Popular Science Monthly/Volume 23/September 1883/Fire-Proof Building Construction
By WILLIAM E. WARD.
IF society is indebted to the restless spirit of progress for most of its modern comforts and conveniences, it certainly is not yet a debtor for any methods which guarantee immunity against calamities from fire. While other departments of industry have received the benefits of improvement, the persistent use of combustible material for exposed portions of buildings has limited the intrinsic elements of the art of building construction, and confined improvements only to matters of design.
Incombustible materials are easily obtained, and, for every apparent reason, much better adapted to the purpose. Doubtless, the question of increased cost, both in money and in time required for more thorough construction, may be in a measure responsible for the tardiness in adopting safer methods; and, in addition to greater expenditure, there may have been a want of confidence in the fire-proof methods which have been offered to the public for adoption. The importance of this question induced the writer, in 1871-'72, to make some experiments in a new and special direction, for the purpose of ascertaining whether a practically fire-proof building could be designed and constructed at a comparatively moderate cost.
The incident which led the writer to the invention of iron with béton occurred in England in 1867, when his attention was called to the difficulties of some laborers on a quay, trying to remove cement from their tools. The adhesion of the cement to the iron was so firm that the cleavage generally appeared in the cement rather than between the cement and the iron.
The experiments which followed were confined exclusively to working up the reciprocal value of béton, in combination with iron, in the construction of beams which were designed for supporting floors and roofs made of the same material. In this particular the facts were conclusively developed that the utility of both iron and béton could be greatly increased for building purposes, through a properly adjusted combination of their special physical qualities, and very much greater efficiency be reached through their association than could possibly be realized by the exclusive use of either material, separately, in the same or in equal quantity.
Experience had long ago proved that unprotected iron, associated with combustible materials, is altogether unreliable for building purposes when exposed to a severe fire-test; but it has been demonstrated, that, if iron is well protected by a heavy clothing of béton, its integrity can be safely depended upon in almost any emergency.
When all doubts concerning the reliability of the several combinations of materials required in the construction were removed, a building, embracing the following radical new features, was erected, for dwelling purposes, near Port Chester, New York: Not only the external and internal walls, cornices, and towers of the building were constructed of béton, but all of the beams, floors, and roofs were made exclusively of béton, re-enforced with light iron beams and rods.
Furthermore, all the closets, stairs, balconies, and porticoes, with their supporting columns, were molded from the same material; the only wood in the whole structure being window-sashes and doors, with their frames, mop-boards, and the stair-rails, thus excluding everything of a combustible nature from the main construction.
Béton can be used in any form of construction, and will serve the requirements of any architectural or decorative effects. All the exterior portions of this house, which are more or less ornamental in their functions, were made of béton in place during the progress of the work. In the interior of the house, the cornices, stiles, and panels of the ceilings are formed of béton, and covered with the hard finish usual in such work. There appears to be no limit to the reproduction in béton of any form used in stone masonry or in stucco. The proportions of material composing the béton for the work varied in strength to meet the requirements of the different parts of the structure: the heavy walls needing the least proportion of cement, while the beams, floors, and roofs required a much larger proportion. Only the best quality of Portland cement, clean beach-sand, and crushed blue-stone, were used in combination with iron for constructing the building.
The proportions used for the heavy wall-work were one part of cement to four parts of sand and fine gravel, thoroughly mixed dry, and dampened with only sufficient water to give it the consistency of well-tempered molding sand.
A finely crushed and screened, hard, blue limestone was found to be better adapted for use in combination with the béton than a coarse sized stone filling, because small-sized stones pack closer than large ones, thereby realizing a proportional saving in cement. The tests made to ascertain the comparative transverse strength of different compositions proved that the bond was stronger in béton made with small stone. In breaking test-sections made of béton in the form of bricks, the fracture of those filled with small stone was almost invariably across the stone lying in the line of fracture, while the fracture of the test-bricks made with a filling of stone three or four times larger showed a frequent tearing away from the bond between the béton and the larger stone filling, the composition of the béton being the same in both cases.
The proportions of cement and coarse beach-sand and gravel, used in re-enforcing iron beams for floors and roof-supports, were one part of cement to two parts of sand and gravel. The size of the iron beam, selected for an experimental test, was a four-inch I-beam of lightest pattern, twelve feet long, weighing thirty pounds to the yard, and its safety load was limited to eleven hundred and fifty pounds. A plank mold was made the length of the iron beam, twelve inches deep by five inches wide, in the bottom of which a layer of béton was first moderately tamped down to an inch in thickness; then the iron beam was laid on the course at equal distances from each side of the mold, and settled down on the surface of the course of béton to a good bearing. This brought the top surface of the beam seven inches below the top of the mold. The work of filling and tamping the courses was then continued until the mold was filled.
The reason for placing the iron beam so near the bottom of the mold was to utilize its tensile quality for resisting the strain below the neutral axis when this composite beam was exposed to heavy loads, while the béton above this line was relied on for resisting compression from load-strain. The béton became thoroughly hardened in about thirty days, when the following tests of transverse strength were made: It was placed upon suitable supports, with a bearing of three inches at each end. A lever was adjusted so as to bring the testing-load on a knife-edge bearing at the center of the beam. Weight was then applied to the long end of the lever, until the stress on the center of the beam reached nine thousand five hundred pounds. Under this load there was a deflection at the center of the beam of seven sixteenths of an inch, but not a sign of rupture appeared at any point.
The load was then removed, and the beam returned to the original line it occupied before the test, showing that the combination possesses the essential quality of elasticity in addition to the enormous increase of capacity to resist strain over that which was possible for either material to sustain if used separately, and in the same quantity.
It is suggested that for future construction an inverted ⊥-beam would furnish a more preferable distribution of iron in the composite beams than the I-beams which were used.
The result of this experiment demonstrated the reliability of the composite beam of iron and béton, and showed that the adhesion of the cement to the iron could be depended on under heavy strains. This warranted the adoption of béton, re-enforced with small rods, for the floors and roofs.
The beams for supporting the floors throughout the house were placed at such convenient distances apart as to insure perfect safety to the floors, and at the same time afford ample opportunities for producing the best effects in deep, paneled ceilings.
All the beams were molded in the positions where they belonged, both for the floors and roofs, and by the same method as the experimental beam above described. The iron beams varied in width and weight per yard in accordance with their length and the prospective load, the largest being nineteen feet long by seven inches wide. When the combination beams were completed and ready for the floors and roofs, heavy planks were firmly placed in position and securely supported between the beams, the upper surface of these plank foundations being adjusted on a level with the top surface of the molded beams. These planks served as the bottom of the floor-molds, and, after the béton forming the floor was hardened, they were removed.
Channel-ways had been molded in the walls, on a line with the top of the beams, for the purpose of supporting the outer edges of the floors.
Before the floors and roofs were laid, care was taken to cover all the supporting surfaces with paper, to prevent the adhesion of floor and roof sections to their supports. This precaution was necessary, to permit the movement of the floors and roofs that would unavoidably take place under varying temperatures and loads.
A part of the experimental system contemplated an attempt to warm the house by passing currents of heated air between the floors and ceilings, and up through flues, made in close proximity to each other, for that purpose, in the interior walls of the building; and it was necessary to core out a liberal area of lateral openings through the upper portion of the beams, in order to permit a free circulation of heated air. The ceilings rested upon flanges projecting from the lower portion of the beams, as shown in Fig. 1.
Instead of using sand and gravel, or both, in combination with cement, for floor and roof construction, the preliminary experiments that proved the superior value of broken blue-stone for massive work, led to the adoption of washed, fine screenings from the same material for the floors and roofs, because its greater angularity insured a stronger bond in the work than could be realized by using sand and gravel.
The proportions of materials used for this purpose were, one part of Portland cement to two parts of the fine stone screenings. The preparations being completed for laying down the floors, a thin course of the béton was first put on, and evenly tamped down, to about an inch in thickness, over the whole space intended to be covered. Then rods of iron, five sixteenths of an inch in diameter, were placed both longitudinally and laterally, at a uniform distance of eight inches apart, over the whole surface. Then, on this, a final layer of two inches in thickness was carefully tamped down. In about eight hours, the béton was hardened sufficiently to allow the application of the top surface, which was floated down with a half-inch coat of cement and fine beach-sand mortar, made of equal parts of each. This completed the finish, and made the whole thickness of the work three and a half inches. It will he observed that for the same reason as in beam construction, and as before explained, the iron rods for re-enforcing were placed near the bottom of the work, so as to resist the tensional stresses due to the load, while that due to compression in the upper
Fig. 1 represents a Re-enforced Beam of Béton supporting a Section of the Floor:5-inch iron rods; b, same sized rods running at right angles with a a a; c, lateral openings through the beams for the circulation of heated air; d d, flanges supporting ceilings; e e, re-enforcing beam of wrought-iron.
portion would be sustained by the béton alone. In this manner, and by this process, over thirteen thousand square feet of flooring and roofing were constructed in the building.
The only test of any consequence upon the combined strength of the floors and beams together was made on a section of the widest floor in the house, where the beams are eighteen feet span and six feet between centers. Casks of plaster were placed upon the floor over the beam, forming a triangular load of thirty tons, which was sustained without any injury to the floor, or measurable permanent deflection. The dimensions of the beam that sustained this load were, seven by sixteen inches, and eighteen feet span, re-enforced in its lower portion with a seven-inch I-beam, weighing fifty-five pounds to the yard.
This test indicates that in addition to its admitted fire-resisting qualities, the re-enforced system of construction challenges comparison with other methods of building in matters of strength and of cost, whether for buildings requiring long or short floor-spans.
Experimental tests were made with several partition walls, to ascertain how thin it would be advisable to construct them where the load was small. The result of the experiments showed that the resistance of partitions eight feet in height made of béton two and a half inches thick, and re-enforced with one-quarter-inch iron rods, was equal to brick walls eight feet high, and eight inches in thickness.
It is the opinion of the writer that for the great majority of houses required for dwelling purposes, a system of thin re-enforced double walls, with a space of from six to ten inches between them, and re-enforced cross-connections every two or three feet apart, to unite the outer and inner walls firmly together, could be built up to thirty or forty feet in height, at a cost not exceeding that of first-class brickwork.
Besides an equal economy in the construction, such double walls would be an incomparably better defense against stormy weather than the best quality of brick-work, because the absorptive capacity of béton is so much less than that of brick.
Thus, all things considered, the thin double-wall system commends itself as embodying the desirable qualities, essential to the outer and inner wall construction of dwelling-houses, furnishing as it would a sure protection against both fire and dampness, and the means for thorough ventilation. Besides the special fitness of the re-enforcing system for floors, roofs, beams, and thin walls, it is an interesting question whether the same system may not be also applicable, and advantageously extended, to a more general use in many engineering requirements—especially in situations where immense weights must be sustained, and where iron construction alone is difficult of application; notably in such an important work as the Hudson River Tunnel, where its tubular form is constructed with an outer cylindrical shell of flue-iron, and lined inside with heavy brick mason-work. Much of this tunnel rests upon a treacherous bed of silt, and might be made absolutely safe from rupture by settling, induced by vibrations resulting from railroad traffic in addition to its own weight, by adding to a thin brick lining a strong béton sixteen or eighteen inches thick, which should include three or four courses of iron bars, of suitable size, laid longitudinally and in sufficient number to bear any amount of strain that might be brought upon it. Rings of flat bar-iron, interspersed in the béton-lining a few feet apart, would further add to its security.
The re-enforced béton system has also been employed with admirable results, in heavy foundations, for stationary engines. The writer, three years ago, mounted a two-hundred-and-fifty-horse-power, tandem, compound engine, of very heavy pattern, on a re-enforced béton-bed, twenty-three feet long, five feet wide, and seven feet deep. It is apparently as firm and hard as a single mass of granite of those dimensions. The outboard bearing of the main shaft is also mounted on a single block of the same construction. The cost of these foundations was less than the estimate made for the same in first-class brick or stone mason-work.
It has also been used for lining a reservoir of ninety-six thousand gallons capacity, which was blasted into a ledge.
Another great advantage realized in re-enforcing béton with iron, is that the iron overcomes its tendency to check in hardening, within useful limits, however large the surface may be, if the distribution of the iron through the work is made with ordinary good judgment. This is demonstrated in the instance of entire freedom from shrinkage checks in the single section of béton-flooring laid in the drawing-room of the house. Its dimensions are eighteen by thirty-six feet, three and a half inches thick, and after a period of eight years, during six of which it has, in winter, been more or less subjected to unequal strains from the expansion and contraction, caused by changing temperatures, while employed as a transmitting medium of heat for warming the room, there is no trace of a check throughout its whole extent.
The method of heating the house is shown in Fig. 2, where the section exhibits the arrangement of hot-room and heating-flues in the walls and floors.
In the center of the cellar is a heating-chamber, measuring eleven by sixteen feet, and eight feet in height. Within this chamber is placed an ordinary cast-iron heater, of a capacity for burning about three hundred and fifty pounds of coal per day. Openings were made, about twelve inches apart, all around the top of the surrounding walls of the chamber, leading outwardly to the spaces between the first floors and the cellar-ceilings, and also up through the flues within the interior walls, which communicate with the spaces between the second story floors and ceilings beneath them. Vertical iron pipes, of suitable size, are located so as to connect the open spaces between the cellar-ceilings and first floor with a large, closed trunk, or passageway, which extends nearly all around the inside of the main wall foundation, under the cellar-floor, and finally terminates in a large flue, which leads directly under and into the heating-chamber.
This comprises about the whole system of arrangements in the construction for warming the house with heat radiated from the floors and interior walls.
Its mode of operation simply consists in the body of warmed air passing from the heating-chamber upward, through the walls and under the floors, and in its passage giving up its surplus heat to the surfaces of these flues. As the air becomes reduced in temperature, it naturally descends through the pipe and trunk passage-ways provided for its return to the heating-chamber, where it is again recharged with heat. It will readily be seen that, by this method, a continuous circulation will be maintained with the same quantity of air; and furthermore, that the velocity of the current will vary with the difference of temperature of the air when leaving the heating-chamber and when re-entering the heating-chamber.
By this system there are about fifty-five thousand cubic feet of the interior of the house heated by radiation, through about thirty-five
Fig. 2.—Section of House, showing the Method of Heating: A, heating-chamber; B, openings in heating-chamber communicating with flues in the interior walls leading to spaces under the floors. The arrows show the direction taken by the current of air.
hundred square feet of floor and wall surfaces, and the capacity of the heating-chamber is fourteen hundred cubic feet, so there is one cubic foot of heated air to forty cubic feet in the house.
The temperature of the air in the heating-chamber averages, in very cold weather, 170°, and after delivering its surplus heat to the floors and interior walls, its temperature registers 58° in the flue where it re-enters the heating-chamber for reheating, showing that 112° of heat had been given up and utilized for warming purposes. With ordinary care in managing the furnace, a temperature of 68° can be uniformly maintained on the first floor, and from 60° to 62° on the second floor, with a consumption of about three hundred and twenty-five pounds of anthracite coal per day in the furnace.
The temperature produced by this system of heating is free from the objectionable variations so common with other modes of heating. The walls and floor form such large heating surfaces that the temperature is uniform in all portions of the rooms, while the air is not vitiated by escaping gases or heated dust, as is universally the case where furnaces or steam-pipes are used for heating.
It is not asserted that its economic results per pound of coal are greater than those claimed for the steam or hot-water systems, but, if the latter were required to make as liberal provision for the renewal of fresh air in the interest of an equally good ventilation, the percentage of useful results per pound of coal from steam or hot water would average no higher than by this method.
The rain-water falling upon the roof passes through two six-inch iron pipes, which are set in the walls, extend across the cellar, and connect with a béton tank in the rear tower, holding five thousand gallons, whose water-level is thirty inches below the level of the roof. This inverted siphon also forms a distributing system to the various points of consumption in the house, through short branch pipes connected with these mains.
There are also two other tanks made of béton, and holding three thousand gallons, situated under the main tank: one of these sustains a head of over twenty feet of water, and has never given any indications of leakage.
In regard to the important factor of cost involved in this system of béton construction, its average for beams, floors, and roofs, including the supporting platforms for laying them down, was a fraction over sixty cents per square foot. This cost also includes the re-enforcing iron beams and rods. The cost of the heavy wall-work, not including cornices, was about twenty-four cents per cubic foot, which includes the cost of plank molds, required for building up the walls. The advantages that contributed most to these economical results were cheap material and cheap labor.
The bulk of the material required for the work abounds in inexhaustible quantities, and is always obtainable at moderate cost. The skill needed consists in a simple knowledge of the right proportions of material, and of its proper manipulation, which can be obtained in a half-day's practice. The most inexperienced laborers can do all the work of the most elaborate béton construction, excepting only the surface-finishing, and this, with all the other work, can be superintended by one competent, experienced builder.
Along with the foregoing data, it may be well to include an account of some experiments that were made to test the heat-enduring qualities of béton. A number of large test-bricks were made of the same proportion of materials used in the construction of the walls, and, in subjecting them to heat of different intensities to see how much they could withstand before breaking up, there was no perceptible difference observed in the tendency to fracture, whether the bricks were exposed to a gradual or rapid heating. Not one of them broke when subjected to a white heat. Several were heated to a bright-red heat, and then plunged into a bath of cold water. They withstood this test without showing a decidedly damaging fracture, and one of the bricks was exposed to an alternate heating and cooling three times before breaking up.
These results were a surprise, and they suggest the advantage of using such a material for the walls of buildings, as a sure defense against uncontrollable conflagrations. The facts that appear to be established by the line of experiments are:
1. That a system of iron beams re-enforced with béton can be made to sustain weights many times greater than the iron beams alone can bear without re-enforcing.
The intrinsic worth of béton construction appears most valuable in furnishing the elements of fire-proof construction, and thus inaugurating a reform in the prevailing system of building based on the principle that safety can be more economically realized through reformation than by exclusive dependence on insurance indemnities for losses by fire. The amount of capital destroyed by fire appears almost fabulous, and has been estimated by insurance authorities to be over one hundred million dollars annually in this country. This enormous estimate takes no cognizance of the losses due to the disturbance of business relations and labor by such enforced interruptions of industry, but the sum of the losses accounted for seems to be enough to awaken an interest in the discovery of some effective remedy for reducing them.
Yet, if the remedy is only to be found in building more thoroughly, its adoption may remain doubtful so long as the hazardous method of building, and the rates for insuring hazardous property, occupy their present relations to each other. Such radical departures from conservative ideas of building as are herein described must necessarily find a slow recognition.
However destructive to the material wealth of the country may be the vast losses of property by fire, they sink completely out of view when compared with the terrible sacrifices of human life that are constantly resulting from unsafe building construction. Against these fearful consequences, humanity can reasonably protest, and claim, for the sake of human welfare, that such structures as hotels, theatres, public schools, and all other places of public resort, shall be made invulnerable to fire.
The writer has heretofore declined to make any public statement concerning the experiments herein described, for the reason that he considered that they ought to undergo thoroughly satisfactory tests of severe weather exposures, and varying temperatures, through a period of time long enough to determine their true and relative value.
In conclusion, it is to be hoped that these experiments may shed enough additional light on the fire-proof building question to make the way easy for reducing re-enforced béton construction to a system, that will deserve public confidence, and ultimately find general adoption.