87S 



NA TV RE 



[December 30, 1922 



The Corrosion of Ferrous Metals. 



IN 1916 a committee was formed by the Institu- 

 tion of Civil Engineers under the chairmanship 

 of the late Sir William Matthews, with sixteen members 

 of the Institution to investigate the "Deterioration 

 of Structures exposed to Sea Action." The project 

 was, in the first instance, submitted to the Depart- 

 ment of Scientific and Industrial Research, which gave 

 it everv encouragement and promised the committee 

 substantial financial assistance which has already 

 amounted to several thousand pounds. 



An important part of the committee's investigations 

 is that connected with the corrosion of iron and steel 

 structures exposed to sea action. In an exception- 

 ally well-illustrated paper, read before the Institution 

 of Civil Engineers on April 4, 1922, Sir Robert Hadfield 



• • ■ 



* . ~ - 













1 





*'/> 



'.. — Mild steel (with 0-7 per cent, manganese). 

 Longitudinal section X 100 (untreated). 



gives a detailed account of the progress of the work 

 down to that date. 



The committee decided to expose fourteen types of 

 ferrous material to sea action in various parts of the 

 world and to determine by quantitative measurement 

 their relative powers of resistance towards corrosion. 

 The metals comprised " Armco " iron, Swedish char- 

 coal and wrought irons, lour types of carbon steel, 

 cupriferous, nickel and stainless steels, and two 

 samples of cast iron, cold and hot blast respectively. 

 With the exception of the cast irons, the various 

 metals were prepared in the form of rolled plates 

 measuring 24 inches in length, 3 inches in breadth 

 and 0-5 inch in thickness. The cast irons were of 

 like dimensions, and were prepared by casting in the 

 ordinary way. No further heat treatment was ac- 

 corded the metals, for the committee considered that 

 the tests would be of a more practical character if 

 carried out with the metals in a condition resembling 

 as closely as possible that obtaining in constructive 



NO. 2774, VOL. I IO] 



work. In general the specimens were allowed to 

 retain their outer skin of oxide, normally present on 

 the rolled or cast metals; in two ca 

 additional specimens were prepared from which the 

 skin was removed bv grinding, in order to obtain 

 information as to the effect of oxide layers upon the 

 corrodibility of the metal. 



Specimens of all the metals were subjected to various 

 mi . hanical tests, such as the Izod and Fremont shock 

 tests, and the Brinell hardness test. Tensile tests 

 were carried out on bars cut in the longitudinal direc- 

 tion. Save in the case of the cast irons the bars were 

 marked at regular intervals along their lengths, 

 and, after pulling, their elongations from point to 

 point were carefully determined. This was done in 



*v^ 



Fig. 2.— Mild steel (with 0.7 per cent, manganese). 

 Transverse section X 600 (untreated). 



order to ascertain the effect of strain upon the corrodi- 

 bility of the metal, the intention being to cut small 

 test pieces from different parts of the strained bars 

 and subject them to laboratory corrosion. 



A duplicate set of tensile test bars, machined ready 

 for testing, was prepared for immersion in that condi- 

 tion in the sea at Plymouth. After a prolonged 

 exposure they will be removed and examined with 

 the view of determining whether or not the mechanical 

 qualities of the material are impaired. Very little 

 work has been carried out on this aspect of the subject 

 and the results obtained should prove of particular 

 interest and value. 



In addition to the foregoing, one bar of each material, 

 excepting the cast-iron specimens, was suitably heat- 

 treated in order to obtain test-data representing the 

 physical properties of the materials under optimum 

 conditions. The results obtained are detailed in the 

 Appendix to Sir Robert Hadfield's paper and illustrate 

 in a striking manner the enormous superiority in every 



