Device for Illustrating the Phenomenon of Buckling.

A. SOMERFELD.

(Zeitschrift des Vereines Deutscher Ingenieure, Berlin, 1905, vol. lxix., pp. 1,320-3.)

A straight strip of steel, 27 inches (70 centimetres) long, 1 inch (4 centimetres) wide, and inch (millimetre) thick, is fixed in a vice at one end, the other being pierced by a hole to receive a screw bolt which can be loaded at both ends so as to subject the steel strip to a total central load of 0.55 lb., 1.1 lb., 1.65 lb., and 2.2 lbs. (1, 1, 2, and 1 kilogram). When the unloaded strip is mounted free end upward and set in vibration it swings at the rate of 263 vibrations per minute, but on reversing the vice and allowing the strip to hang down, the number of vibrations increases to 288, the moment of gravity assisting the elasticity of the metal, whereas it opposed the latter in the former case. With the lowest weight of the series the vibrations are 112 and 159 respectively; with the second, 68 and 131; with the third, 37 and 117; and with the heaviest, 0 and 109, the limit of buckling having been exceeded in this case. With the strip held horizontally the vibrations under the same conditions were calculated as 277, 139, 104, 87, and 76 respectively from the length, weight, and moment of inertia of the strip.

The actual limit at which instability commences is under a load of between 1.92 and 1.96 lb. (870 and 890 grammes), whereas, in calculating according to Euler's formula, even when allowance is made for a deduction of 30 per cent. of the actual weight of the strip (which naturally forms part of the total load), the limit corresponds to 2.02 lbs. (920 grammes), so that this formula gives results in excess of the truth.

The Paper also contains a series of mathematical explanations of the various influential factors in the calculation.

C. S.

Determination of the Blast-Furnace Slag contained in Cements. Prof. M. GARY and J. VON WROCHEM.

(Mittheilungen aus dem Königlichen Materialprüfungsamt, Gross-Lichterfelde-West, 1905, vol. xxiii., pp. 1-21.)

Cements are sometimes mixed with blast furnace slag, either before or after calcining, and as in the latter case the admixture cannot be called an advantage, the problem frequently arises as to whether a cement has been adulterated, and what is the amount of adulteration. Fresenius relied to some extent for the detection of slag on the sulphides it contains, which are practically absent in pure cements, but as the percentages of sulphides vary considerably in different slags this method could give no quantitative results. It also

appears that most blast-furnace slags have a specific gravity of less than 3.0, while cements are heavier than 2.9. But here again variations are very considerable, and separation of the two constituents is impossible until all the dust is removed. The Authors have combined these methods, and, judging by the results of their determinations as contained in the tables in their paper, they have at last been able to make reliable quantitative determinations. This, of course, is only possible if the analyses of the sample and of the pure cement and the pure slag are known. The latter are obtained by freeing the sample from all dust, which would hinder separation by the method of flotations in a dense fluid. The coarser residue is sorted by sifting into various grades, and each lot is then split up by means of a fluid, which has to be nicely adjusted, for density, into slag and cement. This splitting up is facilitated if the sample to be analysed is first of all subjected to a calcining heat in the presence of oxygen, because this reduces the density of the slag. The several lots are then analysed and the results compared with the analyses of the original sample, which enables the relative quantities to be calculated.

C. E. S.

Remarkable Tests, indicating " Flow" of Concrete under Pressure. I. H. WOOLSON.

(Engineering News, New York, 2 November, 1905, p. 459.)

The tests were made in the Columbia University Testing Laboratory. A series of short columns, 4 inches in diameter and 12 inches long, were constructed by filling steel tubes of that size with a finely crushed stone concrete and allowing it to set for 17 days, when it appeared very hard. The metal of the tubes varied from inch to inch in thickness. The columns in the thicker tubes carried a load of 150,000 lbs. without injury except a shortening of less than inch. Two columns in the thinner tubes began to show a marked deformation at about 120,000 lbs., and this gradually increased as the load increased to a maximum of 150,000 lbs., when they were compressed 3 inches and 3 inches respectively, their diameters having increased to about 5 inches. When the tubes were sawn apart and removed, the concrete was found to have taken the exact shape of the distorted tube, and was as solid and perfect as possible showing that the concrete had actually flowed under pressure like any plastic material.

A. W. B.

Effect of Moisture on Reinforced Concrete.

(Revue Technique, Paris, vol. xxvi., pp. 453-7.)

MAYNARD.

Little is known as to the variation in volume of mortars when exposed to moist air or immersed in water. When a cement is gauged the mass at first contracts in volume owing to a re-arrangement of its particles, but subsequently it begins to expand, owing, probably, to the conversion of lime into calcium hydrate. This latter action is all the more marked when the mortar is immersed, as soon as set, in water, where, at the end of 7 days, as much as 6.77 per cent. of water has been known to combine with the mortar, and the absolute increase in volume has amounted to practically 10 per cent. This increase chiefly went to fill the voids in the mass, and the apparent or measurable increase was only about per cent.

A series of curves is given showing the variation in volume of a Portland cement mortar when placed in (1) dry air, (2) damp air, (3) a limited quantity of fresh water, (4) fresh water frequently changed, (5) a limited quantity of sea-water and (6) sea-water frequently changed. In the first case the volume gradually decreased with age; in the second, third and fifth cases, it increased more or less rapidly; while in the fourth and sixth cases an increase was at first apparent but soon attained a maximum, after which a steady decrease took place, owing to the dissolving out of calcium hydrate by the continually renewed water. In the case of structures in reinforced concrete the variation in volume of the iron bars is dependent on stress and temperature alone, and is quite uninfluenced by the presence of moisture. Thus, as the climatic conditions vary, the stresses in the reinforced structure will also vary. If the thickness of the concrete envelope be considerable, the variation in stresses is great; while, if not, fissures are apt to form in the envelope through which the moisture can gain access to the iron which it rapidly corrodes. An instance of this is cited in the case of a tank which was exposed to the action of sea-water, and was purposely demolished after being in use for 3 years. Many of the thin bars were found to have been completely eaten away, a lustrous deposit of magnetic oxide of iron (Fe3O4) alone being left. It would seem that reinforced-concrete structures should be protected with some covering impervious to moisture wherever possible, and the desirability of placing them in contact with water is questioned.

I. C. B.

Tests of Reinforced-Concrete Beams.

(Railway and Engineering Review, Chicago, 26 August, 1905, pp. 624-7.)

About two years ago the Boston Transit Commission began a series of experiments on concrete beams reinforced with steel rods. These beams were 6 feet long, 8 inches deep, and 6 inches wide, and various methods of reinforcement were used. From each lot of

concrete two prisms were made, which were sent to Watertown Arsenal to be tested. The beams and prisms were made of the following proportions:-380 lbs. of Vulcanite cement, 4 cubic feet of coarse sand and 12 cubic feet of broken stone. In mixing the materials and filling the moulds no attempt was made to secure better work than that obtaining in ordinary practice. The beams were left in the moulds about two days and then buried in damp sand until a few days before testing. All loads were applied at the centre. Measurement of deflections and sets were taken in nearly all cases. The deflections of the beams reinforced with corrugated or twisted bars and failing in tension were very large before the ultimate load was reached, in some cases being as much as 4 inches. In the beams reinforced with plain bars the margin between the load at the elastic limit of the steel and at ultimate failure was very much less than in the beams having deformed bars (twisted or corrugated). This is to be attributed partly to the difference between the elastic limits of the steels. The mode of failure of the beams is next described. The article is accompanied by diagrams showing the variation of modulus of elasticity of concrete prisms in terms of the load, strain curves of the steel used in the reinforcement, load deflection curves of the reinforced beams, and the results of tests of Portland cement briquettes made with various percentages of water, and by sixteen reproduced photos of the beams showing the mode of failure and six tables giving the results of the various tests. J. M. M.

Experiments on Concrete at Antwerp. ZANEN.

(Annales des Travaux Publics de Belgique, Brussels, 1905, Part 3, pp. 399-420.)

During the construction of 2,000 metres (2,187 yards) of quay wall on the Scheldt to the south of Antwerp, the Author carried out a series of tests on the resistance to bending of blocks of concrete, and also made observations as to the materials entering into the composition of concrete, the results of which he considered might usefully be made public.

The concrete foundations of the quay-wall were constructed with a toe projecting 2.5 metres (8.2 feet), and the calculated weight of the masonry wall gave a mean pressure on the foot of 2.06 kilograms per square centimetre (29.4 lbs. per square inch).

The trial concrete was composed of three volumes of dry broken well-burnt bricks, three volumes of gravel, and four volumes of mortar, composed of 400 kilograms (880 lbs.) of Portland cement per cubic metre (35.3 cubic feet) of Scheldt sand.

Other blocks were also experimented on, the composition being

The Author describes the tests carried out on briquettes of Portland cement and sand, and tabulates the results, which he summarizes as follows:

(1) The resistance to extension obtained with Scheldt sand was relatively high; but the resistance, at the moment of rupture, was very variable even for briquettes having the same age and composition.

(2) The resistance increases with age.

(3) The mean tensile strength of briquettes made with 600 kilograms of cement was about 60 per cent. higher than with briquettes made with 400 kilograms of cement.

Blocks of concrete 1 metre (3.28 feet) long, and of a section corresponding to the section of the wall, were tested under various conditions, the results being tabulated and described.

The composition of the concrete for this wall was finally settled to be as mixture B, the cost of a cubic metre being 73.61 francs (47 shillings) per cubic yard.

The Paper is illustrated by six figures in the text, and has numerous tables giving particulars of the tests.

H. I. J.

Determination of the Porosity of Mortar. MAYNARD.

(Revue Technique, Paris, vol. xxvi., pp. 244-5.)

The determination of the voids in mortars by ordinary methods presents difficulties by reason principally of the alteration which takes place when the substance is either dried or soaked in water. A new method of making this determination forms the subject of the article under consideration. In this method the specific gravity, as ascertained in the ordinary way, and the chemical composition alone are required. From the former the weight of unit volume of the mortar can be ascertained, and from the latter can be obtained the weight of unit volume of the substance exclusive of voids, since, the exact specific gravity of each of the constituents being known, and also the proportion which this constituent represents, the volume of each of the component parts in a given weight of the mortar can be at once determined, and consequently the weight in unit volume of solid matter. In making the chemical analysis the substance is dried at a temperature of 100° C. It was at first thought that all uncombined water would be expelled at a temperature of 50° C., but it was found that the percentage of water expelled between 50° and 100° was very variable, depending on the age, the chemical composition, and the length of time during which the sample had been immersed. This loss is high in all undecomposed mortars, or in such as are of considerable age; it is slightly higher in mortars which have been "blowing," or have been burnt at a low temperature, and is very much higher in such as are decomposed without increase