Jeong, H 2010, Source and boundary conditions in finite difference time domain modelling of room acoustics , PhD thesis, Salford : University of Salford.
Restricted to Repository staff only until 31 January 2018.
Download (47MB) | Request a copy
This project deals with several issues on the use of Finite Difference Time Domain (FDTD) method for room acoustic simulation. The main features covered are: source implementation; modeling of frequency dependent boundary conditions; and modeling of arbitrary wall geometry and associated errors in stair-case approximations. A number of techniques for modeling frequency dependent boundary conditions in the FDTD method are presented and are validated against the Boundary Element Method (BEM). The results show that a Z-transform based multi-degree-offreedom model is the most effective for frequency dependent boundary conditions. Also of interest is that, when there is enough random incident waves onto the boundaries in the room, the energy based absorption coefficient can be used in conjunction with a minimum phase approach to obtain a suitable approximation for the time domain boundary condition in FDTD calculations. A thorough analysis on different source implementations is presented, and the cause of hard source problem is revealed. Based on the findings, a Time Limited Sine Modulated Gaussian Hard (TLSGH) source is developed that has the same characteristics of a transparent source, and yet retains the excellent computational efficiency of a hard source implementation. For a room with tilted boundary, results in this thesis show that standard FDTD using stair-case approximation can cause errors that exceed subjective difference limens in calculating room acoustic parameters. To solve this problem, the locally conformal boundary modelling method based on Dey-Mittra FDTD is extended for frequency dependent boundary conditions. Putting all together, FDTD method with TLSGH source and locally conformal frequency dependent boundary conditions shows remarkably good agreement compared to the more precise BEM. It is concluded that the techniques developed in this investigation significantly improve the effectiveness of FDTD in room acoustic simulations.
|Item Type:||Thesis (PhD)|
|Contributors:||Lam, YY (Supervisor)|
|Schools:||Schools > School of Computing, Science and Engineering
Schools > School of Computing, Science and Engineering > Salford Innovation Research Centre (SIRC)
|Depositing User:||Institutional Repository|
|Date Deposited:||03 Oct 2012 13:34|
|Last Modified:||01 Jul 2016 08:29|
Actions (login required)
|Edit record (repository staff only)|