Research on the Relationship between the Anchor Pattern Characteristics and Adhesion Based on Neural Network

Abstract:

The anchor characteristics are important factors in the quality of surface treatment of the tube, a good anchor pattern characteristic will ensure the best adhesion between the coating and the substrate. Surface roughness, micro-roughness spacing of anchor pattern and anchor pattern depth are main factors composed of anchor pattern characteristics. Adhesion prediction model was constructed using the BP artificial neural network , according to the relationships between the adhesion and various factors , which not only can achieve nonlinear relationships with high prediction accuracy between the anchor characteristic and adhesion, but also shorten the test time and reduce testing costs.

• Periodical: Applied Mechanics and Materials (Volume 392)
• Pages: 765-769
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.392.765
• Online since: September 2013
• Authors: Shu Cong Liu, Yan Xing Song, Yan Qin Li, Zhen Jing Yao
• Keywords: Anchor Adhesion Pattern Prediction, Anchor Pattern Characteristics, BP Artificial Neural Network, Epoxy Coating

References

[1] Shi Xiuwei, Cui Chao. Impact factors on treatment quality of buried pipeline coating surface [J]. Coating plating and Electroplating. 2009. 8.

[2] Jin Mingshan, Ma Guimei. Application of neural network in model construction of pump frequency control [J]. Energy . 2003. 3.

[3] Zhang Defeng. MATLAB neural network application design. 2009. 9.

[4] GB / T 5210-2006. Paints and varnishes- Pull-off test for adhesion.

[5] Yao Yifu, Jiang Qinyu, He Guorong and so on. Performance prediction of centrifugal pump based on BP neural network[J]. Mechanical design and manufacturing . 2008. 1.