The Effects Of Sample-preparing Methods On Static Tri-axial Test Results Of Fine Sands On Digital Image Measuring System

Tri-axial test plays an important role in soil mechanics and engineering application. It not only can be used to determine the damage of soil strength and ultimate strength, but also obtain the relations of stress and strain. The tri-axial test results are affected by many factors; the sample-preparing method is one of the most important aspects. So it's essential to make the different test results clear into different sample-preparing methods, and for the static tri-axial test, there are also many problems in the traditional tri-axial deformation measuring method.By the application of computer digital image measuring technique to sample deformation measurement of soil static tri-axial test, non-contact measuring of soil specimen deformation was realized. For Nanning saturated fine sand, loose and dense samples were prepared respectively by dry-rapping method and wet-tamping method controlling the different initial moisture content. Combining the digital image measuring technical with traditional tri-axial deformation measuring method, consolidation drained shear experiments were made under different confine pressures, and analyze of the feasibility of the image measuring method. The end restraint and characteristics of the stress and strain calculated by the results of the top, the middle, the bottom and the whole of the samples were comparatively analyzed. Finally, the influence of relation of stress and strain, deformation and failure forms, different of tangent modulus 5 parameters of Duncan-Chang model and shear bands by sample preparation methods were researched.The test results show that a means with digital image processing technique for deformation measuring of soil specimen in tri-axial test can solve many disadvantages of conventional methods on sub-pixel accurate corner recognition.. It was because the different shear failure forms and the locations shear band that difference of axial strain and deviator stress by image measuring and traditional tri-axial test. Duncan-Chang model 5 parameters were affected not only by the sample preparation method, the initial moisture content, density, but also by the methods of measurement. Comparing the each strain of six lines of square with the strain of whole squares and consider regions of four squares in the middle can avoid the influence of the end restraint. The change of axial strain field can describe the initial formation, development and fully-developed of the sample shear band accurately. The digital image method can measure the local axial and radial strain field at any time, but it still has a limitation and need to be further refined and improved.