서지주요정보
Optimal design of the well-based sound diffusers = 동공형 음향 확산체의 최적 설계
서명 / 저자 Optimal design of the well-based sound diffusers = 동공형 음향 확산체의 최적 설계 / Boris Mondet.
발행사항 [대전 : 한국과학기술원, 2016].
Online Access 원문보기 원문인쇄

소장정보

등록번호

8029966

소장위치/청구기호

학술문화관(문화관)B1층 보존서고

MME 16079

휴대폰 전송

도서상태

이용가능(대출불가)

사유안내

반납예정일

리뷰정보

초록정보

Well-based diffusers have greatly developed since the introduction of Schroeder diffusers forty years ago. Nevertheless, common designs found nowadays still have a lack of efficiency, especially at low frequencies when arranged in arrays. The present study therefore aimed at proposing a method to design optimal well-based sound diffusers. Based on simulations performed with the boundary-element method, the different effects of the parameters determining the geometry of a well-based diffuser were first investigated so as to identify the key variables to take into consideration. According to these findings, an optimization strategy was adopted and optimal designs were suggested, which were then compared with a chosen reference diffuser. Finally, a method to improve the poor performance of one of the optimal designs in a specific frequency band was introduced.

40년 전 슈뢰더 음향 확산 구조 (Schroeder diffuser)가 개발된 후 동공형 (Well) 음향 확산체는 지속적으로 사용되고, 발전되어 왔다. 하지만, 최근 주로 사용되는 배열 구조의 경우 낮은 주파수 범위에서 효율성이 떨어진다. 본 연구에서는, 기존의 음향 확산체가 지는 문제점을 극복하고자, 동공형 음향 확산체의 최적 설계를 수행하였다. 구조의 외형을 결정하는 각각의 주요 변수들에 따른 특성 변화를 경계요소법 (BEM) 을 기반으로 모사 실험하였고, 이를 바탕으로 최적 설계 방법을 제시 및 수행 하였다. 또한, 상업적으로 판매되는 음향 확산체 성능과 최적화된 모델을 비교, 분석하였다. 마지막으로, 기존 확산체의 성능이 취약한 특정 주파수 대역에서의 성능을 개선하기 위해 동공형 음향 확산체의 최적 설계 방법을 제시하였다.

서지기타정보

서지기타정보
청구기호 {MME 16079
형태사항 v, 57 p. : 삽화 ; 30 cm
언어 영어
일반주기 저자명의 한글표기 : 보리스 몽데
지도교수의 영문표기 : Jeong-Guon Ih
지도교수의 한글표기 : 이정권
공동지도교수의 영문표기 : Ik-Jin Lee
공동지도교수의 한글표기 : 이익진
Appendix: Simulated random incidence diffusion coefficients of the optimal designs
학위논문 학위논문(석사) - 한국과학기술원 : 기계공학과,
서지주기 References : p. 52-53
QR CODE

책소개

전체보기

목차

전체보기

이 주제의 인기대출도서

Performances of commercialized diffusers for random incidence. -0-: QRD 734; -X-: Modf- fusor.

Cross-sections of the commercialized diffusers.

Performance loss in periodic diffuser arrays. -O-: 1 QRD 734 unit; -X-: 6 QRD 734 units.

Performances of aperiodic diffuser arrays made of QRDs (left) and Modffusors (right). Single diffuser unit; X-: Optimized array.

Diffusion from two plates with a factor 6 between their dimensions.

Diffusion of a thin flat plate. -0-: 0.6m wide; - X-: 3.6m wide

BEM simulation configuration. O: Sources; X: Receivers; -: Diffuser.

Simulation errors in the random incidence diffusion coefficient.

Meshes used for the diffusers modeled; f = 100Hz, X/6= 57.3cm. +: Nodes; X: CHIEF points.

Meshes used for the diffusers modeled; f = 5000Hz, X/6 =1.15cm. +: Nodes; X: CHIEF points.

Comparison between meshes with different minimum numbers of elements per segment. -0- 1 element; - X-: 3 elements; 4··: 5 elements.

Effect of the receiver angular resolution on BEM simulation results. -X-:1 resolution; -0- 50 resolution; 스 ·: 10o resolution.

Random incidence diffusion ofa thin flat plate. -o-: Measured data [10];-X-: BEM simula- tion.

Source and microphone positions with respect to the center of the diffuser.

Source output power spectral density with the two excitation signals. -o-r=1m -X-: r=1.89m.

Source, microphone and diffuser positions during experiments

Experimental repeatability at the four measurement positions. -o-:1st measurement; -X-: 2nd measurement.

Experimental and numerical results with white noise excitation. -0-: Measurement; -X-: Simulation.

Numerical results with different numbers offrequency components in each 1/3-octave band -0-: 3 components; X-: 10 components.

Experimental and numerical results with sine sweep excitation. -0-: Measurement; -X-: Simulation.

Geometry parameters of a well-based diffuser.

Effect of the number of wells on a diffuser with constant depth. ↔ N=6 wells; -X-: N=9 wells; -0-: N=11 wells.

Cross-section of the arbitrary diffuser.

Effect of decreasing well depths. L ·: Reference diffuser; -X-: -1 diffuser; -ㅇ-: diffuser; -O-: -3 diffuser.

Effect ofincreasing well depths. L.·: Reference diffuser; -X-: +1 diffuser; - ◇-: +2 diffuser; -0-: +3 diffuser.

Effect of the well widths. -◇-: wn = 0.051m; -0-: Wn = 0.070m; -X-: Wn = 0.093m: 스..: Variable widths.

Effect of the separation width. A·: Wsep = 0.012m; -X-: Wsep = 0.024m; -0-: Wsep 0.036m.

Effect of rounded edges. -0-: a = 0 (straight edges); -X-:Q三 Wsep/2 (semicircular edges).

Levels of the different parameters studied.

Groups for ANOM (well depths and well widths): definition and number of runs.

ANOM for well depths.

ANOM for well widths.

ANOM for separation width and edge radius: t-test results.

Average relative closeness coefficients

Optimization attempt with Taguchi methods. -0-: Best diffuser from OA; -X-: Taguchi diffuser.

Ranges of the design variables for different numbers of wells.

Reference (top left) and optimal designs obtained from orthogonal arrays.

Comparison ofoptimaland reference designs within optimization conditions. ··..... Modffusor (reference); X-: Optimal design with 7 wells; -◇-: Optimal design with 8 wells; -0-: Optimal design with 9 wells.

Responses of the optimal and reference designs within optimization conditions.

Diffusion coefficients of the optimal and reference designs. 스 ·: Modffusor (reference); -X-: Optimal design with 7 wells; -ㅇ-: Optimal design with 8 wells; -0-: Optimal design with 9 wells.

Responses for the diffusion coefficients of the optimal and reference designs.

Normalized diffusion coefficients of the optimal and reference designs. A ·: Modffusor (reference); - X-: Optimal design with 7 wells; -◇ -: Optimal design with 8 wells; -0-: Optimal design with 9 wells.

Responses for the normalized diffusion coefficients of the optimal and reference designs

Ranges of the design variables for different numbers of wells.

Performance ofthe optimal design with 5 wells compared to the reference and best designs obtained. A··: Modffusor (reference); -X-: Optimal design with 7 wells; -0-: Optimal design with 5 wells.

5-well optimal design with a 6th well added on its left side (left) and on its right side (right)

Normalized diffusion coefficients of the optimal design with 5 wells and designs with 6 wells. L.·: Optimal design with 5 wells; -X-: Design with 6th well on the right; -0-: Design with 6th well on the right.

Responses for the normalized diffusion coefficients of the 5- and 6-well designs.

6-well optimal design with a 7th well added on its left side (left) and on its right side (right).

Responses for the normalized diffusion coefficients of the 6- and 7-well designs.

Normalized diffusion coefficients ofthe optimal design with 6 wells and designs with 7 wells. 소: Optimal design with 6 wells; X-: Design with 7th well on the right; -0-: Design with 7th well on the right.

Diffusion coefficient responses of the optimal designs.

Ranking of the optimal designs based on the diffusion coefficient.

Normalized diffusion coefficient responses of the optimal designs.

Ranking of the optimal designs based on the normalized diffusion coefficient.

Diffusion coefficient.

Normalized diffusion coefficient.