242 



NATURE 



[August i8, 1923 



Hardness Tests. 



EVERY one has a general idea of what is meant by 

 hardness — that the diamond is harder than steel, 

 and steel harder than copper. The workman judges 

 of hardness as the resistance of a material to the action 

 of his cutting-tools or files. But there is as yet no 

 rational definition of hardness. A property connected 

 with hardness is resistance to abrasion or wear. As 

 Sir Robert Hadficld has said, rails are demanded 

 which will not wear out quickly and tyres which will 

 not need renewing every few months. It was entirely 

 for these reasons that modern qualities of steel were 

 produced. To some extent hardness is opposed to 

 ductility or toughness. Very hard materials are 

 generally brittle. The engineer requires a material in 

 which hardness is obtained without too great a sacrifice 

 of toughness. 



The earliest scale of hardness is that proposed by 

 Moh. He selected ten minerals arranged in order such 

 that each would scratch the one next below it in order 

 and be scratched by the one above it in order. On 

 this scale talc has a hardness i and diamond a hardness 

 10 ; iron has a hardness of 4-5. But the scale is qualita- 

 tive only and arbitrary. Prof. Turner has used a 

 balanced lever turning on a knife-edge. The free end 

 carries a diamond. The surface to be tested is 

 polished. The hardness is taken to be the weight in 

 grams on the diamond necessary to produce a definite 

 scratch. The method is useful, but there are practical 

 difficulties in applying it. Recently Mr. Hankins, at 

 the National Physical Laboratory, has modified this 

 test. He uses a diamond shaped so as to produce an 

 indentation furrow rather than a scratch. 



The diamond is loaded with weights and drawn over 

 the surface to be tested. The widths of the scratches 

 with different weights is measured, and it is found that 

 the square of the widths plotted against the weights 

 fall on a straight line passing nearly through the 

 origin. Hence Mr. Hankins takes as the hardness 

 number the quantity 



k = 



,,,2 - n' 



where P is the load on the diamond, w the width of 

 scratch, and p and q small constants not depending on 

 the material tested. 



Various investigators have used an indentation 

 method for determining hardness. Such a test is very 

 suitable for ductile metals, but how far it is applicable 

 to brittle materials is uncertain, though this is not of 

 practical importance. The indenting tool has been a 

 knife-edge, ball, cone, or pyramid. 



In 1895 ^^d 1900 Lieutenant-Colonel Martel com- 

 municated two very interesting papers to the Paris 

 Congress on Testing Materials. He used chiefly a 

 falling monkey with various forms of indenting points 

 and various heights of fall. He concluded that (i) 

 for a given material the work of indentation is pro- 

 portional to the volume of the indentation and in- 

 dependent (within limits) of the form of indenting tool ; 

 (2) that the pressure causing indentation is at each 

 instant proportional to the area of the indentation 

 normal to the pressure. If V is the volume of the 



NO, 2807, VOL. 112] 



indentation, P the weight of the monkc}, .... i > 



height of fall, then Mattel's hardness number 1 



in kilogram-millimetre units. 



About 1900 Brinell introduced the indentation test, 

 which has been most widely used. A ver>' hard steel 

 ball 10 mm. in diameter indents the material by a 

 gradually applied load of 3000 kilograms, which rests 

 on the ball for some seconds until the indentation is 

 complete. The radius of the indentation is measured 

 by a microscope. If P is the load, a is the radius of 

 the indentation, and r the radius of the ball, then 

 Brinell's hardness number is 



H = 



2jrr(f - \/r* - a^) 



The quantity in the denominator is the spherical sur- 

 face of the indentation ; and the units are kilograms 

 and millimetres. In practice it is necessary to use a 

 smaller load for soft materials and sometimes to use 

 a smaller ball. Then the hardness number obtained is 

 not the same unless the load Pj and the ball radius r^ 

 satisfy the condition 



This is Meyer's law confirmed by Mr. Batson, of the 

 National Physical Laboratory. If the law is complied 

 with the indentations are geometrically similar. 



Prof. Ludwik uses a right-angled cone instead of 

 a ball, so that the radius and depth of the indentation 

 are equal and the indentations for different loads are 

 similar. He also takes the hardness number to be 

 the load divided by the conical area of the indenta- 

 tion. 



Prof. Foppl placed two cylinders of the material to 

 be tested at right angles and pressed them together in 

 a testing machine. The pressure per unit of flattened 

 surface is taken as the hardness number. Prof. 

 Henderson, of Greenwich, has introduced a similar test, 

 the material being in the form of square prisms. 



For ordinary materials of construction, Brinell's test 

 has proved most useful. It rather fails for very hard 

 materials from the smallness of the indentation and 

 the distortion of the ball, and efforts have been made 

 to find another test or to revive the scratch test for 

 such cases. 



A new instrument which appears to be very sensitive 

 has been introduced recently by Messrs. E. G. Herbert, 

 Ltd., of Manchester (see Nature, April 28, p. 583). 

 This consists of an arched pendulum weighing 2 or 4 

 kilograms. At its centre is a ball i mm. diameter of 

 ruby or steel. By adjusting screws the centre of gravity 

 of the instrument can be made to coincide with the 

 centre of the ball. A weight over the ball can be 

 adjusted to lower the centre of gravity of the instrument 

 to o'l mm. below the centre of the ball when the time 

 of swing on a very hard surface is 10 sec. A 

 level tube over the ball is graduated from zero at one 

 end to 100 at the other. Two scales of hardness are 



