This work proposed a method to reconstruct the 3D bubble shape in a transparent medium utilizing the three orthographic digital images. The bubble was divided into several ellipse slices. The azimuth angle and projection parameters were extracted from the top view image, while the formulas for dimensionless semi-axes were derived according to the geometric projection relationship. The elliptical axes of each layer were calculated by substituting the projection width into the formulas. All layers of slices were stacked to form the 3D bubble shape. Reconstruction accuracy was evaluated with spheres, ellipsoids, and inverted teardrops. The results show that the position contributes greatly to the reconstruction accuracy of the bubbles with serious horizontal deformation. The method in Bian et al. (2013) is sensitive to both horizontal and vertical deformations. The vertical deformation has little influence on the method in Fujiwara et al. (2004), whereas the horizontal deformation greatly impacts its accuracy. The method in this paper is negligibly affected by vertical deformation, but it does better in reconstructing single bubbles with large horizontal deformation. The azimuth angle affects the accuracy of the methods in Bian et al. (2013) and Fujiwara et al. (2004) more than the method in this paper.

The time-dependent characteristics of transparent soil strength, composed of magnesium lithium phyllosilicate, is important for applying a thixotropic clay surrogate. The gas injection method was employed to obtain the strength, represented as cracking pressure, which was then correlated to variables including rest time, disturbance time, and recovery time. Three concentrations (3, 4, and 5%) were tested. The results show that the strength was directly proportional to the rest time, recovery time, and concentration while the disturbance time reversed. The calculated limit strengths for 3%, 4%, and 5% transparent soils were 3.831 kPa, 8.849 kPa, and 12.048 kPa, respectively. Experimental data also showed that the residual strength for higher concentration transparent soil was more significant than the lower ones. The elastic property immediately generated partial strength recovery after disturbance, while the viscosity property resulted in a slow recovery stage similar to the rest stage. The strength recovery rate was also sensitive to concentration. Furthermore, the strength with 3%, 4%, and 5% concentrations could regain limit values after sufficient recovery, which were calculated as 4.303 kPa, 8.255 kPa, and 14.884 kPa, respectively.

No abstract