Surface roughness prediction model in ultrasonic vibration assisted grinding of BK7 optical glass

Pre-knowledge of machined surface roughness is the key to improve whole machining efficiency and meanwhile reduce the expenditure in machining optical glass components. In order to predict the surface roughness in ultrasonic vibration assisted grinding of brittle materials, the surface morphologies...

Full description

Saved in:
Bibliographic Details
Published in:Journal of Central South University 2018-02, Vol.25 (2), p.277-286
Main Authors: Zhao, Pei-yi, Zhou, Ming, Zhang, Yuan-jing, Qiao, Guo-chao
Format: Article
Language:English
Subjects:
Abrasive wheels
Brittle materials
Ceramics industry
Engineering
Gaussian process
Grinding wheels
Machining
Mathematical models
Metallic Materials
Optical glass
Regression analysis
Surface roughness
Ultrasonic vibration
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Pre-knowledge of machined surface roughness is the key to improve whole machining efficiency and meanwhile reduce the expenditure in machining optical glass components. In order to predict the surface roughness in ultrasonic vibration assisted grinding of brittle materials, the surface morphologies of grinding wheel were obtained firstly in the present work, the grinding wheel model was developed and the abrasive trajectories in ultrasonic vibration assisted grinding were also investigated, the theoretical model for surface roughness was developed based on the above analysis. The prediction model was developed by using Gaussian processing regression (GPR) due to the influence of brittle fracture on machined surface roughness. In order to validate both the proposed theoretical and GPR models, 32 sets of experiments of ultrasonic vibration assisted grinding of BK7 optical glass were carried out. Experimental results show that the average relative errors of the theoretical model and GPR prediction model are 13.11% and 8.12%, respectively. The GPR prediction results can match well with the experimental results.
ISSN:2095-2899
2227-5223
DOI:10.1007/s11771-018-3736-5