Vibation Reduction by Shaping the Terrain Topography

Abstract: MAX-lab is a synchrotron radiation laboratory run in cooperation between Lund University and the Swedish Research Council. It currently consists of three stor-age rings, however, construction has begun of the fourth. MAX IV will be located outside of the Lund University Faculty of Engineering campus, in the northeastern part of Lund. Induced by the development in material technology at the nanometer level, the state-of-the art facility will become a landmark with its large storage ring that has a circumference of approximately 500 m. The instruments operating at MAX IV are extremely sensitive to vibrations, which has led to the establishment of a strict upper level limit. A number of actions has been taken on site, including the implementation of a landscape of hills and valleys surrounding the facility, that should distort and damp out external vibration induced waves traveling in the ground. One of the most concerning sources of vibrations from the surroundings is the highway, E22, passing 100 m to the west of the facility. As the shaped landscape of hills and valleys is not currently extended to cover the area between MAX IV and the highway, this Master's Dissertation has investigated the possible gain in vibration reduction by making this extension. Finite element models in two and three dimensions have been established, onto which a variety of shapes has been applied. Harmonic unit point loading has been used in the steadystate analysis, spanning from 5-20 Hz (as higher frequency waves are damped out quickly). With the low magnitude loading, visco-elastic isotropic behaviour has been applied to the clay and bedrock, modeled with properties derived by commercial contractors and consultants on the MAX IV site. All shapes applied are derived from a set of arcs with constant curvature, and a practical limit in slope angle of 30 degrees. This has led to a maximum difference in altitude of approximately 20 m. On the far side of the shape zone, 100 m from the loading point, a main evaluation point has been investigated, supported by a number of additional evaluation points. The main focus of this project has been vertical displacements. In the 2D analyses some clear tendencies could be seen. The application of fewer hills and valleys of large magnitude provided the best results compared to a at terrain, with reductions in RMS vertical displacements of more than 20%. Reductions were seen for all of the applications with altering hills and valleys as long as the net change in material was positive or zero. Reductions in vertical displacements were also seen for all applications when only hills were applied. A negative effect was seen when applying only valleys or having an altering pattern with more valleys than hills for smaller magnitude level differences. Another tendency seen for the altering applications was that a hill seems to perform better when it is preceded by a valley, i.e. a smooth transition rather than an abrupt. In the 3D analyses that corresponded to the 2D analyses, the same tendencies could be seen, with reductions of the same magnitude. From visualizations, it could also be seen that a continuous hill actually captures the waves and directs them in the longitudinal direction of the hill. A better caption was obtained when rotating the hills horizontally to not run transversally across the model, but in an angle. For this case, however, the risk of transmitting an intensified wave front in the direction of the evaluation points emerges. A checkered pattern of elevations and depressions showed the same tendencies - the waves tended to follow paths of elevations. This explains why some confgurations of checkered patterns provided reductions in vertical displacements whereas others increased the displacements at the evaluation points. Ultimately, it became clear that some of the tested shape confgurations are safe choices, whereas others seem to possibly reduce vertical displacements to a larger extent locally but with the risk of intensifying them elsewhere.