1850.] 



THE CIVIL ENGINEER AND ARCHITECT'S JOURNAL. 



61 



mental girder, similar to the Dee Bridge girder, from whicli it appeared 

 that the tension rods when acting at an angle could camber the girder. 

 Girders are made in separate pieces on account of the difficulty of cooling 

 large masses, and the inconvenience of conveying them. Tliinks it would 

 be imprudent to make larger castings than those recommended for girders. 

 Has had failures. Although there is a considerable variation in the 

 strength of iron, there is a remarkable approximation to an average stan- 

 dard. Practically an engineer is not justified in going to any great expense 

 to get a particular quality of iron. A difference of 20 per cent, in samples 

 of iron is not of much consequence when the girders are made to bear six 

 times the load that comes upon them. Does not consider tiiat any injury 

 can arise to a girder from the bending of the joists supporting the platform ; 

 in many cases has had the bearing secured close to the central web. I're- 

 fers, instead of depending on one girder, having two bolted together, with 

 a baulk of timber between. Prefers a n ooden platform to one resting on 

 iron beams; does not apprehend danger from the vibration, but the noise 

 is so unpleasant that some soft medium should always be interposed. 



Joseph Lode, Esq., M.P., Civil Engineer. — The strength of iron will de- 

 pend upon mixtures. Prefers a mixture. The mixture of hot blast iron 

 with cold blast considerably increases the strength. Understands that a 

 mixture of wrought iron with cast adds to its strength. Considers it bet- 

 ter to trust to the knowledge and experience of iron founders of high 

 character than to specify for particular mixtures. Has generally made the 

 breaking weight of girders from three to four times tlie greatest load ; but 

 the load is supposed to be dead weight, whereas the shocks in railway 

 bridges m«y increase it to within half the breaking weight. On railways 

 from the levels the most convenient form of girder must sometimes be 

 adopted in preference to the stillest. Would not prove a girder with more 

 than double the greatest load. Thinks dead weight a more self evident 

 mode of proving girders, but that the hydraulic press is a very convenient 

 and good mode. Resting the weight on one of the bottom tlanges produces 

 torsion to some extent. Thinks it might be desirable in some instances to 

 test the girders with weights applied as in practice. Has known cases 

 where the test was applied to one ilange as in railway practice. Never 

 kne%v a flange break ort' a girder. To prevent the girders getting out of 

 the perpendicular, makes the baulks supporting the rails fit tight between 

 the girders, and connects the bottom flanges by the rods. Has never ob- 

 served any injury arise to girders from being subjected to permanent 

 "Weights for a length of time, or to changes of temperature. Would allow 

 girders on railway bridges to deflect from the -jTjth of au inch to the :^th 

 of an inch per foot linear. But the amount would depend on the form of the 

 girder. Some forms admit of more deflections than others. Does not 

 like too much deflection in a railway beam. For the forms of girders 

 adopts the large bottom flange. According to his present experience, 

 would limit cast iron girders in one piece to 45 feet long, but he may per- 

 haps go further. Would always prefer an arch if possible. Dislikes 

 cast iron in flat girders at all times and in all spans. Would never use it 

 if he could avoid it. Does not object so much to wrought iron, but would 

 not use that when it could be avoided. Is not favourable to girders com- 

 bined of separate pieces. Would use the bowstring bridge for large spans. 

 Does not approve of combining wrought and cast iron as done in girders of 

 the Dee Bridge class. But does not wish it to be inferred that there is no 

 combination of which he would approve. Objects chiefly on account of 

 the difl'erent rates of expansion of wrought and cast iron. Does not think 

 that in compound bridges well put together the vibration and impact irom 

 trains would affect the joints and rivets ; but if badly put together, or the 

 roadway were not in good order, the joints would sooner or later be 

 affected. Does not consider that the deflection of one girder before that of 

 the other in skew bridges would produce oscillation to any injurious ex- 

 tent. When the roadway is good there is very little difference between 

 the deflection due to weights at rest and that due to the weights moving 

 ■with velocity. A bad joint is much more serious than an increase of 

 Telocity. Has known the deflection to be less wiih velocity. Vt'hen there 

 is any great diflerence, attributes it to bad joints. Conceives that per- 

 petual concussions might change the texture of wrought iron. Does not 

 think the same elTect would be produced in cast iron. A. cast iron beam 

 which had been in use for a long time in the Blackwall Railway was taken 

 out and broken ; it bore a very large weight with reference to its calculated 

 breaking weight. Would observe that axles broke more frequently when 

 crank axles were in use. The fractures he has seen appeared to be the 

 work of time. He has seen nothing in the fractures to induce him to be- 

 lieve they were the result of a change of structure. Considers one ton per 

 foot in length as the greatest weight that comes on railways ; is opposed to 

 increasing the weight of engines. Thinks the plan of having wrought iron 

 box girders a very sound one. They have been long in use fur steam 

 engines ; prefers them in moderate spans to cast iron. M'ould never em- 

 ploy a flat girder unless compelled to do so. The effect of the vibration of 

 trains, however slight, is ultimately to separate the parts, while in an arch 

 the parts are always clinging faster together ; if a general rule is to be 

 adopted, let it be in favour of the arch. 



Charles H. Wild, Esq., Civil Engineer. — In testing the compound girders 

 for the bridge over the Ombrone, an initial strain of 5 tons per square 

 inch of section was put upon the wrought iron ties; by the adjusting 

 pieces, any amount of initial strain can be put on the ties. By that means 



the beam can be cambered 



........ v«.. u.. ^.-,1 uii luc Ilea. By that means 



If, in compound girders, the ties are applied 



in a neutral state, they are of very little practical use. The lies have au 

 initial strain put upon them, but does not believe that any change will 

 take place in the ties to require re-adjustment. If the strain put upon 

 them is far within the limits of elasticity, they will retain that strain. If 

 an extension of the ties were likely to take place, this sort of bridge 

 should be given up. The ties being strong enough to allow for extra weight 

 to come upon them, will never require to be adjusted after being put up. 

 The bridge over the Arno is of compound girders, with the ties lying 

 horizontally along the bottom flange. The ties are in four pieces, and 

 adjusted to the required initial strain by means of gibs and keys at the 

 junctions. This bridge was tested by taking out the dowels connecting 

 the castings, and allowing the whole strain to come on the tension rods. 

 If the ties are put on in a neutral state, the elongation when the weight 

 comes on is so small that the strain would only be about IJ ton per inch; 

 the initial strain can be so adjusted that the tie can take the whole of the 

 tensile strain of the girder, or half, or any proportion. The bridge is 

 96 feet span, and was tested with 40 tons in the centre, trusting entirely 

 to the ties. Has experimented on compound girders with the tie, and 

 when the tie was removed, and has found the stiffness increase with the 

 amount of initial strain put upon the ties. In breaking compound gir- 

 ders, never saw the bolts give way ; no strain can come upon them so long 

 as the joints do not open. Looks upon the dowels and bolts as only 

 useful during the course of erection. Would not like to test the Arno 

 girder without the tie, but thinks the bolts and dowels might be taken 

 away without interfering with the strength. In an experimental girder 

 made for the tie to be adjusted, either horizontally or at an angle, like in 

 the Dee Bridge, it was found to be almost equally efiicacious in three 

 difl'erent positions — viz., when the ends were, 1st, higher than the top 

 flange of the girder ; 2nd, level with the top flange ; 3rd, horizontal. The 

 effect of the difference between the extreme cold of winter and the extreme 

 heat of summer would be to add about half-a-ton per square inch to the 

 existing strain upon the tie. The useful effect of the tie, when the girder 

 is bearing a load, depends on its area, upon the strain upon it per square 

 inch, and upon its depth below the centre of compression ; hence, if the 

 ends of the girder came in so as exactly to counterbalance the extension 

 of the lower part of the girder (a point never reached in practice), if there 

 were an initial strain on the tie, it would still be doing useful work. It is 

 a popular fallacy that there is a disadvantage in having the ends of the 

 ties above the top flange ; the raised ties give greater facility for putting 

 on the initial strain than the horizontal ones. The initial strain is mea- 

 sured by means of an instrument called an extensometer, fixed on to the 

 tie bar, which shows the actual amount the bar is extended ; and having 

 found the rate at which similar iron extends with certain weights per 

 square inch of section, the strain on the bars due to the extension is 

 known. The higher the tie is put the less increase of strain comes on it 

 from the passing load. Has never known the tie slacken. Does not con- 

 sider that a wrought iron bar cast into the bottom flange of a girder is a 

 good method, as no initial strain can be put on it. If, by means of the 

 weight, the bar was extended the y^jVo"' of ''* length, there would be a 

 strain on it of 10 tons per square inch, but has never known them extended 

 beyond :nnni"^ J hence the strain would only be 2\ tons per square inch. 

 The cast iron would break before the tie was doing much work. The 

 above forms of trussed girder are the only ones that have been adopted. 

 Would have the platform of a bridge firmly united to the girders, and 

 sufficiently deep to prevent any twisting in the main girders. The bearers 

 for the platform in the Arno bridge are Sandwich girders. Mr. Stephen- 

 son is using strips of wrought iron, with timber between, for purlins for 

 roofs. They are very stiff'. 



Thomas Cubitt, fsy.. Builder.— Has found variations in the same descrip- 

 tion of iron; experiments by difftrent persons do not give corresponding 

 results from similar makes of iron. Does not trust to experiments made 

 on a small scale. The quality of iron is only affected by the hot or cold 

 blast so far as materials unfavourable to the production of good pig iron 

 are present. Care should be exercised in the selection of hot blast iron. 

 Makes girders strong enough to bear three times the greatest load that 

 could come upon them; these girders are for buildings. Proves girders 

 by the hydraulic press; proves them to double ihe greatest load that could 

 come upon them, or two-thirds the breaking weight. From the liability of 

 girders to internal flaws, would rather prove a girder nearly to the extent 

 that would break it than not prove it at all. In buildings, the weight is 

 more frequently put on the botiom than on the top flange, but has never 

 thought it of importance to apply the proof weight to the bottom flange. 

 The deflection of a girder depends on the shape, section, and qualiiy of 

 metal. In two girders, the length of one being double the lenglh of ihe 

 other, but the section and depth being constant, the longest girder would 

 deflect four times as much as the other. Considers the stiffest iron best for 

 steady weight. Weighted a girder wiih a load equal to two-Uiirds ot its 

 breaking weight, and left it on for 36 hours; tlie deflecljou did not 

 increase, and the permanent set was not uiore than that which had been 

 observed after the Hist application of the weight. Makes the area of the 

 top flange to the bottom (me as 1 lo 3jor4. The bottom flange is equal ia 

 width to about half the greatest depth of the girder ; diminishes the depth 

 of the girder at the ends to about half the depth in tlie centre ; considers 

 it of great importance not lo do anything which woufd lend to inaf^e the 

 girder unsound when cast, or cause un.quaf strains in cooling. Shoes or 

 sockets tend to create flaws, by allowing dirt and sand to accumulate, and 



