CHAPTER XX.

THE DESIGN OF ENGINEERING STRUCTURES

IN CONCRETE STEEL.

THE design and erection of large structures in reinforced concrete, or, as it is perhaps more appropriately termed, concrete steel, is attracting the attention of nearly every civil engineer of the present day, both European and American. The enormous possibilities of the art are only as yet merely touched upon, and very little understood. Concrete steel must not be confused with concrete used together with steel joists and sections for large fireproof buildings. In this chapter we are going to consider how steel, intimately blended," so to speak, with the concrete in a structure, is going to relieve that concrete of the greater part of its tensile stress, with resulting economy and safety in the structure. The use of rolled-steel sections used in Steel Construction (Chapter XXXIV.) is practically entirely eliminated by rods and bars of a comparatively small sectional area. These are always bedded in the concrete, and are invisible to the eye. The uses of concrete steel would have numerous applications, and sometimes odd ones too. Bridges, large buildings, reservoirs, tanks, water-towers, canal locks, sewers, culverts, are all common applications of the art, to say nothing of barges, telegraph poles, gates and boundaryfences, etc., which have been tried in actual practice, and given satisfactory and economical results.

The first thing to do is to investigate the essential qualities of the two ingredients, viz. concrete and steel. Concrete itself was considered in Chapter XIII., but as it is of special importance in this chapter we shall say something more about it.

The Aggregate.-When ballast is used it should be well washed and quite free from loam. It should contain a fair percentage of sand. Broken lime. stone makes a very good aggregate. Its disadvantage is that it cannot be considered as fireproof, the lime being subject to calcination at high temperatures. Sandstone is not to be recommended. Those crystalline rocks called Diorite are excellent (Chapter XVIII.), as also is crusted granite. Broken brick can be well recommended. It has great fire-resisting qualities, and will not splinter at high temperatures. It must, however, be very well washed before use, and used wet. Blast-furnace slag will be found satisfactory if it is selected with great care. It must, however, be porous, and not too glassy, and absolutely free from sulphur. The presence of sulphur in any aggregate must be entirely eliminated, and is only done so by repeated washings and weathering.

Sand. The sand used in this particular work requires very careful selection. In the first place, it must be a happy medium between dust and coarse gravelly

sand. It must be clean. Regarding the fireproof qualities of aggregate, furnace slag appears to have given unsatisfactory results. Broken brick and granite gave good results, coke breeze and furnace clinker gave the best.

When concrete is used for walls, -in. aggregate is ample. It is the sharp sand which is the best in practice, because it is usually of a siliceous nature. Test mixtures of the concrete should always be made in the office. Proceed

as follows:

Procure a piece of iron tube and make a mixture similar to that which we are proposing to use. Very carefully note down its composition, then mix it and ram it into the piece of tube, which was previously sealed at one end by means of a screwed cap. Carefully measure its height in the tube when fully rammed. Then clean out the tube again and make a fresh mixture, varying the proportions of sand and broken stone, but still having the same total weight of the two combined. If less or more, it will be a guide to the future adopted proportions. Again note the height of the sample when rammed. Repeat this process a few times. The lowest height proves the best mixture, but they must be very accurately proportioned if they are to be of any practical value. Now, for a mixture of concrete the following proportions, according to the class of work, have been found to answer well in practice :

The aggregate will, however, shrink during mixing. For example, say we have 4 cub. ft. of Portland cement, 9 of sand, and 20-4 of aggregate. The resulting mass, which might be supposed would yield 34 cub. ft. of concrete, will only yield in practice about 21.4 cub. ft. in mass.

The use of salt water in concrete works has the tendency of retarding the setting power. All water used should be carefully selected; also it must be clean, free from acid and strong alkalies. It must not have much lime present, and be free from organic impurities. The average quantity required would be from 21-24 gallons per cub. yd. of dry mixture.

Concrete should not be tipped from a height, as might be considered an advantage, nor should it be done in frosty weather. If, however, this should be quite unavoidable, as in the case of cold climates, 1% of salt may be added with advantage, or a warm mixture used. The average composition of a cub. yd. of concrete would be as follows, according to the proportions used to make 1 cub. yd. of concrete :

Regarding forms for concrete work, any cheap wood will do. It should not be seasoned. They must be well coated with soft soap or crude oil before use, to prevent them sticking. In case, however, the particular work has to be afterwards plastered, this will not do; in such a case they must only be wetted. Cement mortar may have the proportions of 1-2, 1-3, or 1-4, according to circumstances. The addition of to 1% of lime putty is a distinct advantage, especially for plastering work, as it will tend to make the mortar flow easily, and cause it to be impervious and very adhesive.

All steel used in concrete steelwork must be surrounded by at least 1-2 in. of concrete, according to the size of the structure. Concrete which is made from cinders, as we before stated, gave good fire-resisting qualities. It has, however, the objection when used in conjunction with steel that the free sulphur which it is liable to contain would be likely to corrode the steel. The whole success of concrete steelwork, apart from the scientific design, depends on proper mixing of the concrete. This may be hand done, but when it is, it wants very careful doing by willing workmen. The machine mixer is a much more satisfactory appliance, and is now used on all contracts of any magnitude. They may be hand or power driven, of the rotary type. Taylor's or Ransome's are perhaps the best known as giving satisfactory results. Gravel mixers are also in vogue. The principle is, that by means of baffleplates the mixture thrown in at the top descends by a devious path from one side of the machine to the other, water in suitable quantities being added meanwhile. A gain of about 11% on machine mixing can be relied upon. must always be borne in mind, however, that gravel concrete is only 88% as strong as that made from crushed granite.

It

Regarding the strength of concrete in tension, it may be taken anywhere between 200 and 300 lbs. per sq. in., principally according to age. For instance, it has been found by experiment that a 1-2-4 mixture will give the following average results :

:

28

360 tensile strength, lbs.

Designing Beams.-The modulus of rupture of a concrete beam may be taken as

=

=

where W total load; 7, b, and d length, breadth, and depth in inches. This will give inch-lbs. The modulus of elasticity of concrete E, in tension = 2,000,000 lbs., and in compression 2,700,000 lbs. per sq. in. This is an important factor, because the greater the alteration of the concrete due to the load, the more load will consequently come upon the steel.

Again, we have to consider the coefficient of expansion of concrete. It may be taken as 0000055.

We must now consider the steel used in concrete steel work. In Chapter XIII. we also made some remarks thereon. The ultimate strength of the steel will vary according to the amount of carbon present. A few examples will serve to illustrate our assumption. They are based upon Baushinger's experiments.

In all cases the shear may be taken as 77% of these values, the usual practical values taken being

Now on steel we have to consider the effects of high temperatures. The tensile strength of steel will gradually diminish between the limits of - 4° and 150° Fahr. An increase will then take place up to 450°. It then, however, will rapidly fall. Alternation of stress (vide Strength of Materials) will have the usual effects, of course, of causing the steel to become brittle and useless. For comparison with concrete, the coefficient of expansion of steel may be set down as '0000066.

Now, from what we have previously said, we can infer that in practice concrete is about ten times stronger in compression than in tension. But its adhesion to steel may be taken as anything up to 600 lbs. per sq. in. This, however, is too high for practical consideration. The average practical limit may be set at 250 lbs. per sq. in. Now, of course, we have to make this bond between the concrete and the steel of greater tenacity than that of the bar, to effect which a certain proportion of the bar will have to be embedded in the concrete apart from practical considerations. To find such a length in inches we have a formula based on our previous reasoning. If Fs tensile strength of the steel, d = diameter of the rod in inches, ƒ = 250.

=

As adhesion strength,

We know that the object of the reinforcement is to reduce the tensile stresses, but it will also be capable of taking some of the shear too; and it must be borne in mind that the greater value of E of the materials, the less will be the deformation under a given load. Then assuming thorough bond between the concrete and steel, the permanent set of the whole beam must be equal everywhere. But, however, owing to the permanent deformation being greater in the concrete than in the steel, we will have induced stresses therein, and hence a compressive stress in the concrete. Now the ratio of "E" in concrete and "E" in steel may be set down as 10. Then, considering a load applied to a reinforced concrete beam, assume the bond to be perfect. But we know that steel will elongate more than the concrete; the result then is that the load gradually comes off the concrete and goes on to the steel. In fact, we may safely conclude that the behaviour of concrete steel under load is