Great Neapolitan Earthquake of 1857/Part I. Ch. XII

by Robert Mallet

CHAPTER XII.

EFFECTS OF SHOCK UPON BUILDINGS—THE 4TH MODIFYING CONDITION—RELATIONS OF FLOORS AND ROOFS.

---

Let us proceed now to the relations of floors and roofs to the shock, and the effects of both in modifying its action, immediate and final, upon the other parts of the buildings.

The common construction of provincial Neapolitan floors and roofs, has been already briefly described. Were the floors of an earthquake-shaken house perfectly homogeneous, formed, as if of a single parallel plate of cement or beton, self-supported and free from all timbering, the lines of fracture would be almost perfectly regular, and follow in accordance with those of the side and end walls; so that with a normal or subnormal wave, fractures would extend crossways, parallel to the walls, and generally uniting the fissures wherever these were situated in the latter; and with an abnormal or subabnormal wave, the fractures would be diagonal, at right angles to the line of transit of the wave, when viewed in plan upon the floor, and generally uniting the wall fissures situated in the adjacent walls, or diagonally opposite; while with a vertical or very steeply emergent wave, fractures would be found crossing each other in an horizontal plane at right angles, or nearly so, but having any degree of obliquity to the planes of the walls, dependent upon the direction of the line of wave transit, if very steeply emergent, and upon the form and relations, as to horizontal figure and height, of the building—or to the latter conditions only, if the transit were perfectly vertical.

The three cases are illustrated in Figs. 70, 71, and 72. In either of the two latter, large portions would become

detached, and would fall, leaving other portions, as in Fig. 73, still adherent and supported by the walls; while in the first case, if the width transverse to $a$—$b$ (Fig. 70) were sufficient, the segments detached by the parallel fractures, would break again transversely, by their own weight at or near the mid-length, and also close to the walls, and then fall.

So far, account has been taken only of direct seismic forces in the plane of the floors, but in the subnormal, subabnormal, and vertical waves, the inertia of the floor itself is brought into play transversely to its own plane, and all the displacements by fall just mentioned as due to gravity alone acting after fracture are produced upon an exaggerated scale by the introduction of this force conspiring with inertia at the moment of shock vertically downwards.

Now these are, in fact, precisely the forms of fracture and destruction, observable in these heavy floors of concrete and tiles, so far as they are left free, in a limited degree, by the constraint of the planking and joists beneath.

The joists and planking, are as one mass, and move together, and parallel to their respective lengths, under forces parallel to either, whether horizontal or slightly emergent. The union, however, is not sufficiently complete, and the jointing of the planking is too rough and open, to prevent "racking" by diagonal forces, as in Figs. 74, 75; and as the several quoins give out unequally, in the case of diagonal wave transit, the buildings are no longer truly rectangular in plan, and so in this the floors follow the walls. If the wave be normal or subnormal, and the joists lie parallel to its line of transit, the wall at the end first reached by the wave draws off from the ends of the joists embedded in it. The floor itself, more or less as a whole, follows the wall, and the other ends of the joists draw from the opposite wall, all during the first semi-vibration of the wave. During the second semi-vibration, the whole floor returns by inertia, in the opposite direction to its first movement, and following in the direction of transit of the wave, thrusts back again the previously drawn ends of the joists, into their sockets in the wall $b$ (Fig. 76), and the mass is stopped by coming into contact with the interior face of this wall, against which, it strikes with enormous violence, the edge of the planking and of the concrete striking the whole length of the wall, almost at the same instant and at the same level. This only happens when the direction of shock is pretty nearly in the line of the joists, as $a$, $b$ (Fig. 76). In such a case, the floor is almost

certain, if of any considerable magnitude, to bring down the end wall, upon which it strikes with the power of a "battering ram."

The sockets of the joists, (inserted, as, has been stated, they commonly are, into the bare masonry without tossal or bond timber,) become partially occupied by fragments fallen into them, and on the return into them of the drawn-out ends, these thump out holes, right through the wall, or shake the bond of the masonry effectually. The wall $b$, for the height of the story below, falls outwards, or in the direction of movement of the floor and wave, but for that portion above the shaken floor, it falls inwards, or upon the floor itself, and, at the moment after the whole support, of the floor beneath at that end has been withdrawn. The end $b$ of the floor therefore sinks and falls, and the other end of the joists, so far as they still remain in their sockets, act as levers, thus loaded, and prize the wall at the end $a$ asunder, so that it falls, likewise, very commonly outwardly, but, by possibility, in the direction towards the floor also. This chain of events, one of the most potent in the destruction of domestic buildings in Italy by earthquakes, is illustrated in Figs. 76, 77.

When there are two or three floors, over each other, that thus carry away the walls, they all (walls and floors) go towards $b$, the walls wholly falling outwardly. It will readily be conceived, what an inextricable mass of confused ruin, houses thus thrown down present, for the side walls that run parallel with the joists (or beams), fissured before transversely to the line of wave transit, are always more or less shaken down also, by the tremendous descent of the floors and walls, upon them falling.

If the line of transit of the wave, be in an orthogonal direction however, parallel to the planking, and therefore transverse to the direction of the joists, these having hold of the walls through the intervention of their sockets, the opening of the fissures in the walls on which the joists rest (due to their own inertia), is augmented by the inertia of the floor. The joists, or some of them, in advance of the fissures towards the end $b$ in the side walls, going forward with the flooring and side walls, drag the planking from the remainder of the joists, tearing out or bending partially the spikes, or breaking short the trenails; and the main weight of the flooring, thus freed from constraint of the side walls, runs forward as before, and though to a less extent, induces the same train of events, that have already been described.

And when the line of transit of the wave is thus transverse to the direction of the joists, the same set of phenomena result from both the first and the second semi-vibrations of the wave, and with effects proportionate to the velocities in each semiphase. The concrete and tiles of these floors adhere but very slightly to the planking on which they are laid, and the bond is so destroyed by the first slight movement, that the thick and heavy laminum often slides whole, or in large fragments, upon the planking, and batters the walls, already inclining outwards, independent of any constraint from the timbering.

When the wave is vertical, or of very steep emergence, the heavy tiled roofing, generally comes down upon the upper floor, almost at the instant that the inertia of each floor, acting at its centre of gravity in the opposite direction to the wave transit, tends to bring it down also. The impulse of the suddenly-imposed load of fallen roofing, conspiring with the effect of the shock, from which the resilience of the joists (if able to have done so at all), have not had time to recover, bring down the upper floor; the united load falls upon those beneath, and the whole are carried away in succession to the ground.

But, circumstances of building and of shock may be such, that the floor does not give way, but that, as an elastic plate or beam, supported and encastré at opposite ends (for the planking and concrete, we must bear in mind, have no insertion in the walls, and but slight connection with them), it merely bends downwards by its own inertia and that of whatever be upon it, under the upward stroke of the wave, as by a load suddenly applied over its whole surface. In this case, on the commencement of the second semi-vibration of the wave (or downward stroke), the bended timbers commence to straighten themselves again; their resilience, as a constant force acting through the versed sine of curvature, carries them upwards beyond the line of their passive position, at the same moment, that the second semi-vibration relieves the load upon them, by its inertia acting against gravity. The consequence is, the middle parts of a tolerably large floor, spring up with amazing velocity and power, beneath those who may be upon it, furniture is thrown upwards and towards the walls, tiles are projected upwards from the floor surface, masses of the concrete sometimes dislodged, and persons standing or moving on the floor are thrown upwards, and lose their balance.

Such is commonly the source of the strong impression of those who have experienced steeply emergent shocks, of an upward movement, unbalanced by any corresponding downward one, to which the title "sussultatoreo" is usually given, and which, when very violent, is called "sbalza" by the Mexicans. Of the latter, some remarkable examples will be recorded, as narrated to me.

Floors are not always, sources of increased injury however; they may occasionally act the part of props, to walls that if unsupported for their entire height, would have been prostrated, by normal or other waves of the first four classes.

This can only occur, when the line of transit is transverse to the direction of the joists, and when the end wall is in advance of the wave, $b,$ and, consequently, the edges of the floors also, are intersected by a wall whose plane, is parallel with the line of transit, as in Fig. 78; or when some other such fulcrum, resists the forward motion of the floors themselves, and enables them to hold together the side walls by the insertion of the joist's ends, and so to save the wall at $cc$ $dd,$ by cancellating the structure, in vertical and horizontal directions. The planking connecting the joists, it will be seen, here plays the part of a connecting stay to the side walls $cd$, $cd$, at the return

stroke or semi-vibration of the wave, tying them together by the inserted ends of the joists, whose distances they themselves fix.

When the floors have given way as described, the building is usually too far destroyed, to be of any use for seismometry; beyond this, that a clear comprehension of the mode of fall, will always enable the general direction of shock to be roughly inferred; but where the floors, being heavy and good, have not wholly fallen, nor the walls, but that these are fissured, and have given out unequally, and the floors also are fissured, but not completely displaced, very valuable indications may be obtained from them, as, for example, in Fig. 79, where the walls $fceg$, are fissured and given out. We obtain excellent measures of the extent, of this in the directions $fc$, $ge$, by measurements at the edges of the concrete or tiles, and inside the walls, controlling those of the fissures, which sometimes. (though rarely) are not measurable at all, and from the directions of the fissures of the floor we may obtain evidence, of another wave movement when occurring in the direction $c$—$d$, transverse to the principal one $a$—$b$. In fact, the observation of the floors, is only second in importance to that of the walls. The one illustration given, however, must suffice to indicate a large and very varied class of questions to which they may be made to give response.