[1847. 



THE CIVIL ILNGlNliliR AND ARCHITECTS JOURNAL. 



207 



pieces being placed over the joint, also appear to be Injudiciously ar- 

 raugpd, and are not so good as when an increased depth is given to the 

 casting at the joints, as adopted in some otiier bridges. 



The annexed wood-engraving (Sg. 1) is an elevation of a portion of the 

 brolven girder ; part of one of the lengths, which was not broken, is cut 

 short for want of spare. The abutment end of the girder is that portion 

 which laid on the Saltney abutment, and had a bearing of 5 fl. 6 in. on 

 the masonry. There are two fractures — one in the length nearest to the 

 Sallney abulment, and which was of considerable extent, 3ft. Sin. 

 wide at the bottom, the fracture running along the web, just on the lop of 

 the lower flange, and then upwards in a slanting direction on one side and 

 perpendicular (he other side. Fig. 2 is a section of the iron at this frac- 

 ture, svhich shows two bolt holes at the top, made for tising on an eagle 

 ornament, and doublless considerably weakened the girder, as at these 

 bolt holes ihe flange was found to be completely crushed. From the ap- 

 pearance of this fracture, upon the whole, we are inclined to assign it as 

 the part that first broke. The other fracture is nearly iu the centre of the 

 middle length of the girder, and takes a diagonal direction across the gir- 

 der upwards, to the extent of 4 feet horizontally. 



From experiments that have been made since the accident, the deflection 

 of the girders under different loads is from 1 to 2 inches— the greater the 

 velocity of the train the greater is the deflection : this shows that we must 

 not fix the proportions of a girder at three times the breaking weight ; but 

 considerably more must be allowed — it ought to be at the least four, if not 

 five, times. 



The question that suggests itself, from the failing of this bridge, for con- 

 sideration among engineers, is whether a girder, containing the same quan- 

 tity of metal of wrought and cast iron together, 140 square inches in the 

 section, could not be better arranged than the one before us, so as to form 

 the requisites of crossing over a road or river without interfering with the 

 headway below. From the best consideration that we have bestowed upon 

 the subject, we are induced to adopt a girder of the proportions and form 

 shown in the annexed engravings, figs. 4 and 5. The flanges at the joints 



-JG.O- 



Figs. 4 and 5. Elevation and Plan of Propcsed Girder. 



tobe wide, and of the form shown in fis- 3, and the surfaces planed ; the 

 connecting bolls to be of as large a diameter as the melal flange will 

 allow, the lower bolts being at least 2Hnches diameter; particular atten- 

 tion must be paid to (he fixing of these bolts, and the keying of them, 

 to prevent the nuts loosening by vibration. 



/ 2-lG6nd\ 

 According to Hodgkinson's formula* I W = -, I, the breaking 



weight of the Dee Bridge girder is 60 tons— that is supposing the tension 

 bars to be of no service ; whereas, the breaking weight of our proposed 

 girder is 110 tons, and contains four tons less metal than Ihe Dee Bridge 

 trussed girder. The weight of 110 tons is, as near as can be, the strength 

 required for the Dee Bridge span of 98 feet, which will be equal lo 220 

 Ions for a pair of girders : taking a fourth of this weight as the safe strain, 

 ii will give 55 tons. The calculated strain upon the girder at the time of 

 the accident was 54 tons. 



It is the joints of these girders that require the especial atleution of 

 the engineer, as we shall next proceed to show. 



* W weiglit in tons, n the area of the lower fiaoges in square inches, d the total depth 

 of the girder in inches, and I the length in feet in clear ot the bearings. 



Fig. 3. Section of Proposed Girder at Joiat 



