114 SUMMARY OF CURRENT RESEARCHES RELATING TO 



Metallography, etc. 



The Quenching of Steel.* — Seniozay gives a lengthy account of his 

 investigations into the effect upon the hardness of steel and the position of 

 the critical range, caused by varying the conditions of quenching, such as 

 temperature, duration of heating, rapidity of cooling, and dimensions of 

 the mass quenched. Hardness was measured by the Brinell method. 

 The critical temperatures were determined by taking cooling and heating 

 curves, the junction of the thermocouple being inserted in a central 

 hole. In the first series of experiments the samples (40 mm. by 10 mm. 

 by 10 mm.) were heated in an electric furnace to the required tempera- 

 ture, and quenched. In the second series the steel was heated to a 

 temperature above the critical range, cooled slowly to the required tem- 

 perature, and quenched. Curves showing the relation between quenching 

 temperature, and hardness are given. In the third series the length of 

 time during which the steel had to be maintained at a given temperature 

 to produce the same degree of hardness on quenching was determined. 

 In the experiments upon the effect of variation in the size of the speci- 

 men, heating was carried out in a lead bath of large capacity. The 

 steel was quenched in air, in oil, or in water. The author gives his 

 conclusions at considerable length. A notable omission from the data 

 given is the analysis of the steels employed in this research. 



Corrosion Grooves in Boiler Plate.j — C. Fremont and F. Osmond 

 point out that the existence of local corrosions in boilers can only be 

 accounted for by irregularities either in original structure or resulting 

 from conditions of construction or service. Corrosions have been 

 classified according to form, (1) in spots (" pustulaire ") ; (2) in grooves. 

 The former are probably due to original non-homogeneity of the metal, 

 such as inclusions of slag, sulphides, etc. It might be supposed that the 

 mechanical stresses to which a boiler is subjected when in service cause 

 local strains which lead to irregular oxidation. The authors show how 

 maximum stresses are localised at certain points and lines owing to the 

 method of construction of a boiler. It is along these lines that corrosion 

 grooves are found. The authors advance arguments which tend to show 

 that strain effects in the metal do not account for the grooves. A much 

 more probable explanation is that the plate first becomes covered with a 

 layer of oxide ; this oxide is not deformable, i.e. is brittle, and the 

 slight elastic bending of the plate causes the oxide to crack along the 

 line of maximum stress. A clean surface of steel is thus exposed, and 

 oxidation goes on more rapidly. The small corrosion groove thus set up 

 is a line of weakness : stresses are still more localised along it, the 

 re-formed oxide is cracked, again exposing a line of bright metal for 

 oxidation. Corrosion grooves are much more dangerous in plates of 

 poor quality than in good material. The authors give an account of the 

 investigations, carried out on four old locomotive boilers, which led to 

 the above conclusions, and illustrate their paper with a number of photo- 

 micrographs. 



* Eev. Metallurgie, ii. (1905) pp. 737-74 (46 figs.). 

 t Tom. cit., pp. 775-88 (25 figs.). 



