Please use this identifier to cite or link to this item:
http://acervodigital.unesp.br/handle/11449/74897
- Title:
- Monitoring in precision metal drilling process using multi-sensors and neural network
- Universidade Estadual Paulista (UNESP)
- Universidade Federal de Uberlândia (UFU)
- 0268-3768
- 1433-3015
- This paper presents a new method to estimate hole diameters and surface roughness in precision drilling processes, using coupons taken from a sandwich plate composed of a titanium alloy plate (Ti6Al4V) glued onto an aluminum alloy plate (AA 2024T3). The proposed method uses signals acquired during the cutting process by a multisensor system installed on the machine tool. These signals are mathematically treated and then used as input for an artificial neural network. After training, the neural network system is qualified to estimate the surface roughness and hole diameter based on the signals and cutting process parameters. To evaluate the system, the estimated data were compared with experimental measurements and the errors were calculated. The results proved the efficiency of the proposed method, which yielded very low or even negligible errors of the tolerances used in most industrial drilling processes. This pioneering method opens up a new field of research, showing a promising potential for development and application as an alternative monitoring method for drilling processes. © 2012 Springer-Verlag London Limited.
- 1-Apr-2013
- International Journal of Advanced Manufacturing Technology, v. 66, n. 1-4, p. 151-158, 2013.
- 151-158
- Artificial neural network
- Drilling process monitoring
- Hole diameter
- Surface roughness
- Cutting process
- Development and applications
- Drilling process
- Experimental measurements
- Monitoring methods
- Neural network systems
- Sandwich plates
- Cutting tools
- Errors
- Process monitoring
- Neural networks
- http://dx.doi.org/10.1007/s00170-012-4314-x
- Acesso restrito
- outro
- http://repositorio.unesp.br/handle/11449/74897
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.