Applying a Modified Dynamic Cone Penetration Test to Detect Vertical Soil Structures

• Main Authors: Ming-Hung Cheng, 鄭名宏
• Other Authors: Wei-Li Liang
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/x726j4

碩士 === 國立臺灣大學 === 森林環境暨資源學研究所 === 106 === Soil depth and soil–bedrock interface locations are critical indexes for land utilization and management. The soil–bedrock interface is also a crucial boundary affecting slope stability and hydrological dynamics. Therefore, obtaining complete information on underground soil structures would improve the grading of land utilization and the evaluation of potential landslides. Generally, the soil–bedrock interface is believed to be the potential destruction location in shallow landslides; however, few empirical studies validated this hypothesis. Furthermore, properties of this interface remain unclear. In other words, potential destruction locations of shallow landslides warrant further investigation. This study analyzed the location and characteristics of a landslide destruction interface in a landslide event by using pre-existing data of soil structure before the landslide event. By investigating distributions of soil resistance and variations in surface topography, we found that the entire area was falling down and that the topography and slope of the soil–bedrock interface strongly influenced soil movement. A core penetration test (CPT) is a direct measurement technique for detecting soil depth and obtaining soil resistance (N_h) on a hillslope. However, detecting spatial distributions of soil depth using a CPT requires a large sample size and much time. In this study, we developed a dynamic cone penetration test (DCPT), a form of CPT, to reduce the investigation time in the field. The DCPT can obtain the soil resistance (N_pd) through penetration. The DCPT features a knocking engine instead of the knocking weight of the CPT. We used an experimental device containing layers of varying hardness to evaluate the DCPT and CPT; the vertical distribution of penetration resistance as measured using the two tests corresponded well to each other. The relationship of the penetration resistance measured using the DCPT and CPT can be expressed through a linear regression—N_pd = 0.0197 N_h—with an R2 value of 0.84. On applying the DCPT in a field and digging a trench close to the penetration cone, we obtained the following results. In a simple environment, the DCPT and CPT successfully yielded the same soil depth, soil–bedrock interface, and characteristics of soil resistance. Similarly, in a complex environment, the DCPT and CPT yielded the same characteristics of soil resistance and identified the soil–bedrock interface; however, the DCPT overcame more obstructions caused by small rocks to detect deeper structures in shorter time. These results demonstrate that the DCPT greatly reduces survey duration without compromising on portability. Thus, the DCPT is an efficient method for obtaining underground information in the field.