PERIODICA POLYTECHNICA SER. CiVIL ENG. VOL. 35, NOS. 3-4, PP. 265-293 {1991}
LOAD TESTS ON THE NEW RAILWAY BRIDGES AT TUNYOGMATOLCS AND BANREVE
A. SZITTNER, M. K,,(LL6 and 1. KORONDI
Department of Steel Structures Technical University, H-1521 Budapest
Received: January 20, 1992.
Abstract
In this paper, the results of load tests performed on two railway bridges opened to traffic recently are demonstrated. In the course of static measurement, the deflection of the bridge and the stresses in selected points were measured. Dynamic measurements com prised natural frequencies, dynamic factor and dynamic excess. The measurement results obtained were compared with the calculated deflections and stresses. Comparison of model test results for the Simontornya bridge in the design stage and measurement results for the structurally similar Banreve bridge is of special interest. K eywo7'ds: static, dynamic. load test, railway bridge.
The Department of Steel Structures at the Technical University of Bu dapest performed the examination of the above bridges in Hungary prior to opening them to traffic.
The Railway Bridge over the River Szamos at Tunyogrnatolcs
The bridge is a continuous, welded truss girder open above with cross girder distances of 5750 mm and a span layout of 4x46 m. The structure was built with box-girder profiles and HSFG field joints (Fig. 1). The floor structure was constructed with two stringers at a distance of 1500 mm from each other and a continuous flange plate above, which is connected also to the flange-plate of the cross girder. The upper flange is an orthotropic plate. The floor structure works together with the bottom chord of the main girder through the joint at the cross-girders and through the built in brake structures. This contribution was not taken into consideration regarding the main girder, but the floor girders were dimensioned for the overloading from the contribution effect. The tie-plates (Geo plates) were bonded to the flange of the stringers. 266 A. SZITTNER et al.
The bridge was designed by the planning bureau UVATERV, manu factured and assembled by the factory GANZ-MAVAG. The field erecting in the axis of the bridge took place on the embankments, and then the whole bridge was pulled longitudinally into its ultimate place. The task of the Department was: examination of the structural behaviour during the erection, completing the static and the dynamic load test.
Fig. 1, Main girder of the bridge at Tunyogmatolcs, and the locomotives used for the load test.
Examination of the Structural Behaviour during Erection
In the course of the pull-in operation" the most unfavourable situation from force theory point of view was developed when the front part of the bridge was functioning as a cantilever in 7 frame-joint length. This state was controlled by disassembling the supporting piles before starting the pull in operation. Before the disassembling of piles, measuring stations were developed in the middle of the bottom and upper chords at the support for strain-gage measurement of Pfender-type. After disassembling the piles, the following average stresses were measured: BRIDGES AT TUNYOGMATOLCS AND BANREVE 267
13579no~~~~~~v~~~~~~~~~~ ~~~~~m~ 012345678 91011121314151617181920212223242526272829303132 ~~5750 ~I. 8·5750 + 8·5750 .1_ 8·5750 '1 ~~ Span 11. 4. 46 Cl 171 Span Ill. Span IV. ..j ccfE Q~ l£.~~tfl @N ~~Qj~11~1OCX)~~~ __ I ~~~ 1F-6~'2¥00~k¥N~16~'2?00~k~N*I=====I====~======~======9ICD M62 M62 I 1#i#i#i#:! @N IF=====~i~~I~~i====I==~======~I@ @N I#HW~ @N F======*======~I~~~~~~I===*!==~I® I I (Z)N F=====~======~======!F~~~~~~I® I @N '"lli-#t#H# ~t-ffi-#i! @N F=----~~----~~IF===~I====FI uy~m~~~=pI======~IQD 1#i##t#t I i#j#'1#iit F====1==~1~~=1~~1======91~~~~1@ I @N F=nH#~~I~~~IF#H#~~~~~IF=~,r===F======9I@ F=====~~~~'~#Hft~~!F~~~~~==FI ==~==~I@ I I N F======*======~=#H#~~%H#==~I~==~#Htt~~I® oof- Fehergyarmat TRAIN INFLUENCE-LINE M62 -I> M62 The train-set consisting of two M 62 locos ~'-mitt slops on the cross-girders with its front-axle FI~I~I ~!~'~~i~~!~!~I~~====~~======F======~IQID o 1 2 34 5 6 7 8 9 10 etc. cross-girder M62~-~ 1~1~!~i~~i~!~i~!~======7=~======~F======9I~ 1 11 21 ~ 415161 7181 etc. cross-tie 1 locomotive M62 slops on the selected cross-tie with its front-axle
Fig. 2. Load positions applied with the load test. 268 A. SZITTNER et al.
Table 1 Railway Bridge at Tunyogmatolcs Comparison between the measured (M) and calculated (C) stresses in the course of static load test I\umber Loaded Span Support Span Support IV m-IV III II-III of load span 7-9 8-10 15-17 16-18 23-25 32-34 position Total load A A A A A A 232.4cm2 297.6cm2 166cm2 21.5.2cm 2 215.2cm 2 193.3cm2 1 . Span I -0.042 0.037 -0.126 0.103 0.195 -0.623 2.: C -0.0427 0.037 -0.126 0.103 0.19.5 -0.623 3 0.211 -0.181 0.608 -0.500 -0.946 3.028 C 0.211 -0.181 0.608 -0.500 -0.946 3.028 Ni 0.06 0.92 0.43 -0.69 -0.19 3.16 M 0.08 -0.36 0.43 -0.72 -0.41 3.23 6 Span III -1.029 0.908 -3.089 2.542 4.244 3.028 C -1.029 0.908 -3.089 2.542 4.244 3.028 2.: M -0.53 0.92 -2.10 2.24 2.83 2.07 NI -0.50 0.96 -2.18 2.30 3.00 2.35 8 Span IV .5.139 -4.017 -1.681 3.083 -7.483 -0.623 C 5.139 -4.017 -1.681 3.083 -7.483 -0.623 2.: M 3.32 4.01 -2.41 2.70 -0.84 -0.37 M 3.30 3.78 -2.47 2. -0.91 -0.25 11 -0.042 0.037 -0.126 0.103 0.195 -0.623 Span II -1.029 0.908 -3.089 2.542 4.244 3.028 C -1.071 0.945 -3.215 2.645 4.439 2.40.5 2.: M -0.52 1.30 -2.14 2.69 2.91 1.80 M -0.50 1.11 -2.27 2.67 2.88 1.78 12 Span II 0.211 -0.181 0.608 -0.500 -0.946 3.028 Span IV 5.139 -4.017 -1.681 3.083 -1.483 -0.623 2.: C .5.350 -4.198 -1.073 2.583 -2.429 2.405 M 3.30 4.16 -1.95 2.00 -1.18 1.57 M 3.32 4.017 2.02 -1.29 0.171 14 Span I -0.0142 0.037 -0.126 0.103 0.195 -0.623 Span II 0.191 -0.170 0.573 -0.471 -0.91 2.848 2.: C 0.149 -0.133 0.447 -0.368 -0.696 2.225 M 0.04 0.27 0.34 -0.41 -0.25 2.01 M 0.04 0.15 0.28 -0.34 -0.18 2.00 15 Span II 0.211 -0.181 0.608 -0.500 -0.946 3.028 Span m -1.029 0.908 -3.089 2,542 4.244 3.028 2.: C -0.818 0.727 -2.481 -2.042 3.289 6.056 M -0.40 0.98 -1.61 1.88 2.29 .5.47 M -0.35 -0.89 -1.74 2.03 2.47 5.47 16 Span m 1.094 0.967 -3.283 -2.702 4.345 2.848 Span IV 5.139 -4.017 -1.681 3.083 -1.483 -0.623 2.: C 4.045 -3.050 -4.967 5.785 2.862 2.225 M 2.79 -2.99 -4.85 4.73 2.09 1.79 M 2.70 -2.89 5.06 4.6.5 2.17 2.53 BRIDGES AT TUSYOG.lfATOLCS AND B.4NREVE 269
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 jSlVVVVVSI"){VN\l\I\lV'){V\N\/\I\l\j\/\7 1 3 5 7 9 11 13 15 17 19 21 23 25 2729 31 33 35 o 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 A _4 C?lculated train influence- 1-M62 2-M62 2+3 IBar 15-17j 1?Ii~18rr? -Ime for 2-M62 1+2+3+4 ---/-2- 1~4 1213' 9~6 - 3 ...... --., , ..... " - 2Jl;I, 3 - 8"+'" 7 N /' '- 4 --- ,""- " 11 14 E -2 /.: - '""''' '---- ..... " _...... ::;.--..::::--- ". 1+2+3+4 1$4 II /,' fi;.$.o •• ' .-~~"""_-- '<:----0:.<...... ~:,~ 1+4 ' Z -1 ' .-- •• -2 2, 3 ~ O~~~~~~~~~~~~-L~~~==~~~ 1" M62 Calculated train ------··5i·~·54 A I 11 I I I -I> _ ....I.l0fluence-line for 2-M62 52,Lj.l'53 -~ ~~~~~~~~~"~"'+.-+. -::;---~-~---*- ~~~~-,--~I>- N 1 IBar 7-91 66 67 68 69 § 2 1 Z 51 ~54 'l~T631 i 59,$,,56 ~ 3 52 53 61 64 58 57 4 21 *2~Bar 16-"ij29 +0 26 5 22', 23 28 '4!.27 O~~~~~-r-r'-'-~~",,~~~~~~~~~~~~'~
IBar 23-251 116 117:n8 119 101~104ilE~fGlil09~,107 102~03 JP - 108'T'110 111 114
Fig- 3, Train influence lines of the stresses :11easured in the chords 270 A. SZITTNER et al.
- on the upper chord bar No. 50-52, with a nearly uniform stress dis 2 tribution 0"=14.32 kN / cm , - on the bottom chord bar No. 51-53, 2 on the bottom exterior fibre: 0"=-8.57 kN/cm , 2 on the upper exterior fibre: 0"=-5.25 kN / cm , - in the same section of the stringer, 2 on the bottom flange plate: 0"=-5.99 kN/cm , 2 on the upper flange plate: 0"=-0.38 kN / cm . The deflection of the cantilever-ends were 236 mm and 214 mm, respec tively. With the satisfying results, the pulling-in operation took place ac cording to the plans.
Static load test
The static load tests covered shape measurement, measurement of deformations and stress measurements. The load test took place using 2 X 2 pieces of M62 locomotives having a mass of 120 t each. As a first step, the locomotives were positioned to the most unfavourable place with respect to the mains, later train influ~ ence lines were recorded first for the mains, using 2 locomotives (the front axle positioned above each cross-girder, successively), then for the stringer, using one locomotive (front axle placed above each tie, successively), see Fig. 2. By the shape measurement, the shape of the bridge was recorded prior to and after the static load test. The maximum value of residual deformation of the main girders prov ed to be 3.3 mm and 2.3 mm, respectively. The deflection measured under the influence of train load was about 80 % of that obtained by calculation. Stress measurement was performed in the vicinity of the positive and negative maxima on the bottom and upper chords of the main girder, as well as on the floor structures by using Kyowa-type foil resistance strain gages of 5 mm base bonded on with an adhesive of cyanoacrylate. The measuring system was composed of Hottinger-type units with a connection to the measuring stations ensured by a personal computer, which also pro vided the control of switch-over to the measuring points and that of data acquisition process. The measured and calculated results related to the most unfavourable static load positions are shown in Table 1. Measured stresses are smaller then the calculated ones with the exception of those in the bottom chord BRIDGES AT TUNYOGMATOLCS AND BANREVE 271 is greater with the bottom chords (CC 30 %). With the bar 15-17, there is a considerable deviation at one of the load positions, however, this deviation will occur with the train influence line to be dealt with later on. The results obtained with the train influence lines are in full harmony with those described above (Fig. 3). Thus, e.g. on the bottom chord 7-9, the calculated and measured train influence lines are conform with a train of two locomotives taken as a basis, however, the measured stress is lower by about 30 % than the calculated one. The forms of the measured and calculated train influence lines of stresses in the bar 15-17 do not agree with each other. According to the measurements, the maximal stresses will occur when the train is moving within the span No.l, while the train influence line processed on the basis of the calculated bar force influence line will yield the maximal value within span No.2. The measured stress influence lines of the upper chords No.16-18, 32-34 and 8-10 are in good agreement with the calculated ones'with both the one and the two locomotive train. The comparison of the calculated and measured stresses on the bottom chord 23-25, or on the stringer contained in this section can be seen in Fig. 3.
Dynamic Measurements
In the course of dynamic measurements, the following values were mea sured: - natural frequencies of the unloaded bridge in the horizontal and the vertical plane, - motion of tie-plates, - compression of plastic layers under tie-plates, - midspan deflections, - midspan horizontal vibration. The measured quantities were recorded by instrumentation tape-recorder and an 8 channel strip chart recorder at the same time. Recorded data were processed later by computer, via AID conversion. Digital signals were frequency-analyzed, scaled, filtered, according to the needs of processing requirements. - In Fig. 4, the power spectrum of free oscillations, i.e. natural frequen cies in vertical and horizontal directions, - in Fig. 5 the compression diagrams of the plastic-mortar under tie plates and that of the full rail-plate, - in Fig. 6 the horizontal transverse oscillations of the fourth and third spans, excited by moving load, - in Fig. 7 the deflection recorded in the fourth span, 272 it SZITTNER er al.
("'l CTl. c) N
o
b)
o 10 Hz Fig. 4. Power spertrum of natural frequency Illeasured ill a) horizontal, and b) vertical directions
- in Fig. 8 the stresses measured by the two gages located on the bottom chord 7-9 in the case of quick train motion can be seen.
Bridge over the River Saj6 at Banreve
The bridge is a simple-supported, welded, trussed closed bridge with a span of 66 m and an inclined portal frame, as well as a clearance gage for electric traction. The horizontal distance of the joints is 8250 mm which is reduced by setting in two intermediate cross girders. The distance between the stringers is 1500 mm. The stringers and cross girders are of the same height, their upper flange plates are connected in a single, cross-shaped plate of 40 mm in thickness. The ties are fastened to the stringers in both longitudinal and transversal directions. BRIDGES AT TUNYOGMATOLCS AND BANREVE 273
11. E 40 :J..
20 0)
s Ii>
-20
-40 " Pi. E 80 - b) :J..
60 -
40 !-
20 -
!- al"''\. ~l If ~l 1 I~ ) 10 2~ \30 V40 s - lr-' -2 0-,
Fig. 5. a) Compression of plastic mortar under the tie plates; b) total compression of the whole tie plates
The profiles of the welded, trussed main girder bars are of I-shape, and the inner flange plates of the chords at the joints are transduced to the flange plates of the truss members in an arched form (Fig. 9). The joints are riveted with the exception of the upper wind brace which is bolted. The main girders and the floor beams are connected together rigidly. With calculations of the main girder, this contribution was not taken into consideration, however, the floor girders were dimensioned for the addi- 274 .4. SZITTNER et al.
A c) 0.5 0.48
E E
-0.5, I I I I I I I I I 2 4 6 8 10 12 14 5 16
A
05 b)
E E
- 0.5 , 0.51
I I ! I I I I I 2 4 6 8 10 12 14 5 16 Fig. 6. Horizontal, lateral oscillation a) in the fourth span, and b) in the third span
tionalload resulting from the contribution between the structural elements mentioned above. The bridge was mounted fully on the embankment, and afterwards it was pulled in - first in longitudinal, then in transverse direction - into the place of the former bridge. The bridge was designed by the planning bureau UVATERV, while it was implemented by the MAv (Hungarian State Railways). The investigated bridge structure is similar to that of the new bridge designed for service over the river Si6 at Simontornya, the plans of which drafted also by the UVATERV were checked by the Department of Steel Structures earlier. The process of checking took place by calculation, and - for developing the joints of the truss and determination their static be haviour - by a so-called 'hybrid' method with a photo elastic model, in BRIDGES AT TUNYOGMATOLCS AND BANREVE 275
A. E 20 E
-5 A. E 20" E
15
10
5
0 J " Fig. 7. Deflection III the middle of the fourth span during the running forward and reverse which the stress distribution was determined by resistance strain gages on the curved flange plates and on the web in the places of stress concentration pointed out by photoelastic methods (Fig. 10). Using these results, the measuring points were selected prior to the load-test of the bridge, and afterwards the measurement results obtained in the course of loadtests could be compared with those obtained during model experiments. 276 A. SZITTNER " al.
3
2 Detecting element No. 53
16 18 s
-1., lA
3
2 Detecting element No.58
-1 V Fig. 8. Stresses measured in the bottom chord bar No. 7-9 during fast running
Due to the extremely short stop peace-time static load tests and dy namic tests were performed. For the load tests, 3 M62-type locomotives coupled to each other were stopped at the most unfavourable position for joint No.8.(the main point of the upper chord bar No.7-7'). During the record of the train influence line, also three coupled locomotives were moved, while the dynamic load test took place with 3 coupled locomotives, and then with a single locomotive. BRIDGES AT TUNYOGMATOLCS AND B.4NREVE 277
Fig. g. One of the joints of the raiiway bridge at Bann§ve
Static Load test
In the course of the load test, the midspan deflection and the stresses or strains, respectively, were measured in the characteristic points. In each case - taking the significant subsidiary stresses into consideration - the nor mal and bending components of the stresses acting on the cross-section were calculated from the measurement results, and then the measured stresses were compared - where it was possible - with tfte stresses calculated on the basis of the load applied. The resistance strain gages were positioned in the characteristic cross sections of the main girders and the stringers, on the arched flange plate at joint No.8., on the sections of the tangents subtending an angle of 0°, 20°, 40° and 60°, with respect to the results of our earlier model experiments. On the sections with tangent-angle 20° and 40°, a rosette was bonded on at a distance of about 20 mm from the web-ta-flange fillet weld on both sides of the web. To determine the effect of the deflection caused by horizontal forces, strain gages were positioned on the edge of each flange plate in three sections of the end cross-girder No.O. (Fig. 11). The measuring system applied here was the same as that used with the bridge at Tunyogmatolcs, but for dynamic measurements, a seven-channel FM tape-recorder of type TEAC was used. 278 A. SZITTNER et al.
55.7
"T' I
! Upper surface ~ ~ \ l' \ : (Unfolded) i : i I r~r==='=t -1~~1\1 I y I I I I I ..L i , 1 ...... : '! : @ @~CD.:;;:-1_....., ~ I I I I I I I I I + : Bottom surface"f -; : : : : -I- + I -~-===-=--:'-:::f.:-~-=--=-:=r.:-::--=-t Layout of resistance stroin f= I I I I I -gouges with anarched : + f + ! flange-plate : : I I : ..!.. I I I ...L ...L 1 ~ ______~6~1.~9 ______~
Upper side Upper side I I 0 I I I -32 I I I I I I I I I I I I J I I I I 12 I 34 I I I I ! I I I I I I I I I I I I I I I 123 I I -98 I _.-+.-- 12L ._.-j.- _.-j-.-- t-- G) @ CD G) Underside
I I I I I I I I I 1 I : -17 I I I I I : I 59 ,. I : '181 I I I _.-i==-~=--:-.:t::::.'C--:---..io::::.=.~ ---+.- ~77 '-179 -215 Longitudinal Lateral axial stresses on the actual structure (MPa)
Fig. 10. Summary of the results gained by model experiments for testing one of the joints of the planned railway bridge at Simontornya having a layout similar to that of the bridge at Banreve BRIDGES AT TUNYOGMATOLCS AND BANREVE 279
~~~~~~~~~~~~~~~~~~~~ "0 6'2425 21 04' 2100 2' 2100 8250/2 ..J Layout of the measuring-stations r----'=:.:::.!..!"----,VlU (Eastern main girder) --j\-- -_._--+_._._._. 1/ '~ IVIII \ I '~ i '\ I \:, ~ ! // ) !l(/~ IX XI /(VII 1 60 11140 I ' -'-+'-'Tz70 ·-~-·I---t·-·- IX ~ J VI! 8250/2 @ , '. 105 I 105 I 105 ..; Layout of the resistance strain-gauges 240 ..J::.. 24cJ <>t"-- 1 on the eastern main girder 8 5;>\53 47 t;2 u. 373839 323334 6 5-P959 54 5556 31 35 1 3 Inner -46. t. o81,41U" ,84 36._._.48 'TrRosette:r-. I 2'I I I Outer 'd (W) 82 83 U" 1, ,4 side (E) SI e . . _ 200 400 600 Upper r W 18 ,9 111-111 IV-IV V-V 9 7 chords 77 7927 29 r 78 28 1-1 0° f,130 0 1':<:60 0 . VI-VI VII-VII VIII-VIII IX-IX X-X 11-11 1\" ° 11-;" ooi~e'Y i \Y.. 13(23) Layout of the resistance strain-gauges 16(26~4(2.4) 12(d1f21 ) on the eastern stringer XII 15(25) ~~======;;;;;==~X;bl =::;:;;:;====:;:==~::S: Rosettes I I I ® 32C!fl~~ 717?73 616263 ®XII Layout of the resistance strain-gauges on the end , I ~ 6~ 600750 750 600 cross-girder towards the Ozd-side 7668--70 -r+---t'-ofo-r>f---+:t- 74 75 w XIII i XIV iXVi ! 111114 106109 101104 XI-XI XII-XII
IPlfXIII X V xvCl 112III 113 107108102 103 XIII-XIII XIV-XIV XV-XV
Fig. 11, Positioning of the measuring points during the load test of the bridge at Banreve. 280 A. SZITTNER et al.
The up-to-date structure used with the bridge at Banreve cannot be considered a simple one from measurement technique point of view, either. While with the truss girders of a traditional layout, the bars are in a uniaxial state of stress, the statical behaviour of this bridge owing to the sometimes drastic changes of the profiles, the arched flange plates, the relatively rigid joints, the bending of the chords caused by the intermediate cross girders, the considerable thickness (40 mm) at some places of the web plates and flange plates, the contribution of stringers, the rigid welded joints of the structure, etc. is only a faint reminder of the traditional force condition of the theoretical. (jointed) trussed girders. The effect of the disturbing moment, the lateral bending of the arched flange plates, as well as that of the contribution between the main girder and the stringers seems almost follo\ved exact calculation.
! 1 . x-x [ ,", ~ 500-40 I >< I ' <.0 ~I <.0 1 d~ H \~i500-40 f
v-v
Fig. 12. Distribution of the stresses measured on the arched flange plate at joint No. 8
The most important experiences gained from the evaluation of the stresses, or internal forces, respectively, measured at the most unfavourable load-position, and those calculated for checking them were the following (Table 2): a) The normal stress calculated from the stresses measured on the edges of the flange-plates, in the middle of bars in the upper chord falls Table 2
Measured Measured Load Gage aauer CTN Nmell., Neal e CTM,upper Mmea3 Meal c element cross- position position fiangeElate NLmm2 NLmm2 kN kN NLmm2 kNm kNm upper -51.5 1- Load I -42.14 -2:318 9.:36 104.4 tu ii~ bottom -:32.78 :u o..""::j ;;: tJ 0.. ... I -25:3:3 Cl ;::l 0 upper -51.15 t'J {j Load 11 -42.26 -2269 9.89 110.:3 ;;: bottom -31.38 ,.'" "l upper 10.2 "l Load I 24.;3 1628 14.1 306.9 ~ >< bottom 38.4 '"< T 2458 0 >< upper 10.4 Cl - Load II 24.1 1616 13.7 298.2 ;.. bottom 37.8 "l "...0 upper 12.0 C) Load I 19.5 1306 7.5 163.2 '" 00 >< bottom 27.0 ;.. I I 2458 '" >< upper 12.0 ~ "E Load II 19.42 1:W1 7.42 161.5 tu 0 bottom 26.85 ;... _
~ .....00 IV 00 IV
Tahle 2
~1,'a';lIn'd fli,'aslIl't'd Load ( ~ ilP;P OH l1f'" aN N I1I" tl"j NClllf- (Tf\t,tlJll't;1" Ivl rllca .'I MCPr 12.07 '- v ':.l I..., Load I1 21.1 :Wi 9.0:l 27.0tl !:n >ll 8 ho t.t.o 111 :10.12 ':-::";J1-.------:: ';\ IIppN '27.12 :F. ?- '~ Load 12.8:1 :IS2 14.2H -Ha.5 hol.lo\1I -1.-17 In ( -la.OO) :=j ~ 11 ppet' 2G.!)S '" 5' Loa.d II 12)l(j :18:1 Itl.12 -H2A "'l tn bott.om ~_~~_w,~. ~ __ __ -1.27 ;"
~ BRIDGES AT TUN)'OGMATOLCS AND BANREVE 283
Train influence-line at the measured points of the main girder (stresses) Bonreve Ozd l' 3' 5' 7' 7 5 3 }\/\/\~~ Load: 3·M62 ~ Running direction-I» 51 B 2 0' 2' 4' 6 4 o Joint o 15.5 30.5 45.5 60.5 75.5 90.5 105.5 120.5 130 140 sleeper -50 ------0)
N -40 .... ------• --...... 21~13 E "" ...... - ----...... E-30 ,---:-----" --.-ON VIII-VIII :2-20 _--::::-:::.... ~ dl+0S 11 ,15 -10 ~~~;-.-., 2 ---z.- 7-7 o
Fig. 13. Train influence lines of stresses for some points in characteristic cross-sections of the chords
behind the calculated one by about 10 %. Simultaneously with the normal force there is a considerable bending effect, too. The deviation 284 A. SZITTNER et al.
Train influence -line at the measuring - stations of the stringer Sd.nreve l' 3' 5' 7' 7 5 9 1 Ozd Load: 3 x M62 Running -I> IVVVVVVV\ direction 0' 2' 4' 6' 8 6 4 2 0
15.3 30.5 0°
N 0) ' ~~","'" ...... ~/ E 20 , '" ...... A ...... •• . • , E \ "\ / \ ,", ON ... " ~./ \ I \ ... Z 40, \...... / '---- °74 _20£ 669 61 62 63 0 ,~- ... --... --- ... ------... ------O!> 6~4 b) ...... ·····················ci~··· N b) 6~7 E 20 .§. 68 69 70 z 40
Fig. 14." Train influence lines of stresses for the characteristic points of the stringers
can be attributed to the deviation of the actual dimensions from the nominal ones, the non-uniform stress distribution along the cross section (the stresses arisen in the flange-plate in the cent er line of the web plates are greater than the average stress), the contribution of the upper wind brace with the main girder. b) The bar force calculated on the basis of the stress measured on the bottom chord falls short considerably of the calculated one across each cross-section, which can be explained by the contribution character of the main girder with the stringers, but the axial force calculated on the basis of the stresses measured on the stringers does not cover the shortage of the chord force, the drastic changes in the cross sectional areas of the web plate and flange plates especially those in the vicinity of changes exert a disturbing effect on stress distribution, the non uniform stress distribution in the flange plates, the bottom wind-brace taking also part in the load-bearing of the bottom chord, in addition to that effect of the stringers. c) The axial force and the bending moment calculated on the basis of the stresses measured on the stringers can be only hardly compared - due to the contribution of different loading elements - with the calculated results gained by help of influence lines. BRIDGES AT TUNYOGMATOLCS AND BANREVE 285
d) Similarly, the comparison of the stresses measured and calculated at the end cross girder N 0.0 can be drawn only with much difficulty. e) We tried to measure in detail the stress distribution in the arched plate functioning as an upper flange-plate of the bottom chord-bar No.6-8. The stresses measured at the points with tangents 0°, 20°, 40° and 60°, as well as in the place of the change in the thickness of the flange plate (from 40 mm down to 20 mm) are illustrated in Fig. 12. For the results to be evaluated, it should be noted that a considerable lateral bending will occur in the chords caused by the change in thickness and bending occurring with the arch of R = 300 mm. Since only the longitudinal strains were measured, the stress calculated on the basis of the measured strain can be considered correct only at the edges of the flange plate (free edge - uniaxial state of stress). At the points above the web plate, the strain will increase due to lateral bending. This increase - according to the measurements performed on the model of the bridge at Simontornya - amounts to 20-30 %, due to which the actual stress will be smaller by 20-30 % than that calculated on the basis of the unidirectional measurement. f) From the measurement results obtained on the rosettes positioned on both sides of the web plate at a distance of about 20 mm from the web-to-flange weld in the section of a tangent of 20° and 40°, it is to be seen that in the section of a tangent of 20°, the directions of the principal stresses are nearly horizontal and vertical, in the section of a tangent of 40°, the direction of the principal tensile stress is about -15° , the principal stress in the '20°-section' is greater than that in the '40°-section'. From among the train influence lines recorded with the use of three M 62- type locomotives coupled together, the following will be displayed (Fig. 13): - upper chord bar No.7-7' of the main girder (section VIII-VIII), - bottom chord bars No.6'-8 and 6-8, respectively (sections: I-I, II-II, VI-VI, VII-VII, IX-IX and X-X). It can be seen clearly with each stress train influence line that the bars are subject to a considerable bending in addition to the axial force. In the bottom chord, bending is verified by the eccentricity due to the contribu tion effect of the floor beams as well as the bending effect caused by the intermediate cross girders. The ground of the bending force arisen in the upper chord is unknown. In Fig. 14, the train influence lines of the stresses arisen in the middle of the section 6'-8 [M~;;),:l and those arisen in the cross-section of the sup port at cross girder No. 8 [M~;;-;;:;:)l in the upper and bottom extreme fibres, as well as the influence line of the average stresses (normal components) 286 A. SZITTNER et al. are illustrated. From the comparison drawn between the two influence lines it can be seen that the normal stress at the support is smaller than in the middle of the length, however, due to the larger cross-sectional area at the support the normal forces in the two places are almost identical. The de velopment of the moments reflects well the difference between the influence lines related to the middle of the length and to the support, respectively.
Dynamic Measurements
In vertical direction
I !i> o 6 8 Hz 10 b.
I In horizontal direction
o
Fig. 15. Spectrum range of the vertical and horizontal natural frequencies
In the course of dynamic measurements, the natural frequency, the deflection of the bridge and the changes in stresses were investigated. The results of the dynamic measurements this time, too, were recorded continuously and were processed later by a computer. BRIDGES AT TUNYOGMATOLCS AND BANREVE 287
lA ~:~~
00 10 20 30 40 50 5 60 l»
lA E 30 E 20 10
O)~~~~~----~~----~----~~~~~~~l»
Fig. 16. Deflections by the effect of the loading train passing through at different speeds (downstream side of the bridge, Joint No.S, Load: 3 X M62).
Two trains: 3 coupled and one single M62 locomotive served as dy namic load, passing through the bridge at different speeds. The calculated first vertical bending natural frequency - using the 1 2 simplified formula fo = 5.6 x e- / , where e is equal to the deflection due to the dead load, this case 2.43 cm - is 3.59 Hz. The measured first two vertical natural frequencies proved to be 3.7 and 4.8 Hz (Fig. 15), while the first horizontal bending natural frequency is 2.9 Hz. 288 A. SZITTNER et al.
lA 0.8 v =5 km/h
- 0.4 -0.8 ' ____~I~--~I----_I~--~I----~I----~--~I 10 20 30 40 50 60 5 70
5 10 15 20 25 5
I I 2 4 6 8 10 12 5 14
lA. 2~ !\ V=50km/h o f\ !\ I""v.A fA... I\. 1\ t\ !>eA t\ I!> ~v~ V~\j V4;;; -2 V E E -4 , I I I I I I 2 3 4 5657
Fig. 17. Horizontal motion in lateral direciion of the middle cross-section (Joint No.S.) by the effect of loading trains (3xM62) passing through the bridge at different speeds
The dynamic deflection of the main girder by the effect of 3 locomo tives passing through at speeds of 5, 15, 30, and 50 kmjh is illustrated in Fig. 16. The measured dynamic surplus in succession were: 1.6, 5.0, 4.2 and 4.2 %. BRIDGES AT TUNYOGMATOLCS AND BANREVE 289
The lateral motion of the main girder at the above speed values can be seen in Fig. 17. The maximal values:
v [km/h] 5 15 30 50 lateral movement [mm] 3 locomotives 0.75 2.1 3.1 4.9 1 locomotive 1.1 1.9
v =5 km/h
I 70 5 ~
~ :t ~:~15km/h 010~~~~5~----~1~0-----'1~5----~20~----~25~~s--+~
~~km/h 2 4 6 8 10 12 14 5 I!'>
Fig. 18. Change of the stresses measured in measuring point No.55. positioned on the main girder under the influence of the loading train (3xM62) passing through at different speeds. 290 A. SZITTNER et al.
In Fig. 18, the stress time history, recorded by strain gage No.55 can be seen, as an example of the dynamic stress-measurements. Concluded from the measurement results, the dynamic surplus (%) was the following:
3 locomotives 1 locomotive measuring v(km/h) v(km/h) point 5 15 30 50 5 50 48 1.2 4.8 5.0 4.9 18 3.3 4.4 4.4 4.8 58 3.6 5.1 6.9 7.6 78 4.05 5.4 4.7 5.4 28 2.6 3.9 3.6 5.3 55 3.7 7.1 3.6 6.3 6.2 6.2 43 1.9 4.4 4.4 4.2 2.5 4.3
The DIC-ORE recommendation for the magnitude of the dynamic surplus will be given by relationship:
rP = 0.65 . K/(1 - K K2), where K can be calculated by the formula if v =speed [m/sec] L =span [m] f =natural frequency [Hz]
K = v/(2· L· f).
The DIC-ORE gives also the standard deviation for the formula recom mended, as follows
s = 0.025[1 + 18· K/(1 - K - K2)].
The dynamic surplus occurring with a probability of 95.7 % will be:
rP96 = rP ± 2· 5, which involves 9.4 % for the main girder of the bridge examined at a speed of 50 km/h, a greater value than any obtained by measurement. According to the Railway Bridge Standard of 1951, the dynamic factor for the bridge at Banreve
fL = 1.1 + 15/(L + 21) = 1.27, which provides a value even higher than that contained in the DIC-ORE recommendation. BRIDGES AT TUNYOGMATOLCS AND BANREVE 291
lA a.0 40 ::E 30 V=5km/h 20
10
00
_.' ~ ::~/\f\!\!\I\t\ot .-~~. V~15kmlhI~ o 5 10 15 20 25 s 30
~ V =30km/h I 1'0- , !I> 2d 4 6 8 ID 1~ s
A a.0 ::E I\N\t\/\j\ V=50km/h
:~~ f I I I ~ = , 1 , .~ I !I> 00 1 2 3 4 5 6 7 s 8 Fig. 19. Change of stresses in the stringer in measuring point No.75, under the infiuence of the loading trains (3 X M62) passing through
3 locomotives 1 locomotive measuring v(kmjh) v(kmjh) point 5 15 30 50 5 50 62 2.8 3.6 4.3 5.5 4.8 7.4 72 5.5 5.5 5.5 8.1 8.7 23.0* 75 2.7 5.0 5.0 5.0 2.0 7.1
As for the dynamic investigations related to the stringers, an illus tration is given in Fig. 19 and Fig. 20, where data recorded by a strain gage, positioned in a stringer (midspan, bottom chord) are presented, for 292 A. SZITTNER et al. the cases of different-speed 3 locomotive and 1 locomotive passes. After the evaluation of records of several corresponding measuring elements, the maximal dynamic surplus proved to be: The dynamic surplus indicated by * was measured at a stress level well below the most unfavourable one. ~:i ~ v = 5 kmfh
! I ! 'l~, 50 60 5 70 o 10 20 ~ 30l 2: I
20~ V=50kmfh 10~ ~
Fig. 20. Change of stresses in the stringer in measuring point No.7.s, under the influence of a single loading M62-type locomotive passing through
The DIe-ORE recommendation calculates the dynamic factor for stringers using the formulae:
if; = 0.033· v,
s = 0.066(1 + 0.01 . v). The calculated dynamic surplus, with a probability of 95.7 % and suppos ing 50 km/h speed, will be 27.1 %, which is greater than any value obtained by measu:·ement. BRIDGES AT TUNYOGMATOLCS AND BANREVE 293
From the analysis of the dynamic measurements with respect to fa tigue, the following important conclusions can be reached: In the case of a train consisting of 3 M62 locomotives nearly totalling the full length of the bridge when passing through it - each train-pass involves one maximum (i.e. one fatigue-cycle) for the mains, similarly, each pass involves one maximum for the stringers at the cross-section at the supports (M-max), however, in case of stringer midspan stresses, each group of axles (which consists of 3 axles with the M621ocomotives) causes one maximum (M+max), the stress-amplitudes on the cross-girders are small due to their dense layout.
Address: Dr AntaI SZITTNER Dr Mikl6s KALLO Dr Lasz16 KORONDI Department of Steel Structures Technical University, H-1521 Budapest, Hungary