Vibration measurements on small to medium single-span railway bridges

Due to the need for increasing train speeds several existing small to medium span bridges in the track Linz-Wels (Austria) were re-evaluated. In a preliminary numerical calculation, considering conservative values for the dynamic parameters, very high vertical accelerations were computed for some of those structures. An experimental program was thus carried out in order to get a better estimation for the dynamic behaviour of the bridges, concerning mainly the first vertical eigenfrequency and the corresponding viscous damping. The paper reports on the results of this experimental investigation and identifies some areas where further research is necessary. the accelerations on the rails and in the ballast were also carried on, although those results are not presented here. 2 DESCRIPTION OF THE BRIDGES 2.1 Reinforced concrete frames A set of four monolithic reinforced concrete frames with spans varying from 2.5 to 5.0 meters were measured. The corresponding geometrical characteristics are shown in Table 1 and Fig. 1. Bridge 2 is the only one of the ‘box’ type and is composed by two twin boxes laying side by side, one for each direction of railway traffic. The other bridges are of the ‘frame’ type and are composed by only one monolithic structure for both railway tracks. Table 1 – Geometrical characteristics of the reinforced concrete frames according to the structural layout Bridge Span L [m] Width B [m] Thickness dRF [m] Thickness dRR [m] Thickness dS [m] Structural Type 2 2.50 2x4.25 0.18 0.18 0.18 box 4 2.50 9.30 0.30 0.25 0.25 frame 5 5.00 9.0 0.40 0.38 0.40 frame 10 5.00 9.0 0.32 0.32 0.32 frame (a) box type (b) frame type Figure 1 – Structural layout of small span bridges 2.2 Single-span slabs Another set of six bridges with spans varying from 5.75 meters to 23.5 meters have the common characteristic of being composed of single-span simple supported twin slabs, laying side by side, one for each track. The geometric characteristics are summarized in Fig. 2 and in Table 2. The ballast depth has an average of 0.60 m, varying between 0.55 m and 0.65 m depending on the slab thickness, and spans over the entire width of the twin decks. The structural layout of the prestressed concrete decks (Bridges 1, 3, 8 and 12) corresponds to one-span simply supported slab with slightly variable depth. In the case of Bridge 7 the simply supported slab is made of HEB360 steel bars filled with concrete. Bridge 11 is also simply supported and made of reinforced concrete. The support conditions, although defined generally as simple supports, are of two types. In bridges 1, 3, 8 and 12, the bearing supports, two at each extremity of the deck, are made of steel pots filled with rubber material and can be considered free to rotate. For the other structures no specific apparatus have been used and the slab lies directly on the top of the abutment. There is no continuity of the slab over the supports to the abutments, except the one materialized by the superstructure composed of ballast, sleepers and rails. It is also to emphasize the fact that in bridges 1, 3 and 8 the line of supports is not collinear when considering both decks, as illustrated in Figure 2-b, and that the line of supports in Bridge 12 is skew relatively to the axis of the bridge. Table 2 – Geometrical characteristics of the slab-type bridges Bridge Span [m] Width [m] HL [m] H [m] HR [m] α [o] Type 1 23.50 5.14 0.92 1.14 0.91 90o Prestressed 3 19.50 6.49 0.97 1.10 0.84 90o Prestressed 7 9.00 4.52 0.43 0.43 0.43 90o Mixed 8 21.00 4.23 1.05 1.15 1.05 90o Prestressed 11 5.75 4.44 0.40 0.40 0.40 90o Reinforced 12 11.44 4.54 0.7 0.90 0.70 63.9o Prestressed span width α=90° HL H HR (a) Sinle-span slab (TW1 or TW2) (b) Plan view of the twin slabs Figure 2 – Structural layout of generic medium span bridges 3 MEASUREMENT PROCEDURES 3.1 Measurement layout The main concern of the measurements was to identify the first eigenfrequency for which the respective eigenmode is characterized mainly by vertical deflections of the deck. Although for the slab-type bridges this corresponds to the lowest frequency, it would not be the same for the frame-type bridges, as it was already evidenced in the numerical models. In that sense, for the frame structures, the vertical accelerations along the middle span line (Fig. 3-a) and the horizontal acceleration at about middle height of the vertical slabs were measured. For the slab type structures, according to the expected eigenforms and to the symmetry conditions, only one half-side of each deck was instrumented (Fig. 3-b). Although a maximum of eight channels were available only a number of them were used to capture the vertical accelerations of the deck at mid-span and at 1⁄4 of the span (see Fig. 3-b and Table 3). In this case the remaining sensors were used to measure accelerations on the rails, in the ballast and under the twin deck. The response data was acquired using the Bruel & Kjaer PULSE multi analyser platform and recorded for post processing [4]. In addition, the type of train, number of carriages and velocity measured with a speedometer were manually recorded [8].