62 



THE CIVIL ENGINEER AND ARCHITECT'S JOURNAL. 



[Febri-ahy, 



prevent the equal flow of metal. It does not follow tliat the theoretical 

 form of greatest strength is the best one to adopt. His attention has been 

 principally coiilined to beams subjected to weights at rest. As weight in 

 such castings is not of so niucli importance, he is guided in the selection 

 of irons by the market price. Always mixes irons; is inclined to think 

 that good will result from Rlr. Merries Stirling's endeavours to increase 

 the strength of irun by an admixture of wrought iron. Thinks the manu- 

 facture of iron below the olher mauufaclures of the country ; believes that 

 in trance they roll out bars heavier than we do. Thinks that, if the plan 

 adopted in Belgium of the manufactures exhibiting qualities of iron every 

 year were followed it would improve the manufacture. Considers that the 

 qualitj of iron depends, tirst, upon the raw material, then on the fuel and 

 care in manufacture. Thinks investigation into the manufacture of iron 

 desirable, and that it would be advantageous to ofl^er premiums for the best 

 iron. 



Isambard Kinydon Brunei, Esq., Civil Engineer. — Has a preference for 

 the AVelch and Staffordshire irons. Endeavours to obtain a small propor- 

 tion of hot blast iron in mixtures. Does not like a large proportion of hot 

 blast ; thinks one-fifth advantageous. Takes the greatest possible load 

 that can by any accident come upon a girder, and assumes that as one- 

 third or two-fifths of the breaking weight: but takes the breaking point 

 lower than it is generally taken. As a general rule, would prove a girder 

 with a load a little greater than the greatest that could come upon it, and 

 examines its appearance under that load. Actual weight is the preferable 

 mode of testing girders. Although, strictly speaking, the same load cannot 

 be borne by a girder when resting on one flange as if applied at the top, on 

 account of the torsion, yet, by endeavouring to bring it as near to the 

 centre as possible, has not perceived any sensible difl'erence. If circum- 

 stances made it desirable to construct a girder to carry a load on the 

 flange, at some distance from the centre, it might then be desirable to cal- 

 culate the strength of the girder ; would certainly test it in that manner. 

 Such cases have not been suflicieutly frequent to require a special pro- 

 vision. X)oes not believe that any appreciable difTereiice is caused in the 

 power of resistance of the girder. Considers that, with his form of girder, 

 and with a large dead load of ballast, &c.,the torsion is inappreciable. A 

 soft substance between anything that produces vibration and cast iron is 

 advantageous, but wooden sleepers to support the roadway should not be 

 so elastic as to press on the edge of the flanges of the girders. Does not 

 consider that a moderate weight left on a girder will ever injure it. Has 

 not observed temperature produce any effect except expansion and con- 

 traction. Considers that no weight, except that approaching the bieaking 

 one, will permanentty aft'ect cast iron. The deflection of a girder does not 

 merely depend on the length. In a girder 30 feet long, 15 inches deep, 

 would allow ^(ijth of an inch to a foot. The deflectiou must depend on the 

 form. About half the before-mentioned deflection would t.e allowed in a 

 very stifl' girder. Wakes girders of the inverted T-eection with a very 

 large bottom web, and swelling at the top of the vertical web. The length 

 of cast iron girders is limited by what would insure a sound casting; at 

 present considers it to be 30 or 35 feet. When girders are required for 

 spans beyond the limits of simple cast iron girders, would prefer not using 

 cast iron at all. Would prefer timber or wrought iron, or both combined. 

 Would apply wrought iron to increase the tenacity of cast iron framing. 

 Has adapted that method in machinery. In large spans, assuming there 

 is no diliicully in obtaining an abuiment, would prefer cast iron in the 

 shape of an arch. Does not think that in a work put together by a good 

 mechanic, with ordinary judgment and proportionable strength, that any 

 vibration would aflect the bolts. Considers that the introduction of 

 wrought iron plates into the construction of bridges is the most important 

 step that has lately taken place in engineering ; believes that with ordi- 

 nary care and the improvements which have been introduced into riveting, 

 that the joints may be equal to the other parts of the structure. Does not 

 think that vibration can have any efl'ect on well-made riveting. Hivets 

 should not act as pins or bolts, but like clamps, and hold the plates together 

 by the friction of the one on the other ; in that manner the plates may be 

 insured not to break in any part contiguous to the rivets. Coiisiders that 

 the crjslalline fractures observed in bars broken by a succession of blows 

 is not the consequence of any internal change in the metal, but that iron 

 breaks with a crystalline or fibrous fracture according to the circumstances 

 under which it is broken ; produced several pieces of iron broken, some 

 with a crystalline fracture \>y a short sharp blow, others with a fibrous 

 fracture by means of a slow heavy blow. The same efl'ects may be pro- 

 duced by varying the temperature of the bar. Considers that when the 

 rails are well laid the defler lion will be less from a moving weight than 

 from that weight at rest. Some new engines weigh as much as 35 

 tons, and occupy a length of 20 feel or 1 \ ton to the foot run. Believes that 

 cast as V ell as wrought iron vaiii s its strength with the temperature ; the 

 colder it is the easier it will break. Thinks that suspension bridges might 

 be applicable to railways. Has once proposed one under very peculiar 

 circumstances. Considers the Indian tension bridge inferior to ordinary 

 suspension bridges. Would only use a lattice bridge when lie could not 

 get materials for the component parts exceeding a certain length: if he 

 were obliged to make a bridge of great length with short slicks, it might 

 be one mode of meeting the difficulty. 



Edwin Clarke, Esq., Civil Engineer. — Has superintended the Conway 

 Bridge for filr. Stephenson. It is a wrought iron tube made of boiler 



plates riveted together as in iron ship building: the span is 400 feet, the 

 extreme depth at the centre is 25 ft. G in., breadth 15 feet; the internal 

 breadth and depth are 21 ft. b in. aud 14 ft. 3 in.; the depth at the ends is 

 S feel less than at the centre. It was constructed on a limber platform on 

 the beach of the river Conway, 200 yards from its permanent site, and was 

 floated to its position on six pontoons of 350 tons each, and raised 17 feet 

 to its position by hydraulic presses; its weight is nearly 1,300 Ions. It 

 has a bearing at each end of 12 feet, and rests on bed plates and rollers to 

 allow of ils expansion from change of temperature. It was commenced 

 at the beginning of 1847 and finished in March 1846. The original idea 

 arose from considering whether a beam could be made large enough to 

 cross a sjian of 460 feet. Mr. Stephenson had formed beams of separate 

 pieces united by bolts, and had also applied tension rods to some beams 

 formed of separate castings. A cast iron arch was proposed but aban- 

 doned, partly on account of interference with the navigation of the straits. 

 Two beams side by side with an ordinary upper and lower flange would 

 make a space, through which if large enough a railway caniage might 

 pass. The first experiments were on round and oval tubes ; they changed 

 their shapes when loaded ; rectangular tubes did not ; that form was there- 

 fore adopted. Experiments were made to determine the resistance of 

 wrought iron to compression, that the actual strength of a large tube 

 might be calculated ; the power of wrought iron to resist compression 

 increased as the cube of the thickness of the plates: the strength of the 

 tube varied as the square of the linear dimensions. A model tube one- 

 sixth the real siz;; was made at Mill Wall, anil broken five or six times, 

 and strengthened at the part it had broken at after every time, till it was 

 considered that the strength was everywhere proportioned to the strain. 

 The thickness of the sides of the lubes appeared to produce very little 

 comparative efl'ect. The difference of elasticity rendered it difficult to 

 apply cast iron to the top of the tube. A bar of cast iron yields twice as 

 much under Ihe same weight as a similar bar of wrought iron, though its 

 ultimate resistance to compression is four or five times as great. If the 

 top of the tube were made partly of wrought and partly of cast iron, the 

 wrought iron would have to bear more than its share of pressure. Cast 

 iron must also be castlhick, which increases its weight, and the places of 

 junction require heavy flanges. The Mill Mall model it was assumed, if 

 increased to six times ils linear dimensions, should he 30 times as strong 

 and 210 as heavy. The bottom of the tube was considered as a chain, and 

 the plates were lapped over to make the chain as strong as possible ; the 

 rivets were proportioned so that the section of the rivet to be sheared 

 through equalled Ihe section of the plate it connected. The shearing 

 strain of a rivet is as it? tensile strain. Cells were put in the bottom of 

 the tube as being the most convenient way of getting suflicient area of 

 section of iron. The cells are kept stiff by angle irons. There are five 

 rows of cells in the bottom of Ihe tube. The bottom has great strength to 

 resist lateral pressure, as Ihe wind. The sectional area of Ihe bottom is to 

 that of the lop as 5 to G. The area of the bottom is 508 square inches; the 

 area of the lop 608 square inches. In the small experiments Ihe top had 

 always failed by buckling, but the strength of plates to resist buckling 

 varied as the cube of Iheir thickness, aud the lop might therefore in the 

 large tube have been of Ihe same area as the bottom ; but as Ihe top had 

 always been the part to fall, and Ihe data for calculating the resistance to 

 compression were not so complete as those for the resistance to tension, a 

 little was added to the top ; 12 tons to the square inch is as much com- 

 pression as wrought iron can be safely subjected to. At 10 tons per 

 square iuch most iron begins to be perceptibly altered in shape. The first 

 experiments were made before February, 1840. The last Mill Wall expe- 

 riment was made in April, 1847. The sides of the tube were considered a 

 mass of trellice work so thickly interwoven as to become a solid plate ; at 

 every 2 feet two pair of angle irons were placed face to face, aud running 

 from top to bottom of the tube, one inside and the olher out, like vertical 

 pillars, to keep the lop and bottom apart. The side plates are 2 feet broad. 

 These pillars appear to give suflicient rigidity, as Ihe sides of the tube 

 have never exhibited the least alteration of shape. For a distance of 60 

 feet from each end vertical plates have been added to strengthen the sides, 

 where the strain was considered greatest. At the ends, to prevent any 

 crushing of the sides, strong cast iron frames have been inserted. The 

 side plates in the centre are half an inch thick, but towards the ends 

 igths of an iuch thick ; the bottom plates are half an iuch thick in the 

 middle, and a quarter inch thick at the ends : on the principle that the 

 strain on the bottom varies at each point as the rectangle of the segments 

 into which the tube is divided at that point. M'heu the side of the model 

 tubes were thin near the ends, they invariably buckled there. The resist- 

 ance of the top cells to compression was never exactly ascertained ; 

 wrought iron will not bear above 12 tons compression per square inch. 

 The first cells experimented on were oval ; the square and circular were 

 then tried ; the iron when thin puckered, but a certain thickness of plate 

 answered equally well to prevent the cells either oval, circular, or square 

 from buckling, aud the iron crushed. The cells were made square not be- 

 cause the square form is best to resist compression, but because there were 

 many diflicullies in filling a circular cell in the lop of the tube, and lateral 

 streuglh was wanted to resist the wind, and also all the parts could be 

 more readily got at ; the cells are 1 fool 9 inches square, and the plates 

 three-fourths of an inch thick. As regards tension, rivels weaken plates, 

 but rivets increase the strength of plates to resist compression. Plates 

 riveted together generally break at the rivet, though they derive some 



