Statistical characterisation and responses of acoustic screens and two-layers systems

The present dataset contains measurements of the acoustic response of screens in an impedance tube with different backings. It includes as well the characterised parameters for the screens and backings. License This dataset is released under Creative Commons with Attribution (CC BY). Any reutilisation of the present dataset in scholarly works must properly cite this work and its authors: Statistical characterisation and responses of acoustics screens and two-layers systems
M. Gaborit, L. Jaouen, O. Dazel, P. Göransson
DOI: 10.5281/zenodo.3358921
Protocol Two different films (one woven and one non-woven) were sampled 40 times in order to characterise their parameters in an impedance tube. Sampling is manual, the tube inner diameter is 44.44mm and the samples are cut to a slightly large size in order to be clamped inside the tube for measurements. The films are tested in different configurations including some with backings. Two types of backings are considered (melamine and glass wool). These backings are also characterised using 3 and 2 samples respectively. When measuring the response of the film on a backing only 20 samples of each film were used. All measurements span the [4, 4520] Hz band. The setup is calibrated anew before each measurement session. Films' characterisation The films are clamped in the seam between the two parts of the tube, a 200mm air plenum is formed downstream and ends with a rigid backing. The samples for each type of film are measured within 4h. The characterisation procedure implements the method described in the following reference: Jaouen L. Bécot F.-X. JASA 129(3) pp. 1400 - 1406, 2011 Backings' characterisation The sample and backing piston are placed to give a rigid backing and have the sample's surface flush with the seam between the two parts of the tube. In the case of the glass wool, the material has a slight (below millimetre) embossing and the supporting surface is placed flush with respect to the seam (the top of the embossing is not considered as the sample surface). The characterisation procedure implements the method described in the following references: Olny, X. & Panneton, R. JASA 123(2) pp. 814 – 824, 2008 Panneton R. & Olny, X. JASA 119(4), p. 2027, 2006 Response measurements To measure responses of the film/backing systems, the backing are placed as described in the previous paragraph. The film is clamped in the seam as described in the "Films' characterisation" section. A plastic grid with thin struts is placed upstream to prevent the film from bending. This setup allows for easily changing the film without replacing the backing sample. Films' samples 1 to 20 are used for response measurements. Measurement of the grid response on a rigid backing and an air plenum of 200mm are provided for reference. Hardware Mecanum 3-microphones tube, diameter: 44.44mm Software Matelys TubeCell (rig control) Matelys RoKCell (characterisation) The piece of software controlling the rig implements a 3-microphones technique described in the following references: Iwase T. Izumi Y. J. Acoust. Soc. Jpn J 52(6) pp. 411–419, 1996 (in Japanese) Iwase T. Izumi Y. Kawabata R. in proc. Internoise, 1998 (Christchurch, New Zealand) Structure of the datasets and files The present dataset comprises two parts. The actual traces and associated measurement parameters as exported from TubeCell are found in the directory measurements. The results from the characterisation process as exported with RoKCell are in the characterisations folder. These results were augmented with direct measurements of the weights and thickness of the samples (both for the films and the backings). Measurements The measurements are stored in the measurements directory. There are several sub-directory for each of the measurement sessions: backings: measurements for the caracterisation of the two backings (all samples, with and without grid) grid: measurements of the grid alone on a rigid backing and with a 200mm air plenum nonwoven_alone: measurements for the caracterisation of the non-woven film on a 200mm air plenum, no grid (all 40 samples) woven_alone: measurements for the caracterisation of the woven film on a 200mm air plenum, no grid (all 40 samples) nonwoven_on_glasswool: measurements of the response of 20 non-woven film samples clamped with a grid upstream and the same glass wool sample downstream on a rigid backing nonwoven_on_melamine: measurements of the response of 20 non-woven film samples clamped with a grid upstream and the same melamine sample downstream on a rigid backing woven_on_glasswool: measurements of the response of 20 woven film samples clamped with a grid upstream and the same glass wool sample downstream on a rigid backing woven_on_melamine: measurements of the response of 20 woven film samples clamped with a grid upstream and the same melamine sample downstream on a rigid backing Each measurement is stored into a .irk and a .rok file. The first one contains some meta-data on the measurement process, the second contains the actual traces. Structure of irk files The .irk files contain the following meta-data: Date Medium ID Thickness Size of the air plenum Measure parameter l Measure parameter s Sampling frequency Tube radius [m] Average 1 Average 2 Temperature [°C] Atmospheric pressure [Pa] Humidity [%] High cut-off frequency (generator) [Hz] Low cut-off frequency (generator) [Hz] Position of the cavity microphone Comments Structure of rok files The .rok files contain the following vector (column-major). The first line is to be omitted when reading the files, it contains the columns title. The columns are these: Frequency (Hz) Absorption coefficient Real part of the reflection coefficient Imaginary part of the reflection coefficient Real part of the normalised surface impedance Imaginary part of the normalised surface impedance Real part of the transfer function Imaginary part of the transfer function Real part of the normalised density Imaginary part of the normalised density Real part of the normalised compressibility Imaginary part of the normalised compressibility ... frequency response functions used to compute the previous quantities Characterisations The directory characterisations contains YAML 1.2 files created from the output files of the characterisation software, exploiting the measurement data. There are three subdirectories: backings contains the characterisation of the backings (each sample and an average file) nonwoven contains the characterisation of the non-woven film (each sample and an average file) woven contains the characterisation of the woven film (each sample and an average file) More information For further inquiry write to gaborit at kth dot se

本数据集包含阻抗管中不同衬背条件下声学薄膜的声学响应测量数据，同时涵盖声学薄膜与衬背的表征参数。
本数据集采用署名知识共享许可协议（Creative Commons Attribution, CC BY）发布。若在学术著作中复用本数据集，请务必正确引用本研究及作者：《Statistical characterisation and responses of acoustics screens and two-layers systems》，作者M. Gaborit、L. Jaouen、O. Dazel、P. Göransson，DOI: 10.5281/zenodo.3358921。

实验方案：针对两种不同薄膜（机织与非机织），分别采样40次以表征其阻抗管内参数。采样采用手工方式完成，阻抗管内径为44.44mm，样品裁切为略大于管内径的尺寸以固定于管内完成测量。薄膜在多种配置下进行测试，其中包含带有衬背的场景。本次实验采用两种衬背材料：密胺泡沫与玻璃棉，两种衬背分别通过3组与2组样本完成表征。当测量带衬背薄膜的响应时，每种薄膜仅使用20组样本。所有测量的频率范围均为[4, 4520] Hz，每次测量会话前均会重新校准实验装置。

薄膜表征：薄膜被固定于阻抗管两部分的接缝处，管下游形成长度为200mm的气室，末端为刚性衬背。每种薄膜的样本均在4小时内完成测量。表征流程采用以下文献所述方法：Jaouen L., Bécot F.-X. 《美国声学学会杂志》（Journal of the Acoustical Society of America, JASA）129(3), pp.1400-1406, 2011。

衬背表征：将样本与衬背活塞放置为刚性衬背结构，使样本表面与阻抗管两部分的接缝齐平。对于玻璃棉衬背，材料存在毫米级以下的轻微压花，支撑表面需与接缝齐平（不将压花顶部视为样本表面）。表征流程采用以下文献所述方法：Olny, X. & Panneton, R. JASA 123(2), pp.814–824, 2008；Panneton R. & Olny, X. JASA 119(4), p.2027, 2006。

响应测量：测量薄膜/衬背系统的响应时，衬背的放置方式如前文“衬背表征”所述，薄膜固定方式如“薄膜表征”章节所述。在薄膜上游放置带有细支柱的塑料格栅，以防止薄膜发生弯曲。该装置可在无需更换衬背样本的前提下便捷更换薄膜。本实验使用编号1至20的薄膜样本完成响应测量。同时提供了刚性衬背与200mm气室条件下格栅的响应测量数据作为参考。

硬件与软件：
硬件：麦卡纳姆（Mecanum）三麦克风阻抗管，内径：44.44mm
软件：Matelys TubeCell（装置控制）、Matelys RoKCell（表征分析）
装置控制软件采用以下文献所述的三麦克风技术：Iwase T., Izumi Y. 《日本声学学会杂志》（J. Acoust. Soc. Jpn J）52(6), pp.411–419, 1996（日文）；Iwase T., Izumi Y., Kawabata R. 载于《Internoise 1998会议论文集》（新西兰基督城）。

数据集与文件结构：
本数据集包含两部分：从TubeCell导出的原始测量轨迹与相关测量参数，存储于measurements目录；通过RoKCell导出的表征流程结果，存储于characterisations目录。上述结果补充了薄膜与衬背样本的重量与厚度直接测量数据。

测量数据：
测量数据存储于measurements目录下，每个测量会话对应一个子目录：
- backings：两种衬背的表征测量数据（所有样本，含与不含格栅的场景）
- grid：仅格栅在刚性衬背与200mm气室条件下的测量数据
- nonwoven_alone：非机织薄膜在200mm气室、无格栅条件下的表征测量数据（全部40组样本）
- woven_alone：机织薄膜在200mm气室、无格栅条件下的表征测量数据（全部40组样本）
- nonwoven_on_glasswool：20组非机织薄膜样本（上游加装格栅）搭配同组玻璃棉衬背（下游刚性固定）的响应测量数据
- nonwoven_on_melamine：20组非机织薄膜样本（上游加装格栅）搭配同组密胺泡沫衬背（下游刚性固定）的响应测量数据
- woven_on_glasswool：20组机织薄膜样本（上游加装格栅）搭配同组玻璃棉衬背（下游刚性固定）的响应测量数据
- woven_on_melamine：20组机织薄膜样本（上游加装格栅）搭配同组密胺泡沫衬背（下游刚性固定）的响应测量数据

每组测量存储为.irk与.rok两种格式的文件。其中.irk文件包含测量过程的元数据，.rok文件包含实际测量轨迹。

.irk文件结构：
.irk文件元数据包含以下字段：测量日期、介质、样本ID、厚度、气室长度、测量参数l、测量参数s、采样频率、管半径[m]、平均值1、平均值2、温度[℃]、大气压强[Pa]、湿度[%]、高通截止频率（发生器）[Hz]、低通截止频率（发生器）[Hz]、腔室麦克风位置、备注。

.rok文件结构：
.rok文件包含以下列优先存储的向量。读取文件时需跳过第一行（该行为列标题），各列依次为：频率(Hz)、吸声系数、反射系数实部、反射系数虚部、归一化表面阻抗实部、归一化表面阻抗虚部、传递函数实部、传递函数虚部、归一化密度实部、归一化密度虚部、归一化压缩率实部、归一化压缩率虚部、……用于计算上述物理量的频率响应函数。

表征数据：
characterisations目录包含由表征软件输出文件生成的YAML 1.2格式文件，基于测量数据构建。该目录包含三个子目录：
- backings：衬背的表征数据（含单样本文件与平均文件）
- nonwoven：非机织薄膜的表征数据（含单样本文件与平均文件）
- woven：机织薄膜的表征数据（含单样本文件与平均文件）

更多信息：
若需进一步咨询，请发送邮件至[e-mail]。