Muzakky, Chezta Ahmad (2024) Optimasi Multi Respons Proses Drilling Glass Fiber Reinforced Polymer Menggunakan Minimum Quantity Lubrication Dengan Metode Backpropagation Neural Network-Genetic Algorithm. Masters thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 6007221008-Master_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
6007221008-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2026. Download (6MB) | Request a copy |
Abstract
Proses drilling merupakan salah satu operasi pemesinan yang penting di bidang manufaktur. Proses drilling sering diterapkan pada material komposit, seperti glass fiber reinforced polymer (GFRP) di perakitan komponen pesawat terbang. Parameter proses pemesinan di proses drilling pada material komposit adalah kecepatan makan (Vf), kecepatan putaran spindle (n), dan jenis pahat drilling (Tt). Ketiga parameter proses tersebut mempengaruhi parameter respons drilling seperti gaya tekan (Fz), torsi (Mz), delaminasi masuk (Fde), dan delaminasi keluar (Fdo). Proses drilling dilakukan dengan menggunakan teknik pendinginan minimum quantity lubrication (MQL). Desain eksperimen pada penelitian ini menggunakan faktorial penuh dengan dua replikasi. Jenis pahat drilling memiliki dua level, yaitu HSS dan HSS-cobalt. Kecepatan makan sebesar 12 mm/min, 24 mm/min, dan 36 mm/min. Kecepatan spindle sebesar 500 rpm, 700 rpm, dan 900 rpm. Karakteristik kualitas dari gaya tekan, torsi, delaminasi masuk, dan delaminasi keluar adalah smaller-the-better. Backpropagation neural network (BPNN) digunakan untuk memprediksi hubungan antara parameter proses dan parameter respons. Arsitektur jaringan BPNN dibuat untuk masing-masing respons. Genetic algorithm (GA) kemudian diterapkan pada hasil pemodelan BPNN untuk mendapatkan kombinasi level parameter pemesinan proses drilling yang menghasilkan respons yang optimal. Hasil optimasi GA memberikan pengaturan parameter proses berupa kecepatan makan 12 mm/menit, kecepatan putaran spindle 888 rpm, dan jenis pahat HSS-cobalt. Berdasarkan eksperimen konfirmasi, nilai optimal gaya tekan sebesar 23,06 N, torsi sebesar 0,10593 Nm, delaminasi masuk sebesar 1,053, dan delaminasi keluar sebesar 1,206.
=================================================================================================================================
The drilling process is one of the most vital machining process in manufacturing industry. Drilling process is often conducted on composite materials used in airplane’s component assembly, such as glass fiber reinforced polymer (GFRP). The machining parameters in drilling composite material are mainly feeding speed (Vf) measured in mm/min, spindle speed (n) measured in rpm, and tool geometry or tool type (Tt). Those are found to affect thrust force (Fz), torque (Mz), peel-up delamination (Fde), and push-out delamination (Fdo), hence the name multi-response. Drilling process was performed with minimum quantity lubrication (MQL) to help alleviate the temperature which gave better surface quality, observed from the delamination responses (Fde and Fdo). The design of experiments was full factorial with two replication. The selected tool type was HSS and HSS-cobalt, therefore having two levels. Feeding speed was set on 12, 24, and 36 mm/min, while spindle speed was set on 500, 700, and 900 rpm. The quality characteristic of the responses is all smaller-the-better. Backpropagation neural network (BPNN) was used to predict the relationship between the machining parameters and the responses. Each responses had their own BPNN architecture. Genetic algorithm (GA) was then applied to optimize the prediction result of BPNN. GA gave the optimal settings for machining parameters which were as follows: HSS-cobalt for the tool type; 12 mm/min for feeding speed; 888 rpm for spindle speed. Based on the results of foregone confirmation experiment, which used the same settings and gave optimal values of 23,06 N for thrust force, 0,10593 Nm for torque, 1,053 for peel-up delamination, and 1,206 for push-out delamination, it was apparent that the optimized settings done by BPNN-GA method is not significantly different.
| Item Type: | Thesis (Masters) |
|---|---|
| Uncontrolled Keywords: | Optimasi multi-respons, Drilling, Komposit GFRP, Backpropagation Neural Network, Genetic Algorithm, Optimization, Multi-response, Drilling, GFRP, Composite, MQL, Backpropagation Neural Network (BPNN), Genetic Algorithm (GA) |
| Subjects: | T Technology > T Technology (General) > T57.5 Data Processing T Technology > TJ Mechanical engineering and machinery > TJ1225 Milling machines numerically controlled T Technology > TL Motor vehicles. Aeronautics. Astronautics > TL240.5 Composite materials |
| Divisions: | Faculty of Industrial Technology and Systems Engineering (INDSYS) > Mechanical Engineering > 21101-(S2) Master Thesis |
| Depositing User: | Chezta Muzakky |
| Date Deposited: | 09 Feb 2024 16:45 |
| Last Modified: | 09 Feb 2024 16:50 |
| URI: | http://repository.its.ac.id/id/eprint/106602 |
Actions (login required)
 |
View Item |