In order to acquire the optimal working clearance for non-contact detecting stress of steel members with magneto-elastic stress sensor, a magneto-elastic sensor probe with E-shaped structure is adopted for carrying out the relevant research. Firstly, the principle of non-contact stress detection is discussed based on magneto-elastic effect, and the magnetic circuit of the magneto-elastic stress sensor is analyzed for deducing the basic output voltage equation of sensor when tested pieces (low carbon steel Q235) is loaded with uniaxial stress, on the basis of ferromagnetism and presented references, the technical parameter of sensor is determined for designing non-contact stress detection system. After that, focusing on the design of the testing program with different excitation frequencies and air gap, actual experiments are carried out to optimize working clearance when tested pieces are loaded with uniaxial stress. Results of the test show that this kind of sensor is not only simple in structure but also valuable with non-destructive, convenient and fast measurement of stress in application.
Phase-shifting methods for 3-D shape measurement have long been employed in optical metrology for their speed and accuracy. For real-time, accurate, 3-D shape measurement, a four-step phase-shifting algorithm which has the advantage of its symmetry is a good choice; however, its measurement error is sensitive to any fringe image errors caused by various sources such as motion blur. To alleviate this problem, a fast two-plus-one phase-shifting algorithm is proposed in this paper. This kind of technology will benefit many applications such as medical imaging, gaming, animation, computer vision, computer graphics, etc.
Considering the characteristics that point cloud data obtained by laser scanning system has the same coordinate values in
some dimension, the triangular B-B surfaces reconstruction technique is proposed. For the preprocessing of the initial
point cloud, the points that don't belong to the surface are deleted, and the methods of splitting, sorting and slicing are
used. The least square non-uniform B-Spline method is used to smooth and interpolate the section point cloud data, then
the data of reconstructed surface can be achieved after a secondary sampling. An example of the mouse is given to testify
that the method is effective and feasible.