Observation of unsaturated seepage behavior
Unsaturated seepage examination was carried out by using MRI. The MRI photography results of horizontal seepage examination using dry Toyoura sand are shown. The absorbing water process is shown at intervals of 50, 231, 531, and 920 seconds in Fig. 4a, b, c, d. Water is supplied at the left edge and spreads to the right edge. Measuring the time for the examination begins when the resin valve is opened. The high-water-content part is shown as a whiter image. Laboratory dishes with an outside diameter of 50.7 mm (and an inside diameter of 47.4 mm, and a height of 14.1 mm) are placed under the ring cells, to measure the sample water content (w = 25%, 20%, 15%, 10% and 7.5%). The soil that was adjusted for each water content is filled in those dishes. These are used as an index for estimating the water content of the sample by MRI. For photographing in the MRI device, iron sand must be removed from the Toyoura sand (approximately 0.4% overall at mass ratio). The dry density is around 1.6 mg/mm3. At this time, the dry density of the sand in the laboratory dishes should be of the same value. Hardly any image appears when the water content is less than 10%.
The seepage proceeds approximately perpendicular for a short time after the start of the water supply, but over time, becomes faster in the lower part of two cylinder than in the upper part. These can be confirmed from Fig. 4. It is revealed that the water movement is affected by gravity even with 30 mm differences in height. There is a clearer white and black boundary in the area of the seepage front. Hereafter, this boundary will be called the seepage line. It can be confirmed that a seepage line is a disturbance (Fig. 4c). It is suggested that discontinuous water content change occurs on the seepage line. There is a black circle in the cylinder around 210 mm position (Fig. 4d). This is considered an air trap.
The MRI photography results are shown by horizontal seepage examination after packing Kaolin clay into the right part of the cylinder and Toyoura sand into the left part. The absorbing water process is shown at intervals of 735, 1185 and 3654 seconds (Fig. 5a, b, c) and the cross-sectional imaging result is shown 5072 seconds later around the 193 mm point (Fig. 6).The laboratory dishes filled with Kaolin clay over the acrylic cylinder with a water content of 50, 35, 25% are shown on the left. The laboratory dishes filled with Toyoura sand under the acrylic cylinder a water contents of 25, 15, 10% are shown on the left in Fig. 5.
It is confirmed that seepage advances from the bottom as in the case of the cylinder filled with only Toyoura sand (Fig. 5a). The seepage line is almost orthogonal to the penetration direction when the seepage line reaches the Kaolin clay, and, it can be confirmed that the seepage front line of the Toyoura sand and the Kaolin clay are in a similar position (Fig. 5b). In the wetting process, it is because the seepage speed through the Toyoura sand layer is equal to that through the Kaolin clay layer. In addition, a black part can be identified on a border of the Kaolin clay and the Toyoura standard sand. As seen from Fig. 5c) it can be confirmed that the light and shade are not a uniform border of the left side and the bottom of the Kaolin clay and the Toyoura sand. The seepage line moves to the bottom before it reaches the Kaolin clay. It is shown that the moisture holding ability of the Kaolin clay is very much greater than that of the Toyoura sand.
As can be seen from Fig. 6, a black part towards lower left direction on a border of the Kaolin clay and the Toyoura sand. This seems to be an air trap, and the possibility that such an air trap occurs between materials of different quality of soil.
Water content estimation of the sample by MRI
It is extremely useful if the seepage behavior from a MRI image can be evaluated continuously and in a non-destructive manner. It was shown that the logarithm approximation curve and the water content adjustment were the highest numbers of the pixel level of the laboratory dish on the cylinder in Fig. 7. Because the correlation coefficient R2 had a value of around 0.9, which is a high value, there is a high correlation between the pixel value and the water content. It should be noted that the water content was 0 but the pixel value of the MRI image was not 0, so a logarithm approximations was used.
Figure 8 shows the water contents vs the distance in the Toyoura sand. The water contents were evaluated from the MRI image and a laboratory test (Araki et al. 2011) (Hereafter, the indicative method) respectively. In the MRI method, the water contents were evaluated from the average of the vertical section. As can be seen there is an excellent agreement between the MRI method and the indicative method.
An evaluation of the unsaturated seepage behavior by the MRI method is shown in Fig. 9. These calculation results of the Fig. 9 from Fig 4a, b, c, d respectively. It was possible to evaluate the water content and distance relations continuously by using the MRI method. It was confirmed about the water content that the seepage time increases at the same distance (Fig. 9). The water content of the part which is near to the water supply section was evaluated as having a small value, in the water content and distance at 50 seconds later. It is thought that this is affected by the boundary conditions. The seepage line reaches the position of approximately 25 mm 50 seconds later, the position of approximately 90 mm 231 seconds later, the position of approximately 180 mm 531 seconds later, the position of approximately 250 mm 920 seconds later. It is understood that it becomes difficult for the seepage line to go ahead as time passes. All water content tends to decrease when it passes a certain spot. It is estimated that the air trap of Fig. 4d, produces the low water content part at 210 mm 920 seconds later as shown in Fig. 9.
