The dimensional accuracy evaluation and forming quality control are mainly rely on the accurate measurement of the springback of the micro-bent parts. In this paper, micro W-bending is proposed for the first time. Considering the small feature size and extremely thin thickness of the W-shaped micro-bent parts, a high-precision 2D image measuring instrument, Mitutoyo Quick Scope, was used to capture the clear image of the micro parts. QSPAK software and MATLAB image processing program were then adopted to measure the springback angle. Comparison of final bending angles obtained by QSPAK software and MATLAB program indicated that both methods could meet the accuracy requirement. Finally, the springback radii were measured using the similar MATLAB image processing program. The investigation addressed in this paper could provide an access to achieve the accurately control and prediction of the dimensional accuracy and forming quality of the W-shaped micro-bent parts.
A new on-machine measurement technique is proposed to observe the quality of micro-hole in micro Electrical Discharge Machining (EDM) progress. Images of micro-hole are captured by means of a simple configuration use a light-emitting diode (LED) illuminator, optical fiber, and a camera charge-coupled device (CCD).The device does not make contact with workpiece and acquire the image from the contour of the finished micro-hole. Because of using the optical fiber to transmit the real-time image, the device can woke in the narrow and small place. Based on the detected feature points, the geometrical features of the finished micro-hole such as its aperture can be obtained in real time. Meanwhile, the measuring result shows the micro-hole is whether or not qualified and can optimize the process parameters.
Compared to the function-oriented tolerancing rules for micro monolithic components, the lack of specific tolerancing rules for micro non-monolithic ones results in difficulties in bulk production and quality assurance. In order to regulate micro non-monolithic components in micro/meso-scale, a mathematical model of power function is adopted to forecast the tolerance values of nominal sizes in 10~10,000 microns by the linearized regression analysis. The goodness-of-fit qualifies the regression with the power function model and the forecasting results are reasonable in the view of relative accuracy. It is hoped that the improved numerical value table of tolerance can provide some beneficial proposals for the establishment of new tolerancing rules for micro non-monolithic components in micro/meso-scale.