Date of Graduation

12-2017

Document Type

Thesis

Degree Name

Master of Science in Civil Engineering (MSCE)

Degree Level

Graduate

Department

Civil Engineering

Advisor/Mentor

Coffman, Richard A.

Committee Member

Bernhardt-Barry, Michelle L.

Second Committee Member

Oommen, Thomas

Keywords

Close-range Photogrammetry; Laboratory Apparatus; Local Deformation; Photogrammetry; Triaxial Testing; Volume Measurement

Abstract

Accurate measurement of axial, radial, and volumetric strain parameters are critical to the understanding of phase relationships and the constitutive behavior for saturated and unsaturated soils. The use of photographic monitoring techniques for laboratory-based measurement of these parameters have become common. A novel technique that utilized camera instrumentation located within the triaxial testing cell was developed and validated. By placing the instrumentation inside of the cell, instead of the instrumentation being located outside of the cell, the technique eliminated cumbersome corrections required to account for optical distortions due to 1) the refraction of light at the confining fluid-cell wall-atmosphere interfaces, 2) the curvature of the cylindrical cell wall, and 3) the deformation of the cell wall induced by changes in cell pressure. Digital images of various soil and analog (brass, acrylic) specimens were captured within the triaxial apparatus during testing. The images were processed using the principles of close-range photogrammetry to construct three-dimensional models of the specimens. The models were analysed to determine surface deformation and total volume of the specimens. Additionally, the models obtained from triaxial tests performed on the soil samples were compared to quantify deformation and volume of the sample as a function of axial strain. Sensitivity studies and evaluation of measurement accuracy for the internal, close-range photogrammetry approach are documented herein. Specimen volume, as obtained using the approach, was compared with volume obtained from four other techniques, including: DSLR camera photogrammetry, 3D structured light scanning, manual measurements (caliper and pi tape), and water displacement. A relative error of 0.13 percent was assessed for the internal photogrammetry technique. The viability of determining total and local strains, volumetric changes, and total volume at any stage of triaxial testing was demonstrated through undrained triaxial compression and extension tests. Results from all tests are presented herein. The use of the internal, close-range photogrammetry technique is recommended.

Download

Share

COinS