1850.] 



THE CIVIL ENGINEER AND ARCHITECT'S JOURNAL 



53 



feet, recoiiimends the wrought iron tubular or box girder. Being a strong 

 advocate for simplicity in mechanical structures, he would not recommend 

 compound girders where they can be dispensed with. Approves of 

 •wrou};ht iron tension rods to girders only in cases of necessity, and where 

 the top flange is enlarged, but prefers girders all of one material, even if 

 formed in parts. A\'ould rather give strength to a cast iron girder than 

 assist it by a wrought iron truss; the two materials are so widely different 

 in character that it is safer to keep them separate. By screwing up the 

 tension rods a strain is thrown either on the girder or on the tension rods 

 themselves ; an ignorant person might do injury without being aware of 

 it. M hen not limited by expense or levels, would prefer for narrow spans 

 a simple girder ; for moderately wide spans, the arch ; for spans exceed- 

 ing 70 or 80 feet, the wrought iron girder. Thinks that no vibration to 

 ■which railway bridges are subject can injure the joints or rivets, unless 

 the work is shamefully executed; nor would impact bave any effect on 

 the joints of a well-made cast iron girder. Does not think any effect is 

 produced by the load in skew bridges being alternately nearer one side 

 than the other. It is the opinion of some practical men and philosophers 

 that iron when hammered at a low temperature undergoes a complete 

 change in its internal structure, and that this effect is due to percussion, 

 heat, and magnetism, and time, which is an element in every process of 

 crystallisation. The application and abstraction of heat operates more 

 powerfully than probaljly any other agency ; too much influence is pro- 

 bably attributed to the other-mentioned causes ; a bar of the best wrought 

 iron, heated red hot and plunged into cold water, is changed from a fibrous 

 to a crystalline body ; heating and cooling will produce this effect in a de- 

 gree proportioned to the intensity of the heat applied ; by annealing the 

 iron its fibrous texture is restored, and sonielinies made more tougli than 

 before. Thinks magnetism may have some effect ; but often whei e causes 

 are inexplicable we fly to electricity for the solution; heating iron to a 

 high temperature deprives it of its magnetic powers which are restored by 

 cooling. Doubts that vibration changes the fibrous structure to a crys- 

 talline one, but thinks that each blow produces injury. Axles of a loco- 

 motive engine are subjected to repeated shocks from irregularities in the 

 rails and lateral action in passing curves, fiom a body weighing 18 or 20 

 tons, moving at 40 miles per hour. Each percussion tends to bend the 

 axles, and from the injury being continued many thousand times, it is 

 evident that time alone will determine the moment of fracture. If the 

 axles were so rigid as to resist the efl'ect of percussion, no injury could 

 ever take place or crystallisation appear. A bar bsat with a small ham- 

 mer is not altered at all, but the blows of a large hammer produce a 

 change of form which renders it brittle, not probably crystallising it. Is 

 of opinion that a fibrous body cannot be changed to a crystalline one by 

 any mechanical process, except when percussion is carried on lo the ex- 

 tent of proilucing a considerable increase of temperature. Fibres may be 

 shortened by continual bending, and the parts be thus made brittle, but 

 fibres cannot be changed into crystals. These clianges apply to all mate- 

 rials subjected to repeated alterations of form. Has not traced the break- 

 ing of null work to the change of internal structure. The shafts usually 

 break eventually from getting out of line. It would be interesting and 

 useful to experiment on the above points. The greatest weights on rail- 

 ways may be reckoned at IJ tons per foot linear for a single line, or 

 two tons per foot for a double line of rails. Considers that recommenda- 

 tions made by the Commission as to particular forms for bridges would 

 probably not be followed, but that experiments would be very beneficial. 



Joseph Glynn, Esq., Civil Engineer. — Was engineer-in-chief to the But- 

 terley Company. Cast iron is always combined wilh earths, as lime and 

 silica, as well as with carbon ; the more puje the iron is the stronger it is. 

 Never saw pure iron. Iron cast from the air furnace of a mottled or of a 

 clear grey fracture, bears ihe greatest weight. Iron cast from the air 

 furnace is stronger than from the cupola. Doubts the utility of mixing 

 wrought with cast iron for increasing the slrenglh of iron; doubts the 

 complete union of the two. The quality of iron depends, to a certain ex- 

 tent, on the ore, fuel, and flux used ; and an experienced person can gene- 

 rally tell what the produce will be. From a reverberalory furnace the 

 required mixture can be invariably produced. The length of time iron 

 remains in the furnace affects the quality. In the air-furnace it is 

 weakened by remaining too long. The best mixture for girders is about 

 one-third of strong crystalline Vtelch iron with two-thirds of the softer irons 

 of Derbyshire, Yorkshire, or Shropshire. The hot blast of itsell produces 

 no effect on iron. It may be used to smelt stubborn uniractable materials 

 that would not afford strong iron, and could not be otherwise smelted. In 

 the west of Scotland inferior iron has been produced by means of the hot 

 blast. There is no certain mode of detecting the difference between hot 

 and cold blast iron, but iron of a dark grey colour and very fine in the 

 crystal is generally hot blast. The difference is more marked as iron is 

 harder. Loam castings are stronger than open sand. Casts machinery 

 required to be very strong from the air-furnace in dry sand. A shaft cast 

 in an upright position is stronger than one cast horizontally on account of 

 the impurities floating to the top, and the density being increased. Adopts 

 the H form lor the section of girders, the bottom flange being largest. 

 Would not make a simple cast iron girder more than 50 feet long. Where 

 spans have exceeded that, has always used the arched form. Built an 

 arched bridge of 70 feet span over the Aire, at Haddesley; and one of 100 

 feet span over the Trent on the Midland Counties Railway. M ould in- 

 variably employ an arch when possible. Would not employ wrought iron 



as an auxiliary to cast iron in point of strength. Would only employ 

 it for bolts; on a large scale the workmanship cannot be so accurate that 

 each will bear its share of the stress. For spans beyond 50 feet, would 

 give the girder as much depth as possible, and join the pieces by bolls and 

 dowels. Would not have a wrought iron truss. When the workmanship 

 is good, does not consider the vibration and impact can affect the bolts and 

 rivets. Believes that the internal structure of iron becomes altered hv 

 being submitted to a succession of slight blows at a low temperature. E^is 

 seen many axles broken which presented a coarse crystalline fracture. 

 The continual succession of blows induces fracture, and changes the inter- 

 nal structure of fibrous iron to crystalline, the crystals increasing in size 

 as Ihe effect goes on. Crane chains made of fibrous iron break in a few 

 years with a crystalline fracture. Considers the same effect takes place 

 in cast iron. Shafts in mill-woik break. And there appears to be a limit 

 as to time in the durability of wheels. The fractures in these cases exhibit 

 an increased size of crystals. Considers that a stationary weight would 

 deflect a beam more than a moving one. Never made large girders of 

 wrought iron plates; the method is adopted for paddle beams of steam- 

 vessels, vibration has not affected those made for steam vessels, nor did 

 the rivets become loose. Considers that the strength of a wrought ii-on 

 girder is diminished by riiets. 



William Henry Barlow, Esr/., Civil Engineer.— Is resident engineer tc 

 the Midland Railway. Has found so much difficulty in obtaining the 

 quality of iron specified that he now simply specifies the dimensions of 

 the girders and the test to which the iron is to be submitted, leaving the 

 mixture to the founder. Objects to the inferior qualities of cinder iron 

 and hot blast iron generally ; though, at times, hot blast iron exhibits good 

 rfesults. Some specimens of hot blast are as strong as cold blast. Hot 

 blast iron seems more liable to abuse in manufacture than cold blast. Is 

 not aware of any mode of telling hot blast iron from cold blast. Specifies 

 that girders should bear a given weight wilh a given deflection. Would 

 make a girder so that the breaking weight should be four time the greatest 

 load. Considers that safe for weights moving at high velocilies. Proves 

 a girder to half the breaking weight. It gives the girder a permanent set, 

 but does not consider that it injures its strength. The proof is proportioned 

 to what the girder has to bear. Tests them by dead pressure by the 

 hydraulic press. Has not tried impact, during the test but thinks it might 

 be desirable when the breaking weight of the girder is nearly approached ; 

 but, practically, would give a large amount of surplus strength. Never 

 allows the load to exceed one-fourth of the breaking weight; it is often 

 one-fifth. The pressure being applied in the central plane of the girder. 

 In actual structures the pressure is usually applied to one side of the 

 bottom flange, but does not consider that when Ihe surplus strength is so 

 great and the iron good that it is of importance to apply the test in the 

 same way. A torsion is introduced; it is not, however, so perceptible iq 

 short girders. The effect of a great permanent load on girders is not in 

 operation in railways; but girders do not appear to be deteriorated by the 

 frequent passage.ofa load. Theone-forliethof an inch to a foot is assumed 

 as the amount of deflection that may be allowed in girders, but it is 

 empirical. The short time which a load rests on a railway girder appa- 

 rently renders the weight of less efl'ect than in warehouse girders which 

 bear a large load for years. Observed once on a timber viaduct that a 

 goods train produced a certain amount of deflection; an express train 

 coming afterwards, though with a lighter engine, seemed to produce a 

 wave through the bridge, and evideutly produced a worse effect than the 

 goods train. The point of maximum efl'ect would not be when the load 

 was in the centre of the bridge. And this is probably a reason for allowing 

 girders to deflect less in railway bridges than when exposed to dead pres^ 

 sure only. Has generally adopted iNlr. Hodgkinson's form of girder. In 

 spans of 50 feet, whenever the headway allows, prefers and has adopted 

 arched gilders, which are supported by abutments, and also act as girders. 

 A skew bridge on thai principle is a series of square bridges. The arched 

 girder for the bridge over the canal at Weelock is in two pieces, bolted 

 together in the middle ; Ihe rise is one-tenth. There are cases where OE 

 account of the headway reclangular openings are required, but they are 

 rare; girders have been used to a greater extent than necessity required, 

 from being in fashion. The length for cast iron girders will be limited by 

 the power of casting them; has not used any lunger than 42 feet. The 

 bowstring bridge is the best mode of construction where the spans are too 

 large for simple girders, a cast iron arch with a wrought iron string. In 

 a very large structure the rise of the arch might allow of a pair being tied 

 together at the top. If in combinations of wrought and cast iron, the two 

 inelals are bolted side by side, the different rates of expansion might pro- 

 duce a bad result. Has not found that Ihe impact and vibration lo which 

 railway bridges are subject has produced any bad effect on tlie bulls and 

 rivets of bowslring girders. The girders in skew bridges might, if the 

 deflection were excessive, suffer from the load coming on the centre of one 

 girder before it comes on that of the other. Except at high velocilies the 

 maximum effect will take place when the load is at the centre. To try 

 the effect of impact of trains, whitewashed the rails for a mile on an incline 

 of 1 in 31<i, and watched the effect of a fast train of 12 carriages going 

 down it over Ihem ; in cases of imperfections at the joints, there were 

 spaces 5 inches in length untouched by any wheels in the train. The rails 

 weighed 7slb. and were on wooden sleepers. Used to use felt as an inter- 

 posing medium to diminish vibration, which answered for light engines ; 

 the present ones are so heavy that any substan.'ie is soon crushed out. 



