Abstract
This thesis examines vibration transmission in timber-framed buildings. The work includes both transmission through the frame taken in isolation, and transmission between the plates and frame of such buildings. Statistical energy analysis (SEA) was the framework used to develop predictive models.
Examination of transmission between two timber beams connected at right angles showed that the junction between timber beams does not behave as a rigid junction. A junction model of beams meeting at right angles was developed and showed that real timber junctions were better modelled as pinned junctions, in which no moments are transmitted. Significant fluctuations in the response were observed due to the low mode count and modal overlap of the beams, which could be taken into account by the model. These theories were applied to the two beam junction and when appropriate corrections for the both junction transmission and the low mode count were considered, there was good agreement between measured and predicted results. Transmission through a large two storey timber frame was then examined using these theories and it was found that standard SEA models could be used to predict the mean response and that limits of fluctuations about that mean could also be estimated.
Vibration transmission between lightweight plates connected to a frame was then examined, for a variety of structures with point or line connections. For point connection models (based on mobility functions) it was shown that the increased transmission observed at high frequencies can be related to an increase in the modal density of the frame, which can in turn be related to the excitation of extra modes in the frame. For line connection models (based on a wave approach) a model was developed in which the connection between the plates and frame is pinned. Results showed that rigid and pinned predictions can provide limits for transmission between panels on the same side of a wall, measured results lying closer to the rigid prediction for shallow beams and closer to the pinned prediction for deep beams.
Examination of transmission between two timber beams connected at right angles showed that the junction between timber beams does not behave as a rigid junction. A junction model of beams meeting at right angles was developed and showed that real timber junctions were better modelled as pinned junctions, in which no moments are transmitted. Significant fluctuations in the response were observed due to the low mode count and modal overlap of the beams, which could be taken into account by the model. These theories were applied to the two beam junction and when appropriate corrections for the both junction transmission and the low mode count were considered, there was good agreement between measured and predicted results. Transmission through a large two storey timber frame was then examined using these theories and it was found that standard SEA models could be used to predict the mean response and that limits of fluctuations about that mean could also be estimated.
Vibration transmission between lightweight plates connected to a frame was then examined, for a variety of structures with point or line connections. For point connection models (based on mobility functions) it was shown that the increased transmission observed at high frequencies can be related to an increase in the modal density of the frame, which can in turn be related to the excitation of extra modes in the frame. For line connection models (based on a wave approach) a model was developed in which the connection between the plates and frame is pinned. Results showed that rigid and pinned predictions can provide limits for transmission between panels on the same side of a wall, measured results lying closer to the rigid prediction for shallow beams and closer to the pinned prediction for deep beams.
| Original language | English |
|---|---|
| Type | PhD Thesis |
| Publisher | Heriot-Watt University |
| Number of pages | 234 |
| Publication status | Published - 2004 |