Dewantara, Abi Surya (2026) Investigasi Pengaruh Parameter Proses Boring Terhadap Stabilitas Chatter Dan MRR Pada Pipa Thin-Wall Aluminium 6063 Berbasis BPNN Menggunakan Sinyal Akustik Pada Kondisi MQL. Masters thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 6007241076-Master_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
6007241076-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (6MB) | Request a copy |
Abstract
Manufaktur komponen berdinding tipis (thin-walled) dari material Aluminium Alloy 6063 memegang peranan penting dalam industri modern seperti kedirgantaraan, di mana presisi tinggi dan bobot ringan sangat dibutuhkan. Namun, proses boring (bubut dalam) pada komponen ini sangat rentan terhadap regenerative chatter akibat kekakuan struktural yang rendah pada benda kerja dan penggunaan boring bar yang panjang. Fenomena ini berdampak negatif signifikan terhadap kualitas permukaan dan produktivitas. Penelitian ini bertujuan untuk menganalisis dan memetakan karakteristik chatter pada proses boring Al-6063 menggunakan pahat karbida di bawah kondisi Minimum Quantity Lubrication (MQL) berbasis Vegetable Cutting Oil (VCO). Parameter proses yang divariasikan meliputi spindle speed, feed rate, dan depth of cut. Hasil penelitian menunjukkan bahwa instabilitas didominasi oleh frekuensi natural boring bar (800 Hz) dibandingkan benda kerja (1047 Hz). Secara fisik, kondisi chatter teridentifikasi melalui geram berbentuk pita kusut (tangled ribbon) dengan tepi bergerigi tajam (secondary saw-teeth), berbeda dengan kondisi stabil yang menghasilkan geram spiral. Analisis sinyal menunjukkan bahwa chatter memiliki ciri khas lonjakan Power Spectral Density (PSD) tinggi dan nilai Spectral Entropy rendah (< 10.5), sedangkan zona transisi ditandai oleh energi frekuensi intermiten dengan entropi tinggi. Dari sisi parameter, peningkatan spindle speed (hingga 630 rpm) memperluas zona instabilitas, sedangkan peningkatan feed rate justru memperluas zona stabil akibat efek process damping. Berdasarkan pelatihan data, model Backpropagation Neural Network (BPNN) yang dikembangkan mampu mengklasifikasikan stabilitas (Stabil, Transisi, Chatter) dengan akurasi validasi 95.8% serta memprediksi Material Removal Rate (MRR) dengan Mean Squared Error (MSE) 0.0061. Penelitian ini menghasilkan peta kestabilan berbasis BPNN sebagai panduan strategis untuk menghindari zona instabilitas sekaligus mempertahankan produktivitas optimal.
================================================================================================================================
The manufacturing of thin-walled components from Aluminium Alloy 6063 plays a crucial role in modern industries such as aerospace, where high precision and lightweight materials are essential. However, the boring process for these components is highly susceptible to chatter due to low structural stiffness and the use of long boring bars. This chatter phenomenon significantly impacts surface quality, tool wear, and productivity. This research proposes a multi-response optimization for the boring of Al 6063 under Minimum Quantity Lubrication (MQL) conditions using Vegetable Cutting Oil (VCO). The varied process parameters include spindle speed, feed rate, and depth of cut. The results indicate that instability is dominated by the boring bar's natural frequency (800 Hz) rather than the workpiece (1047 Hz). Physically, chatter is identified by tangled ribbon chips with sharp secondary saw-teeth edges, distinct from the spiral chips formed in stable conditions. Signal analysis reveals that chatter exhibits high Power Spectral Density (PSD) peaks and low Spectral Entropy (< 10.5), while the transition zone is characterized by intermittent frequency energy with high entropy. Regarding process parameters, increasing spindle speed (up to 630 rpm) expands the instability zone, whereas increasing feed rate extends the stable zone due to the process damping effect. Based on data training, the developed Backpropagation Neural Network (BPNN) model achieves 95.8% validation accuracy in classifying stability states (Stable, Transition, Chatter) and predicts Material Removal Rate (MRR) with a Mean Squared Error (MSE) of 0.0061. The outcome of this research is a BPNN-based stability map that serves as a strategic guide for parameter adjustments to avoid instability zones while maintaining optimal productivity
Actions (login required)
 |
View Item |