58 



THE CIVIL ENGINEER AND ARCHJITBCrS JOUHNAIL' 



[Fedrjabt, 



changes of (cmperature. Allows tott"' "^ "'^ ^P*" ''"' '''^ deflection of 

 a girder. The deflection of the Godstoiie Bridge is ttsti''' "f ""= span, or 

 §ths of an inch. Proposes 10 ff et as the limit for simple cast-irou girders. 

 Used a level and levelling staff for obtaining the deflections of the God. 

 stone Bridge. Considers the girders rest so flrmly on their beds, that the 

 deflection observed is not due to any yielding in that respect. Depends on 

 Mr. Hodgkinson's rules for the form of construction for girders. Has 

 made no experiments on tlie amount of torsion caused by supporting the 

 roadway on the bollom flange of a girder. Considers a girder of separate 

 castings boiled together is a good mode of construction beypnd spans of 

 10 feet. Would not use that method for bridges of 101) feet span. Would 

 limit girders cast in one piece to tO feet span. Does not consider suspen- 

 sion rods a good mode of combining wrought and cast-iron. Would lay a 

 ■wrought-iron rod along the bottom flange. Assistance given to the ex- 

 tended part of a beam is more efl'ective than when given to the compressed 

 part. To avoid a large mass of cast-iron, would lay a wrought-iron rod 

 along the bottom flange. Does not consider that the different rates of 

 expansion would prevent the wroughtiron coming into play. When the 

 bridge gels much load it must come into play. Prefers an arch of cast- 

 iron where expense or height is not a matter of consideration. Is making 

 one over the Surrey Canal of three pieces bolted together. Does not 

 consider the vibration on a railway bridge sufficient to disturb the screws. 

 Does not consider that there is much difference of effect between engines 

 going fast or going slowly. Does not think vibrution so important as is 

 imagined. Fancied he observed an increase of deflection from engines 

 going fast ; there was a great deal of horizontal jar. Which he attributes 

 (0 blows given by the engine on the rails. Some may be due to the torsion 

 created by the weight being on one of the bottom flanges. Has not ob- 

 served any change produced in the internal structure of iron from repeated 

 blows at a low temperature. Thinks the subject an important one, and 

 that experiments could be made best by breaking beams which had been 

 long in use. Or testing girders whose previous test had been recorded. 

 Engines and tenders are being made, weighing together 32 tons. Engines 

 for inclines weigh as much as 30 tons without a tender. In estimating 

 the greatest load for a railway bridge, considers it covered with a train, or 

 a train composed of engines. Considers the Commissioners might make 

 some useful experiments on the Godstone Bridge. Has paid attention to 

 wrought-iron girders. It is desirable in a girder to concentrate the power 

 of resistance as near the lop as possible, and the power of extension as 

 near the bottom as passible, which can be accomplished in a cast-iron 

 girder; but in wrought-iron tube girders the bottom web, which does most 

 work, is a very small proportion of the whole section. Prefers wrouglit- 

 iron, or wrought-iron combined with cast-iron, to resist compression, to 

 cast-irou alone. Considers solid-sided wrought-iron girders an imperfect 

 mode of construction. Thinks the top of tube girders should be of cast- 

 iron. For a large span, considers wrought-iron safest. On account of 

 the uncertainty of cast-iron would make a cast-iron girder 50 per cent, 

 stronger than a wrought iroa one. The relative expense would be about 

 half. 



William Fairbairn, Esq., Civil Engineer. — In early life was a mechanical 

 engineer. Has been employed in engineering works of various descriptions. 

 Thinks Welch cold-blast iron, Blaina, for instance, best for girder bridges. 

 Considers most British irons improved by mixture. A good mixture is 

 two-thirds strong Welch, No. 3, the remainder Scotch or Stall'ordshire, 

 No. 2, with a little old iron. The same mixture is used in girders for 

 railway bridges and girders to support dead pressure only. Thinks Mr. 

 Morries Stirling's patent for mixing wrought iron with c;ist iron gives 

 indications of very superior strength, and states the results of experiments 

 upon it; also cites experiments by i\Ir. Lillie, of Manchester, on the 

 mixture of wrought and cast iron, which proved that the mixture was 

 one-third stronger ihan common cast iron, and one-eighth stronger than 

 wrought iron to resist transverse pressure. Considers the following mixture 

 of cast iron the best, viz. : — 



Lowmoor, No. 3 . 

 Blaina, No. 2 . . . 



Shropshire or Derbyshire, No. 3 

 Good old scrap 



30 per cent. 

 25 per cent. 

 25 per cent. 

 20 per cent. 



100 



This mixture can rarely be obtained on account of the price of Lowmoor, 

 and founders cannot be depended upon for the exact proportions. Prac- 

 tically he doubts any mixture unliss the parties interested were present 

 to witness the selection of the iron, and to see it put in the furnace. 

 Scotch and SlaUbrdshire iron are good for light castings. Good castings 

 depend on the care of the furnace man, the temperature of the furnace, 

 and the heat at which the metal is run into the mould. Recommends the 

 anthracite iron where rigidity and strength is retjuired. The strongest 

 iron should be put in railway bridges. Considers that the hot blast does 

 not improve the quality of \\ elch and English irons; but that its applica- 

 tion in the Scotch furnaces to the reduction of the black band is an im- 

 provement. Scotch hot blast mixes well with Welch irous. The etiects of the 

 hot blast vary with the quality of the fuel and ore, and much depends on the 

 quantity of sulphur present in the coal and coke. The Lowmoor ores 

 were injured by the application of the hoi blast. Fuel is an important 

 element in the manufacture of iron, the nearer it approaches pure carbon 



the better. In the Scotch black band and similar ores the hot blast will 

 bring more iron out of the same mine than the cold blast. The hot blast 

 enables the manufacturer to work up not only poorer ores but cinder 

 heaps, into apparently flue granulated iron. The use of the hot blast at 

 first led to the introduction into the market of a very inferior description 

 of iron. Considers the Scotch iron weaker and more fluid than most 

 English irons; it is equal to Staffordshire, but weaker than Welch and 

 Yorkshire. Scotch iron is an exceedingly fluid and fine-working iron, 

 and well suited to machinery ; it runs well into the mould, an 1 brings 

 out tiie castings with the edges sharp. Does not think the most experi- 

 enced metallurgist could tell the dilierence between hot blast and cold 

 blast iron from the appearance. Considers that liot blast presents greater 

 uniformity than cold blast in its granulated appearance, and indicates a 

 more perfect process of crystallisation, probably arising from the greater 

 heat of tho furnace. In cast iron girders, would make the breaking 

 weight four times the greatest load. In structures exposed to shocks or 

 vibratory motion would adopt live times or six times. It is safer to adopt a 

 light load, so as to make allowance for casual strains which cannot be com- 

 puted. Never proves a girder to more than half the breaking weight, gene- 

 rally one-third; disapproves of testing a girder much beyond the permanent 

 load, the object being to ascertain its soundness and elasticity ; a further 

 test tends to permanent injury. In testing girders, carefully inspects the 

 outward appearauce, and then hangs weight from the centre, and observes 

 the deflection and permanent set. Does not consider that a permanent 

 set given to beams in the early stages of loading injures the strength. 

 Thinks that within certain limits the form of a beam may be distorted 

 without its strength being injured. Considers that to support the load on 

 one side of the bottom flange is wrong in principle, and to a certain extent 

 injurious in practice ; but the method has many conveniences: to meet 

 the requirements of structures, self-evident principles must in practice be 

 sometimes abandoned. When the load is supported on the bottom flange, 

 the bearing should be brought as close as possible to the central web, by 

 casting a fillet or shelf to carry the cross-beams ; bolt holes should be made 

 as near the neutral axis as possible , or when required for bulling wooden 

 bearers to the bottom flange, projections on the bottom flange should be 

 cast to receive them ; bolting the roadway to the girders resists, in a great 

 measure, any lateral strain on the girders ; but the lateral strain is best 

 resisted by broad top and bottom flanges. Considers bolt holes and other 

 perforations in cast iron girders very objectionable, and they should in no 

 case be made, even through tlie neutral axis, without thickening the adja- 

 cent part to compensate for the part taken out. These objections arise 

 from considering the complexity of such a girder and the additional mate- 

 rial required to make it equally strong as if plain. Is an advocate for 

 simplicity of construction in everything, and would only allow distortion 

 of form when inevitable. M^ould prefer supporting the cross bearers on 

 Ihe top flange or suspending them from the bottom flange by hook 

 bolts. Supporting the road on one side of the bottom flange is wrong la 

 principle, but convenient. If the top flange be broad and rigid, that mode 

 of construction is less objectionable. It would be advantageous to seek 

 for a new form of beam ; a narrow top flange, though well proportioned 

 for vertical pressure, is weak to resist lateral strain. The practice of 

 supporting the roadway on the bottom flange is simple, cheap, and conve- 

 nient, and will not easily be abandoned. Recommends a new form of 

 girder to be sought for, to give the girder suflicient stiffness. Has him- 

 self always increased the top flange to resist the lateral strain. In a large 

 span with girders having small top flanges, the lateral deflection, if not 

 resisted by a firm connection of the cross beams to the girders, might cause 

 an outward pressure dangerous to the structure. .Vs girders are generally 

 tested to ascertaui their soundness, it is usual to apply the test to the top 

 flange, but it would be of great value to test them as they are to be used. 

 The test is usually applied to ascertain the soundness of the casting, the 

 strengtli being computed at three or four times the load. The joists which 

 support the roadway when carried on the bottom flange, tend to cause by 

 their deflection a lateral pressure on the girder. This effect takes place to 

 some extent in wooden and Sandwich beams ; from experiments it appears 

 that this latter description of cross beam is weak, and its elasticity so im- 

 perfect as to render it inadmissible for supporting great weights. The 

 Sandwich beam is objectionable and expensive. Is of opinion that a beam 

 loaded with a given weight, even approaching its ultimate powers of re- 

 sistance, would support the load ad injinituni if not disturbed or exposed 

 to changes of lemperature; although time is an element in the change 

 which lakes place in every material, any increase of deflection in a loaded 

 girder may be traced to atmospheric action, vibration, change of load, 

 and teniptratiire : remove these disturbing causes and the deflection will 

 remain fixed. Casl iron of average quality loses strength when heated 

 beyond a mean lemperature of 220"^, becomuig more ductde and less rigid 

 to resist an uniform strain, and becomes insecure at the freezing point or 

 under 32" of Falireuheil. In girders of 40 feet span finch is the maxi- 

 mum allowable deflection, that is, '02 inches per linear loot; '005 inches 

 is preferable. Adopts Mr. Hodgkinson's form of girder modified in the 

 top flange to ensure uniforinily in the casting. Considers 40 feet to be the 

 greatest allowable length between Ihe supports for simple cast iron girders. 

 Knows an instance of a girder 70 feet long, cast in one piece in Holland. 

 Never heard of a girder breaking by its own weight; a properly propor- 

 tioned girder could not do so. For spans beyond the limit of simple cast iron 

 girders which must be passed with ii level soflit to the extent of 100 or 200 



