Evaluation of Strain Variation Within a Triaxial Specimen Due to End Effects

A new triaxial test setup is described that has been developed to test large specimens (150 mm in diameter and 300 mm long). The system was designed to use eight proximitors: two placed on opposite ends of a diameter at midheight of the specimen to measure lateral deformation and six placed in pairs along the specimen and distributed at 120 degrees around the specimen to measure axial deformations. To calibrate this setup and to cover the typical range for highway base materials, nine elastomer specimens were manufactured with three different stiffnesses. A quasi-static test program was implemented with these specimens to evaluate the capabilities and shortcomings of the system, the repeatability and accuracy of the test results, and the optimization of the gauge length to be used for measuring axial strains in the production test system. An axisymmetric computer model was assembled to predict axial and lateral deformations at the measuring points used in the experimental program. The experimental and numerical results are in fairly good agreement except in the vicinity of the loading end platens. The numerical results indicate overprediction of the moduli by about 15 percent when the gauge length used is the whole length of the specimen instead of the central one-fifth. The experimental data show somewhat higher discrepancies (about 20 percent) for the same range of gauge lengths.