• No results found

Large-Scale Triaxial Equipment .1 General Layout



4.2.2 Large-Scale Triaxial Equipment .1 General Layout

confining cell were measured to determine the lateral and vertical stresses applied by the compressed specimen (Hicher, 1998). By using elastic theory applied to the steel wall, the horizontal stresses in the ballast specimen at the level of the gauges were computed.

Eight electrical strain gauges (Showa, 25 mm long) were installed at two levels in two diametrically opposed positions and were identified from 1 to 8, horizontal and vertical, left and right, top and bottom, as presented in Figure 38. These strain gauges were connected to two strain meters (HBM Digitaler Dehnungsmesser DMD) that allowed the measurement of strains (in microstrain) on separate channels. The hoop and longitudinal strain gauge readings were then transformed into lateral and vertical stresses at the selected points by the use of calibration charts that were established prior to testing and discussed in detail in Section 4.3.

4.2.2 Large-Scale Triaxial Equipment

Chapter 4: Test equipment and calibration

V = vertical H = horizontal L = left R = right T = top B = bottom





VLT-1, VLB-3 HLT-2, HLB-4

VRT-5, VRB-7 HRT-6, HRB-8

Figure 38. Position of electrical strain gauges on the consolidation cell walls

pore pressure measurement device and the axial and volume change measurement systems. A schematic diagram of the triaxial apparatus is given in Figure 39, whilst Figure 40 presents a detail of the set-up of the specimen inside the triaxial chamber (Indraratna et al., 1997). Triaxial Chamber

A triaxial chamber enables a vertical cylindrical sample to be confined inside a membrane by a uniform radial fluid pressure while an independent axial stress is applied to the specimen by a piston. In this investigation, the neoprene membrane had an average thickness of 4 mm whilst the specimen itself was encapsulated in a stainless steel cell (615 mm outside diameter and 945 mm in height with a 30 mm wall thickness) capable of resisting confining pressures as high as 4 MPa. This triaxial cell was designed to enable testing of rockfill specimens. Pore water pressures exceeding 1 MPa can be developed and accurately measured. The cell body was provided with three high

Figure 39. Schematic diagram of the triaxial loading unit

Figure 40. Specimen set-up inside the triaxial chamber (Indraratna et al., 1998)

Chapter 4: Test equipment and calibration

strength glass windows for visual observations (Figures 39 and 40). The top and bottom plates of the triaxial cell were drilled for connections which enabled the application of confining pressure, specimen vacuum, back saturation and drainage and facilitated pore water pressure monitoring, as shown in Figures 41. Upper and lower platens, provided with 10 mm thick porous brass plates, allowed the passage of water for saturation of the specimen and the required measurements, but prevented any loss of fines into the drainage or measurement systems (Figures 40). A portable crane was used to assemble the specimen. The entire chamber was moved under and out from the loading frame by mean of three detachable wheels. During testing, the wheels were removed to ensure full contact between the cell bottom platen and the loading frame stand. Vertical Loading Unit

The rigid load frame consisted of four steel columns (70 mm diameter) and a cross-head (Figs. 40 and 41). The frame incorporated a hydraulic loading system, located at the top, capable of applying static forces up to 200 tonnes, equivalent to a 140 MPa pressure on a piston 133 mm diameter. The double action (push/pull = 1987/993 kN) hydraulic jack (Enerpac, RR-2006) had a maximum stroke of 150 mm, so the maximum achievable axial strain was 25%. This vertical loading system could operate in either a constant strain or a constant stress regime. A two-stage directional controlled electro-hydraulic pump (Reuland oil pump), capable of a maximum pressure of 70 MPa, provided the hydraulic pressure to the axial loading unit.

A pressure transducer (Shaevith) was connected to the hydraulic jack cylinder and linked to a digital display (Amalgamated Instruments PM6-IV, reading accuracy of

Figure 41. View of the triaxial cell set for a test

1 kPa) mounted on the control panel. This enabled accurate control of the applied axial pressure on the specimen. The normal load was applied to the test specimen by a 100 mm diameter stainless steel ram located at the top of the cell. Ram friction was minimized by the use of the bond centered bronze (bush) bearing system, a modified system of those employed by others (Lee, 1986; Indraratna et al., 1993). It proved to be effective in minimizing the friction between the ram and the cell collar and as most systems the piston would fall, gradually, under its own weight. Air Pressure and Water Control Unit

Two separate control units were required: the first monitored and controlled the supply of air, whilst the second controlled the supply of the water required for saturation of the

Chapter 4: Test equipment and calibration

specimen and for the application of hydraulic pressure. Air pressure (maximum 1 MPa) was supplied to the transfer chamber (air/water receiver) mounted beneath the control panel from a mobile air compressor unit. For tests performed at a lower range of confining pressure, in lieu of the standard air compressor, air supply could be also provided from compressed air bottles (maximum air pressure of 300 kPa). Air pressure supplied to the air/water receiver was controlled through a pressure regulator and monitored at the control panel on a dial-type pressure gauge with a reading accuracy of 20 kPa. Confining pressure was applied hydraulically from the pressure tank (capacity 20 litre approx.) through the Belofram cylinder to the triaxial chamber and was kept at a prescribed level by a pressure regulator. A similar but smaller capacity setup was used for the back saturation pressure system, which was entirely independent of the cell pressure system.

A pressure transducer (Shaevith) was installed in the cell steel wall (at the top) providing accurate monitoring of the confining pressure inside the chamber. It had a measurement range up to 4 MPa, and for precision of measurement it was connected to a digital display (Amalgamated Instruments, accuracy of 0.5 kPa) on the control panel. For the backpressure or pore water pressure measurement, a pressure transducer (Shaevith) with a limit of 1 MPa was mounted on a ‘T’ block on the inlet tap fitted on the bottom plate of the cell. The back saturation pressure was monitored on the control panel using a digital display of the same make as that for the confining pressure, but with a higher accuracy, enabling measurement of pressure variation as low as 0.2 kPa. Pore Pressure Measurement

The original chamber design used a porous disc (25 mm x 12 mm) inserted in the

middle of the specimen to measure the pore water pressure. It was connected to a pressure transducer and a digital display on the control panel (as for the measurement of back saturation pressure) via 1/8 in. (~3.18 mm) plastic tubes able to withstand high pressure. Due to the aggressive nature of aggregate particles (sharp edges) and a concern that the disk and its connections would be damaged during specimen preparation and testing, it was decided to attach the disk to the bottom of specimen for better protection: the first series of tests was conducted using this setup. The drainage, hence, the dissipation of pore water inside the specimen was facilitated from the top of the specimen. Also the maximum pore water pressure occurs at the bottom of the specimen. However, this arrangement did not provide full protection for the (quite expensive) porous disk, so an alternative solution was sought and found. The water under pressure seeped through the porous plate mounted at the bottom of the cell, to the pressure transducer on the ‘T’ block and flowed along the same channels as those used for the porous disk connection. Tests were conducted with and without the porous disk on the same material specimens with the same confining pressure, and the pore water pressure was monitored during the tests. After comparing the results it was found that the pore water pressure variation could be accurately measured without the porous disk (provided that the correct connections were made). Therefore, it was decided to remove the porous disk altogether for all subsequent tests. Axial Displacement and Volumetric Change Measurements

Bressani (1995) reviewed various procedures of strain measurement during triaxial tests and concluded that if the correct procedures were followed, external measurements could be as accurate as internal measurements. Following this recommendation, axial deformation was measured externally using a Gefran PC (GF-PCM 150S) LVDT

Chapter 4: Test equipment and calibration

(Fig. 39). It was fitted to a rigid Perspex collar (20 mm thick) that was mounted on the axial load piston at the top of the triaxial cell (Fig. 40) to measure relative movement between the cell ram and the cell top. The maximum stroke of the LVDT was 150 mm and the digital displacement meter (Gefran Sensori 250) had a display accuracy of 0.1 mm, equivalent to 0.017 % strain.

A Belofram cylinder, named as a ‘voluminometer’, was used to transmit the confining pressure to the triaxial chamber through equilibrium between cell and Belofram cylinder (Fig. 40). This system had been proved to be accurate in the measurement of volume change during previous studies (modified after Indraratna et al., 1993). It consisted of a very rigid steel structure (20 mm thick walls, internal diameter of 210 mm) and separated into two by a 20 mm thick piston, machined to a close tolerance to fit inside the cylinder. The Belofram membrane sealed the top part of the cylinder from the bottom part. Water flow was possible through a three-way (‘T’) valve fitted on the top of cylinder. This valve permitted connection to the triaxial chamber and bottom part of

‘voluminometer’ through a 3/8 in. (~9.53 mm) plastic tubes resistant to high pressure.

The bottom section of the cylinder was connected to the pressure transfer tank through similar tubes. When the specimen was compressed/dilated the fluid in the cell flowed in or out of the ‘voluminometer’ causing the internal piston to move up or down. By measuring the movement of the piston ram (20 mm diameter) relative to the top of the cylinder and using a calibration chart, the travel of the ram could be converted into water volume that entered or left the ‘voluminometer’ as the test proceeded. This volume quantified the volume change of the specimen itself. The piston permitted a maximum travel of 100 mm, and this was monitored by a Gefran PC (GF-PCM 100S) LVDT connected to a digital display having an accuracy of 0.1 mm, equivalent to 0.0083 %

volumetric strains. As confirmation, a conventional method of specimen volume change measurement was used. This consisted of a 100 mm diameter graduated burette connected to the specimen top drainage lines. The capacity of the burette was 4000 ml and reading could be made to an accuracy of 7.85 ml. Measurements taken using both methods were compared for crosschecking.

4.2.3 Large-Scale Process Simulation True Triaxial Rig