A System Identification and Implementation of a Soft Sensor for Freeform Bending.

Area of Science:

• Materials Science and Engineering
• Mechanical Engineering
• Sensor Technology

Background:

• Accurate prediction of mechanical properties is crucial for industrial processes like freeform bending.
• Current methods often lack inline capabilities for real-time property assessment.
• Closed-loop control requires continuous monitoring of material behavior.

Purpose of the Study:

• To develop a soft sensor for predicting key mechanical properties during freeform bending.
• To establish a foundation for closed-loop property control systems.
• To validate inline hardness measurement for predicting residual stresses and strength.

Main Methods:

• Formulation of a soft sensor integrating a novel hardness-based correlation scheme.
• Implementation of the scheme into an extended Kalman filter (EKF).
• Validation of the ultrasonic contact impedance (UCI) method for inline hardness measurement.

Main Results:

• The developed soft sensor successfully predicts local strength and residual hoop stresses.
• Inline hardness measurements correlate with critical mechanical properties.
• The extended Kalman filter (EKF) effectively integrates hardness data for predictive control.

Conclusions:

• Inline hardness measurement, particularly using the UCI method, is a viable approach for real-time prediction of mechanical properties.
• The developed soft sensor and EKF-based correlation scheme enable closed-loop control for freeform bending.
• This approach enhances process control and material quality assurance in manufacturing.