Prediction of tool wear and surface roughness of waspaloy by using Artificial Neural Network (ANN)
This research focuses on prediction on tool wear and surface roughness of Waspaloy under different machining conditions by using artificial neural network (ANN). Cutting speed and feed rate were the input nodes while tool wear and surface roughness were the output nodes. ANN with various number of n...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English |
| Published: |
2021
|
| Subjects: | |
| Online Access: | http://eprints.utem.edu.my/id/eprint/26952/1/Prediction%20of%20tool%20wear%20and%20surface%20roughness%20of%20waspaloy%20by%20using%20Artificial%20Neural%20Network%20%28ANN%29.pdf http://eprints.utem.edu.my/id/eprint/26952/2/Prediction%20of%20tool%20wear%20and%20surface%20roughness%20of%20waspaloy%20by%20using%20Artificial%20Neural%20Network%20%28ANN%29.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This research focuses on prediction on tool wear and surface roughness of Waspaloy under different machining conditions by using artificial neural network (ANN). Cutting speed and feed rate were the input nodes while tool wear and surface roughness were the output nodes. ANN with various number of neurons, namely 1, 5 and 10 in the hidden layer were created by using MATLAB R2021a. The machining performances between the optimal neural network structure (lowest mean squared error, MSE, mean absolute error, MAE, and mean absolute percentage error, MAPE) and the experimental result were compared. The predicted value by ANN agrees well with the experimental results for tool wear and surface roughness, except in 2-1-2 model. There are three predicted values of tool wear (7th, 8th, and 9th run) in MQL condition and one predicted value in surface roughness (6th run) in wet condition are far away from the experiment result. The error percentage generated is 30.00 %, 32.23 %, 22.78 %, and 22.32 %, respectively. 2-10-2 neural network has shown the lowest MSE, MAE, and MAPE and has been selected as the optimal model. By comparing the dry, wet, and MQL, the lowest tool wear (0.12 mm) located at 8th (MQL) and 9th (MQL) runs while the lowest surface roughness (0.22 μm) located at 9th (MQL) run. MQL is preferable to use in machining to decrease the tool wear and surface roughness especially in machining hard-to-machine materials. |
|---|