ON THE INFLUENCE OF SILICON ON THE PROPERTIES OF STEEL. 275 



due rather to the combined presence of the two hardeners, silicon and 

 carbon. It may be here mentioned that the ductility noticed cannot be 

 attributed to manganese, which was only present in small quantities, 

 about '20 to '30 per cent. 



The material employed in these experiments was made by meltin"- in 

 crucibles good wrought-iron scrap low in sulphur and phosphorus, to 

 which was added in varying and increasing quantities a rich ferro-silicon 

 containing 20 per cent, of silicon. The ingots, 2|" square, were reduced 

 by forging in the ordinary method to 1|" square billets, then rolled down 

 to bars 1^" diameter. 



The effect of sihcon upon iron is as follows, (1), in its forged con- 

 dition ; (2), in its cast condition : — 



1. Forged Condition. — The forge reports that the material A ("24 per 

 cent. Si) did not forge well, cracking somewhat whilst being hammered, 

 but all the other samples B to H ('79 per cent, to 5'63 per cent. Si) 

 when forged at a fair yellow heat required no special care ; thus clearly 

 showing that silicon even up to as high as about 6 per cent, does not 

 destroy the malleability of the metal iron. Upon, however, exceeding 

 this percentage the material is red-short, crumbles at a low heat, and 

 notwithstanding the low percentage of carbon present (-25 per cent.), 

 becomes really a species of cast-iron. It should also be here mentioned 

 that if the carbon had been higher the point at which malleability ceases 

 would have been with a much lower percentage of silicon. 



Nor is such red-shortness removed by the addition of manganese. 



It may also be here mentioned that no return of strength takes place 

 by a further addition of silicon, as is so specially characteristic of man- 

 ganese steel. Any further addition merely increases its resemblance to 

 silicious cast-iron. Nor do gradually increasing percentages, as is the 

 case with manganese steel, destroy the magnetic properties of the alloy ; 

 a 7 per cent, material seems quite as susceptible as ordinary iron 

 drillings. 



As regards the results of the mechanical tests on this steel in its 

 forged state, these are so fully detailed in Table II. (Tensile and Bend- 

 ing Tests), Table III. (Compression Tests), and Diagram 3, that it is 

 only necessary to add a few remarks here. 



Apparently silicon up to I^ or If per cent, added to iron, although 

 increasing the limit of elasticity and raising the tensile strength, does 

 not impair ductility, but after this the further increase of tensile strength 

 noticed is only obtained with a serious loss of ductility. Apparently 

 there is no sharp line of demarcation. After passing 1| per cent, further 

 slight increases cause great changes in the characteristics of tlie material. 

 In this respect, therefore, its action rather resembles that of carbon in 

 contradistinction to the action of manganese, of which much larger 

 amounts are required to effect similar changes. 



The fractures from the tensile test-bars up to ' D ' specimen (218 per 

 cent. Si) are silky, after this completely and coarsely crystalline. As in 

 the specimens in the cast state, neither annealing nor water-quenching 

 seem to have any effect on the structure. 



The annealed flat bending pieces, V' wide by \" thick, gave good 

 results, specimens A, B, C, and D (-24 "per cent, to 218 per cent. Si) 

 bending double cold without fracture, more like dead soft steel, and 

 after being bent double the pieces were flattened close together cold 

 without showing signs of fracture. Specimen E (267 per cent. Si) 



T 2 



