Uncertainty analysis of milling parameters using Monte Carlo simulation, the Taguchi optimization method and data-driven modeling
| Autor: | Habibullah Bilge, Mehmet Fatih Kahraman, Sabri Ozturk |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Surface (mathematics)
0209 industrial biotechnology Materials science Mechanical Engineering Monte Carlo method Mechanical engineering 02 engineering and technology 021001 nanoscience & nanotechnology Data-driven Taguchi optimization 020901 industrial engineering & automation Quality (physics) Mechanics of Materials Surface roughness General Materials Science 0210 nano-technology Uncertainty analysis |
| Zdroj: | Materials Testing. 61:477-483 |
| ISSN: | 2195-8572 0025-5300 |
| DOI: | 10.3139/120.111344 |
| Popis: | Surface roughness plays an important role in the performance of finished structures. The surface quality obtained is enormously affected by cutting parameters. Therefore, the purpose of the present study is to examine the surface roughness value of aluminum 7075 workpiece material during milling operation by considering three steps: (1) the multi-nonlinear regression (MNLR) modeling basis of Taguchi design, (2) optimization based on signal to noise ratio (S/N), and (3) probabilistic uncertainty analysis depending on Monte Carlo technique as a result of depth of cut, cutting speed and feed rate. The depth of cut of 0.2 mm, cutting speed of 900 m × min−1, and feed rate of 0.1 mm × tooth−1 were determined as Taguchi-optimized conditions with a surface roughness of 0.964 μm. In order to justify the surface roughness predicted under optimized conditions in relation to the predicted Taguchi method, three repetitive verification experiments were performed and surface roughness of 0.964 μm ± 0.3 % was achieved. The best-fit MNLR method with an R2 pred (predicted regression coefficient) of 98.02 % is useful for calculating the success of estimating the outcome variable. Monte Carlo simulations were found to be quite effective for identifying the uncertainties in surface roughness that could not be captured by means of deterministic methods. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|