Strain Measurement Research Based On Digital Image Correlation Method

Digital image correlation is a method which can calculate theinformation, such as displacement of the images caught before and afterdeformation, through the related criterion function. Since it has severaladvantages: convenient operation, high accuracy, simple equipmentinstrument and full-field measurement etc., it has been developed and appliedgreatly in recent years, especially in strain measurement. Exploring how toimprove the accuracy of the method is always the hot spot to the researchworkers.In this paper, the author improves the algorithm of the digital imagecorrelation method to control the maximum correlation coefficient into anarrow area before the quadric surface fitting. Thus, it will raise the theorycalculation accuracy through further quadric surface fitting. The author also optimizes the interpolation process during the implementation of program toless time-consuming. Using the simulation speckle images which can becommanded in theory, we can verify the improving algorithm that it willbring the accuracy up to10-3pixel. Besides, this article focuses on thetheoretical calculation error factors. Through analysis, the author gives thebest range for the parameter value to guarantee the calculation accuracy andsave time.Finally, the author carries out a series of experiments, like translationalexperiment, tensile experiment and experiment under complex stress. Withthese experiments we can easily verify the specific applicability and comparethe ordinary DIC method with the improving algorithm DIC method. Itshows that improving-DIC has a better accuracy, practice application,stability and smaller variance. The author also puts forward relatedoptimization methods combined with specific experiment to reducecalculation error and improve the precision of the calculation. This can bemore fit to the actual measurement requirements.