Development of sustainability performance model for turning process by using neural network model

• Main Author: Hadi, Abdul Salaam
• Format: Thesis
• Language: English
• Published: 2020
• Subjects: TJ Mechanical engineering and machinery; TS Manufactures
• Online Access: http://umpir.ump.edu.my/id/eprint/30386/1/Development%20of%20sustainability%20performance%20model%20for%20turning%20process.pdf; http://umpir.ump.edu.my/id/eprint/30386/
• Institution: Universiti Malaysia Pahang

Sustainability concept was introduced by Harlem Brundtland in the 1980s, consists of three evaluation criteria’s; namely economics, environmental and social. However, recent research on the indicator used had increasingly called into question where the indicator is difficult to be assessed and the measurement is indirect. The novelty of the present study is to focus on the development of new sustainability assessment methods based on Malaysia industry scenario, demonstrating the new sustainability assessment model focusing on a turning process and optimized the assessment model to obtain the optimum cutting parameter. In the manufacturing industry perspective, manufacturing costs criteria is known to measure the company economic sustainability. Whilst, environmental criteria is a measure of the impact of manufacturing activities on the environment. The social criteria can be measured by using the production operator health. In the present study, three main sustainability evaluation methods are used after getting feedback from the survey respondents which mostly works in the manufacturing industry. They are the total manufacturing costs, environmental impact, ergonomics impact and combined with energy criteria used during the manufacturing process of a pneumatic nipple hose connector. Energy criteria was introduced because of the implementation of principal component analysis (PCA) disadvantage. The total manufacturing costs consists of six cost assessments which include raw material, tool, coolant, lubricant, energy and manpower. The environmental impact assessments used are chip recycling impact and energy impact. Cutting tool impact, coolant impact and lubricant impact did not take into account as the contribution of impact to the environment is too small when compared to the number of the produced product. The ergonomic assessment used is The Revised NIOSH Weight Lifting Index as the method measures the potential impact of the worker during lifting activities. The index calculation requires raw material mass data as this also used either in chip recycling impact and raw material cost assessment. The present study also highlights the usage of neural network and inversion of the neural network model assessment. The data obtained from both theoretical and experimental methods were compared for their validity which is proved to be less than 12%. The experimental data used for the development of neural network model provides comprehensive results in comparison to the theoretical data. Additionally, inputs data tested using the developed neural network model produced the predicted neural network results for all the four criteria. These data were compared with the experimental data for validation and showed the value of less than 5%. Later, the input and output experimental data used are then inversed with the input is used as an output and (vice-versa) to obtain the optimum cutting parameters by using the inversion of neural network model method. For optimization of cutting parameters, the minimum values from each criteria were selected. These parameters were tested for verification and validation purpose in both experimental and theoretical assessment methods. The targeted percentage difference used at this stage is 5%. The results of optimum cutting speed and feedrate obtained in this project is 55.25 m/min for cutting speed and 0.10 mm/rev for Aluminum 6061 and 82.00 m/min and 0.10 mm/rev for Brass C3604 material. As a conclusion, this study proved that sustainability assessment method can be used to select optimum cutting parameters. Additional energy criteria being introduce able to specifically control the energy data since the summation of all assessment data being used in each criterion. In the future, the proposed method can be applied in other machining process for a better machining parameter optimization in others machining process.