.v. ' V I j-ri.’ _. n REMOTE STORAGE 723 Memes PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE H23 2' 281? 20!: Blue 10,43 p XCIRC,'DareDueForms_2013 rndd pg 5 .T'pT. .1.'. ,. ,4 ( 1 l \ I , \ k An Analysis of the Delta Mills Bridge A Thesis Submitted to The Faculty of MICHIGAN STATE COLLEGE of AGRICULTURE AND APPLIED SCIENCE by \/ ‘ J -. v . .4" LI L‘ :- Charles BE Gates A Candidate for the Degree of Bachelor of Science June 1943 THESIS llllnr Ill lllll‘l rlll Jill- This thesis is respectfully dedicated to my parents, without whose help and consideration it would never have been written. Part No. II III IV VI VII VIII IX XVII L V.» 9 1' .1 I: . L' A. -‘. v 3 r3 .-"" TABLE OF CONTENTS Title Introduction Check or Stringer Size and Spacing Check of Floor Beams Live Load on the Trusses Dead Loads on the Trusses Dead Load Stresses Live Load Stresses Design of Truss Members Check of Lateral Bars and Spacing Lateral Forces TOp Chord Pins Top Chord Pin Plates Pin Design in the Bottom.Chords Hanger Pin Plates Upper Transverse Connection Plates Diagonal Connection Plates Check of Pedestals and Shoes Conclusion Bibliography General Drawing of Structure . 55; 555 35 5:15.. ,5: - ". ,r‘. ' _; 2 .mm saw“ v: [1. Ti? :‘ . :55“ rrrwvrv'ro r ism Swiss Page 1. 3. 10. 10. ll. 14. 17. 26. 28. 37. 42. 44. 48. 49. 50. 50. 54. “1‘. L'fi Ls' «OUI (Pocket in back) INTRODUCTION The bridge to be analyzed in this report is an old highway bridge of the through truss type. It was put up in 1891, by the R. L. Wheaton.Company of Chicago, Illinois. It is located at Delta Mills, Michigan, which is a small community located about six miles to the Northwest of Lansing. The greater part of the load travelling over the bridge is composed of passenger vehicle traffic, for which the structure was undoubtedly designed. However, since the World War II, an increasing amount of heavier traffic has been using it. Gravel and equipment trucks have been re-routed over this road, and the structure gives evidence of failing. It is a two—Span bridge, and has a span length of 120 feet. The span is divided into eight panels of fifteen feet each, and the height is also fifteen feet. The roadway is twelve feet wide, and the total structure is fourteen feet wide. It is a pin—connected truss, and is supported on rock-masonary piers founded in the Grand River. The bridge has not been adequately protected against the weathering effects, and is rusting and Scaling badly. The object of this thesis is to determine whether—or—not the bridge will safely carry a loading as prescribed by the American Association of State Highway Officials as a H-lO loading. This load- ing is caused by a train of trucks of the 10 ton type or an equiva- lent loading. It will be determined just what loading is allowable if this is not safe. Recommendations will also be made for contin— uing the useful life of the structure, to prevent further failure. \\\\\\\\\\\\\\\\\\\\\t\\.\\\\\\ .bcb\.\o xoox k00c§ £01 . «.Qx QNm. bUOV 5.23: CD 32;. 33.0005 xcoso)‘ tour. NV bcou buxokxtuutou 0m. y .3 .0m _ .3 4 .0m a! .0m +2.1 .0m. 0 C , C G O O C (chums. .55..“ _ _:oko\_ (flown. a w w z 9 0 «w M W: nova m gm u s y a «OIW vstiesz wow OT}. Part I CHECK OF_STRINGER SIZE AND SPACING In this part of the check analysis, it is to be determined whether or not the stringers are of adequate size and spacing to resist the forces set up by the K—10 loading as prescribed by the American Association of State Highway Officials' Specifications. 1. Wheel Load. The stringers are 7 x 3—3/L @ 17.5# per foot sec— tions, and are Spaced 2 feet center-to-center. The magnitude of the wheel load is equal to 8,000# per foot of stringer(20,000# x 0.4), which is reduced by the factor 2/6 because of the distribution to adjacent stringers. The wheel load is than 2/6 x 8,000 or 2670# per foot. 2. Impact. The formula that is used to determine the impact ratio is: Impact ratio = .50 L - 125 where (L) is the value ofdthe length of the stringer, in feet. The impact ratio is never taken less than 30%(AASHO). Substituting the value of 15 feet for (L), the ratio comes out 35.714%. 3. Dead Load. The dead load consists of the weight of the stringer plus the weight of 3 inches of concrete roadway slab. A weight of 50# per foot is arbitrarily assumed as the weight of the stringer, and from this the needed size of stringer will be determined, which will later be corrected for the actual weight. The 3 inches of con— crete is equivalent to a load of 37.5# per square foot(150 x %, assuming that the concrete weighs 150# per cubic feet). This makes the total dead load on the stringer 125# per foot. 4. Moments. The bending moment due to the concentrated wheel load is: B.M. flit—2.x 7.5 2 10,000 Foot—pounds The bending moment due to impact equals 10,000 x 0.35714, or 3571.4 foot-pounds. The bending moment due to the dead load is 125 x'% x 7.52 which equals 3515.625 foot—pounds. The stringer must be designed to resist the total of these three moments, which is 17,087 foot—pounds. 5. Section Modulus. The formula used for determining the modulus is: Total.Bending Moment Allowable stress in Steel (18,000#/in27 17,087 x 12 18,000 Section modulus 11.38 Inches3 . The actual section modulus of the 7 I 17.5 is 11.1 inchesB, which is not sufficient for the assumed conditions. However, by inserting the correct weight of 17.5# per foot in place of the assumed 50# per foot, the following results are obtained: Dead load moment = 2601.56 Foot-pounds Total moment = 16,172.963 Foot—pounds Section modulus = 10.78 Inches3 Since the actual section modulus exceeds the modulus required, the stringer is satisfactory at this point. 6. Web Area. The compression flange is supported laterally by the concrete so that the full unit stress is considered to be developed. The maximum shear developed(live load and impact) is equal to ( 8,000 . 2,000 x 1%— ) x .31. x 1.357 The dead load shear is equal to 694 pounds = (37.5 x 2 + 17.5 ) x 7%. Since the web area necessary is found by dividing the total shear by the allowable shearing stress(ll,000#/in2), the required web area is found to be 0.401 square inches. The available web area is 5-3/8 x 0.345 or 1.854 square inches, which is more than adequate for the loading. 7. Summary. In summary, the stringers are of sufficient size and spacing to withstand the loading of a 10 ton truck or the equivalent Part II CHECK 0F FLOOR.BEAMS The floor beams are built—up members constructed as shown in the sketch. They must be checked for section modulus, web area, and adequacy of connections and Splice plates. [4' T1. s-axfa 4.1“ i G M5 %”Th/ck Web Splice SHi-‘fcner lg . 'I_““‘—‘———————__________ e A 11 3x274 a4./* _1 I. S A 31:31—40 4.91! Conn ecfion Anglos 1. Depth. The minimum depth to be used is equal to l/25 the length of the floor beam or 1/25 x 14 x 12 = 6.72 inches (AASHO). 2. Loading. The maximum concentration on the floor beams at the middle stringer comes frem the 10 ton truck when one 8,000# wheel load is directly over the floor beam and a 2,000# load 14 feet away. The concentration equals 8,000# plus 2,000# x 1/15 or 8, 133.33 pOunds. 3. Bending Moment. The maximum bending moment occurs at this point ‘Ihen similar loads are placed as near it as the axle Spacing of 6 feet and a clearance of 3 feet will permit. The result is a central loading, with only one train of trucks on the span. 633.333”3 6’ _ 8/33.333# r r 3 "Concrcfje 5 lab L .7 Sirlngerrfl L\ v The bending moment at the critical section is the same as if the wheel loads were applied to the floor beams directly, and they need not be brought down through the stringers. The bending moment due to the live load is equal to 32,533.33 foot—pounds(8133.33 x 4). The bending moment due to impact is obtained by using the following formula: _(_2000 _ L.) (1600 +1..) of the floor beam. The ratio is never taken less than 30%(AASHO), Impact ratio =._%_.x ; (L) being the length and equals 34.562%. The impact moment is 32,533.33 x 0.34562 or 11,244.170 foot—pounds. Assuming that the floor beam weighs 200# per foot( to determine the approximate size of floor beam needed ), the other portion of the dead load on the floor beam is due to the 3 inches of concrete road- way. The equivalent dead load is (l%<3___+ 8.75) x15 1. 200 = 893.75 Pounds The bending moment due to the dead load is found as followsz Bending moment = 7(—§—9—3—é—Zi—)x l4 - ( 893.73 ) x g. - 21,896.875 Foot—pounds The floor beam must be designed to resist the total bending moment of 65,674.377 foot—pounds. 4. Section Modulus. The section modulus required to resist the above moment is 43.783 inchesB, which requires a l2x5§ I @ 40.8# per foot section. Since the actual beam is a built up section, the section modulus is calculated according to the formula: Section modulus = i 3 where (I) is the moment of inertia of the cross section, and (c) is the distance from the neutral axis to the extreme fiber of the member in inches. The moment of inertia of the web plate is equal to 1/12 bd3 and is 182.6 inches4. The mom— ent of inertia of the flange angles is obtained by use of the transfer formula with the distance between axes being 8.01 inches, and is equal to 309.802 incheSA. The distance (0) is equal to 9 inches. The section modulus available then equals (182.25 + 309.802) e 9 or 54.672 inchesB, which is greater than the required modulus. The floor beam is thus safely designed at the center. Since the bending moment becomes smaller as the end of the beam is approached, the section may be reduced and still be safely designed. 5. Rivet Spacing. The first check will be made of the rivets in the flange angles. The object is to determine the pitches of 7/8" rivets in a built—up floor beam with the following loadings: a uni- form load of 729.291# per foot, two truck wheel loads of 8,133.33#, and a 36% impact. The wheel loads are spaced 6 feet apart. The value of one rivet in bearing in the 3/8 inch plate is 7875#, (7/8 x 3/8 x 24,000-assuming an allowable stress of 24000#/in2). The distance between lines of rivets is 11.0 inches for the first three feet of the floor beam. The decrease in the dead load shear per panel is 729.291 x i; = 670 Pounds The decrease in the live load shear per panel is 8,133.33 x 1.36 x 2 x 11 14 x 12 = 1450 Pounds The vertical component of the live load is 81133033 X 1.36 x 2 XL;— x 623.7: 14 12 12 - 850 Pounds The gross moment of inertia of the beam at a point 3 feet out from the end of the beam is equal to the moment of inertia of the web plate plus that of the flange angles. The moment of inertia of the web is 1/12 bd3, which equals 95.270 inchesA, and that of the flange is 2( 1.1 + 1.19 x 6.262 ) . 95.466 InchesA, making the total moment of inertia 190.736 inches4. The horizontal shear is equal to 493.2# and is obtained from the formulas Bendingqfloment(tan a + tan b) Shear increment ' Total shear ‘ Distance between Rivet Lines where (a) and (b) are the angles of inclination of the bottom and top flanges with the horizontal. The value of the horizontal shear is Shear =__§hear X Flange Moment of inertia Dist. between Rivet Lines x Gross I (umw—zwjx9i@6 11 x 190.736 493.2 Pounds The resultant shear is equal to the sum of the squares of the com— ponent parts and equals 976 pounds. The pitch then required is found by dividing the allowable bearing by this resultant shear and equals 8 inches. The maximum pitch used is 4% inches for this type of member, and since the actual Spacing is 3% inches, it is more than adequate. The spacing is investigated at two other points, at 4% feet out from the end and at the mid-point, and the results are Distance from Required Actual End of Floor Beam Spacing Spacing 3 Feet 4.5 Inches 3.5 Inches 4 " 4.5 " 4-0 " 7 " 4-5 " 4-5 " The Spacing in the bottom flanges is the same as that in the top flange, and both are sufficient for the loading. 6. Web Splice Plate. The plate is used at the mid-point of the floor beam to transmit the load from one web section to the other. The dimensions and rivet spacings are as shown in the sketch below. /’ 7/9 ic/(ness =52. Rlvcf Diamefer 35” 663/ I4” aw! —€r«9- —e—e—e—e—e—e—e- The gross moment of inertia of the plate is 95.5 inches4 as found by using the formula 1/12 bd3. The moment of inertia of the holes in the plate is equal to 53.5 incheSA. The moment resisted by the web plate is equal to fI/c, where (f) is the allowable stress(16,000#/in2), (I) is the net moment of inertia, and (c) is the distance from the neutral axis to the extreme fiber. Resisting moment = 16-003 2553.5 - 180,000 Inch—pounds The eccentricity is equal to the resisting moment divided by the summation of the vertical and horizontal distances to the rivets, squared. The summation of the vertical distances squared is 224, and of the horizontal distances 14 x 12 or 14. The eccentricity is then 180,000 9 238 or 756. The horizontal component of the stress in the critical rivet is 756 x 6, which equals 4,536 pounds. The vertical component is 756 X 4.5 + 1150207 a 3753 Pounds, and the resultant stress is 5900 pounds, which is the actual stress that the most critically stressed rivet must resist. The allowable stress that the rivet is capable of withstanding is 7,031 pounds, assuming an al- lowable stress of 30,000# per square inch in bearing for double shear. The web Splice is of sufficient design to withstand the forces set up, 7. End Connection. The end connection is with the use of two 3 x 3 x % angle section, 12 inches long, which contain five rivets, Spaced as shown; 2% inches between rivets and 1% inches for edge margin. The check is made to determine the Spacing needed for the floor beam to resist the end reaction, using unit stresses of 11,000 in shear and 24,000 bearing(double shear), and using 7/8 rivets. 9. The maximum end reaction occurs when the truck load is placed as close to the left reaction or left end as is possib1e(one foot). 6/33.333 6/33.333 ¢ECZ.S' 126205 , 262.5 2a.: are: , 262.5 Izmir, I - 2' J. ,a' 1. 2f 2' J- .2 41 a ng 1? T‘ T‘ t l 3 .. Concrc fc S/a b I \‘fl/fl/ To obtain the magnitude of the reaction, take moments about the right end or Left reaction = 16,480 Pounds 8395.8 x 13 + 262.5(11 +91 + 8395.8 x 7 + 262.5 x3 + 662.6398 14 The effective bearing length of one pair of angles is equal to 2 x 3 or 6 inches.. The required thickness is found by dividing the end reaction by the product of the bearing length and the allowable stress in bearing, and is equal to 0.115 inches. The actual flange or angle thickness is 0.25 inches or more than double the required 0.115 inches. ~The number of rivets required in the web legs of the angles is equal to the end reaction divided by the allowable stress in double shear, which is 5400 pounds, making a total of 3.03 needed in each angle. The actual number present is 5, which is an excess of about 2 rivets per angle. The number of rivets required in the stiffener plate, as determined by bearing in the 3/8 inch plate, is 3000 x 3/8 x '7/8 or 7875# divided into the end reaction. This number is 2.14, and Since 3.0 are used, the plate is safe. 10. This concludes the check analysis of the floor beams, and the struct— ure is so far capable of withstanding the loading with a fairly high factor of safety. Part III LIVE LOAD ON THE TRUSSES The live load for the trusses per lane is 320 pounds per linear foot of lane plus a single concentrated load of 13,000 pounds(AASHO). l. Lanes. The number of lanes is 1.33 or 1 + 3/9. 2. Load per Panel. The live load per panel per truss is composed of the uniform loading and the concentrated load. The uniform load is 160 x 1.33 x 15_ = 3192 Pounds 2 The concentrated load produces a load per panel per truss of 13,000 x 1.33/2 or 8645 pounds. 3. Impact Load. The impact ratio for the chord members is equal to 1 2000 — 120 x 1.33/2 _ 3 3‘(1600 + 1200 x 1.33/2 “ 26'7% Part IV DEAD LOAD ON THE TRUSSES The dead load brought down from each floor beam is equal to ( 693.75 + 22) x 7 = 5012 Pounds 11. The weight of each truss, including bracing, per foot is approxe imately equal to the length in feet plus one-ninth the total load per foot from the floor system. The weight per panel is 5012 J, M x 1.267 8 2 l 0 x 15 + 9 — 2550 Pounds. This makes the total dead load per panel 2550 t 5012 or 7562 pounds. ar _ DEAD LOAD STRESSES In this section, the computations will be made for a few members to illustrate the method of procedure, and the remaining values will be placed in tabular form. 1. Load Distribution. The dead load is assumed to be distributed between the top and bottom chords at a one-to-two ratio. The top chord is considered to take one-third of the total load per panel, and the bottom chord takes two-thirds. The magnitudes are 2520# per panel in the top, and 5040# per panel in the bottom chord. 2. Member (a). The vertical component of the stress in this member is equal to the reaction at the left end, since there is no other vertical member at this point. The reaction is equal to 7/2 x 5040 + 2520 or 26,460 pounds. Since the truss is one that is the same height as panel length, the angle at the supports or panel points is 45 degrees. The stress in (a) is then 26460 x 1.414, which is 37,500 pounds, compression. 3. Member (b). The stress in (b) is equal to the horizontal component of the stress in member (a), because these two members are the only members acting at this point that are capable of taking a horizontal loading. The stress then equals 24,460#, tension. 4. Member (c). Since the only load acting on the member (0) is the panel load of 5040 pounds, the stress in this member is 5040 pounds, tension. 5. For the stresses in the remaining members, refer to the accompanying table of values on the following page. Notezr Member (m) is made up of two cylindrical rods with an unbraced length of 15 feet, and thus it cannot take any compression whatsoever. To take compression, the members (n) are provided and are called counters. These members do not directly take the compress- ion, however. The force that would produce compression in the member (m) produces a tensile stress in the counters, and they are so designed to take all of this stress so that none of it will act on member (m). bush webm aflhhzhuu \Cbo + donate: I no Simona?) «O t... he. that‘s)... MEMBER 13. Table of DEAD LOAD STRESSES MAGNITUDE 0F STRESS 37§500 Pounds 26,460 5,040 45,360 26,700 26,460 13,860 16,000 45,360 56,700 60,480 7,240 5,350 66,600 9,900 96,320 '1 TYPE OF STRESS Compression Tension Tension Compression Tension Tension Compression Tension Tension Compression Compression Compression Tension Tension Compression Compression Part VI LIVE LOAD STRESSES The uniform live load of 3192 pounds per panel is treated the same as the dead load in calculating the stresses caused by it. One-third or 1064 pounds per panel is taken as acting on the top chord, and two-thirds or 2128 pounds on the bottom chord. The stresses caused by the moving live load of 8645 pounds will be determined by the use of load influence lines. 1. Uniform Live Load. Since the method of analysis of the stresses produced by this load is the same as for the dead load, it will not be necessary to repeat the procedure. The values of the stresses can be found in the table on the following page. 2. Moving Live Load. The method of influence load lines consists of placing the load of unity at various panel points and determining the stress resulting from this position in the member being invest- igated. The stress is then multiplied by the actual value of the moving load, 8645 pounds in our case, to determine the actual maximumvstress. The influence line for member (a) is a sloping line going from zero at the right end to 12,224 pounds at the left end, assuming that load enters from the right. This procedure is continued for each member in the truss and points of maximum stress and reversal of stress are noted. The maximum stress is that stress that must be used to determine the design stresses. The influence diagrams for the various members are as indicated in the following sketches and the stresses are indicated in the table. hash NOLD harshness»- X28. 1.. write: I .. hoarse-«mo: a a. > .3. 2» -n «.3» 43. or“ r on '\ I * u ‘3’. M To «4 . .3 , w s, a O a a o . . 1 3 q r \I , 1 a - r 7 o r i \CV 12*» is?» 1.3.3» . .1..qu «O h... E0. \{hxotfl . \hl , A .Dl . I” . . . _ y l . . n; _. a m h . . 1 n r 4. . v . .t f z . . . H. . N m It a, . . s . d r ., 7.. n . , . ., . s. ' ; y m .. a . 1 . a . . . : v . t t. .2244. MEMBER l5. f1:abl§_ of LIVE LOAD STRESSES MAGNITUDE OF STRESS 15,800 Pounds 11,172 2,128 19,152 11,280 11,172 5,852 6,780 19,152 23,942 24,536 22,292 841 33,850 4,308 48,924 I! TYPE OF STRESS Compression Tension Tension Compression Tension Tension Compression Tension Tension Compression Compression Compression Tension Tension Compression Compression 57/?56655 DUE 7'0 Coggggvrne 7351:.» .f , 1.. .s. 4" 1' ,e .1; .. die-«69 d J, A T"""‘ ‘" ’ " " " ‘ " . 7 8@/5=/zo’ fiSCO/C — /;’= 20, 0007* 7 1 1 1 8645' @645 1 MGmbe/vs 1 x /’ 1 b97cfl // 1 \\fMemberC M- —— \ M // Men? tier 0 /2,224- 43988 “Mb” d '\ 1 M ; Mid“ ; \ Member D.“ Member 4 §\<¢/” \ I .r /6,z 'x - L 5' W113 A. fim\:\\ ‘44.! £465,162? 5..9# QT: Al/ Mcm hers (TR us: TRUIt+ F400 at 34.401, e- ~-~~h~. 'fi $sz rm” A «A SrRr/vssns 7r3$xgjmm# Spaced Z ’ cac GENERAL DESIGN OF 0:; TA SCALE - Au. V’s-w: M445, M/CH/CAAL7-Hflougfi 7kg; IINcH 2 if??? SCALE /”=5’ JUNE, l743 Char lea B. Gafed‘ ‘-J:-§-»'-f«3(gfi ‘Ipl - w 1..— ‘CT'T'IV I ' rv'l'lV '3erva _.“. ‘ s 7 Fri .., n . ' 15'4““ “‘7' .I: out-wan“... . F‘bT TATE UNIVERSITI LIBRM PIFS III II HI" I III II 93 0306 m .\ ‘70—..- I I ? z j T I I 'I I 5 I . . t- 1' ‘— . ‘ ;.' I I - ‘ \ _' , , ‘, .' _ )A ' — - ‘ i ‘ a . I i . ". ‘ . :