However, the whole subject is very uncertain, the increase or decrease of strength by re-meltings depending evidently on the particular mixture of iron used. A test in each case would be the only reliable criterion. It must be remembered, too, that re-meltings mean much additional expense.

Bower- Barf f

Process.

Iron Cement.

Re-melting

Cast-iron.

If cast-iron be kept melted in the furnace for a long time - from one to three hours - it can be greatly increased in tensile strength, some experiments showing nearly doubled strength.

Cold-blast iron seems to be stronger than hot-blast (in tension), varying with the temperature of the test piece, the cold-blast iron being nearly twenty-five per cent stronger than iron made with a blast at 250° F.

Mr. Thomas Box has collected all the data obtainable as to the effect of thickness on cast-iron, and from them is obtained the formula:

Effect of Thick ness.

(100)

Where b = the least thickness of a casting, in inches.

Where t = the ultimate tensional stress, per square inch, of a similar casting, whose least thickness is one inch.

Where t1 = the ultimate tensional stress, per square inch, of a casting, whose least thickness is b inches, and

(101)

Where b = the least thickness of a casting, in inches. Where k = the ultimate modulus of rupture, per square inch, of a casting whose least thickness is one inch.

Where k1 = the ultimate modulus of rupture, per square inch, of a casting whose least thickness is b inches.

Accordingly if the transverse or tensional strength of a casting one inch thick be =1,0 we should have :

Proportionate Strength of Castings.

 For 1/2 inch thick = 1,350 For 1 inch thick = 1,000 For 1 1/4 inch thick = 0,907 For 1 1/2 inch thick = 0,838 For 1 3/4 inch thick = 0,783 For 2 inch thick = 0,739 For 2 1/2 inch thick = 0,671 For 3 inch thick = 0,620
 For 4 inch thick = 0,547 For 5 inch thick = 0,497 For 6 inch thick = 0,459 For 7 inch thick = 0,429 For 8 inch thick = 0,405 For 9 inch thick = 0,385 For 10 inch thick = 0,367

According to Hodgkinson the tensile strengths of cast-iron bars 1", 2" and 3" thick would be as 1,0 is to 0,8 and to 0,7 7 or slightly more than the above.

With wrought-iron the strength, according to Clay, increases with each rolling to the sixth; and then again decreases; or if the strength (tensional) of the muck-bar be =1,0 then that of a piece piled and rolled three or four times = 1,36 ; and if six times=1,41; decreasing again to == 1,0 at the twelfth re-working. This, however, as already stated is shown to be very doubtful with American irons, and largely dependent on the nature of the muck-bar. Kirkaldy has shown that welding does not make as strong a joint as the original iron. He made some eighteen experiments on welded bars of wrought-iron from 1 1/4' to 3/4" diameter, the result being that none were equal to the original strength. The maximum strength attained was 97,4 per cent; the minimum 56,2 per cent and the average was 80,62 per cent of the original strength. Welded joints should, therefore, be taken at only about four-fifths of the strength of iron in calculating for tension or transverse strains. It is usual, however, to call for sufficient extra metal in all welded joints, so that the pieces must invariably break elsewhere than at the joint, when tested.

In wrought-iron the skin seems to add some strength to the metal and for this reason, partly also on account of the more intricate interlacing of fibres in bar-iron, which is worked on all sides, bar-iron is generally stronger than plate-iron; plate-iron being about 84 per cent of bar-iron. The plate-iron itself is always stronger with the grain than across it, resembling wood in this quality. Experiments have shown the tensile strength to be nearly 10 per cent greater along the grain, than across it. Shearing would show even greater difference, being some 25 per cent easier with the grain than across it.

Wrought-iron should always be heated before hammering or work-ing.

Effect of Re-rollingWrought-iron, Bar Iron Strongest.

Its tensile stress is greatly reduced by cold hammering; experiments (according to Box) showing the reduction to be or an average loss of 27 per cent by cold hammering.

 a loss of 32 per cent if 2" thick a loss of 36 per cent if 1 1/2" thick a loss of 13 per cent if 1" thick

By annealing the iron the strength can be partially restored but not entirely, the loss in the above experiment still being after the annealing Effect of Cold Rolling.

 some 14 per cent in the 2" thick iron some 4 per cent in the 1 1/2" thick iron some 8,6 per cent in the 1" thick iron

or an average loss of about 9 per cent.

Unwin, however, claims the exact opposite. According to tests quoted by him, wrought-iron and mild steel increase in tensional strength and have a greater elastic limit, if worked cold, but suffer a large loss in ductility. The working however must be uniform on all parts and not local otherwise the material is weakened by uneven resistance. Annealing, according to Unwin, removes the effects of cold rolling or cold working and should be used where this is only local or on certain parts, as in the case of punching of rivet holes, where the part around the hole is hardened by cold hammering and the rest left unhardened. He cites the process of wire-drawing, which is similar to cold rolling and certainly adds greatly to the tensional strength of the material.

He finds one very curious fact, however, and that is that while neither cold nor hot working hurts wrought-iron or mild steel, working them at an intermediate or "blue-heat," that is at about 4 70° to 600° F., is positively harmful. In many tests it was found that samples of either, about 3/8 inches thick, could be bent back and forth on an average some twenty times, without breaking, if they were either cold or red hot. But if heated to an intermediate blue-heat they all broke readily after being bent back and forth only two or three times. Annealing ordinary wrought-iron means a loss of from 5 per cent to 10 per cent in tensile strength. The strength is not affected, however, by low temperatures nor ordinary variations in temperature, nor very much at higher temperatures. At high temperatures up to from 600° to 900° F. the tensile strength of wrought-iron increases somewhat. Beyond this it decreases rapidly. Pure wrought-iron will increase up to a higher temperature than impure iron. Neither cast-iron nor steel are affected much by low temperature nor the ordinary variations in same.