Cutting Capability Assessment of Highly Porous CBN Wheels by Microrelief of Plane Parts from 06Cr14Ni6Cu2MoWТi-Sh Steel Using Artificial Intelligence System

Soler, Yakov Iosifovich ( Technology and Equipment of Mechanical Manufacture Department, Irkutsk National Research Technical University, 664074, Irkutsk, Lermontov St., Bldg. 83)
Tiem, Nguyen Manh ( Technology and Equipment of Mechanical Manufacture Department, Irkutsk National Research Technical University, 664074, Irkutsk, Lermontov St., Bldg. 83)
Le, Nguyen Van ( Technology and Equipment of Mechanical Manufacture Department, Irkutsk National Research Technical University, 664074, Irkutsk, Lermontov St., Bldg. 83)

Journal of Engineering and Technological Sciences
Vol 48, No 4 (2016)

The abrasive tools are the weakest element in the grinding process system, to which great attention is being paid by scientific and industrial collectives. Eleven highly porous wheels (HPWs) were tested: CBN30 (B76, B107, B126, B151) 100 OVK27-КF40; CBN30 B107 100 OVKC10-КF40; CBN30 B126 100(M, L) VK27- (КF25, КF40); LKV50 (B107, B126) 100 (M, O) VK27-КF40. Assessment of the surface topography was carried out by roughness parameters Ra, Rmax, and Sm(GOST 25472-82), which were considered random variables with their position and dispersion measures. Two artificial intelligence systems – fuzzy logic (FL) and neural networks (NN) – were used to analyze the HPW’s cutting capability (CC). In both cases, the best CC was predicted for grinding with CBN30 (B76 and B151) 100 OVK27-КF40 and LKV50 (B107) 100 OVK27-KF40. In the absence of a training process in FL modeling, the assessments for the wheels with a low CC were less reliable.