90 



THE CIVIL ENGIKEER AND ARCHITECT'S JOURNAL. 



[March, 



Remarks on some of the leading Resiills in the foregoing Abstract. 



1st. The bars in Tables I., II., and III. were of the same sec- 

 tidiial area, leiif^tli, and weiglit nearly, but differed in tlie form of 

 their transverse section. They were jjlaced on supports at the 

 same distance (l.'Ji| feet) asunder, and struck horizontally by the 

 same ball, 60.3 lb. weight, suspended by a radius of 17 ft. 6 in. 

 From the results given it appears that the beam, 3 in. scpiare, and 

 the rectangular beams, G X I5 i". sections, struck on the broader 

 and narrower sides respectively, had all very nearly the same 

 strength to resist impact. The conclusions are drawn from a mean 

 between two experiments in each case. In Table XV. si.\ bars, 

 each 2 X I inch section, and 5 ft. long, were laid on supports 4.5 ft. 

 asunder, and all struck by the same ball 75j lb. weight, with arcs 

 of a radius 17 ft. C in. Three of them were struck on the broader 

 and three on the narrower sides, and their mean chords of impact 

 to produce fracture were 70 in. and 71 "07 in. respectively, or nearly 

 the same, agreeing with the results of the experiments upon the 

 former bars. 



2nd. In Table IV. the bars were of the same dimensions in sec- 

 tion as those in Table 1., or 3 in. .square, but the distance between 

 the supports was reduced one-half. The resulting breaking deflec- 

 tion, 1 '23 in., was somewhat greater than one-fourth of that in 

 Table I., or 4'875 in. and the vertical descent to produce fracture 

 was nearly one-half, but rather more, the depth fallen through in 

 the two cases being -639 in. and 1'238 in. Comparing, in like 

 manner, the half and whole bars in Tables V. and II., the depths 

 are '5521 in. and 1"2071 in. respectively. This result, coupled with 

 the former one, shows that the depth fallen through to break the 

 half bar is nearly half of that required to break the whole one. 

 ( 'om])aring the results in Tables VIII. and XII., and also Tables 

 X. and XIII., it appears also that a bar of half the length of another 

 resists with nearly half the energy, but somewhat more. 



3rd. The experiments in Tables I., II., III., IV^, and V, afford 

 illustrations of some of the conclusions in the large generalisation 

 of Dr. Young, deduced from neglecting the inertia of the beam. 

 CA'dt. Phi/., Lecture XIII. J "The resUience of a prismatic beam, 

 resisting a transverse impulse, follows a law very different from 

 that which determines its strength, for it is simply proportional to 

 the bulk or weight of the beam, whether it be shorter or longer, 

 narrower or wider, shallower or deeper, solid or hollow. Thus, a 

 beam 10 ft. long will support but half as great a pressure without 

 breaking as a beam of the same breadth and depth which is only 

 o ft. in length; but it wiU bear the impulse of a double weight 



striking against it with a given velocity, and will require that a 

 given body should fall from a double height in order to break 

 it." 



4th. The experiments in Table VI. were made to compare the 

 effects of striking a bar midway between the centre and one sup- 

 port with those of striking similar bars at the centre, as in Table 

 IV. The great impacts, so near to the support in these cases, 

 would necessarily cause it to yield slightly, and thus increase the 

 resisting powers of the bars to sustain impact. In experiments 

 made by the author several years ago, given in the Fifth Rejjort of 

 the British Association, page 112, on bars 1 in. square — some sub- 

 jected to impacts in the middle, and others at half the distance 

 between the middle and one support — the chord of impact neces- 

 sary to produce fracture was nearly equal in the two cases. The 

 ratio of the deflections, from equal impacts at the middle and at 

 one-fourth span, was nearly constant under different increasing 

 degrees of impact; the deflections at the middle from equal im- 

 pacts being to those at one-fourth span, as 10 : 7 nearly. The 

 relative ultimate deflections of the beam in the middle, and at a 

 point half way between the middle and one end, ought to be as 

 10 : 7"5 nearly. 



5th. The bars in Tables VI 11., IX., and X. were all of the same 

 iron and size, and the only difference was in the weights of the 

 striking balls. The distances fallen through, and the work done 

 by the balls to produce fracture, being respectively -3139 and 

 190-488 with the 603 lb. ball ; 1-2856 and 194-417 with 'the 151ilb. 

 ball; and 3-0506 and 230-32 with the 75j lb. ball, affording a good 

 illustration of the resistance from the weight of the bar. 



6th. The bars in Table XI. were of the same iron, Blaenavon 

 No. 2, as the others, but re-melted, to ascertain the effect of melt- 

 ing this iron a second time without mixture upon its power to bear 

 impact. The strength to resist blows was increased, but the iron 

 was harder and much more unsound than before. The work done 

 by the ball to break the beam in each case was increased in the 

 ratio of 261 to 194. 



7th. The deflections in cast-iron beams were always found to be 

 greater than in proportion to the velocity of impact; whilst in 

 wrought iron they were nearly constant with impacts of very dif- 

 ferent velocities. This fact shows that there is a falling off in the 

 elasticity of cast iron through impact, analogous to that through 

 pressure. The difficulty of obtaining a satisfactory theory of the 

 power of cast-iron beams to sustain impact is considerably increased 

 by this falling off in elasticity; but it is hoped that the varied 

 nature of these experiments will tend much to reduce it. 



Abstract No. IV. 

 Abstract of Results on Vertical Impacts upon Loaded Bea.ms of Cast-Iron. 



All of the beams were of the same weight and strength nearly. 

 They were placed on supports at a constant distance asunder, and 

 struck in the middle by the same ball, falling through different 

 heights. The object of the experiments was to obtain the effect 

 of additional loads, spread uniformly over the beam, in increasing 



its power of bearing impacts from the same ball. The beams were 

 of IJlaenavon Iron No. 2, cast to be 14 ft. 6 in. long and S inches 

 square. The mean weight of beam, 410-7 lb. ; mean weight of 

 beam between supports, 382 lb. nearly; distance between supports, 

 13 ft. 6 in.; weight of ball, 303 lb. 



