2.1 INSTRUMENTATION
Instruments Research Area Measurement
Miniature tip tensiometer Field, Lab - Matric Suction < 80 kPa Heat Dissipation Water Field, Lab - Matric Suction < 2500 kPa Potential Sensor
Soil Moisture Probe Field, Lab - Soil Moisture
Quantum Sensor Lab - Radiation energy
Pore Water Pressure Pullout Test - Pore water pressure Transducer (PPT)
Strain gauges Pullout Test - Force distribution along root model Load cell Pullout Test - Total pullout resistance of root model Linear Variable Pullout Test - Root model displacement
Differential Transformer (LVDT)
Table 1: Instrumentation
2.2 TEST PROCEDURES
The hydrological part of the project mainly involves the field investigation and laboratory test.
As mentioned in Chapter 1, the field investigation is done in a specific site where the re-compacted slope is constructed. The laboratory work mainly works to support and correlate the data with field monitoring under controlled parameters. Besides, suction behavior understood from site and laboratory enhanced the application level of root reinforcement effect.
In both areas, the same plant species (root modeling in physical testing), soil type and instruments are used, the only difference is site condition and laboratory condition with sufficient control of different parameters.
Soil used in the tests is Complete Decomposed Granite (CDG), which is commonly found in Hong Kong. GEO 2007 states that approximately 90% of the rock in Hong Kong is granite, and due to the weathering of granite, a lot of CDG has been formed. CDG occupies around 80% of developed areas in Hong Kong. At least a 95% Degree of Compaction (D.O.C) of soil has to be attained to achieve a higher shear strength for slope stability in Hong Kong (CEDD, 2006).
Therefore, the soil used in the tests is compacted to this level.
2.1.1Hydrological Part
There were two types of vegetation used in this area, consisting of Cynodon Dactylon (grass) and Schejjlera Heptaphylla (small tree), which are commonly found in Hong Kong and in other parts of Asia such as Malaysia, India and Vietnam, were used in the test. They are mainly used due to their high survival rate.
Cynodon Dactylon, as shown in Figure 2, is also known as Bermuda grass. The blades are a grey-green color and are short, usually 2-15 centimeters long with rough edges. The erect stems can grow 1-30 centimeters tall. The stems are slightly flattened, often tinged purple in color. It has a deep root system; in drought situations with penetrable soil, the root system can grow to over 2 m deep, though most of the root mass IS
less than 60 em under the surface.
Bare ground the fundamental understanding of root-soil-water interactions, a compacted embankment was formed to develop a practical and environmentally friendly reliability-based preliminary design framework for "integrated bioengineering live slope cover" for shallow soil slopes in Hong Kong.
The compacted embankment is formed as a trapezoidal sloping ground by completely decomposed granite (CDG) with 95% degree of compaction. The dimensions are 30m x 18m x 2m. The volume used for the embankment is 540 m3. The study was obtained on 33° slope located at the north part of the embankment.
Site Monitoring Plan
• Calibration and Installation of Instrument
• Data collection from site
• Without rainfall: per 3 days
• With rainfall: per 2 days
• Data Analysis
• Comparison with laboratory and Centrifuge result
Figure 3 Flow of Site Monitoring Plan
Before site monitoring, installation of instrument was carried with proper calibration to ensure a high quality of instrument and accuracy. Collection of data included suction in different cover and depth, rainfall and total evaporation. After instrumentation, site monitoring with data collection from site was implemented per 3 days without rainfall and 2 days with rainfall. Data collected from site was analyzed and compared with the laboratory and centrifuge result.
Therefore, a deeper understanding of suction induced behavior from vegetation can be achieved.
The flow of site monitoring plan is shown on Figure 3.
2.1.1.2 Laboratory Test
In general, the experimental procedure for the project is categorized into three sections, consisting of Pre-test, Test and After Test. The flow chart can be seen in Figure 4. Before the test begun, the soil moisture probe (Tensiometer) was saturated to remove air retained in the wall of ceramic cup, and this is important to ensure accurate measurement. Also, Datalogger Software-PC 400, all instrumentation and rainfall simulator must be installed and ready for the specific test series. For instance, the initial room condition must be assured as this will affect the induced suction in the experiment.
• Satnration ofTensiometers
• [m;tallationofDatalogger Software-PC400
• lru;tnunental Set Up
• Rainfall simulator; potential evaporation
• Apply rainfall (Intensity: I 00 nunfhr)#2
• Conclitiom: WettingandDry:ing
• [nstallation of tensiometers mld water potential sensor (pasted with kaolinite)
• Daily monitoring and data collection
• Cany out next test
• Data analysis: Graph Plotting
Figure 4: General Procedure for Laboratory Test
The saturated tensiometers were installed at different depth of the test box - 40mm, 80mm, 140mm and 210 mm respectively. Test boxes were exposed to the environment and suction would be observed before conducting any experiment. Once the initial suction had reached around 50~60 kPa, rainfall of intensity 100 mm/hr was applied on the plants. Next, readings were taken during rainfall. Daily monitoring, data collection and data analyses were also conducted Figure 5: Atmospheric Controlled Room after the test, for the period of around 1 to 2 weeks.
2.1.2 Mechanical Part 2.1.2.1 Pullout Test
SGI SGZ
SGJ SG4 SGS
Figure 6: Root Model for Pullout tests
Root Models for pull out tests were made of Viton Rubber tubes with outer diameter of 19mm and inner diameter of 13mm. It was 300mm long, with strain gauges installed at positions shown in Figure 6.Strain gauges are sensors measuring material strain or bending moment. In order to get the result, four strain gauges were installed around the same location.
In order to prevent the damage of strain gauges during the test and simulate real root surface, the main root and branches was cover by 1 mm thick silicon grease. It effectively protected strain gauge from shearing as well as potential damage of water and ensured the stability of strain gauge results.
Figure 7 Test Setup
An aluminum box with a dimension of 541.8x541.8x380 mm was placed in a big testing box.
The hydraulic jack was installed on two metal bars fixed on the sides of testing box. The root model was connected to the end of hydraulic jack by a metal clamp. A load cell was placed above the clamp to measure the whole pullout resistance of root model. The L VDT was at the back of hydraulic jack, which measured the displacement of hydraulic end instantaneously. The signal wires of LVDT, Load cell and strain gauges on root model were connected to the server beside the testing box, which record the data throughout the testing process. Three PPTs were placed at different heights in the soil sample, measuring the matric suction value.
Soil was compacted with root model vertically buried. After two-day equilibrium, the values of PPTs were recorded for soil matric suction. The server was turned on, measuring the values of L VDT, load cell and stain gauges. The valve of oil supplier was opened to supply oil to the hydraulic jack. Hydraulic jack started to pull out the root model in a constant speed no more than 10 mm/min. The slow speed ensured that the process was in drained conditions, and dynamic effects were minimized. The tests continued until root model was fully pulled out. The values of LVDT, load cell and strain gauges were recorded and memorized by the server instantaneously.
These data recorded was to be analyzed as test results.