54 



THE CIVIL ENGINEER AND ARCHITECT'S JOURNAL, 



[February, 



Mr. Glynii names one-third strong iron from South Wales and two-thirds 

 of the more fluid metal of Yorkshire, Derbyshire, and Shropshire. Mr. C. 

 Fox, Mr. Grissell, and Mr. Charles Jlay, however, all concur in staling that 

 mixtures of iron practically depend very much upon the commercial question 

 of cost, and it is generally ailmitted that engineers have no guarantee that 

 the mixture for which they may have stipulated in a contract shall be that 

 used bv the founder ; hence Mr. Fox recommends that engineers in contract- 

 ing for a number of girders should stipulate that they should not break with 

 ess than a certain weight (leaving the mixture to the founder), and cause 

 one more than the required number to he cast ; the engineer might then 

 select anv one to be broken, and if it broke with a less weight than had 

 been agreed upon, the whole should be rejected. Mr. Glynn considers that 

 the strongest castings are those cast from the air-furnace in dry sand, and 

 castings in loam are stronger than those in open sand. The metal is more 

 dense and more free from impurity when cast upright. Mr. Fox aud Mr. 

 Fairbairu also prefer the air-furnace. With respect to wrought-iron, Mr. 

 Morries Stirling considers the process adapted in its manufacture as capable 

 of great improvement. Mr. E. Clarke states, that wrought-iron from the 

 same maker is not always the same, and though there is not much difference 

 in the ultimate strength of iron, that some qualities extend much more than 

 others before breaking. 



Proporlion of Load to Breaking Weight, in Girders. — There appears to he 

 a considerable difference of opinion as to the proportion between the greatest 

 load which a girder should be allowed to bear and the breaking weight. 

 There are two conditions under which the weight may be applied, viz., first, 

 when stationary, as in the case of water-tanks, floors, &c. ; second, when 

 the weight moves so as to cause concussions and vibrations, as in railway 

 bridges. In girders required for the first case Mr. Fox and Mr. T. Cubitt 

 consider that the breaking weight should be three times the greatest load ; 

 Mr. P. M'. Barlow four times; and Mr. Glynn would not make it less than 

 five times the load. 



In girders for railway bridges Mr. Brunei states that he allows the load 

 to he one-third or two-fifths of the breaking weight ; but he considers that 

 the rule he adopts for calculating the dimensions of his girders gives more 

 than the nsual strength. -Mr. Grissell and Mr. Charles May consider one- 

 third to be suflScient; Mr. Eastrick, Mr. P. W. Barlow, Mr. E. Stephenson, 

 and Mr. Joseph Cubitt adopt one-sixth; Mr. Hawkshaw prefers one-seventh, 

 except in cases where great care is exercised in the selection of materials 

 and workmanship, when a smaller proportion would suflice; aud Mr. Glynn 

 considers that in structures exposed to concussion and vibration the ultimate 

 strength of a girder should be ten times the greatest load. 



Testa for Girders. — The general opinion as to the amount of test which 

 should he applied to girders is that the test should amount to twice the 

 greatest load. Mr. Joseph Cubitt would employ three times the greatest 

 load, or half the breaking weight ; and Mr. Thomas Cubitt considers it safer 

 to test a girder almost to the extent that would break it than not to prove 

 it at all, as the testing of girders is the only means of discovering defects 

 under the surface, and concealed from the eye. Mr. Brunei, however, 

 thinks that a girder should not be tested with a weight exceeding the 

 greatest load, as the object in testing is to ascertain the soundness of the 

 casting, which may he judged of by its appearance under the load, and all 

 risk of permanent injury should be carefully avoided. Mr. Eastrick, Mr. 

 Glynn, and Mr. Joseph Cubitt recommend that blows be applied to cast-iron 

 girders when under the testing load. Mr. Hawkshaw and Mr. P. W. Barlow 

 consider that where actual weight is used sufficient vibration is given to the 

 beam by throwing the weight into the scales used in testing. It is stated 

 that, for convenience sake, girders are usually tested by means of the 

 hydraulic press; but Mr. Fairbairn, Mr. Locke, Mr. Brunei, Mr. Joseph 

 Cubitt, and Mr. Fox prefer using actual weight, on account of the uncer- 

 tainty as to the actual pressure the hydraulic press brings upon the girder ; 

 though the latter gentleman considers that all liability to error in the press 

 is obviated by an improved construction which he has adopted. Mr. C. 

 May states that, as girders are bought at the lowest possible price per ton, 

 the manufacturer is compelled to adopt the most convenient and not the 

 best mode for testing them, or ten times his profit would not pay him for 

 the experiment. 



Loads on the Bottom Flange. — It is admitted that the mode of supporting 

 the roadway on the bottom flange of a girder causes torsion in the girder, 

 though Mr. Eastrick and Mr. Locke do not consider that the strength is 

 diminished by the pressure being so applied; and Mr. Stephenson does not 

 think the torsion is of sufficient consequence to be noticed. In order to 

 guard against any ill effects which might arise from the torsion, Mr. Locke 

 fits in transverse pieces of timber between the two girders which support a 

 line of rails, chocked perfectly tight, and he ties the bottom webs together 

 with tension bars. Mr. Fairbairn and Mr. Hawkshaw consider it would be 

 advantageous to alter the form of girders to enable them to withstand the 

 torsion. Mr. Fairbairn thinks the cross beams should either lay on the top 

 flange, or be suspended by hook bolts from the bottom flange, in whicli 

 opinion Mr. Glynn concurs. Mr. Hawkshaw would increase the top flange 

 of the girder, or would cast shoes or brackets on them to bring the bearing 

 of the transverse joists close to tlie vertical web. Mr. P. W. Barlow has 

 adopted a new form of bridge to avoid this torsion. Mr. W. II. Barlow 

 observed considerable torsion in a girder without auy top flange. Mr. 



Fairbairn and Mr. Hawkshaw are of opinion that wooden cross-bearers for 

 the roadway are liable to increase the amount of torsion by bending ; but 

 Mr. Stephenson and Mr. Brunei state that wood is desirable as a cushion to 

 prevent the noise and vibration which iron on iron would he subject to. 



Length for Simple Cast-Iron Girders. — The use of simple cast-iron girders 

 in bridges appears to be limited only by the power to make sound castings 

 (which arises chiefly from the difficulty of pouring the metal equally, and 

 the inconvenience of handhng large masses. Mr. Eastrick, however, would 

 not put any limit to the length. Mr. Hawkshaw considers that they may 

 safely be made more than 50 feet long; in which opinion Mr. Fox and Mt. 

 Grissell concur, but name CO feet as the limit. Mr. Glynn, Mr. Charles 

 May, and Mr. Joseph Cubitt would make them from 40 to 50 feet. Mr. P. 

 W. Barlow, Mr. Fairbairn, Mr. W. H. Barlow, and Mr. Stephenson state 40 

 feet as the limit; and Mr. Brunei names 35 feet, as he does not consider 

 that sound castings can be ensured to a greater length. Mr. Fairbairn, 

 however, mentions a girder in Holland 70 feet long cast in one piece. 



Form for .Simple Girders. — It appears to he universally admitted that the 

 form resulting from Mr. Hodgkinson's experiments on the tension and com- 

 pression of iron is that which gives the greatest strength; but the actual 

 proportions are generally modified to suit the varying circumstances under 

 which girders are employed. Mr. Stephenson sometimes makes the top 

 flange equal to the bottom one, but usually in tlie proporlion of 3 ; 5, partly 

 to obviate any risk from unequal cooling of the materials, and partly from 

 the necessity of having a large top flange to bolt the flooring to. In prefer- 

 ence to using a single girder, Mr. Stephenson recommends two girders to be 

 bolted together, with a baulk of timber between, to which the rail is fixed. 

 Mr. Hawkshaw, Mr. Fox, and Mr. Joseph Cubitt recommend that the 

 top flange be increased beyond the proportions given by Mr. Hodgkinson, in 

 order to resist the lateral torsion. Mr. W. H. Barlow and Mr. Locke would 

 use the arched form of girder whenever practicable, and the former gentle- 

 man says that straight girders have been in fashion, and consequently more 

 used than practice actually required. Mr. Fox, in girders subject to dead 

 weight only, would make the proportion of the top flange to the bottom one 

 as 1 ; 6; but in railway bridges he recommends 1 ; 4. Mr. Thomas Cubitt. 

 mentions that shoes, or sockets, or any projections cast on girders, have a 

 tendency to create flaws from causing the dirt to accumulate in those places, 

 and he considers that the shape which will ensure a sound casting should be 

 as much considered as the theoretical form of greatest strength. 



Deflection of Girders, and Effects of Permanent Loads and Change of 

 Temperature. — It is considered that girders should not deflect more than 

 from if^th to :rio"' "f their length according to the form of the girder. It 

 does not appear from the evidence that a weight equal to what a girder is 

 constructed to carry will, even if left on for any length of time, cause the 

 deflection of the girder to increase, unless subjected at tlie same time to 

 considerable changes of temperature. Some experiments made by Mr. 

 Fairbairn and Mr. Braidnood show that iron loses a considerable pro- 

 portion of its strength when heated to a temperature of more than 220° 

 Fahrenheit, and that it becomes uncertain below 32°. Mr. Clarke described 

 the effect of the sun coming out and shining on the Conway tubular bridge 

 for half an hour to have been to raise the tube vertically one inch ; and he 

 mentions that at night, from the low temperature, the deflection was always 

 greater than in the day-time. Mr. Fox instances the eflTect of frequent and 

 great changes of temperature on some short girders, 6 feet long, which 

 support the hoods of the forges in his workshops. In the day-time they are 

 so warm that the hand can only just bear the heat ; at night they become 

 cold. The effect is to make the girders sicag, and the swagging appears to 

 be continually increasing. Some have attained as nmch as 3" deflection in 

 the centre; but their strength does not seem to be impaired. 



The general impression of engineers appears to be that the deflection caused 

 by passing a weight at a high velocity over a girder is less than the deflection 

 which would be produced by the same weight at rest; and the increase ob- 

 served, in many instances, is attributed by ilr. Locke, Mr. Stephenson, and 

 Mr. Fox, to the inequaUties at the junction of the rails, or to the jerks of 

 the engine. Mr. Hawkshaw however considers, that the deflections would 

 be increased, and has given some examples of a manifest increase. 



Mr. P. W. Barlow has observed a slight increase, and Mr. M". H. Barlow, 

 in reference to this subject, cites a curious phenomenon which he observed 

 on a timber viaduct, viz.: that with a heavy goods train at a low velocity, a 

 certain amount of deflection was produced ; but an express train passing 

 immediately afterwards, with a much lighter engine, seemed to push the 

 bridge like a wave before it. 



Forms of Girders beyond the limits of simple Cast-iron Girders. — The 

 modes of construction which have been adopted by engineers for crossing 

 spans beyond the limits of girders made of a single casting, are very various ; 

 but the chief forms which have been adopted by engineers for girders of a 

 compound nature in railway bridges may be classed under straight built- 

 girders of cast-iron in separate pieces bolted together; arched girders of 

 cast-iron; trussed girders^ bow-string girders; wrought-iron box and 

 tubular girders. 



The Bmlt Girder is formed of separate castings fitted closely at the joints 

 and bolted together, and is entirely dependent upon the bolts for support. 

 Mr. Grissell instances one of 120 feet span, and states that he should have 



