CHAPTER 1 INTRODUCTION
1.3 Structure
The subject of this thesis is about the procedure and apparatus of the test of Taipei clay at small strain. The skeleton of the thesis is as follows.
Chapter 1 introduces the thesis, including the background of small strain triaxial test and the object of this research.
Chapter 2 describes the common errors in conventional triaxial test apparatus.
For the purpose of performing tests at small strain and getting more accurate results, the fist step is to find the errors sources out and remove them. There are numerous ways to modify the traditional test instrument, and the most effective is to build a new one which modifies all defects that the traditional one has. In order to get precise de-formation data of soils, to abandon the external dede-formation measurements and adopt the local strain transducers is necessarily. And it will be discussed in detail in chapter 2.
After talking about the errors and measurement in chapter 2, the chapter 3 brings up the new pressure controller and loading system which is special designed for tests of Taipei silty clay at small strain. In previous works, the triaxial test apparatus, the product of GDS from England, has some defects on design. For this reason, we started to set up a whole new small-strain triaxial test instrument. In order to ensure the ac-curacy in tests and minimizing the errors from gears, we chose the different transmis-sion and motor. The new motor has advantages of high torque and precitransmis-sion, and the transmission between motor and screw is more direct than GDS triaxial test machine.
The author will introduce the new triaxial test apparatus explicitly in chapter 3.
Chapter 2 and 3 focus on the hardware of the new test apparatus, and chapter 4 will introduce the software, included the test program and data acquisition system.
The test program is according to soil theories, ex: k0-consolidation, SHANSEP con-cept and so on, to perform triaxial test automatically and reasonably. The data
acqui-sition system has quite high resolution to capture the response of transducers better.
And the final objective of this chapter is to create a test program which can perform the test in any way we want.
Chapter 5 assembled the results of tri-axial test at small strain through the new apparatus. The test type of triaxial test is CK0U_AC, meant the sample was consoli-dated under K0 state and sheared with undrained condition. Besides the small strain triaxial test, we use the bender element to make some waves transmit through the soil sample and get the parameters. Accordingly, the author will collect all of the test data and point out the stress –strain properties at small strain of Taipei silty clay.
At last, in chapter 6, we can make some discussions and suggestions base on the results from chapter 5. And we are trying to provide more exact parameters of Taipei silty clay for engineer.
Chapter 2
The Measurements in Triaxial Test Apparatus 2.1 Strain Measurement in Triaxial Test
2.1.1 Errors in conventional triaxial test apparatus
The traditional triaxial test apparatus was unsuitable for tests at small strains as a result of the inevitable errors. Jardine, et al. (1984) said that the standard Bishop &
Wesley triaxial test apparatus has a lot of errors in measuring axial deformation through external transducers, and these errors always produced larger axial strain.
Jardine, et al. also pointed out the errors might cause the unreasonable test results, which is at small strain and using conventional triaxial test apparatus, as follows:
1. It’s difficult to cut the specimen into perfect cylinder.
2. The way to connect load cell and top cap is not good.
3. Bedding errors – the specimen can’t connect tightly with top cap and pedestal due to the un-smooth and porosity on top and bottom plane of specimen.
Goto et al. (1991) mentioned that, for a triaxial specimen, there are several sources of error involved in the external axial strain measurement, which are:
1. System compliance due to the deflection of load cell, top cap, cell, piston and so on,
2. Tilting of the specimen,
3. Bedding error on the top and bottom of the specimen,
4. Strain non-uniformity of the specimen due to the end restraints, leading to bulging of the specimen, and
5. Shear banding (strain localization) of the specimen.
And Atkinson (1993) pointed out the possible errors source as follows:
1. The deformation of top and bottom of specimen.
2. The deformation between loading shaft and triaxial cell.
3. The deformation of load cell.
4. The deformation of triaxial cell.
According to above, we can summarize all possible error sources in conventional triaxial test instruments described as follows:
I. Load cell was placed out from the triaxial cell, and the friction between the shaft and bearing might affect the test results.
II. External axial deformation measurement might include the deformation of triaxial cell, top cap, porous stone, pedestal, load cell, and shaft etc; it doesn’t represent the actual deformation of soil.
III. It’s unavoidable that the top and bottom plane isn’t parallel because cutting the specimen.
IV. Specimen incline and bedding error are hard to avoid in the process of setting up the specimen.
V. The un-uniform deformation caused by top cap rotate.
VI. Shear banding of the specimen.
VII. Environment disturbance (temperature change, vibration and etc.).
The conventional triaxial test apparatus usually use “external” deformation measurement, ex: external LVDTs. The measurement result (Δ) maybe included lots of errors, as shown in Fig. 2.1, and it represent the whole deformation of this system.
Pedestal
In equation (1), the suffix of each term represents: LC=Load Cell, S=Sample, TB=Top Bedding, BB=Bottom Bedding etc. Actually, what we interest in is the really deformation of soil ( ). Therefore, the error due to external measurement can be express as equation (2).
Δs
Pedestal
According to the test for London clay performed by Jardin, et al. (1984), and using both internal and external strain measurement. The test results are shown in Fig.
2.2, and quite different from each other. It’s obviously that the data measured from external measurement covered the behavior of high initial stiffness and non-linearity of North Sea clay at small strain.
Another example is shown in Fig. 2.3 (Goto, Tatsuoka et al., 1991) in which the results of consolidated undrained compression test presented the extremely large discrepancy between external and internal measurements. And the axial strain is apparently overestimated while measured by external deformation transducers. In Fig.
2.4 presented by Diego C. F. Lo Presti et al., (1995) show how the errors in measuring axial strains influence the Young’s modulus of Pisa clay. It can be seen that the Young’s modulus determined from external axial strains with a large-capacity LVDT are quite unreliable forεa <0.01%.
2.1.2 Local displacement transducer
As mentioned above, the external strain measurement will caused serious errors.
Therefore, how to erase the errors is an important task before investigating stress-strain behavior of soils at small strain. In past three decades, many researchers were involved in how to measure the deformation of soils more precisely. And the best one of these methods is to use the “local” displacement transducers, because all errors can be almost obviates.
2.1.2.1 Electrolytic liquid levels
Cooke and Price (1974) describe the use of electrolytic liquid levels as horizontal inclinometers for the measurement of vertical displacement around experimental piles in the field. Daramola (1978) and Costa-Filho (1980) measured relative displacement between two reference footings using local strain devices over a central length of a sample of clay and sand in laboratory tests, respectively. These devices, strictly speaking, are suitable only for very small-strain levels as a result of they are unwieldy and can suffer from jamming and damage at large strains. Although those devices are not good enough, but they create the new options of measurements and influence the
subsequent development of devices.
Burland and Symes (1982) made use of electrolytic levels to develop an axial displacement gauge, which has the resolution less than 1μm over a range of 15mm.
This gauge is the first one that can measure a wide range of axial displacement, and test results demonstrated the gauge is simple to use and applicable for wide spectrum of soils. The arrangement using an electrolytic level is shown in Fig. 2.5. The liquid level consists of an electrolyte sealed in a glass capsule. Three co-planar electrodes protrude in a line into the capsule and are partially immersed in the electrolyte. The resistance between the central electrode and the outer ones varies as the capsule is tilted. The principle can be used for measuring the changes in length.
Fig. 2.6 shows an arrangement, which converts a change in height Δh to a change in slope Δθ. There is a spring-loaded ball hinge at C; the hinge at D rotates in two orthogonal vertical planes. Pads C and D are glued to the specimen membrane. A vertical deflection gauge, which can be used in a triaxial apparatus, is shown in Fig.
2.7. The input voltage of the gauge is 5V supplied by an AC power, and the excitation frequency is 5 kHz. The gain was adjusted to give an output sensitivity of approximately 0.4 V per degree tilt, with a scatter of ±0.002 V. The electrolyte level is sensitive to the changes in temperature, so it should be operated in the conditions, which are controlled to within ±5 . Local measurements have been measured to an ℃ accuracy of ±2μm.
The typical results from an undrained test with fixed end platens on an undisturbed sample of London clay is shown in Fig. 2.8. Although the fixed end platens were used to reduce the influence of rotation of platens, the effect of sample tilts during early stages of shearing was still greatly significant. The two deflection gauges show clear evidence that this is due to sample tilt. The average of the readings from the two deflection gauges gives reasonable results that are not affected by the bedding errors.
The deflection gauge developed by Burland & Symes (1982) was improved by Jardin, et al., (1984) and carried out a series of undrained triaxial tests through making use of the gauge and measured the stress-strain behavior of geomaterials, including chalk, clay and sand at small strain. The accessories of the gauge are shown in Fig.
2.9, and the resolution of this gauge is less than 1μm.
It is beneficial that using local displacement transducer to measure the axial displacement of the sample. Nevertheless, do the readings from the transducers on the specimen represent the actual deformation of the specimen? Is it possible that relative movement occurs between the specimen and membrane? Gens (1982) used an optical technique to demonstrate that the membrane only moves in relation to the sample when large strains are developed
2.1.2.2 Hall Effect Semiconductors
Clayton and Khatrush (1986) developed a local strain device made of the semiconductor material, which make use of the Hall Effect, discovered by E.H. Hall in 1879. Hall Effect means that consider a metallic or semiconductor plate through which electrical current is flowing. If this is placed in a magnetic field where the flux lines are perpendicular both to the plate and to the current flow, the charge carries will be defected. A voltage will thus be produced across the plate in a direct normal to the current flow. This voltage is termed the Hall voltage. Hall Effect semiconductors are used widely as switches and to measure magnetic flux density.
Clayton et al. (1989) described the development of Hall Effect semiconductors in geotechnical measurement. Fig. 2.10 shows the temperature compensated semiconductors manufactured by Micro-Switch have been used since 1985. Four basic configurations, shown in Fig. 2.11, of sensor and magnets have been used in geotechnical instrumentation at the University of Surrey. Fig. 2.12 shows the influence of varying both the separation S between the magnets and the gap G between the magnets and the semiconductor for a double magnet, bi-polar side-by configuration. Up to a separation S approximating to the minimum face width of the pole of the magnet, an increase in the gap results in an increase in the linear range of the device. Decreasing the gap G can increase the sensitivity of the output. Therefore, in order to get highly sensitive displacement measurement the spacing is reduced to zero, and the gap is reduced to as little as possible.
The first use of linear-output Hall Effect semiconductors was to control radial strains during K0 triaxial compression test on sand at the University of Surrey in 1983.
The prototype of the local axial strain gauge is as shown in Fig. 2.13. And Fig. 2.14 shows the design of the Hall Effect local strain gauge in Clayton and Khatrush (1986).
And this gauge is consisted of two major parts:
I. A spring-mounted pendulum that holds two bar magnets each had 3mm×3mm square cross-section area and separated by 3mm. This is suspended from an upper pad fixed to the specimen by pins, and bounded to the membrane by adhesive. The spring allows relative motion between the pendulum and the fixing pad.
II. The Hall Effect semiconductor encapsulated in epoxy resin within a brass container, which is mounted on the specimen by means of a pinned fixing pad.
In modern soil mechanics laboratory equipped with the Hall Effect semiconductor transducers, the resolution of the gauge is better than 1 µm, which is equivalent to an axial strain of less than 0.002%.
2.1.2.3 Local deformation transducer
There is another axial displacement measurement, LDT (Local Deformation Transducer), shown in Fig. 2.15 and developed by Goto, et al., (1991). Each LDT consists of a thin, hence flexible, strip of phosphor bronze and a couple of pseudo-hinged attachments. As shown in Fig. 2.15, four strain gauges, two on each side, are glued at the central part of the strip. The working principle of LDT is such that, when the specimen is deformed, the change in distance between the two attachments triggers the increase (or decrease) of the gauge strain, which in turn determines the local axial strain over the gauge length. The axial strain is the average of those measured using a couple of LTDs that are instrumented right opposite to each other (see Fig. 2.15). The resolution of the LDT is 0.6 µm, and strain levels range from 10-6~10-2.
2.1.2.4 Proximity transducer
The proximity transducer contains one metal target and the transducer. The working principle of the proximity transducer is such that, the deformation of sample
produced the axial displacement, and the magnetic field between the target and the transducer would be change, and this resulted in the change of the voltage. The variance of voltage represents the axial displacement of sample. Kokusho (1980) and Hird and Yung (1989) used the proximity transducer to measure axial displacement in dynamic and static triaxial tests, respectively. The resolution of the proximity transducer can reach as small as 1μm or less.
Matsumoto et al., (1998) made use of the proximity transducer (or so-called gap sensor) to measure the local strain of the sample in the triaxial tests. The arrangement of proximity transducers on sample was shown in Fig. 2.16. Generally speaking, the resolution of the proximity transducer is 0.5~1 µm. But, in recent year the researchers adopted the new LVDT, which is rather small and light, and to replace the proximity transducer. The resolution of the new LVDT is similar to the proximity transducer.
2.2 The Pore Pressure Measurement in Triaxial Test
The most common pore pressure measurement is making use of the electrical pressure transducer to measure the pressure of sample. However, the ordinary arrangement of pore pressure transducer is to connect the transducer to the bottom of the sample through nylon tube. In other words, the pore pressure measured by transducer was the pressure at the bottom of the sample, and there were lots of connectors and tubes in the path of measurement. Therefore, there are two major possible errors in measuring the pore pressure using the above method:
I. The pore pressure measured from the bottom of sample cannot represent the true pore pressure of the sample, due to the un-uniformity of the distribution of the pore pressure in soil samples, especially for clay.
II. The measurement path consisted of many connectors and nylon tubes. In the process of measuring, there might be a few air bubbles hide in connectors.
The bubbles will cause un-estimative errors in pore pressure measuring.
In order to reduce the errors and to measure the pore pressure at different positions of sample, we adopt one kind of miniature pore pressure transducer to achieve the above objects. The feature of the miniature pore pressure transducer is the very mini size, so as to place it in any position we want.
2.2.1 Standard Pore Pressure Transducer-PMP 1400
The base pore pressure measurement is making use of the pore pressure transducer PMP1400 from Druck. The range of pressure measurement is -1 bar to 4 bar. The word “standard” is meant that the transducer is commonly used in conventional triaxial test apparatus to measure the base pore water pressure. In addition to use the base pore pressure transducer, we also adopt the miniature pore water transducer to measure the pressure at the middle of sample. Therefore, the pore pressure measured at different position through two kinds of pore pressure transducer can be monitored and get comparison.
2.2.2 Miniature Pore Pressure Transducer-PDCR81
The miniature pore pressure transducer used in this research is PDCR81 from Druck. The PDCR81, shown in Fig. 2.17, consists of a 0.09-mm-thick, single-crystal, and silicon diaphragm with a fully active strain gauge bridge diffused into the surface.
A high air entry porous stone is placed in the tip of the transducer, just overlying the diaphragm. One side of the diaphragm is exposed to the atmosphere via the transducer cabling, while the other side is exposed to the pore water via the porous stone. The deformation of the diaphragm causes a change in voltage measured across the strain gauge, which is equated to pressure.
The small size of the PDCR81 allows it to be inserted easily into soil samples while causing minimal interference. The small size also leads to a quick response time since only a small amount of fluid is required to flow into or out of the device for a given change in pressure. This quick response time allows the PDCR81 to be better used for real time monitoring of pore pressure changes during testing, including dynamic events. Therefore, miniature pore pressure transducers, such as the PDCR81, are currently used in a variety of geotechnical testing applications for measuring pore water pressure.
The kit supplied comprises the following components: Druck PDCR 81 pressure transducer, 8.5mm drill and 3/8" UNF Tap, Cutter and Rubber Block, Flanged Grommet, Pair of o-rings and o-ring stretcher, Right Angle Bracket. Their purpose and installation is described below.
1. Druck PDCR 81 pressure transducer. The transducer is provided with a small ceramic tip. This is the sensing end of the transducer and is placed
1. Druck PDCR 81 pressure transducer. The transducer is provided with a small ceramic tip. This is the sensing end of the transducer and is placed