THE MEMPHIS BRIDGE
21
The form of the structure is such that reversals of strains occur inthe web members of both central and intermediate spans and in thechords of the central span. Wherever this occurs the member was pro-portioned to resist the sum of compression and tension on whicheverbasis (tension or compression) there would be the greatest strain persquare inch; and, in addition, the net section was proportioned to resistthe maximum tension and the gross section to resist the maximum com-pression.
The floor beams and girders of the floor system were calculated onthe basis of the moment of inertia, the strain being limited to 10 000pounds per square inch in extreme fibres. In the floor beams the rivetholes in cover plates and flanges were deducted. In the stringers, wherethere are no cover plates, and pains were taken to avoid rivet holes inthe flanges, the gross section was used.
The rivets, all of which are of steel and in drilled or reamed holes,were proportioned on the basis of a bearing strain of 15 000 pounds persquare inch and a shearing strain of 7500 pounds per square inch, andspecial pains were taken to get the double hear in as many rivets aspossible. This was the requirement for shop rivets. In the case of fieldrivets the number was increased one half. In the splices of the chordsof the central span, wiiich is strained both in tension and compression, butin which the faced ends of the several sections were carefully fitted toclose bearings, the number of rivets was based on the sum of the twostrains ; but the increase in the number of field rivets over what wouldhave been required for shop rivets was made 25 instead of 50 per cent.The chords of this span are of uniform section throughout, and all theriveted joints are alike.
The pins were proportioned on the basis of a bearing strain of 18 000pounds per square inch and a bending strain of 20 000 pounds per squareinch in extreme fibre, the diameters of the pins being never made morethan one inch less than the width of the largest eye bar attaching .o them.Special pains were taken in packing the pins to divide the strains from themembers on one side among those of the members on the other side, thegeneral principle being to take the strain in each eye bar by itself, andconsider that one half of it goes into the eye bar on either side leading inthe other direction. With this system of packing the bending strain inextreme fibre is a matter of little importance.
The weight on the rollers of the expansion joint on Pier II is 40 000pounds per lineal foot of roller, or 3333 pounds per lineal inch, the rollersbeing 15 inches in diameter.
DECK SPAN.
The deck span is a single triangular truss divided into six panels,the floor being sustained at the half panel points by vertical posts. Thefloor is kept uniform throughout, the east pair of stringers being rivetedto the end floor beam of the Continuous Superstructure, this floor beambeing directly over the centre of Pier IV. As the length of the floorpanels is the same as in the Continuous Superstructure (27 ft. 2f in.), thelength of this span is 338 ft. 9 in. The only peculiarity about this spanis the fact that the support is taken in niches on the west side of Pier IV;this made it necessary to shorten the end post and to deflect the bottomchord bars upwards. At the west end the full depth of the truss is main-tained, and an expansion bearing on rollers similar to those used overPier II is used. 1
The actual weight of the deck span, including the floor system andexcluding end bearings, etc., is 3600 pounds per lineal foot, this beingsomewhat in excess of the weights estimated in the first calculation. Thestrain sheet given on Plate 55 is calculated on this basis. It will be ob-served that the actual sections are in some instances slightly less than thetheoretical, but the maximum aggregate strains in tension never exceed13 000 pounds per square inch, and those in compression never exceed15 000 pounds per square inch.
MATERIAL.
As the sections of the superstructure were necessarily unusuallyheavy, and the strains from dead load were greatly in excess of those frommoving load, it was thought best to use a slightly higher steel than isnow generally used for lighter structures, and to work this steel withoutpunching, all holes being drilled. A somewhat softer steel was used inthe floor system, lateral connections, and other lighter parts.
The details of the requirements both for steel and manufacture aregiven in the Specifications dated January 4th, 1890, which accompanythis report, marked Appendix L. The principal requirements whichwere to be obtained as the results of tests on samples cut from finishedmaterial were as follows:
Maximum Ultimate Strength, lbs. per sq. in..,Mimimum “ “ “ “ “
“ Elastic Limit, lbs. per sq. in.
“ Percentage of Elongation in 8 in...** " “ Reduction at fracture
High-grade
Medium
Soft
Steel.
Steel.
Steel.
. 78 500
72 500
63 000
. 69 000
64 000
55 000
. 40 000
37 000
30 000
18
22
28
38
44
50
A piece of each sample bar was also required to be bent 180°, and toclose up against itself without showing any crack or flaw on the outsideof the bent portion.
The specifications as originally drawn provided for a preliminary teston a | inch round bar, and allowed the steel to be made by either theopen-hearth or Bessemer process. So much difficulty was experienced,however, in getting a satisfactory Bessemer steel, and the requirementsfor the preliminary test on the round bar were so much reduced as toamount to little, that all steel was required to be made by the open-hearthprocess. These requirements appear in the supplementary specificationsdated May 6th, 1890, which is attached to this report and marked Ap-pendix M.
After the first tests had been made on full-size eye bars it appearedexpedient to adopt for this purpose a steel midway between the highgrade and medium steel of the former specifications, and steel of the fol-lowing requirements, denominated Eye-bar Steel, was prescribed there-after :
i i . Eye-bar Steel
Maximum Ultimate Strength, lbs. per sq. in. 75 006
Minimum “ “ << “ . 66 000
Elastic Limit, lbs. per sq. in..... 38 000
“ Percentage of Elongation in 8 ins. 20
“ “ “ Reduction at fracture.... 40
These requirements were provided for in a supplementary specifica-tion dated January 1st, 1891, which accompanies this report, and ismarked Appendix N.
The results showed that this material was thoroughly satisfactory formost of the purposes for which it was wanted. It was specially so in r the10 in. eye bars which form the tension members of the anchorage andcantilever arms, and of the webs of the central span. The smaller eyebars which suspended the intermediate points did not give quite as satis-factory results. Tests were made of 56 full-size eye bars and the resultsare given in Appendix O attached to this report. An inspection of thislist shows the excellent character of the 10 in. bars.
Owing to the difficulty of getting satisfactory small bars, the sizes ofthe suspenders which support the floor beams at the intermediate pointswere increased from 6 X 1^-g- in. to 7 X 1 T 3 T in., and from 6X 1{ in.to 7 X l T 3 g- in., and a softer steel was accepted than the specifications re-quired. A corresponding change was made in the diagonal eye bars ofthe intermediate span last manufactured, the width of these bars being in-creased from 8 to 9 in., with no other change.
The results of these tests and observations on other work seem to in-