634 PROCEEDINGS OF SECTION H. 



■sign ; for example, it was observed that a piece of the lacing 

 buckled, as in compression, at one stage and straightened out 

 tightly afterwards. But the collapse occurred with suddenness 

 in each case. Inferences may, however, be drawn, as to the order 

 in which failure occurred in the various parts. 



The columns experimented with were all 48 inches long by 4.8 

 inches square, this transverse dimension being measured outside 

 the four longitudinal L bars placed at the corners. The material 

 was mild steel. That in the longitudinals showed a not very well 

 defined yield point at 38,000 lb. per square inch, and the specimen 

 tested, one inch in length, failed by the L opening out when the 

 stress was 57,100 lb. per square inch. The sectional areas of the 

 L bars were ascertained by calcfalation from the weights of known 

 lengths, taking 4*533 ounces as the weight of a cubic inch. The area 

 of cross section of the L bar which was used in the longitudinals 

 of all the columns and in the flanges at their ends was found in this 

 way to be 0.145 square inch, so that the four longitudinal L bars, 

 which supported the load, had a joint area of 0-58 square inch. 

 The rivets were of diameters roughly proportioned to those that 

 would be used in real girder work of similar design. Those used 

 for attaching the flanges at the ends of the columns were 3-16 in 

 ■diameter. This size was also employed for attaching the lacing 

 members to the longitudinals in columns 5 and 6. The lacing in 

 columns 2, 3, 4, 2A, 3A and 4A was attached with rivets 0T2in. 

 in diameter, and in column No. 1 the diameter was 0*08 in. 



All the lacing pieces were steel L bars. As will be seen from 

 the photographs, the columns numbered 1 to 6 had a single system 

 of lacing on each of their sides, each piece making an angle of about 

 45 degrees with the longitudinals. One rivet at each end formed 

 the attacliment for each piece of lacing to the longitudinals. 

 Naturally those lacing bars which failed in tension tore at the 

 rivet holes ; but those that failed in compression did so, in most 

 cases, by deflecting laterally before damage was done at the riveted 

 ends. The fact that so many lacing L bars failed in compression 

 shows that fiat bars are quite unsuited for lacing columns. The 

 lacing was lightest in No. 1 and heaviest in No. 6, the variation in 

 Nos. 2, 3, 4 and 5 being gradual. 



The photographs taken of column No. 1 after failure show 

 how the lacing was practically all destroyed, being unable to hold the 

 four longitudinals together. Two of the longitudinal bars diagonally 

 opposite each other bulged apart, while the other two became 

 bowed towards each other. The load at collapse was 9150 lb. 



In column No. 2 the lacing failed mainly in compression and 

 the longitudinals became bowed, but not to the same extent as 

 in No. 1. This column failed at 17,320 lb. Inspection of Table A 

 makes it clear that neither 1 nor 2 were sufficiently strongly laced 

 to enable the full strength of the longitudinals to be developed. 

 By strength of longitudinals in this case is meant the strength of 

 those portions of them between lacing points, for, as will be seen. 



