A sensor-based process modeling approach to increase the reliability of hot bar rolling

A sensor-based process modeling approach to reduce seams through off-line robust design of rolling parameters, and on-line adjustments through seam detection, is discussed. The system integrates inspection and measurement systems with an inference engine based on probability of fault generation to detect the severity and frequency of defects. This inference engine can be used for off-line design of robust rolling schedules that are insensitive to uncontrolled noise factors. OG Technologies' (OGT) machine vision technology provides solutions to critical inspection and measurement challenges. OGT has developed a noncontact machine vision system, HotEyeTM, that can identify and measured rolled seams on-line. The sensor based process modeling approach provides imaging system with an intelligent advisor for rolling process design and control. A minimum sensitivity approach is used to explore robust parameter design by perturbing the reductions in flat passes.