same way, if brick, stone and terra-cotta cannot be made to protect, efficiently, the metal framework of the buildings which we are trying to make fireproof, something else, that will accomplish this object, will assuredly be employed, no matter how much our traditions may have to be upset for the purpose. For example, there is no practical reason why a steel skeleton should not be clothed first with cement mortar, on iron lath, and then cased entirely over with copper, or brass, or aluminum, riveted at the joints. With a little grouting between the steel-work and the casing to give a substantial backing to the sheet metal, a building of this sort would be almost indestructible. Fire and water would expend their force upon it in vain, centuries of exposure to the weather would only increase the beauty of its patina, and, if well anchored to the ground, and cross-braced with moderate skill, no earthquake could bring it down. It would be a queer-looking affair, no doubt, and the critics would have plenty of sport at its expense, but we may remember, in our efforts to accomplish, by the light of our own intelligence, the results that are required of us, that it was the ugliest of the ducklings that grew up to be a swan, and that in the solution of the most important, most difficult, and yet most imperious problem that confronts us certainly lies, to a great extent, the future of the art of architecture in America. Mr. Jenney's paper, the reading of which was omitted last evening, is inserted as a contribution to the discussion on high buildings. THE WIND PRESSURE IN TALL BUILDINGS OF SKELETON CONSTRUCTION. [The examples under consideration are from the FORT DEARBORN Building, now under construction southwest corner Clark and Monroe Streets, Chicago, JENNEY & MUNDIE — Architects.] In considering the effects of wind pressure on a steel skeleton building, we assume that the mechanical effect is similar to that on a beam fixed at one end and uniformly loaded from the outer and towards the point of support. In the case of a building the support is from the first floor through the basement to the footings. The lowest point to which we consider the wind pressure to extend is the average height of the surrounding buildings, which are so situated as to break the wind. We can properly consider the building as a truss fixed at one end; the different stories corresponding to the panels. The actual effect of the wind on the building must be calculated in detail in order to know just what amount of wind pressure the building itself will safely sustain. The balance of the wind pressure must be provided for by means of special wind bracing. In the calculations, we will consider the bending at each floor with the assumption that every thing below this floor is rigid. In general we can assume the total bending effect on the connections between columns and beams at any floor as the product of the total shearing force, multiplied by the proper lever arm the total shearing force being the total wind force on the exposed surface that reaches from the top of the building down to a point one half way from that floor to the floor next below, the lever arm being in general one half the length of the column below the floor under consideration. F a с Owing to the rigid construction of the floors and the close fit and riveting of all the connections, the tile arches and tie rods, we can assume that this wind force at any one floor will be transmitted to all of the columns supporting that floor. From the manner of construction, the columns and the beams between them can be considered as fixed at both ends-for illustration-In the case of the portal bracing of a bridge where the ends of the posts ab and cd are fixed, the bending moments at each of the joints a and c are each equal to one-half the force F into one-half the arm ab. Fone-half of the wind pressure on the side ab of the bridge. b a d b (Fig. 1.) d The exaggerated distortion would be as shown in Fig. 2. We find from experience in general that it is reasonable to assume that the interior construction of a building, such as the Fort Dearborn, will resist of the total wind pressure or 10 lbs. per square foot of exposed surface. The maximum wind pressure being assumed at forty pounds per square foot. The following example will show on what data the assumption is made. Take a building 75x90 ft. in area, 150 ft. high from grade and the average height of the surrounding buildings 50 ft. and height from floor to floor 11 ft. We will assume that the interior connection takes of the wind load. Following is the column spacing: (Fig. 2.) Let the connections considered be in a floor 100 ft. from the top of the building and at the height of the surrounding buildings. The total wind force at that floor will be 100x75x40 lbs.=300,000 lbs. 4 of this =75,000 lbs. As the columns are equally spaced and there are 20 interior columns the force on one will be 75000 and multiplying by 1⁄2 the length of the column from that floor to the one below or 66 inches, the bending moment on each column will be 247,500 inch lbs. Take column 25. The wind coming from the direction shown, it seems evident that the effective connections are those connecting column 25, with the beams 2425 and 25-26. Sketch of these connections and connections distorted. The connections of beam A to the column will take the bending moment and the connections of beam B the other half. Take the angle b connecting B to column. The stress on the four rivets e. e. etc., will be the bending moment on that side of the column 247500 divided by the lever arm "M” = 9 inches=13,650 lbs. This gives for one rivet 3,410 lbs. which a 3/4 inch rivet in shear will take care of. The force on the 4 rivets f, etc., will be the bending moment 247500 divided by the arm n-12 inches=10,000 lbs.=2,500 lbs. for each rivet which a 34 rivet in tension can take care of. This shows that our assumption that the interior connections will take care of the wind force is not arbitrary. The bending moments here given at the bottom of the surface exposed to wind will be the maximum for the interior columns. Above this floor the bending moments will decrease with the decreased area exposed to the wind. Below it they will be the same as at the floor first considered which is at the bottom of the surface exposed to the wind. We have shown that the interior construction will take care of 4 of the total wind pressure; this leaves 4 or 30 lbs. per sq. foot of exposed surface to be resisted by some special construction. There have been several ways of doing this. The main feature of all these has been the construction of vertical trusses in the interior of the buildings. This was done by connection of two or more tiers of columns from basement to attic by means of cross tie rods running from the top of one column to the bottom of the next, and vice versa, and setting these up by turn-buckles. The objection to this has been that the tie rods were very apt to be in the way, either coming where it was desirable to have a door, or else coming where they would be unsightly. This objection has been partly overcome by using a portal bracing between the columns, but even this is often objectionable. Lately in the New York Life Building, finished last May, and in the Fort Dearborn Building (now under construction) we have adopted a different method, the idea for which was, we think, first suggested in general terms by the late Wm. H. Sherzer, an engineer of the Carnegie Steel Co. and Keystone Bridge Co., which we worked out in detail. This method is to make the beam connections at the outside columns such that they will resist the wind pressure not taken up by the interior connections. At the Fort Dearborn Building, now in course of erection, this was accomplished by using channels between the outside columns to carry the curtain walls, and connecting them to the columns by means of gusset plates. (See illustration.) These gusset plates are used to make the lever arm of the resisting bending moment of the connection as large as possible, and reducing thereby the direct force that is taken by the rivets. The bending moments of course being the product of the direct force into the lever arm. The action of the wind force produces a bending moment at the end of the channel. As the greatest bending moment from the floor and curtain wall loads is generally at the middle of these beams or channels, the section which will resist this bending moment from the floor loads in the middle is generally sufficiently large to resist the bending moment at the end from the wind force, and in general no increase in weight of metal is required. At the upper floors the decrease in the bending moments will render the gusset plates unnecessary. The column below is increased to take the vertical components of the wind force. Referring to the example already considered, Fig. 3. Note that this system of wind bracing is entirely concealed in the masonry. The connections between the exterior columns and channels from 1-7 and from 36-42 will take up the total bending moment of the wind pressure equal to 100X75X30 lbs. into a lever arm of 66 inches. The exterior columns are so designed that the dimensions parallel to the direction of the wind pressure which they resist shall be as large as practicable, thereby increasing their resisting moment. The foregoing method of resisting the wind pressure is based in part on the assumption of the complete rigidity of the floors. When the floors are constructed as in the New York Life Building, Chicago, in which the beams are coped to fit closely to the girders, cast iron separators between double girders, the whole secured by hot rivets and tie rods, and the floor formed by hollow tile arches, set close in both directions in cement mortar, and concreted above to top of the beams, we believe this assumption to be justifiable. |