Acoustic Performance Study Of 3d Printed Micro-Perforated Panel

Kamarulzaman, Muhammad Amir (2021) Acoustic Performance Study Of 3d Printed Micro-Perforated Panel. Project Report. Universiti Sains Malaysia, Pusat Pengajian Kejuruteraan Mekanik. (Submitted)

[img]
Preview
PDF
Download (986kB) | Preview

Abstract

In this study, the optimum configurations for producing micro-perforated panel with targeted hole roundness is discussed through measurement analysis and construction of artificial neural network. The development of micro-perforated panel using additive manufacturing as an alternative method have been proposed. However, recent cases have found that the method can cause imperfection holes which affect its actual performance. In a further attempt to solve the problem, several perforation hole samples were designed using SolidWorks software and fabricated by using PolyJet machine. The sample parameters (hole diameters and sample thickness) were measured and analysed the deviations produced. The analysis showed that higher sample thickness and lower hole diameters produced more consistent and lower deviations throughout the fabricated model. The perforation holes images were captured using high resolution microscope (Alicona Infinite Focus). The capture images were processed in graph-cut image segmentation algorithm. The circularity of each hole is measured using the processed images and evaluated based on the probability outcomes of specified circularities. The evaluation showed that lower thickness and lower hole diameter sample has a higher tendency to produce better hole circularity. The measured parameters and designed parameters were constructed and trained in two-layer feed forward network model. The model was verified by conducting a reliability test. The trained network model achieved satisfactory prediction results as the mean accuracy for both hole diameter and sample thickness predictions were 1.5588% and 0.3328%, respectively. Based on the experimental results and analytical modelling, it was found that the artificial neural network model can predict the optimum configurations for producing the MPP with the best hole roundness.

Item Type: Monograph (Project Report)
Subjects: T Technology
T Technology > TJ Mechanical engineering and machinery > TJ1-1570 Mechanical engineering and machinery
Divisions: Kampus Kejuruteraan (Engineering Campus) > Pusat Pengajian Kejuruteraan Mekanikal (School of Mechanical Engineering) > Monograph
Depositing User: Mr Mohamed Yunus Mat Yusof
Date Deposited: 04 Dec 2022 05:36
Last Modified: 04 Dec 2022 05:36
URI: http://eprints.usm.my/id/eprint/55891

Actions (login required)

View Item View Item
Share