Search Results

In recent years, spatial phase-shift technique has been applied in digital shearography to obtain interference phases, solving the problem that temporal phase-shift technique cannot be used in dynamic measurement. In this article, two optical setups of spatial phase-shift digital shearography are compared. The issue of optical setup parameter optimization is also discussed.

Temporal phase-shifting and spatial-carrier techniques are the two dominant phase extraction methods used in digital speckle pattern interferometry (DSPI). Temporal phase-shifting technique enjoys the advantages of precise phase extraction and capability of high-quality phase map formation, but suffers from the limitation in its use in dynamic measurement due to the time cost for carrying out the phase shift. To meet the rising demand of dynamic measurement of deformations in modern industry, spatial-carrier technique is developed to extract the phase information from a single speckle interferogram, yielding less time cost during a test. In this paper, both temporal phase-shifting and spatial-carrier techniques are investigated in theory as well as experiment. The experiment results from measuring a same deformation behavior using both techniques are used to compare the performances of the two phase extraction techniques.

Laser triangulation sensors suffer from the varied reflected light intensity which is due to the different object surface reflection characteristics. This varied reflected light intensity causes asymmetry of light spot shape, shifting of light spot center, and so on. This condition decreases the accuracy of measurement. Furthermore, it causes the linear CCD inside the sensor to go out of dynamic range, which is called blooming. Unlike plane CCD, linear CCD shows an asymmetry overflow phenomenon when blooming occurs. This phenomenon will limit the dynamic range and decrease the accuracy of laser triangulation sensor that is based on the linear CCD. However, we found the leading edge of light spot image on CCD array kept stable while scattering light intensity varied and the center of the light spot image shifted. Thus we designed a novel triangulation sensor which used the leading edge to indicate the location of light spot on the CCD, rather than using the center of light spot.

Adhesive bonding has many applications in the automotive industry. The single-lapped bonded joint is the most typically used among various bonding types. This paper presents experimental research for determining the strain field of the single-lapped joint under tensile loading. The materials for the joint are epoxy-based structural adhesive and low-carbon electrolytic zinc steel plate. In the study, a DIC (digital image correlation) system was adopted to measure the strain distribution of the bonded joint during a tensile test. The bonded steel coupons in the tensile test were prepared according to the ASTM standard. During the measurement, images of the coupon joint were taken before and after the deformation process. Then the DIC system measured the strain of bonded joint by comparing two consecutive images. The measured data from the DIC was compared to data taken simultaneously from a traditional extensometer.

A self-calibrating phase-shifting technique using a Michelson Interferometer is presented to measure phase distribution more accurately in Digital Speckle Pattern Interferometry (DSPI). DSPI is a well-established technique for the determination of whole field deformation via quantitatively measuring the phase distribution of speckle interferograms that use the phase shifting technique. In the phase shifting technique, the phase distribution in a speckle interferogram is quantitatively determined by recording multiple intensity images (usually four images) in which a constant phase shift, e.g. 90 degrees, is introduced between each consecutive image. A precise phase determination is greatly dependent on the accuracy of the phase shift introduced. The popular methods to minimize the error resulting from inaccurate phase shift use various algorithms and need to record five or eight images (rather than four images).