3

metal between the rivet holes and the edge of the sheets, gave way by tearing out, leaving the rivets and intermediate metal uninjured, and this consists with the preceding estimate which makes this element of the joint the weakest of the three. Inasmuch, however, as they approach so nearly an equality of strength, and each may occasionally give way before the others, it may be as well to take the average of their resistances for an expression of the strength of the joint, 1.125 × 0.925 + 0.936 and this should be 1-015, or say one square inch of metal for every 12 inches (6 inches of strip and 6 in. of space) in the length of the boiler, th of a square inch, per running inch of the same; and this is the measure of the strength of that part of it over the legs. Now the elements of the calculation to determine the strain that a cylindrical vessel will bear, from the outward pressure of an elastic fluid, consists of the diameter of the cylinder, the thickness of the material composing it, and the modulus of the strength of that material. Thus, if D be the diameter of the cylinder, t the thickness of the iron (both in inches,) P the average force in pounds that will tear asunder a square inch of boiler, iron, and x the steam pressure per square inch on the boiler sufficient to burst it, we have the 2 Pt

equation D x = 2 P t, where x =

of an inch. Consequently

=

D

no

Now D=132 inches - P = 55,000 lbs. t = 208 lbs., which would be the pressure of the steam per square inch required to burst the boiler, if it were a continuous hollow cylinder without seams or joints to reduce the quantity and resistance of the metal composing it. But this is not the case in any boiler. The joinings of the plates, if there were holes in the boiler, would necessarily reduce the strength to about two-thirds of the entire strength of the sheets, were it not for the support they yield each other at the laps, where they are doubled upon each other. In the Merora's boiler, it is seen above, that for every 12 inches in the length of the cylinder over each leg, there is but one square inch of metal resisting rupture, instead of three square inches, which there would be if there were no rivet holes, or spaces, between the strips of iron. So the strength of the boiler over the legs is reduced to the one-third of it full strength in other places, where there are so seams, or perforations; consequently, we must divide the value of x, as above obtained, by 3, to get the real pressure, per square inch, that the Then boiler was capable of sustaining. 208-33

69-44 pounds per square inch,

the utmost strain that this boiler could have borne, without giving way at this, the weakess place. The pressure here spoken of, is of course the effective pressure, or the excess of that of the steam over that of the atmosphere.

Let us now proceed to estimate the pressure which could have been produced upon the boiler of the Medora by loading the safety valve to the utmost with the weights which were attached to it, and intended so to be used, when occasion should require the maximum pressure, considered by the engine builder to be safe. The diameter of the valve was 15 inches at the bottom, with a mitre of 12 inches, making its top diameter 17 inches. The levers of the valve were of the second order, and two in number. The primary lever, operating immediately on the valve, had a total length of 35 inches, and from the fulcrum to the centre of the valve disk, was 12 inches, making a ratio of The secondary lever, operating on the end of the primary one, had a total length of 67 inches, and from the rod connecting the two, to the fulcrum of the former, the distance

was 10 inches. There were two weights of cast iron, on the secondary lever, the largest, nearest to the end of it, weighing (by estimate) 200 pounds, and the smaller 56 pounds. When these two weights, which slid as usual on the arm of the lever, which they were perforated to receive, were in contact, and pushed out to the end of the lever, the distance of their centre of gravity from the fulcrum of the long lever, would be 57,6 inches; so that the ratio of this lever would 1

be and the ratio compounded of those 5.76

this will increase the whole pressure there to 4820,77 lbs., or to 23 lbs. per square inch of the valve and boiler.

This appears to be the highest pressure which could be brought upon the boiler, by the system of weights and levers, belonging properly to the valve.

The strength of the boiler, in its weakest places, being then estimated, as above, at 69.44 lbs. per square inch, amounted to just three times the extreme pressure which the engine builder appears to have intended it should be called on to bear, and in proportioning the strength of his work to the duty it was to perform, he would seem to have been sufficiently prudent. How, then, did the boiler explode, when guarded by a safety valve of such ample dimensions, designed to give way at one third of the bursting strain?

That the free action of the valve must have been interfered with, would be a natural conclusion, and is corroborated by the testimony of one of the assistant engine

The nu

But the boiler may have been in fact weaker than it has been estimated, from a deficiency of strength in the iron of which it was made, and which has been assumed at 55,000 lbs. per square inch. merous experiments that have been made upon the strength of this metal (of which those recorded in the Journal of the Franklin Institute, vols. xix. and xx., 2nd series, are the most extensive and satisfactory that I have seen, and also the most applicable to the present case, as they were made upon boiler iron,) show that its cohesive power as often exceeds as it falls short of that measure, which may be taken as a fair average. If the iron of the Medora's boiler was, in fact, of bad quality, it may, however, have possessed a tenacity far within that just given; but it scarcely could have descended as low as the one-third of 55,000 lbs., or to 18,000 or 19,000 lbs. per square inch, which it must have done to have yielded to the steam pressure of 23 lbs. per square inch, supposing the preceding estimates of strength and pressure to be correct. My examination of the iron, as it appeared on the torn edges of the sheets, did not impress me with an unfavourable opinion of its quality, although it showed the distinctly laminous structure which most sheet iron exhibits. Its strength in the ruptured parts could not have been diminished by over heating, for those parts were far under the lowest level to which it is in the least degree probable that the water could have fallen, even had it declined below a safe and proper height. Moreover, the strips-the giving way of which caused

the bursting of the boiler-were so situated that the fire could not have been at all t contact with them had the boiler been dry, as they occupied the spaces over the legs, and, consequently, must have been always immersed either in water or steam. There is, however, no evidence that the water was deficient in quantity at the moment of explosion; on the contrary, many witnesses declared that the gauge cocks showed a full supply, and, considering the vast size of the boiler, the shortness of the time between the lighting of the fire and the occurrence of the explosion, it seems not likely that the evaporation could have sunk the water to a dangerously low level, if, as is to be supposed, it was properly filled at first. I could, indeed, discover no trace of injury to any part of the boiler by burning of the metal, and my examination of the uppermost tubes, and the top of the back smoke chamber, was very careful, and if these parts of the boiler were overheated, they, nevertheless, stood firm, and left the rupture to take place elsewhere. No incrustations likely to impede the transmission of the caloric to the water, and thus render the iron liable to burn, anywhere appeared, and were not to be expected in a perfectly new boiler. The degree of heat imparted to the parts of the boiler covered with water, would not, at all events, have exceeded the temperature due to the effective pressure of about 4 atmospheres, or 69,44 lbs. per square inch, which has been above estimated as sufficient to burst the vessel. This temperature (see Journal of the Franklin Institute, vol. xvii., page 291, 2nd series,) is not more than 300 degrees of Fahrenheit, and the experiments recorded in the same journal (vol. xx., pages 24 to 31, 2nd series, and curve traced in plate X.,) show that the tenacity of iron is increase by heat, until a temperature of at least 400° Fahrenheit is surpassed, when it begins to diminish. The boiler, then, could not, in my opinion, have burst from overheating the metal in the parts where the rupture actually took place to wit, in the strips over the legs, uniting the segments of the cylinder between them. In estimating the strength of these strips, I have supposed, from inspection and measurement of those that remain, that they contained one-half of the original quantity of metal in the sheets, which was further reduced to one-third of that quantity by the rivet holes. In this estimate there is room for mistake, and, possibly, I may have in this assigned more strength to the boiler than it in fact possessed. In both the quality of the iron and the amount of metal in the strips, then, there may be room for considerable reductions upon the preceding calculations of the ability of the boiler to withstand the pres

66

sure of the steam. Again, although the metal did not, on account of its greater strength than that of the strips, give way where, if at all, it must have overheated-viz. in the top of the smoke box, or in the upper flues-there is a possibility that these parts may have been laid dry by the falling of the water, and have become so hot as to generate suddenly a larger body of vapour (or, who knows, an explosive gas,) than the safety valve could vent with sufficient rapidity to save a rupture, even if that had been loaded to less than the strength of the boiler. This supposition is, however, in the face of the evidence, though, unfortunately, too little credit is due to such testimony in cases of this kind.

It is stated, by several witnesses, that the safety-valve did not blow at all while the steam was getting up, and that from the time of making the fire, up to the moment of the explosion, not more than two or three short puffs proceeded from it. These would, indeed, suffice to show that it could not have stuck fast in its seat from the action of some adhesive force-an instance of which kind is on record. But it also proves that if, after all, the valve was not overloaded, at least the engineman and his assistants, &c., were strangely indifferent to the unusual absence of that audible evidence of its free action, which always attends the starting of a steamboat.

Although doubt must continue to rest upon the true and special cause of the explosion of the boiler of the Medora, and I am not prepared, in relation to that cause, to offer more than the preceding facts and inferences, for the judgment of others, yet some general conclusions may, I submit, be satisfactorily derived from the circumstances of this catastrophe. First-the boiler was too large in its diameter for the strength of metal employed, looking to the risk of bad material and workmanship. The thickness of the sheets was a quarter of an inch-a thickness, by the way, almost universally employed in the construction of steamboilers of all diameters, as if there were some magic in this particular dimension, which made it most pliantly applicable to all cases, however varying. Referring to 2 Pt

the formula x=" we find that iron one

D

quarter of an inch thick would require a pressure of 573 lbs. per square inch to rupture it in a cylinder of three feet diameter, (a diameter of usual occurrence in locomotive and other high-pressure boilers,) if the modulus of its strength be 55,000 lbs. per square inch, and a deduction be made for the seams, of twenty-five per cent. While

13

the Medora's boiler, of eleven feet diameter, and the same thickness and strength of sheet iron, would have borne but 156 lbs. per square inch, with a proportional reduction of one-quarter for the joints, (saying nothing of the still greater subtraction of strength due to the apertures over the legs,) can there be any propriety in using the same thickness of plate for each of these widely differing diameters? Yet it is done under the influence of the apparently prescriptive right of the quarter inch iron to be employed in all cases whatever. It is true that the small diameter boilers are high-pressure, and subjected to the greater strain, but their excess of strength is, at the same time, vastly greater than that of the large low-pressure boilers. The usual high-pressure strain is, perhaps, about 100 lbs., and the low-pressure strain, 25 lbs. per square inch. The excess of strength in the three feet highpressure boiler is, then, 473 lbs. per square inch, and in the eleven feet low-pressure boiler of the Medora it would be but 1264 lbs. per square inch. It may be said, indeed, that the ultimate strength of the former is but 5.73 times its usual strain, while that of the latter is 6.25 times its ordinary stress; also, that at the high temperatures accompanying high pressures, a given increase in the temperature causes a more rapid rise in the pressure than at the low temperatures of the lower pressures, so that there is occasion for more excess of strength in the former than in the latter cases, to guard against accidental augmentations of temperature. Also, that the consequences of explosion at high pressures are more disastrous than at those of a lower grade, and should, therefore, be more carefully guarded against. There may be some propriety in the two first of these suggestions, but not so much in the last, as some of the most fatal explosions have occurred in low-pressure boilers; and still I think that low-pressure boilers, whose diameters are generally from eight to nine feet, are usually too weak when made of quarter inch iron; and this opinion is held à fortiori in regard to the Medora, with her boiler of eleven feet across. Braces are indeed used in these large boilers, but with often only partial effect, and in the Medora, as has been already remarked, they could not be applied in the vertical, or radial direction, in which they would have done most good, on account of the positions of the tubes.

Second-a boiler of the colossal size of that of the Medora, presents a bold and striking aspect, and seems fitted for the generation of a vast supply of steam; and so, doubtless, was the boiler in question, the fire surface and steam room of which

was of unusually ample extent. But the same fire surface will be as effectual if distributed among two or three boilers, and with great increase of security against explosion. The one large boiler will, perhaps, cost less in the manufacture, and occupy less room in the boat, but it will be much more difficult to move, in placing and displacing it; and if accident happens to it, the supply of steam is wholly cut off; while with more than a single boiler, each of which can be insulated from the other, it may be kept up, and the engine worked at a lower speed, till the injured boiler is repaired.

Third, the design of a boiler resembling the Medora's, is deficient in strength at the junction of the belly with the legs which support it, and form the sides of the fire-place. The perforations of the cylinder must be large enough to permit the water and steam bubbles to pass freely up and down, in the necessary circulation of them through the boiler. Small holes over the legs would not allow this; the circulation would be checked, less steam would be generated, and the legs of the boiler become unduly heated for want of the constantly required supply of cold water, which, in a boiler with no obstruction to circulation, is constantly descending towards the fire, from the upper and cooler parts of the vessel - towards which last, the steam bubbles are simultaneously rising to the steam chamber; the two currents thus running contrary to each, without mutual interference, as their very different specific gravities maintain an easy separation between them. Thus the vertical has an advantage, in regard to circulation, over the horizontal tubular boiler; the tubes of the former interfering much less with the passage of the two counter currents, and its fire-place being at the bottom of the boiler, instead of at one end of it, horizontally, gravitation gives more assistance to circulation. The maker of the Medora's boiler was right, therefore, in giving wide passages for water into the legs, but, as to do this necessarily weakened his boiler so much, it shows a faultiness in its plan. It is not, indeed, easy to see why he departed from the usual mode of building boilers of this character by arching the fire-places, and so arranging his flues as to permit numerous ties in a radial direction across the boiler, connecting the outer shell with the firearches, as well as the flues. Here no cutting of holes would have been required, and any degree of strength given without interference with circulation. The ties introduced into the Medora's boiler, are not in the most effective position. A tie, or brace, in a cylindrical boiler, should never, if possible,

occupy any other position than that of a perpendicular to the surface supported by the tie or brace. Here they are diagonal to the ends and roof, and as chords, less than diameters, across the cylinder.

Some other obvious remarks suggest themselves, in conclusion, and are generally applicable to steamboats when under trial, on such occasions as this ill-fated vessel was about to begin.

Most of the persons on board the Medora, at the time of her explosion, were, as before stated, workmen who had been engaged in her construction. It was not safe to leave the boat as it was left, in the hands of these men, most of them, probably, reckless of danger by character, and fired with the false ambition congenial to the occasion. The safety valve and mercurial gauge should have been constantly under the eye of the engine builder, whose machine was being submitted to proof. The valve lever and its attached weights, and the means of moving them, should have been so constructed, and surrounded by guards, as to make it impossible that more than the extreme pressure designed to be put upon the valve, could be applied without doing violence to the defences of the apparatus. The boiler (as should all steamboat boilers) should have been provided with a small pumping steam engine, to keep up the water, as at such times, especially, the boat has often to wait a good while before starting, the water gets low, and the engine man and his assistants (viz., the crowd around him) are too excited, and anxious for a quick and favourable commencement of the trip, to go to work at the drudgery of pumping up by hand. Frequent trials of the state of the safety valve should be made, and it should not be left to blow of its own accord, at a safe pressure, but should be raised by force, and the surplus steam permitted to escape. This would prevent the valve from adhering to the seat, as it has been known to do, in consequence of the rusting of the iron, or the introduction of some glutinous or cohesive matter thereto. The propriety of these and similar precautions, need not be enlarged upon. It is remarkable, and yet a clear consequence of the laws of mechanical momentum, that, in all these explosions, the rents made in the vessel should be so much greater than necessary to vent the steam and water with the rapidity that one would suppose far more than sufficient to relieve the pressure upon all other parts of the boiler than the part first ruptured, so much as to save them from injury. While the confined fluid is quiescent, however powerful its effort to escape, it does no harm, except in the preparation it is making to

force an outlet, by gradually increasing its
strain on the metal, until an equilibrium is
attained between the two forces of impact
and resistance. But the instant that equili-
brium is passed, upon a single square inch
of the vessel, and the imprisoned fluid be-
gins to put itself in motion, it breaks its
way out, in a hundred directions instead of
one, and we see parts of the vessel, of ex-
tremely unequal strength, giving way at
the same moment. Thus in the Medora's
boiler, several pieces were blown out at the
back, which were unquestionably far stronger
than the strips of metal over the legs. The
sudden escape of the steam and water, also,
acting on the sides and bottom of the boat,
give to the boiler that projectile force which
throws it, en masse, into the air, and inflicts
other great injuries in its fall. In all con-
siderable explosions, indeed, the boiler has
been more or less displaced.

Thus is demonstrated the unsoundness of
the opinion that simple pressure, steadily
increasing, within a steam boiler, ought to
open the seams, as being the weakest part
of the vessel, and thus provide a safety valve,
for the escape of the vapour. To account
for the fact that boilers never do give way
in this manner, the hypothesis of explosive
gases has been paraded, with, as I humbly
conceive, very little foundation in fact, al-
though rare instances of that description
may have occurred under particular circum-
stances.

ABSTRACTS OF RECENT AMERICAN PUB-
LICATIONS.

OF

[Selected and abridged from the Franklin Journal
for October and November 1842.]
IMPROVEMENTS IN THE PROCESS
HARDENING STEEL.-Perry Davis.-The
method of hardening steel as practised by
the patentee is as follows:-"Instead of
plunging the steel to be hardened into cold
water, as is usual, it is plunged into a com-
position of borax, oil, and charcoal, to
harden it, which mode of hardening renders
the metal malleable with the same degree of
hardness that was obtained by the old me-
thod; so that the metal, after being har-
dened in this way, can be straightened or
bent without any danger of being broken."

Claim." What I claim as my invention,
and which I desire to secure by letters pa-
tent, is the mode of hardening steel so as to
render it flexible by means of the composi-
tion of oil, charcoal, and borax, as set
forth."

What influence the above described mix-
ture can exert upon the part to be hardened,
we are at a loss to perceive, and apprehend
that the method frequently practised of using

15

oil alone, would have the same effect with
the compound above indicated. We have
been in the habit of hardening steel, and
know of many devices which have been re-
sorted to for the purpose of communicating
toughness, but believe them all to be founded
in error. The less rapid the cooling, the
less will be the hardness, and, of course, the
greater the toughness of the article operated
upon.

IMPROVEMENT IN PROPELLING BOATS.
Meredith Mallary.-There is to be a com-
mon water-wheel, such as is used for driving
mills, on each side of the vessel, or boat,
immediately in the rear of the ordinary pad-
dle wheel, and the shafts of the two sets of
wheels are to be geared together; the water
which is thrown up by the common paddle-
wheels is to fall into the buckets of the
water-wheels, and is there to lend its aid in
the business of propelling.

Claim." Now what I claim as my in-
vention is the applying the water lifted or
thrown up by the paddle-wheels of steam-
boats, so as to produce an auxiliary pro-
pelling power, in the manner described, or
in any way analogous thereto.'

That these auxiliary wheels will exert some
influence we have no doubt, but we are
much mistaken if they do not prove to be-
long to the class of consumers only, and
that of the worst kind, as they will not pay
for what they devour.

IMPROVEMENTS IN THE CONDENSER AND
APPARATUS FOR SUPPLYING STEAM-BOIL-

ERS. Joseph Echols.-The apparatus, as
described, for supplying boilers, consists of
a receptacle, communicating with a cistern
and with the boiler, by means of pipes
governed by a four-way cock, which alter-
nately forms a connexion with the cistern,
from which to receive the water, and then
with the boiler to supply it with the water
thus received. The pipes forming these
connexions are narrow and high, and the
one leading to the boiler has an inclination
towards and opens into it above the water
line. Thus, when the communication be-
tween the boiler and receptacle is open, the
steam from the former passes into the latter,
and from this the water runs down the lower
inclined surface of the tube into the former.
The cock being then turned, cuts off this
communication and opens it with the re-
servoir, which allows the water to run into
the receptacle and condense the steam by
which the water was expelled in the previous
part of the operation. For low pressure
engines, it is said that the top of the reser-
voir must be closed, and a connexion formed
between it and the condenser.

The condenser is divided into two com-
partments, separated by a diaphragm pierced