laso.] 



THE CIVIL ENGINEER AND ARCHITECT'S JOURNAL. 



it fails reject tlie whole. In girders not subjected to vibration, considers 

 that the greatest load should not exceed one-third of the breaking 

 weight; in girders for railway bridges, one-fourth. Proves girders to 

 double the greatest load. For girders of new forms applies the proof by 

 dead weight ; but in known forms, uses the hydraulic press as being more 

 convenient, observing the amount of deflection. Considers the objection 

 to the hydraulic press obviated by the use of cylindrical instead of conical 

 valves. The load on one of the bottom flanges is not objectionable pro- 

 vided the girder does not cant. Tesis girders sometimes by a weight ap- 

 plied to one of the bottom flanges. Considers a span of about 50 feet as 

 the limit for simple cast iron girders. For girders to support a quiescent 

 load would make the section of the top flange one-sixlh that of the bottom. 

 In a railway bridge, where the lop table would not be supported laterally, 

 would make the area of the top table one-fourlh that of the bottom. In a 

 railway bridge, where the top table is supported laterally, makes the area 

 of the top flange one-fifth that of the bottom. The top flange of a girder 

 being subject to compression may be compared to a column ; and if bent, 

 its liability to break will be increased. If circumstances required, would 

 make a girder of more than 60 feet long in one piece ; roving bridges over 

 the New Birmingham Canal are 80 feet long, cast in one piece. In well 

 constructed bridges the deflection of the platform should not cause any 

 injury. Considers the smallest weight applied impairs the elasticity of a 

 beam, and that a girder exposed to change of temperature and vibration 

 will swag, and that this effect will go on increasing; but he considers 

 that the only diminution of strength from this is due to the diminution of 

 the sectional area of the bottom table ; but that in cases where a beam is 

 not subject lo change of temperature, it would retain its original position. 

 Instances some girders feet long for supporting hoods to smiths' foiges, 

 which are warmed by day and allowed to cool at night; they swag nearly 

 3 inches in the centre. Considers that in the alteration in the arrange- 

 ment of the particles of iron caused by a change of temperature the weight 

 takes advantage of the change. Does not consider that remoung and re- 

 placing a weight on a beam continually would have quite the same efl'ect; 

 mentions that anchors when tested take a week to regain their original 

 position ; considers alteration of temperature more likely to produce swag- 

 ging than vibration. Thinks that railway girders will gradually swag and 

 must be exclianged, and that few which have been ten years in use have 

 not swagged, but that their strength is only impaired to the extent men- 

 tioned above'; the greater the inertia of the bridge the longer would the 

 action be delayed. Considers the mode of supporting the roadway on one 

 side of the girder to be wrong. The deflection of a girder should not be 

 considered with reference to the span. For large spans, prefers cast iron 

 on the principle of compression. Would make straight girders for large 

 spans of several castings bolted together with wrought iron tension rods 

 fixed horizontally along the bottom flange, and put considerable initial 

 strain upon the w rought iron bars, that the cast iron may come into opera- 

 tion when the wrought iron is under a considerable degree of tension, so 

 that the ultimate efl'ect from the two may be obtained. The expansion 

 produced by changes of temperature being only a difl^erential quantity, 

 would be small in a length of 100 feet; and the wrought iron being more 

 elastic than cast iron, should bear it. The bow string girder, with a bow 

 of cast and a string of wrought iron, would be cheap and safe. A bridge 

 for crossing the Arno is being made of straight girders on the above men- 

 tioned principle ; the wrought iron bars are under a tensile power of 6 

 tons per square inch. In process of time the wrought iron would stretch; 

 wrought iron would stretch ^th inch in 10 feet, with a weight of 10 tons. 

 Would not let rails rest on the top of a wrought iron riveted girder without 

 a piece of wood between. Girders made of separate castings should, in 

 addition to bolts, have a wrought iron tie bar. .Soft timber between the 

 rails and girders will prevent danger from vibration. Considers alteration 

 of temperature as likely lo subject wrought iron girders lo a great deal of 

 undue compression and extension. Thinks experiments on impact and 

 vibration desirable. Believes that wrought iron is rendered crystalline by 

 a succession of slight blows at a low temperature, and has observed that 

 the older axles are, the more crystalline they are ; also remarks, that if the 

 thread of a screw be cut on a bar of fibrous iron, the tapped part will 

 break with a more crystalline fracture than the other. Shafts in mill 

 work break and exhibit a crystalline structure. Thinks cold hammering 

 injurious to axles from tending to make them crystalline, and also from 

 producing a strain like Ihat produced by straightening castings by ham. 

 mering. Would prefer their being finished at a high temperature to being 

 annealed. Cold hammered axles may be detected by their appearance. 

 Ihinks experiments on long continued deflection are very desirable. In 

 estimating the strength of a girder, adopts as the greatest weight li Ions 

 per foot per single line of way ; that is ^lon per foot for weight of platform 

 and 1 ton per foot for weight of train ; for two girders of 40 feet span, would 

 take the weight at 60 tons distributed, equal to 30 tons in the centre. 

 \\ ould calculate the breaking weight of each girder at 60 tons in the 

 centre and prove them to 30 tons. Considers that with a carefully laid 

 road the deflection due to rapidly moving weights is less than that due to 

 such weights at rest, from the shorter time allowed to overcome Ihe vis 

 mertia of the bridge. There has been a great want of fixed principles in 

 the construction of railway bridges: no general principle has been laid 

 down ; whilst one engineer is satisfied with one amount of proof, another 

 adopts SIX limes as much. In making contracts for railway chairs, stipu- 

 lates that the mixture he uses when cast into a bar of a certain form shall 



break with a specified weight. Is inclined to think the castings from the 

 air furnace better than those from the cupola, but the diflference is rery 

 minute. 



Henri/ Grissell, Esq., Iron-founder and Machinist. — Amongst other large 

 works, is at present constructing a built girder bridge for a span of 121 

 feet ; it is 12 feet high, and weighs 100 tons; it has been proved lo IDS 

 tons distributed over it. Has not studied the chemical constitution of 

 iron. Prefers a mixture of iron for castings. The mixture depends on 

 the state of the markets ; and from old iron being so plentiful in London, 

 pig-iron is not considered so much as in the country. Mixes Scotch iron, 

 old iron, cold-blast Welch iron, the proportions being dependent on the ap- 

 pearance of the fractures ; for cylinders a larger proportion of cold-blast 

 iron is used than for girders. Considers London castings 15 per cent 

 stronger than country ones, from the use of old iron. Hol-blast iron when 

 mixed is as good as cold-blast, but alone it is not to be depended on. The 

 proportions lor mixtures are so dependent on the qualities of iron, that he 

 is guided by the appearance of the fractures in determining them. Con- 

 siders he could mix iron so as to make a casting bear any weight in rea- 

 son. Could not tell hot-blast from cold-blast iron from the fracture. The 

 proportion of stress to strength varies with the section of the girder and 

 the strains to which it is subjected ; generally considers the load should 

 be one-third the breaking weight for railway bridges. Handed in the rule 

 he adopts in calculating the strength of girders. Has made simple and 

 compound girders. AV'ould make a girder in one casting 50 or 60 feet. 

 Considers a level top flange a waste of metal. In designing a girder, 

 judges by the eye of the probable strains it would be subject to, and then 

 calculates the strength, and alters the form so as to obtain the greatest 

 strength with the least quantity of inetal. Adopts the double T section, 

 the bottom flange being largest. Girders may be proved by a lever or an 

 hydraulic press; the latter is what he usually adopts, and it is as certain 

 as the lever when correctly made. Does not think a girder will bear the 

 same weight if applied only on one flange as if applied to both equally. 

 Proves girders to find out whether the casting is sound, and so applies the 

 proof to the top. Has never noticed that length of time or change of 

 temperature makes beams swag. For compound girders prefers the built 

 girder. Considers half an inch deflection may be allowed in every 20 

 feet of length ; can regulate the deflection by the mixture of iron he uses; 

 would not consider a beam injured by a deflection of | inch in 20 feet, 

 if it returned lo its original position. For large spans when not tied by 

 expense or height, would generally prefer a built girder. But thinks that 

 an arch is a stronger form than a straight girder, but more expensive, 

 Would guarantee a straight girder with top and bottom flange lo bear any 

 amount of pressure. Would not hesitate to use one for a span of 200 

 feet ; thinks it would bear any weight th^t could come on. Does not 

 think impact and vibration would affect large bolts and rivets, but that 

 where no more than just the necessary sirenglh is put in, every jar would 

 tend to loosen them. Thinks vibration dangerous to wrought iron ; vi- 

 bration takes much more effect on wrought than on cast-iron. Has ob- 

 served in crane chains an alteration in the structure of the iron, after a 

 few years' use ; instead of its breaking with a black tensile appearance, 

 it breaks short and white like cast-irun; it is changed from beautiful 

 malleable iron lo the appearance of very good cast-iron. Cold hammering 

 will also produce this efl'ect on cast-iron, but it can be restored very 

 nearly to its original texture by annealing. Feels convinced wiought-iron 

 girders will become altered to a crystalline texture by vibration. Knows 

 no case of cast-iron becoming altered, or breaking from vibration alone. 

 Has not given his atteution to axles. Has made numerous experiments 

 on iron of all sorts and mixtures. Considers that if the form of a girder 

 be given him, he could mix the iron for making it to such a degree of 

 nicely, 'that he could guarantee any amount of deflection, and carry any 

 load required in moderation. ,\ttaches the greatest value to old iron, but 

 not to differences in pig-iron ; considers all Scotch iron to be much of the 

 same quality, except one or two sorts, which are very superior. The 

 metal for mixtures must be selected with great judgment. Does not con- 

 sider it necessary to try the relative strengths of the different sorts of 

 metal before mixing, but judges of the proportions by the fracture. A 

 good mixture would be one-third hot-blast iron, one-third old iron, one- 

 third Blaenavon Welch iron, but he does not confine himself lo one parti 

 cular mixture. 



Peter William Barlow, Esq., Civil Engineer. — Has been employed 

 chiefly latterly on the Soulh-Eastern Railway. Has not observed much 

 difference in the strength of castings. Has always made the breaking 

 weight of girders six times the greatest load fur railway bridges. For 

 other works four times would be sufficient. Proves girders to one-third 

 of the breaking weight, or double ihe greatest load. Prefers proving 

 them with actual weight, and giving some vibration lo the beam by 

 putting ou the weight. Girders will not bear the same weight when 

 resting on the bottom flange as if applied at top. Has adopted 

 another form of girder, the object being to make the bridge one 

 complete plate. Considers 40 feet as the limit for such a bridge. 

 Has made one over a railway at Tonbridge wells. Finds that the 

 deflections are less than he calculated, from the assistance one part 

 affords lo another. Has not observed any injury from the bending 

 of the joists which carry the roadway between two girders. Has 

 not noticed any increase of deflection from a permanent load or from 



9 



