PSI - Issue 84

Francesco Ballio et al. / Procedia Structural Integrity 84 (2026) 151–158

153

stream-wise profile of the free surface, from which it was possible to determine the backwater rise induced by the obstacle to the flow. A transparent bridge model was installed and was compound of just a rectangular deck with stream-wise and vertical dimensions of 0.17 and 0.06 m, respectively. The lower side of the bridge was 0.15 m from the flume bottom. The bridge deck was a bit narrower than the flume because it was not anchored to the latter’s walls but, instead, it was

Fig. 1. Driftwood accumulation after run 2 described below (side and top views). Note the vertical bars holding the bridge deck from above.

hanging (Fig. 1) on a trolley that could slide over a frame, in turn anchored to the flume banks. The stream-wise movement of the trolley was prevented by a force sensor attached to the supporting frame; the sensor directly furnished the total hydrodynamic force acting onto the bridge. The driftwood material was made of stick-like, branchless natural wood, preliminarily cut in pieces with lengths of 0.20 and 0.15 m. The average diameter of the pieces was equal to 0.9 cm. During an experiment, driftwood was supplied at the upstream end of the channel, after being placed in water buckets for some time. The mixture of the wood supply was defined by a parameter p 1 , corresponding to the proportion of 0.20-m sticks in the supplied material; the experiments realized in the present campaign had p 1 = 1, 0.5, and 0. The facility was equipped with five cameras operating at 30 fps: three cameras shot the system from above, and were placed at some distance upstream of the bridge, over it, and at some distance downstream of it; one took a side view of the region just upstream of the bridge; finally, one shot a panel with vertical pipes connected to the piezometric probes that were present along the channel. 2.2. Measurement methods for the driftwood accumulation The images of the driftwood accumulations were processed as follows. First, the driftwood (that was darker then the background) was recognized applying image filtering based on an appropriate threshold grey level. Second, the number of pixels occupied by wood in a movie frame was determined and converted into a projected area on the basis of a predetermined pixel/cm conversion. The same procedure was applied to the movie frames taken from above and from the side, thus obtaining the temporal evolution of the projected areas of the accumulation (from above and from the side for upstream of the bridge, from above for downstream of it). Determining an accumulation volume from two-dimensional views is, in principle, possible assuming the accumulation shape. However, in this work, we preferred to determine the accumulation volume on the basis of the wood mass conservation. The two cameras placed above, at some distance upstream and downstream of the bridge, were used to measure by Particle Image Velocimetry (PIV) the instantaneous wood transport rate. Upstream of the bridge, the latter corresponded to the wood feeding rate; downstream of the bridge, it was the rate of the driftwood that left the accumulation, escaping from the flow recirculation region. These measurements, suitably manipulated accounting for the distance between the PIV measuring section and the driftwood accumulation, furnished the rate at which material joined to or left the accumulation and, in turn, the accumulation volume. Compared to the measurements described in the previous paragraph, this one had the advantage that furnished the volume of the accumulation rather than its projected area, and the disadvantage that it was impossible to divide the accumulation volumes upstream and downstream of the bridge.