exposed in Canada, the plug failed before the walls of the shell, the freezing water, instead of bursting the shell found relief by extruding a column of ice through the fuse hole. If a bottle is filled and not too tightly corked, it will present the same phenomenon when the water is frozen. Evidently, if \.^ 24 HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. the pores at the surface of the brick are completely sealed by the first formed ice, the ajnouiit of damage done depends plugs or spicules beyond the immediate surface of the brick. Mosely^^ determined the tenacity of ice as about 100 pounds to the square inch. This was measured as the ice was melting. Andrews^^' measured the resistance offered by a block of ice to penetration by a heavily loaded iron rod at different temperatures. He found that the ice was very hard and resistant, allowing scarcely any penetration, at the temperatures from minus thirty to ten degrees F. At ten degrees it became slightly softer, gradually soften- ing until about twenty degrees was reached. The softening now became more rapid until near the melting point, when it bcame vei\y rapid. Tammann^'^ found that the plastic- ity of ice is relatively small but increases near the melting point. It is evident that the resistance offered by the ice plugs when they are first formed is quite small. As the temperature falls, however, their rigidity increases rapidly and when the temperature is lowered sufficiently, may be- come rigid enough to withstand the pressure of the freezing water in the interior of the brick. The irregular shape of the pores is an important aid to the plugs in maintaining their position at the outlets of the pores. As the cooling of the brick progresses slowly into the interior, the tem- perature of the ice plugs has probably fallen several de- grees by the time the interior water begins to freeze. Tak- ing into consideration, therefore, the increased strength of the plug due to this decrease in temperature, and the ex- treme irregularity of shape of the pores, it is believed that there is but little relief of pressure due to exuding of the pugs or spicules beyond the immediate surface of the brick. J^Phil. Mag., p. 39, 1870. '"Quoted by Barnes, op. cit. i6Ann. der Physik, vol. 7, 1902, p. 208. HAKDNKSS OF BKICK AM> THKIK KKSISTANCK TO KKOST. 25 Partiul druhiufjc. It is evident that if a pore were ouly niiie-tentlis full, expansion of the forming ice could take place without in- jury to the pore. Xo pressure can arise within the pore until it is more than nine-tenths tilled, for this leaves suffi- cient space for expansion of the ice or water. If, after pores and cavities are filled, it is possible for one-tenth of the water to drain before it freezes, they will be removed from danger of rupture and no damage can result. The two factors that govern the amount of water in the pores and cavities when the plane of frost reaches them are : 1st, the completeness with which the pores are sealed, and 2nd the rate of flow of the water through the pore system. It goes without saying that if brick are not completely sealed on all sides, the pressure within maj'^ be relieved by the flow of the water from the open sides. In freezing tests, as usually conducted, the brick are exposed to cold on all sides alike. As they lose their heat, they cool simultan- eously on all surfaces, thereby freezing and sealing the pores on all sides at the same time, and completel.y enclos- ing the remaining water within the brick. In the case of brick laid in the wall, this is not true. The outside face becomes chilled long before the others, and while the pores on the surface are sealed, the others are left open, offering a passage for the water as pressure in- creases. Consequently the freezing test puts the brick un- der conditions to which those in the wall are never sub- jected. A brick saturated with water and placed in a position where it is possible for it to drain, but where evaporation is prevented, will lose but a very small amount of water. A series of the brick tested were placed, after saturation, in such a position, and at the end of forty-eight- hours had lost but one or two grams of water. Consequently, when the brick are frozen in the freezing can, they do not drain. But as it is impossible to completely saturate brick by soaking, certain parts are free from water. When ice be- gins to form and the ice plugs have become strong enough to resist the growing pressure of the freezing water, relief 26 HA.KDNESS OF BKICK AND THEIK KESISTANCK TO FROST. is had by the flow of part of the water into the vacant spaces. The relative amount of damage done in the brick depends upon the ease and rapidity with which this is ac- complished. As the resistance to flow increases rapidly with the decrease in the effective diameter of the pores, the finer pored brick should suffer the greater loss in freezing. Flow throuijli capillary tubes. The pores of the brick are not of course perfect capillary tubes, but it has been shown that they obey the same laws. The truths, therefore, will be at least approximated by considering them as such. The volume of flow of a liquid through a capillary tube in a unit of time is expressed by the formula in which r is the radius of the tube, p is the pressure, or unbalanced force, at the end of the tube, u is the viscosity, or resistance to flow of the liquid used, and I is the length of the tube. Expressed in words, the formula means that the amount of flow in, say a second, increases sixteen times when the radius of the tube is doubled, is doubled when the pressure is doubled, is halved when the viscosity of the liquid is doubled, and is halved when the length of the tube is doubled. The velocity of flow may be determined by dividing the total flow by the area of the cross section of the tube. Velocity or V=: ^> — j- h- tt r - This means that if the pressure, viscosity of the water, and the length of the pores were equal, a pore twice the size of another would transmit water four times as fast. Therefore the coarser pores of a brick would offer the least resistance to the flow of water through them. Effect of laminations upon the durahility of hrick. The laminations are irregularly scattered through the brick, and as a result, the water is not uniformly distri- buted but the greater portion of it is concentrated in these cavities. The volume of water in the laminations is enor- mously greater than that in the pores. Consequently the HARDNESS OF BRICK AND THEIR RESISTANCE TO KKOST. 2( amount of expansion due to freezing is not only greater, but it is concentrated at one point. Further, the lamina- tions extend over a relatively large area and so weaken the brick that much more. On account of the greater amount of expansion and the smaller resistance offered by the brick at these points, the greater amount of damage is done by water in the laminations. The laminations are connected with each other and the surface of the brick only through the pores. The rapidity with which they are filled and drained, is governed by the size of the pores and that alone. As with the pores, the amount of damage that is done depends upon the complete- ness with which the laminations are filled. Therefore, it follows that the durability of brick depends, not upon the number or size of the laminations, but upon the effective diameter of the connecting pores. As these form the greater part of the total pore-space, it is now clear why the dura- bility of brick does not run parallel to the total pore-space. As the hardness of brick increases, the effective diame- ter of the pores decreases, and tenacity and brittleness in- creases. The effect of decrease in the effective diameter of the pores is to increase the resistance offered to the flow of water through the pores, the increase of tenacity gives greater strength to resist the expansion of the forming ice, and the increased brittleness decreases the limit to which the brick may be strained without rupture. Consequently the harder burned brick may be expected to suffer the greater relative loss in a freezing test, whenever the effects of the smaller pores and the greater brittleness overcomes the increased strength. Even when this is the case, it does not necessarily mean that the harder brick are the poorer, for as a rule their increased strength gives a large factor of safety, and even after they have lost forty percent of their original strength, they are often much stronger than is required. Relative coniraction of hrick and ice. After ice is formed, the further damage it will do to the brick depends upon their relative rates of contraction. The coefficient of 28 HARDNESS OF BKICK AND THEIK RESISTANCE TO FROST. expansion^' of clay wares is .00000457 and that of ice .0000350. This means that ice will contract nearly nine times as fast as brick, and will shrink away from the walls of the pores as the temperature is lowered. Therefore ice can damcujc brick only at the time of its formation. CONDITIONS GOVERNING THE BRICK IN THE WALL. The conditions under which brick are placed varies in different parts of the wall. In the foundation, below the water line, brick are subject to continual immersion in water, and under these circumstances must sooner or later become fully saturated. In this part of the wall, there is no chance for any drainage, and the only factors that are called into play are: the total amount of pore-space, and the greatest strength. These bricks are those that have been burned hardest, and in this situation are, beyond doubt, the ones that should be used. Just above the water line, capillarity determines the amount of water contained. The height at which water stands in a capillary tube, depends upon the diameter of the tube. The smaller the tube, the greater is the height. Consequently, in walls footed in a constant source of sup- ply, water will ascend higher when built of finer-pored brick, than when built of brick with larger pores. The finest-pored brick are the slowest to pass on to the founda- tion, the water received from above. On this account, also, finer-pored brick are saturated to a greater height above the water line than the coarser-pored bricks. On the other as the conditions are practically those of saturation, the gain in strength of the harder burned brick over the coarser-pored ones, more than offsets the advantage of a narrower saturation zone afforded by the latter. Above the zone where capillarity is dominant, the brick obtain water only from atmospheric precipitation, and leakage from pipes and gutters. Here the conditions i^Castell Evans, Phys. Chem. Tables, p. 147. HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. 29 are those of drainage, and the brick that can rid itself of water most quickly is most likely to endure.. The velocity with which water passes through a brick depends entirely upon the size of the pores. Therefore, in this situation, coarser-pored brick should be used, since they transmit water faster. The three different situations in which brick are placed in a wall are : the saturated zone below the water line, the capillary region just above, and the drainage zone above these. (See figure la.) In designing freezing tests, the difference in the situations in which brick are to be placed should be taken into account, and the tests should indicate the brick best suited to each of these positions. This has not been generally recognized, and as a conse- (pieuce many brick that are well suited for certain condi- tions are condemned, because the tests used are the same for all situations. Thus the coarser-pored brick, which are really best for situations where drainage is the dominant factor, are unjustly condemned on account of their high porosity. The claim is constantly made that certain build- ing materials will not withstand the frost, simply because they are porous and absorb water rapidly. It is evident that this material, in a position where it can drain, will be freed from danger of freezing much more rapidly than one that absorbs water slowly, and therefore drains slowly. It is true that the coarser material will pass water through the wall to the interior much more rapidly and abundantly than that which is finer. If the amount trans- mitted is greater than the air in the interior can absorb, the walls will be damp. This is another problem entirely, and has nothing to do with resistance to frost. Therefore, in considering only the powers of a building to withstand weather, the conclusion is obvious that in situations above the saturated zone, coarser-pored brick are better, provided they are burned sufficiently to remove them from the soft salmon class. 30 HARDNESS OF BBICK AND THEIK KKSISTANCE TO FKOST. TRANS. AM CER SOC VOL.IX JONES, PLATE lA I I Zone of -iV^terL/ne Ideal Sketch Showing Normal Distribution of Water in a Brick Wall. HARDNKSS OK HKK'K AND THEIK RESISTANCE TO FROST. 31 FREEZING TeSTS. Recognizing- the different conditions under which brick are placed in a wall, tests should be designed with these conditions in view. The important factors to be de- termined in brick to be placed in the zone of saturation, or in similar situations, are total pore-space and strength. The factor of greatest importance in brick to be placed in the zone of drainage, is rate of flow of water through the pores. As long as the brick possesses strength enough to carry its load with a safe margin, further strength is not necessary. The question here is not one of amount of porc- spacc, but size of pores. These should be sufficiently large to permit proper drainage, thus preventing danmge to the brick by frost. Methods of determining pore-space. The common method of determining pore-space is to place the brick in water, and after a certain time, determine the amount of water absorbed. This is considered as efpiivalent to the total pore-space. When a brick absorbs water, it is neces- sary for the enclosed air to flow out and water flow in through the pores. The rapidity with which this takes place depends upon the effective diameter of the pores. If, during this process, any air remains in the pores and cavi- ties, and is surrounded by the entering water, its only method of escape is by diffusion through tlie water. This is a very slow process as compared with flow. It lias been observed that a bri(k will gain in weight when left in water, even after a iiionth's time.''* Part of this gain is probably due to bacteria, and other minute or- ganisms, that colonize and multiply within the brick. When a surface is exposed to the air, water evaporates, and may deposit salts within the brick. The greater part of the gain, notwithstanding these other sources, is due to the diffusion of imprisoned air, and its replacement by water. As an illustration of this point, two of the series of brick tested were placed in water three inches deep and wwheeler, Mo. Geol. Surv., Vol. 11. 82 HARDNESS OK BRICK AND THEIR RESISTANCE TO FROST. allowed to .stand fortjreiglit hours. After weighing and re-drying, they were placed in water onlj' one inch deep, for the same length of time. Using the amount of water ab- sorbed during deep immersion as one hundred percent, the percentage of water absorbed during shallow immersion was found to be as follows : Kind ot Brick Soft Med. Soft Med. Hard Hard Soft mud shale Wire cut shale 163.2% 95.5% 120.6% 98.5% 91.5% 43.3% 70.9% 46 3% The coarser-pored brick had absorbed much more dur- ing shallow immersion than during deep immersion, while the finer-pored had not absorbed so much. When a brick was deeply immersed, water flowed in from the sides, above the air in the bottom of the brick, before it had a chance to escai)e. Replacement of the entrapped air could only be accomplished b}^ ditfusion. In shallow immersion, the air had a chance to escape through the pores above, and was forced out by the ascending water. Another factor in the experiment is the relation of floAv to pressure. It will be noticed in the formula express- mg the velocity of flow in capillary tubes g — . that the flow increases when the pressure is increased. The pres- sure of the water on the bri(*k was twice as great in th(^ deep immersion as in the shallow. The water, therefoi-e, flowed in twice as fast through th(; pores during deep im- mersion. In addition, the area exposed for the entrance of water was nearly three times as great, and therefore many more pores were available for the entrance of the water. It is this effect that masks the imprisonment of air in finer- pored brick, and it was in spite of this that the more rapid flow and trapping of air took place in the coarser -pored brick. Many investigators have maintained that complete im- mersion is the only natural method by which the relative HARDNESS OK BRICK AND THEIR RESISTANCE TO FROST. 33 porosity of brick sliould be tested, since it is by soaking tliat a brick becomes filled, when it is in a wall. It is true that brick in the zone of saturation, where they are contin- ually in contact with water, become saturated by this method. The length of time thej are in contact with the water is not taken into account, however, nor the fact that the air imprisoned at the first filling of the pores slowly dittiises until only the amount held -in solution by the water is left. Brick in this situation become eventually as thoroughly saturated as if all the air had been removed and replaced by water in the first place. This complete re- placement of the air is not possible by complete immersion for the short time given in the usual absorption test. As the only object in finding the pore-space is to determine its total amount, in order to judge the relative durability of brick in situations of complete saturation, the value of this method is small except as a rough test. Wlieuever the method of soaking is used, the depth of immersion should be adjusted to the rate of flow of the water througli the pores. This is illustrated by the above experiment where the coarser pored brick were more com- pletely filled by shallow immersion, while the finer pored were better filled by deep immersion. The method that undoubtedly gives the most accurate and concordant results is that proposed by Dr. Buckley.^" He placed the specimens to be tested, after drying and weighing, in an air tight jar and, after exhausting the air as completely as possible, allowed l)oiling water to slowly enter and cover the stones. This demands considerable apparatus and is not available for general use on this ac- count. It is possible to approximate the truth by placing the brick, after drying and weighing, in a pan with a small amount of boiling water and boiling them six hours, add- ing more water from time to time, until during the last liour of the test they are completely immersed. As heat decreases the viscosity of both air and water, the flow 'Building and Ornamental Stones of Wisconsin, "Wis. Geol. Survey. 34 HARDNESS OF BRIOK AND THEIR RESISTANCE TO FROST. through the pores is accelerated, and the saturation fairly complete at the end of the time. Care should be taken to use as pure water as possible, in order to prevent the pre- cipitation of salts in the brick through evaporation of water from the exposed surfaces. If this should take place the amount of pore-space determined would be more than the true pore-space. Rain water or distilled water is the best. Calculation of pore-space. Having determined the total amount of water ab- sorbed, it is possible to calculate the pore-space. . In doing this, care must be taken to use the same kind of units throughout. For instance, the common method is to divide the weight of water absorbed by the dry weight of the brick. This cannot give the true pore-space, since it uses the mass of the brick to divide the mass of the water, a sub- stance two and a half times as light, volume for volume. The result thus obtained is too small. To obtain the cor- rect pore-space, the two masses must be reduced to equiva- lent substances, or volumes, and must be expressed in terms of one or the other. The simplest reliable method with which the writer is acquainted is one devised by Mr. Purdy.^^ This is, divide the water absorbed by the difference between the wet weight of the l)rick and its weiglit when suspended in water. This gives the volume of water absorbed divided l>y tlie volume of the brick, or true pore-space. In determin- ing the pore-space of the brick tested, a slightly modified form of this method 'was used which gave, when carefully' executed, nearly as accurate results. The volume was de- termined by measurement of the three dimensions, and this was used as the divisor of the water absorbed. Measuremeut of the rate of floio through pores. As has been stated, the velocity of flow through capil- lary tubes is expressed by the formula ^ , in which r is 20Page 211 of this volume. HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. 36 the radius of the tube, P is the difference iu pressure at the ends of the tube, u is the viscosity of the fluid used, and I is the length of the tube. When a brick is placed in water, the force that causes the water to enter and pass into the brick, is the adhesion or attraction of the water for the walls of the pores. The height that the water will reach, depends upon the mutual attraction of the molecules at the suitface of the water, or surface tension. The force that causes the water to flow is expressed by the formula 2TrrT cos a, in which r is the radius of the tube, T is the surface tension, and a is the angle of contact between the water and the walls of the pore. The opposing downward force, due to the weight of water raised in the pore, is expressed by the formula irrVipg in which h is the height to which the water has risen at any particular instant, p is the density of the water, and f/ is the force of gravity. The unbalanced force that causes movement is equal to their difference. Unbalanced force=P^=2irT cos a — -n-r^pg. Since, in the equation, all of the factors may be con- sidered as constants in the case of the brick and water, ex- cepting r and h, and designating these constants as K and k respectively, the equation may be written P=Kr — h-V,. Substituting this value of P in the velocity equation, we have Velocity=V== gu^ Since 8u is a constant in this particular case, and I equals h we can write the equation K.h "^ K.\h F= "• „ ';• " = ^ (-^ —kr' This means that the velocity of the water entering the pores of the brick through capillary action varies approxi- mately as the cubes of the radii of the pores, and inversely 36 HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. as the lieiglit to which it ascends. Consequently a coarse- pored brick will fill much more rapidly than a fine-pored one. As a test of this theory, the three series of brick used in the freezing tests were placed in water, and the rate at which they filled determined by weighing at intervals of Vi) %? I5 6, 24, and 48 hours. The first set, the soft mud made from surface clay, were placed in water three in^fhes deep. While some differentiation was shown, it was thought best to set the others in water only one inch deep, and allow capillarity to have full play. As was expected, the differentiation was more marked. The results are as follows: Kind of Grade of Hardness Pore Space G ain Total BriCK abs'b'd. 15 miu. 369.0 30min 1 hr. 6hrs. 24hrs. 10.5 48hrs 4.6 Soft mud Soft 33.0% 45.3 5.1 8.7 442.0 Surface Med. s't 26.9% 320 9 2.2 2.1 7.1 7.6 3 6 343 6 Clav ' ' hard 21.2% ii37.9 8.1 1.6 7.7 4.6 3.2 258 9 Hard Soft 10.3% 26.9 4.4 52.7 8.9 37.4 33 2 12.5 17 4 2 106 6 Soft 26. 2 f. 235.1 94.2 50.4 4.53 1 Mud Med. s't 17.8 166.8 44.2 68.0 23.7 5.0 10 310.1 Shale " liard 11.6 17.5 9.1 8.1 68.2 26.4 1.0 180.4 Hard 5.8 4 5 1.0 40.4 1.0 64.1 6.7 19 2 16 6 16 8 9.1 48 4 Wire Soft 27.6% 160.5 111 3 402 Cut Med. s't. 17.1 78.8 19.4 25.7 89.7 13 7 8 4 233.3 Shale " hard 2.1 3.8 0.4 0.9 2.2 2 3 1.6 11.1 Hard 0.9 2.2 0,3 0.4 0.5 0.5 0.7 5.6 Using the total amount of water absorbed b}' each grade as one hundred, the percentage absorbed during each interval was found to be as follows: HARDNESS OF BRICK AND THKIR RESISTANCE TO FROST. 37 Gain Kind of Grades of Hardness Brick 13 min. 30 min. Ihr. 6hrs. 24 hrs. 48 hrs Soft Soft 83.0% 10.3% 1.2% 2.0 2.4% 11% Mud Med. soft 93.4 0.7 6 2.0 2.2 1.1 Surface Med. hard 92.1 1.2 0.7 3.0 1.8 1.2 Clay Hard 25.1 51 9 4.0 11.9 3.4 35 31.1 1.4 Soft Soft 21.0 11.3 2.9 10 Mud Med. soft 54.2 14.3 21.9 7.7 16 0.8 Shale Med. hard 13.4 7.0 6.2 52 3 20.3 8 Hard Soft 9.6 2.1 2.1 13.8 39.7 3 27 Wire 40.0 10.5 15.6 27.5 4.1 2.3 Cut Med. soft 33.9 8.5 11 3 38.8 6.0 15 Shale Med. hard 34.5 3.5 8.0 19 3 20.5 14 2 Hard 47.3 6.4 9.6 10.3 10.3 IH.l Average Soft 58.3 11.0 12.6 13.6 3.1 1.5 of Med. soft 60.5 7.8 11 2 16.2 3.3 10 Grades Med. hard 46.7 3.9 4.9 24.9 14.2 5.4 Hard 27.3 4.2 5.0 19.7 27:0 16 7 Plotting the average percentages absorbed by the sim- ilar grades of the series shows graphically the relative rates- with which the water was absorbed by the different grades of brick. (See figure No. 2.) It is fair to assume that all of the brick in a series had approximately the same number of pores before they were burned. As the burn progressed, the effective size of the pores decreased as is indicated by the diminishing pore space. Therefore, the difference in the rates of absorp- tion is a function of the size of the pores rather than the number. The curves of the soft and medium burned brick show but little difference. As has been stated, the effective diameter of the pores increases during the period of dehy- dration and oxidation. As these brick had been burned only to about the close of this period, their pores should be about the same size, as is indicated in the result. It is proposed to use this method of determining the relative effective diameter of the pores in testing brick that are to be placed in situations where drainage is the domi- nant factor. In using it, the method of procedure will be as follows: 38 HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. <. \ uJ \ a u z o -> 1 \ \ \ i \ i '1 i s^ s^ i^ ;v liJ 01 vO CO ' 'O S^ (D 6 00 in ^ •: "^ , « in 7 2 2 Z .. 1 1 5 5 ^ ffl m to (fl ^ t t g S , o o < < i^ If) ip I I 1 1 1 1 \ \ 1 1 1 \ 1 \ f _A < 5 5 > o o 2 5 <■ \ \ I \ N \ : \ \ \ \ \ \ 1 • \ \ \ \ \ • \ \ \ ; \ \ X J 5 ■ \ \ \ \ \ . u in (T I \ \ u < ^ "^*. %- \ \ 111 z 1- % 1 >■« V. -o n c f ? 5 I .5 ^ s s c o NOUdaOSQV JO 39VJ.N30b3c) \ HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. 39 (1) Dry the brick forty-eight hours at a temperature of 300 degrees Fahrenheit. (2) Weigh, after cooling to room temperature. (3) By room temperature is meant about 75 degrees Fahrenheit. (4) Place the brick on edge in water whose final depth after the brick are placed is one inch. (5) The temperature of the water is important, as its viscosity changes rapidly with temperature, thus chang- ing the rate of flow. If it is kept at room temperature, or near 75° F in every test, the results may be safely com- pared. Otherwise a variable factor is introduced. (6) At the end of fifteen minutes, remove the brick and weigh, after removing the surplus water clinging to the surface. (7) Replace in water for forty-eight hours. (8) Weigh as before. (9) The percentage of water absorbed in fifteen min- utes, using the amount absorbed in forty-eight hours as 100 percent, indicates the relative rate of absorption. Methods of freezing. In testing brick to be placed in a position where they cannot drain, as in a foundation, they should be saturated as completely as possible by boiling, or under the air pump, and frozen while standing in water. This will test them under conditions that are similar to those actually occur- ring. The brick to be placed in situations where they can drain, as in the upper wall, should be filled with w^ater by soaking and placed in a dry can w^hile freezing. These are approximately the conditions under which they will be placed in the wall. At the temperatures just below freezing, overcooling plays a more or less important part. In tests using these temperatures, the results are uncertain to the extent of the unknown value of this factor. As is indicated by the win- ter temperature conditions prevailing in Springfield and Chicago, the drop in temperature after a storm is probably to about ten degrees F. It is probable that only the pores 40 HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. are able to cool to this temperature without freezing. The additional damage resulting from freezing them is at most ver}^ slight, and the differential expansion of the brick it- self is negligible. Cooling to a moderate extent below this temperature would not materially alter the results. It seems to the writer, therefore, that the temperature of the freezing can should be at least as low as that prevailing in nature, and that average temperatures as low as zero F are permissible. It is best for the comparison of results to keep between the limits given. As it is impossible to con- trol the temperature of the atmosphere, or to obtain uni- formly low temperatures during the length of time neces- sary — twenty days — to conduct a freezing test, a refriger- ating plant is a necessity. EXPERIMENTAL DATA. As an illustration, three series of brick were selected, a soft mud made from surface clay, a soft mud made from a shale, and a wire cut made from the same shale. An at- tempt was made in selecting the brick to have in each sSr- ies four grades of hardness which were designated as soft, medium soft, medium hard, and hard. Each grade of the surface clay was represented by five brick, and each grade of the shale brick by eight. These were selected by eye and by the sound emitted when struck, from the different parts of the kilns, each series coming from one kiln, and were as nearly uniform as was possible to obtain under the circum- stances. The selections were made by Mr. Purdy. Four brick of each grade and series were used in the tests of absorption and freezing. The brick were dried forty-eight hours in an air bath at a temperature of 340 degrees F and weighed, after cooling in the bath to room temperature. They were then placed in water three inches deep, and allowed to remain forty-eight hours, after which they were packed, and covered in a can placed in brine to a depth that brought the surface of the brine considerably above the top level of the brick. The brick were left here from ten to twenty-four hours until they were thoroughly HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. 41 frozen. They were then removed and immersed in water at a temperature of approximateh' tifty degrees F, until they were completely thawed out. This cycle of freezing and thawing was repeated twenty times. After the final thawing, the brick were dried as before, weighed and the loss due to freezing determined. The temperature of the brine averaged two degrees F during the tests, with a max- imum range from 18 to 24 degrees. As all brick of a series were frozen together, this extreme variation did not affect the comparison of the results. The brick were then crushed, together with unfrozen duplicates, in an Olsen testing machine of 200,000 pounds capacity. For the crushing test, the brick were first broken with a hammer and the half-brick used, in order to bring them within the capacity of the machine. Where there was but one brick to crush, as in the case of the unfrozen ones made from surface clay, both halves were crushed and the average taken. The brick were bedded in plaster of paris and the plaster allowed to set under an ini- tial pressure of 2000 pounds. It was impossible to get the brick bedded uniformly even by this method, and there was an extreme variation of 400 pounds in halves of the same brick. The brick generally failed quietly, although occa- sionally, especially among the harder ones, they would ex- plode and send fragments flying ten feet from the machine. As the number of the unfrozen bricks varied, the num- ber tested is given in the results. The results given by the frozen brick are invariably the average of four specimens. As it was practically impossible to select specimens in each grade that would have the same crushing strength, and since the perfection of bedding necessarily varied in the different brick, the extreme variation in each grade ranged from 400 pounds in the softer grades to 4000 pounds in the harder brick. Otherwise the tests are quite satisfactory, and it is believed as nearly accurate as was possible to ob- tain under the circumstances. 42 HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. Results of the Freezing Tests. Percent Percent Kind of Grade of Crushin 5 Strength No. Loss Percent Ab- Brick Hardness Tested in Strength Pore Space sorption in 15 minutes Remarks Frozen Unfrozen Soft Soft 1194 1374 1 13.1 33.0 83.0 Frozen Mud Med. soft 3567 3400 1 *4.6 26.9 93 4 Brick Surface ' ' ' hard 4289 5315 1 19.9 21 2 92.1 Scaled Clay Hard . J,, 7377 7260 1 *1.6 10.2 25.1 on Face Soft Soft 2671 2913 3 8 6 26 2 52.9 One Brick Mud Med. soft 46 .'5 6793 3 20.2 17.8 54.2 Broken in Shale " liard 8522 10143 2 16.5 11.6 13 4 Freezing Hard Soft 7606 11470 4 33 8 5.8 9.6 Wire 3729 4637 4 19 6 27.6 40.0 Two bricks Cut Med. soft 6965 8117 4 14.2 17.1 33 9 freezing. Shale " hard 9165 11315 4 19.4 2 1 34.5 Two bricks Hard 115 . 11997 4 4.1 0.9 47.3 cracked in freezing. *Gain. The loss in weight due to freezing was very small, being but a few grams, in most cases, and 2% in an excep- tional one. If these results are arranged in the order of their pore-space, hardness, and rate of absorption as indicated by the percentage absorbed in the first fifteen minutes, the relations of these factors to loss of strength due to freezing is clearing brought out. Arranged in the order of hardness. Grade Surface Clay Soft Mud Shale Wirecut Shale Average of All Soft Med. soft Med. hard Hard 13.1% 4.6* 19.9 1.6* 8.6% 20.2 16 6 33 8 19.6% 14.2 19 4 4,1 13.8% 9.9 18.6 12 1 *Gain. As may be seen, there is little relation between the hardness of the brick and its resistance to frost. The sur- face clay suffered greatest loss when burned medium hard, and their hard burned representatives suffered much less; HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. 4S the soft mud shale suffered most when hard burned, and the softest brick the least; while with the wire cut shale brick, the soft and medium hard suffered most, and the hard burned much less. Upon plotting the average of the three kinds, the curve zigzags very decidedly, indicating that hardness in itself does not determine the amount of resistance the brick will offer to frost. (See figure 3.) Arranged as to Porespace. Porespace Per cent. Loss. Average of Similar Groups. Porespace Per cent. Loss. 33.0% 13.1% 33.0% 27.0 13.1% 27.6 26.9 26.2 19.6 4.0* 8.6 7 9 21.2 17.8 17.1 11.6 10.2 19.9 20 2 14.2 16 5 1.6* 18 11 18 1 7.4 5.8 2 1 0.9 33.8 19 4 4 1 3.0 19 1 *Gain Arranged as to pore-space even greater discrepancy than in the former case is seen. The plotted curves zigzag in every instance. (See figure No. 4.) They should be continuous, if there were any relation between the amount of pore-space and resistance to frost. An interesting series of results along this same line is contained in Dr. Buck- ley's report on the building stones of Missouri-^, in which the same lack of definite relation of amount of pore space and resistance to frost is strikinglv brought out. 21 Buckley, Quarrying Industry of Missouri, 2nd ser. Mo. Geol. Surv., Vol. 2, PI. LIX. 44 HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. TR,ANS.AM.CER.SOC.VOL.IX JONES, PLATE III 30 Soft M.Soft M.Hard Hard Relation of hardness and percent of loss HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. 46 TRANS.AM.CER.SOC.VOL.IX JONES, PLATE !V 30r 40 30 20 lO PERCENTAGE PORE SPACE RELATION OF PORE 5P/VCE AND PERCENT OF LOS5 46 HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. Arranged in the order of their rates of absorption. Percent absorbed in 15 minutes. Per cent. Loss Average of Similar Groups. Per cent. Absorbed Per cent Loss 93 4 92.1 83 4.6* 19 9 13.1 89 5 9.4 54.2 52 9 47 3 20.2 8.6 4.1 50.0 11 40 34 6 33 9 19.6 19.4 14.2 35.0 17 7 26 1 1.6* ^25.1 1.6* 13.4 9 6 ,. 16 5 33.8 10.0 25.1 *Gain. When the brick are arranged in the order of their rates of absorption, as indicated by the amount absorbed in the first fifteen minutes, the individual curves show only slight agreement with the loss caused by freezing. (See figure No. 5.) The surface clay seems to indicate the reverse of the theory developed. The brick that absorbed most slowly, and therefore had the smallest pores, suffered least. The explanation of this probably lies in the number of unfrozen bricly tested. Only one in each grade was available, and consequently the individual variations in strength, perfec- tion of bedding in the machine, and other conditions, the effect of which it is impossible to avoid except by using a number of test pieces, are prominent. Indeed, two of the grades indicate a gain in strength rather than a loss due to freezing, which is not a reasonable thing to believe and, consequently, while the evidence of the surface clay brick is accepted, not as much weight can be given it as that of the shale brick which were tested more completely. With HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. 47 30 TRANS.AM CER.SOC VOL. l> JONES, PLATE V ?o A 1 1 1 10 — ' ^ :/ 1 1 1 1 1 1 o--::^ ^^ e^f V/^ i 1 Q s \ \ > 1 0) >i ? IL 30 z UJ ^ 20 UJ a 10 \ ^- f\ i '-'At /) 1 • (^ A* 1 < N) \00 90 &0 70 eO 50 '^O 50 2.0 \o o PERCENTAGE ABSORPTION IN l5 MINUTES RELATION OF WATER ABSORBED IN 15 MINUTES A.NO PERCENT OF LOSS 48 HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. the latter, the finer-pored brick, even though they were not always the hardest or possessed the greatest pore-space, generally suffered the greatest loss. Upon averaging the groups with approximately the same rate of absorption, a curve is obtained that is continuous from beginning to end. The only exception is one of the questionable values of the surface clay brick. As the individual variations can only be eliminated by the use of a considerable number of speci- mens, it is believed that this curve approximates the truth, and indicates the value of the rate of absorption as an in- dication of the power of brick to withstand the ravages of frost. This is true, since the rate of absorption is governed by the same factor that controls the rate of flow through the pores, and consequent relief from danger of damage by frost. Therefore the results obtained from the brick tested are believed to confirm the theory developed. SUMMARY AND CONCLUSION. It has been shown that the pores in brick originate in the spaces between the grains of clay used in manufacture. The size of the pores depends upon the size of the grains, and upon the manner in w^hich they are packed. The coarser pores result from the larger sized grains and looser packing. Scattered through the brick are numerous cracks and cavities, known as laminations, which are produced in several ways during the process of manufacture. These are relatively much larger than the pores.* As they are not in direct contact, their only connection with each other, and the surface of the brick is through the pores. The pores and the laminations together make up the pore-space of the brick, the laminations furnishing the major part. Tlie method of -manufacture determines the compact- ness with which the grains are packed, and therefore the size of the pores and the pore-space. Other things being equal, the soft mud process will make a more porous brick than the stiff mud process, on account of the greater amount of water and slighter pressure used. During the water-smoking and oxidation period of the HARDNESS OF BRICK AND THEIR RESISTANCE TO FROST. 49 burn, pores are opened and enlarged bv the escaping gases. The bricks have tlieir maximum amount of pore-space at this time, and it remains at this value until the amorphous material begins to melt. When this happens, the grains soften around their borders and run together. This tends to obstruct the pores, and eventually completely closes them at different points along their length. The parts of the pores that are enclosed between the points sealed, con- tain gas that prevents the further closure of the pore. As the amount of glass increases, the enclosed gases form bubbles which become spherical in shape. These bubbles are completely shut off from the system of pores, and the pore-space apparently decreases. Those pores that are not sealed are obstructed, which is equivalent to making them smaller. The strength of the brick increases as it is burned, and also its rigidity or brittleness. This gives the brick greater resistance to strain, l)ut decreases the distance it may be deformed without breaking. A given amount of expansion or contraction may, therefore, rupture a rigid brick, when it would not harm a tougher one. Owing to the universal occurrence and abundance of water, and its property of expanding upon freezing, it is the only substance that under ordinary conditions causes any considerable damage to brick. It expands one-tenth of its volume when freezing, and when it is confined within the brick, may burst it upon freezing. The important factor serving to mitigate the destructive effect of freezing- water is the opportunity generally afforded for a portion of the water to drain, before the brick cools sufficiently to freeze. The temperature at which the brick freezes may be blow the ordinary freezing point of water, owing to the property of capillary tubes which delays the freezing of water within them until a lower temperature is reached. It is not probable that freezing in the larger openings is en- tirely prevented at the temperatures prevalent during the winter months in our northern States. The amount of water that may drain depends upon the rate with which it passes through the pores. This is deter- 50 HARDNESS OP BRICK AND THEIR RESISTANCE TO FROST. mined by the size of the pores, and is much greater in coarse pores than in the finer ones. It was found that water will pass through pores four times as fast if they are doubled in size, and nine times as fast if they are trebled. Although the lamination cracks contain the bulk of the water, they are dependent upon the pores for drain- age. Consequently the amount of damage done in a brick depends on the size of its pores, since this governs the rap- idity with which it will drain. In the foundation at the water line, the brick are con- tinually in contact with water. As any air originally con- fined in them will eventually diffuse, they become com- pletely saturated. The brick just above the water line are filled through capillary action from below, and from the drainage of the upper wall. These, also become completely saturated, and as the frost seldom reaches the water line, it is this part of the wall that suffers most. The amount of damage done the brick in this zone of saturation is pro- portional to the total pore-space, and the strength of the brick. The best brick for this situation is, therefore, the one that is strongest and least porous. In the upper Avail the brick are able to drain. The amount of damage is therefore proportional to the rate of flow through the pores, and the total amoutn of pore-space has little direct effect. As in freezing, tests giving similar effects are present, it is easy to understand why the hard- burned, fine-pored brick suffered more than those softer, but with coarser pores. It is of vital importance to consider the future position and conditions in which brick are to be placed, in making tests to determine the ones best adapted. In the situations where saturation is the controlling condition, as in foun- dations, evidently the brick that contains the least amount of pore-space is the best. In consequence it is necessary to determine the total pore-space, minute pores and all, since these become filled, sooner or later, in saturated situations. This cannot be done by the method of soaking at present used, but may be approximated by boiling or the use of the air pump. t HARDNESS OK IIHU'K AN1> THKIH KKSISTANOK TO KROST. ■)1 In situations where drainage is the controlling factor the brick that will drain fastest is the best, if injury from frost only is considered. The relative rate with which the brick will drain must be obtained. This may be easily done by the method proposed on page IV,K The crushing strength of brick has an important value in foundation brick, as it indicates the relative resistance the brick will offer to expansion of the freezing water. The brick in the upper wall, on the other hand, need only the strength necessary to carry their load. Consequently these three tests, — total pore-space, rate of flow through the pores, and crushing strength,— should give a correct indication of the power of a brick to with- stand frost. In the brick to be used in situations of satur- ation, only pore-space and strength need be determined. The characters of the brick which are altered during burning are: the size of the pores, the amount of pore- space, the strength of the brick, and its rigidity. The harder burned brick have finer pores, a smaller amount of pore-space, greater strength, and greater rigidity. They therefore drain more slowly, contain a smaller amount of water when jftlled, have greater strength to resist the ex- pansion of freezing water, but will rupture with a smaller amount of expansion. Whether or not the hardest burned brick will resist frost best, depends upon the relation be- tween its gain in strength and loss of pore-space on the one hand, and the decrease in the effective diameter of its pores and increase in brittleness on the other. If the for- mer factors are progressively altered more rapidly during the burn than the latter, the harder burned brick will be the more durable. If, on the other hand, the latter factors are the ones to develop more rapidly, the power of resist- ance of the brick as the burn progresses is relatively de- creasing. The quantitative expression of this relation has not been worked out, and must be left for some future in- vestigator. The present investigation, however, has shown the direction in which the relation between hardness and resistance to frost mav be found. n K'%. -'5f,.. V%.' ■»»-, .^'. i^^ -^'^ '^-jifC^j J ^r >\ ^\*- T'^ /■^^'^C^v-'K UNIVERSITY OF ILLIN0I9-URBANA 3 0112 052567101 i?^-v Mfr %fm,