Dam-break flows are likely to occur in area that are not subject to more common floods. Also, the importance of the wave might be such that the flow route is no more directed by the thalweg of the river. The whole valley is involved, and the roads, bridges and urban buildings become obstacles to the flow. The purpose of the paper is to study the influence of such an obstacle on a dam-break wave, the obstacle being an idealized representation of a single building. The experimental set-up consists in a channel with a rectangular shaped obstacle, representing a building, placed immediately downstream from the dam. The building is not centred in the channel, and is not aligned with the flow direction. Flow observation shows that after the violent impact of the wave on the building, the flow is forced to change its direction to pass the building. This implies the formation of hydraulic jumps with the consequence that the water level is locally more important than without building. Behind the building, a wake zone is observed. Then, further downstream, the flow slowly recovers the structure it would have without the building. Several measurement devices were used to characterize the flow. The water level evolution was measured at five different locations by means of water level gauges. At each gauging point, the velocity was measured by means of acoustic Doppler velocimeter. Then, the surface velocity field was obtained using digital imaging techniques. This study is part of the IMPACT European project which aims at investigating extreme flood processes and uncertainty. The experimental data set presented here was used in the IMPACT benchmarking programme, as a validation means for numerical models. such a way that its minimum C2, corresponding to the critical stage, is higher that the uniform condition in the standard section (Fig. 2b). This results in a control section in the constricted section and in the development of a supercritical water profile, which returns to normal conditions through a hydraulic jump. Figure 2: Water profiles along bridge piers in mild-slope cases A similar behavior may be observed in steepslope conditions, where the flow is generally supercritical. In these situations, the water profile presents the two configurations of Figure 3. Figure 3: Water profiles along bridge piers in steep-slope cases If the constriction is limited, only the water surface close to the constriction is disturbed and the effect will not extend farther upstream. If the constriction becomes significant, the minimum possible energy in the constricted section becomes too high to be reached immediately. The specific energy thus needs to increase in the upstream reach, which is only possible in subcritical flow, thus requiring the formation of a hydraulic jump. The resulting upstream water profile extends only over a short distance. The above description of the obstruction effect caused by the presence of an obstacle is based on a one-dimensional flow modeling. If the resulting constriction is important, two-dimensional aspects may predominate and change significantly the location and shape of discontinuities. The second observed consequence of an obstacle concerns the effect of nonlinear alignment, which may be rather complex. In subcritical flows, a transverse superelevation may arise near the outer bank (or along the obstacle). In supercritical case, the outer wall, which turns inward to the flow, will produce an oblique hydraulic jump and a corresponding positive wave front, while the inner wall, which turns away from the flow, will develop a negative wave front, both forming cross waves (Fig. 4). Figure 4: Cross waves in nonlinear alignment (after Chow, 1959) When the flow is unsteady, and above all in severe transient situations (damor dike break for example), this already complex behavior becomes even more complicated. The obstacle now induces reflected waves that in turn may reflect against the banks or other obstacles. Few descriptions of such transient flow phenomena against obstacles are available. However, they could be of great interest in the modeling of dambreak flow. Above all it is important to evaluate the extension in space and time of the influence of an obstacle to limit accurate and refined computations where they are really needed. At the large scale of a real flood event, it is not always possible to distinguish between the different features of the flow. Therefore, it was decided to design an idealized experiment that limits the parameters involved in the flow. A single obstacle representing a building is placed in a channel, immediately downstream from the dam. Through appropriate measurements of the flow, it is possible to identify its specific features. The experiment presented in this paper was used in the benchmarking program of the IMPACT European project (Soares Frazao et al., 2003). This research project, which addresses the assessment and reduction of risks from extreme flooding caused by natural events or by the failure of dams and flood defense structures (Morris, 2002; Morris and Vaskinn, 2002) has also a special focus on floods in urban areas and in the presence of obstacles. 2 EXPERIMENTAL SET-UP The experiments were carried out in the laboratory of the Civil Engineering Department of the Universite catholique de Louvain (UCL) in Belgium. The channel is sketched in Figures 5-7. Figure 5: Experimental set-up (dimensions in meters) Figure 6: Location and dimensions of the building (in meters) Figure 7: Channel cross section (a) in the reservoir and in the main channel, (b) at the dam location It has a total length of 35.80 m and is 3.60 m wide. The upstream reservoir is 6.90 m long. The cross section is trapezoidal near the bed (Fig. 7a). The dam is represented by a gate located between two solid blocks; its cross-section is rectangular and it is 1.00 m wide (Fig. 7b). To simulate a dam break, the gate is pulled up rapidly. The building, located 3.40 m downstream from the dam, consists in a rectangular block with dimensions 0.80 x 0.40 m. It makes an angle of 64° with the channel axis (Fig. 6). After measurement in uniform conditions, the Manning bed friction coefficient is n = 0.01 s m. The channel is closed by a wall at the upstream end. The downstream boundary condition consists in a weir and a chute, but has no influence on the flow during the test duration, which is 30 s. The initial conditions consist in a water level of 0.40 m in the upstream reservoir and a thin layer of 0.01 m of water in the downstream part of the channel. 3 GENERAL FLOW DESCRIPTION After the rapid opening of the gate, the strong dambreak wave reflects against the building, almost submerging it, and the flow separates, forming a series of shock waves crossing each other. A wake zone can be identified just downstream from the building, surrounded by cross waves. The flow rapidly reaches an almost steady state with a decreasing discharge due to the emptying of the reservoir. Also, re-circulation zones can be identified between the building and the walls. This description of the flow is illustrated by means of figures obtained from computed results (Noel et al., 2003). Figure 8 shows the free-surface elevation at t = 1 s, with the two-dimensional spreading of the flood wave. Figure 8: Computed image of the flow at time t = 1 s At t = 3 s, the reflection of the wave against the building has occurred and results in the formation of an oblique hydraulic jump (Fig. 9). The circular front wave also reflects against the side walls of the channel and lateral jumps are formed. Figure 10 shows the flow after 10 s: the upstream reservoir empties and the hydraulic jump formed by the reflection against the building migrates in the upstream direction. The separation of the flow around the building and the wake zone can also be identified. Figure 9: Computed image of the flow at time t = 3 s Figure 10: Computed image of the flow at time t = 10 s