Investigating the feasibility of using dynamic stiffness beam functions in a hybrid structure-borne noise prediction model

Ground-borne vibration and structure-borne noise in buildings associated with nearby railway systems, either above ground or underground, is a common occurrence indensely populated areas. This is especially true where new transportation systems are built in proximity to existing noise and vibration sensitive buildings, and wheresimilar buildings are built close to existing rail systems. With the tendency to build closer to railways both overground and underground, there is a need to betterunderstand how to predict the levels of ground-borne vibration and structure-borne noise in buildings from such systems. A review of the mechanisms through whichenergy is generated from train sources and propagates into buildings, as well as a review of the existing empirical and theoretical models for the prediction of structure-bornenoise in buildings has been undertaken. A proposal for the simplification of an existing hybrid deterministic – statistical model for the prediction of structure-borne noise in buildings has been put forwardand its feasibility investigated. The original model consists of a hybrid finite element – statistical energy analysis tool, where low frequencies are modelled deterministically with finite element, and the higher frequencies are modelledstatistically. In the simplified model, the deterministic elements of the system, i.e. beams and columns, are proposed to be modelled using a dynamic stiffness approach. The analytical mobility functions for free-free beams with six degrees of freedom at each end have been derived from first principles. The results from these were compared against a finite element model for the same beam arrangement. Good agreement was found between the results of the analytical and finite element models.The coupling between beams has been accounted for by using the impedance addition method. Various scenarios were modelled. Good agreement was obtained with a finiteelement model for the two beams in line scenario. Discrepancies were present for some of the degrees of freedom coupled for the beams in L-junctions scenarios.Further works have been suggested to address these. A comparison between the various stages of the full and simplified hybrid prediction models has been provided, along with suggestions for the next steps to further developand assemble the proposed simplified model.

Level dc:type.qualificationlevel

Master's Level (Level 7)

Year dc:date.issued

2019

Author dc:creator

• Filipe, DD

Language dc:language

en

Identifier

oai:salford-repository.worktribe.com:1373651

OAI identifier oai:identifier

oai:salford-repository.worktribe.com:1373651