Quantifying dynamic soil-structure interaction for railway induced vibrations. (2017)

• Record Type: Journal Article
• Authors: Kuo, K.A.; Lombaert, G.; Degrande, G.
• Is Part Of: Procedia engineering. Volume 199(2017)
• Page Start: 2372
• Page End: 2377
• Publication Date: 2017
• Subjects: railway induced vibration -- dynamic soil-structure interaction -- vibration predictions; Engineering -- Congresses; Engineering -- Periodicals; Engineering; Conference proceedings; Periodicals; 620.005
• DOI: 10.1016/j.proeng.2017.09.256
• Languages: English
• ISSNs: 1877-7058

Abstract

Abstract: Predicting the vibration response of buildings due to excitation from a nearby railway is an important step in ensuring the protection of nearby residents and the integrity of manufacturing processes. It is difficult and expensive to undertake measurement campaigns to determine the dynamic soil-structure interaction, that is, how vibrations propagate from the ground and through the building. Numerical models that encompass the track, soil and building are rarely used due to their computational expense and the required detailed parameter inputs. In practice, vibration attenuation through the foundation, floors and spans is accounted for using generic adjustment factors. In this study, the response of a building to railway induced vibration is investigated. Three different configurations of the railway-soil-building system are modelled using a coupled finite element–boundary element model. The first configuration includes both the railway and a nearby building, the second configuration contains only the railway and surrounding soil; and the final configuration contains a building that is excited by nearby impulses on the soil's surface. By comparing the vibration levels in each of these configurations at various locations within the building and on the free surface, the magnitude of the interaction between the railway and the building can be investigated. Factors affecting the soil-structure interaction, such as soil type, building geometry, distance between the railway and building, and excitation type (whether train passage or impacts) are explored. The findings of the study support the separation of the source and receiver terms, demonstrating that the railway can be represented by a series of on-soil impacts when characterising the building response, without significant loss of modelling accuracy. This result means that simplified computational models that do not include a detailed railway structure can be employed. It also suggests that the vibration response of the building can be determined by in situ measurements prior to railway construction.