PSI - Issue 5

Ivan Baláž et al. / Procedia Structural Integrity 5 (2017) 1057 – 1064 Ivan Baláž / Structural Integrity Procedia 00 (2017) 000 – 000

1061

5

were very low (Chladný, Baláž, 1993, Table 3). There were six reasons for the very low and zero values of the rating factors: a) The SNP Bridge was designed according to the older codes with lower loading actions. The shear lag phenomenon was not taken into account in its design in 1969; b) The loaded length of the bending moment influence line for the critical section is 303 m + 54 m = 357 m. Czechoslovak standard (ČSN 73 6205: 1986) did not recognize the principle that as the loaded length of unique large span bridges increases the average load per unit length may be decreased. Load models defined in EN 1991-2 also should be used only for the design of road bridges with loaded lengths less than 200 m; c) The load and resistance factor concept applied to the calculation of the dead load internal forces, which were computed as the sum of the extreme values from 23 erection steps, leads to their unrealistic very high factored values; d) Uncertainty of the code in the question of the impact factor for long-span and cable- stayed bridges. In this case (ČSN 73 6205: 1986) prescribed a dynamic calculation. The value δ = 1.1 was taken as a minimum value from the code; e) According to the British code (BS 5400: part 3, 1982) a maximum of 60 % of the bending moment and/or the axial force in the web may be redistributed to the flanges, provided that the assumed stress distribution, after such shedding, is such that the whole of the applied bending moment and axial force is transmitted and equilibrium is maintained. No shedding is permitted by the Czechoslovak code. The following condition had to be satisfied according to (ČSN 73 6205: 1986) fo r each subpanel of the longitudinally stiffened plate panel: The interaction formula (3) does not recognize the plastic redistribution of bending moment stresses at ultimate state and for the critical web subpanel (shaded area of the web bottom subpanel in Fig. 3) leads to very low rating factors. The critical web subpanel has a thickness 12 mm and can be deemed to restrained. The adjacent bottom flange subpanel is, for construction reasons, 36 mm thick and therefore it is not critical. The very low rating factors were the reason that in (Chladný, Baláž, at al., 1991) the proposal for exception allowance relating to standard for load-carrying capacity calculation of SNP Bridge was justified. Based on this exception the following changes in calculation of load-carrying capacities were done: a) in agreement with (Buckland, 1981) the uniformly distributed carriageway loading for the loaded length 357 m was decreased to 1.94 kN/m 2 ; b) for loaded length 357 m the decreased sidewalk loading 1.5 kN/m 2 was taken into account according to (OHBDC, 1983). See also study (Baláž, Ároch, 1982); c) The load factors γ D,i = 1.0 were used in computing the dead load internal forces related to 23 erection steps; d) factor used to approximate the dynamic effects of moving vehicles δ = 1.0 was taken into account, because it was recognized that for long-span bridges the maximum loading occurs with the stationary traffic; e) it was decided that the left side of the verification condition (3) can be greater than 1.0, but must be less than the value 1.375, which corresponds to the condition under the test loading. The loading tests were performed in 1972 (Harasymiv, 1973). The loading test results (direct stresses in the critical section symbolized by 6 black points) relating to the loading test load 73 kN/m placed in the spans 303 m and 54 m long are shown in Fig. 3. The theoretical direct and shear stresses in Fig. 3 were obtained with computer program SEKTOR. The cross-section of the critical section of SNP Bridge used in calculation consisted of 296 points describing the real geometry including all longitudinal stiffeners (Fig. 2 and 3). For the purpose of SNP Bridge the original SEKTOR was modified taking into account non-linear direct stress distribution in the both cross-section flanges (Fig. 3). The direct stress distribution was calculated from the formula. 6. Comparison of theoretical results with loading test of SNP Bridge 1.0 0.6 2 2                  d q  b d R R  c d  b c  R   (3)

I M N y

 

N

  f y

z

 



(4)

A

  z

y



ef

w

z