| A small batch sheet metal assembly system is proposed. A critical factor for quality sheet metal production is the control of flange angles. For small radius (punch radius/sheet thickness {dollar}<{dollar} 1) pressbrake bending operations, control of angle requires extensive setup time, making small batch runs inefficient. This thesis demonstrates the use of relatively simple measurements of material properties to provide the information necessary to estimate and control springback with accuracy approaching 1/4 degree.; The control method combines the generality of a model-based approach with the accuracy of an on-line angle sensor. A new sensor design provides on-line measurements of bend angle with an accuracy of {dollar}pm{dollar}1.5 arc-minutes. A photo-sensitive diode sensing element provides a cost-effective solution to angle measurement when compared to vision systems of comparable accuracy.; To support the on-line angle sensor, an analytic bending model has been developed which provides insights into the material and geometric variables that affect springback. Estimates are made of the relative sensitivity of springback to changes in these variables. In addition, a set of variable ratios are determined which are linearly related to springback and thus serve as the input data for a linear regression model of springback based on experimental results.; Utilizing this experimental estimate of springback, a bend angle control model is developed. This control methodology provides accurate bend angles (within one-third degree) in the presence of sheet to sheet variations in thickness, hardness, and bend length. The control system is independent of bend location along the die and is insensitive to minor variations in tooling dimensions and deflections in the structure of the machine tool. |
