788 REPORT— 1903. 



each; anifurt/ier explains tlie use of the pendulum ftpparatus, as used by Messrs. 

 Willans and Robinson, with the reasons that led to its adoption. 



Tests by shock or tests for brittleness have from the earliest times been used 

 by engineers, especially on the Continent; and Swedenborsr, in 1734, gives some 

 interesting particulars of the riile-of-thumb tests carried out by purchasers of iron 

 in those days ; and in nearly every case an impact or brittleness test was used, 

 which though only empirical was no doubt all that was required, and gave a deal 

 of useful information. 



At the present time, probably owing to the state of perfection to which the 

 testing machine has been brouglit, there is too much inclination to neglect other 

 properties of material which the usual tests do not detect, but which are quite as 

 important as the usual standard physical tests ; and that this is so is sbown by 

 M. Fremont's paper, published by the Society d'Encouragement pour I'lndustrie 

 Rationale, September 1901, in which he throws an extraordinary light on the 

 brittleness question, and gives several instances where sei'ious fractures in struc- 

 tural and otlier steel were not accounted for by any of the ordinary testing metliods, 

 but were readily explained when tested for brittleness by an impact machine. 



At Messrs. Willans & Ilobinson's, Rugby, it has long been felt that a method 

 of testing such as used by M. Fremont and otliers was wanted to detect the reason 

 for certain fractures which were inexplicable by the ordinary methods used ; and 

 an experimental pendulum impact machine was made, and the tests carried out 

 with it gave promise of such good results that a standard machine was made and 

 is in use at the present time, giving results which gain iu importance witli every 

 series of experiments carried out on it. 



Many types of impfict machines are used, but the pendulum form of apparatus 

 seems to give most satisfactory results : it can be calibrated to give direct readings 

 for energy absorbed, and lends itself to very quick working even by an inexperienced 

 operator. 



The idea of the arrangement is as follows : — A weight is suspended pendulum- 

 wise on a stitf rod, which swings from a centre designed to be as frictionless as 

 possible. This weigiit or tup is then moved out of the vertical and allowed to fall 

 on to the free end of a test-piece gripped by the other end in a vice, tiie specimen 

 being nutchcd to locate the break, the height of fall being always made sufficient 

 to cause fracture with n/ie hloiv. A suitable measuring arrangement is used to 

 record the enei'gy remaining in the weight after fracture of test-piece has occurred ; 

 and this subtracted from the calculated energy in the tup before fracture gives the 

 energy required to break the specimen. 



Measurements are taken of the test-piece, and results are transferred to equi- 

 valent energy absorbed on specimen one inch square. 



The paper is accompanied by drawings of the apparatus and tables of results 

 obtained. 



3. Permanent Sel in Cast Iron due to Small Stresses, and its Bearing on 

 the Design of Piston Rings and Springs. By C. H. Wingfield, 



Some of Messrs. "Willans & Robinson's steam packing rings, each consisting of 

 an outer cast-iron ' ring ' of uniform section, and of the same diameter as the 

 bored cylinder hi which it worked, and an inner cast-iron 'spring' ring of section 

 varying as described in vol. II. of Unwin's ' Machine Design,' under the heading 

 ' Theory of a Oast-iron Spring Ring,' were tested as to equality of their outward 

 pressures per inch of circumference. AVhereas by published formulae they should 

 have agreed in this respect, it was found that they actually differed considerabl}-. 

 Some springs made specially for the purpose, and having different amounts of 

 ' follow,' ' but alike in other respects, were then experimented with by being 

 forced into circular ring-gauges, and being removed and measured after periods 



' By ' follow ' is meant the difference of diameters of the spring when free and 

 when forced into its working position. This difference enabled the combined ring 

 and spring to follow up and compensate for wear. 



