The laboratory method for determination of permeability of a soil is to observe the rate at which water seeps through a cylindrical specimen in the longitudinal direction due to a known hydraulic gradient or controlled loss of head. Either the constant head method or falling head method may be used, the latter being best for less pervious soils. Provided laminar flow occurs, Darcy’s Law applies (i.e. v = ki). The coefficient of permeability, k, may thus be calculated. Apparatus
The Geomechanics Laboratory has apparatus for carrying out three different types of tests:
(a) Constant head tests for measuring permeability of medium to coarse sands, or mixtures of sand and fine gravel (low or no % fines).
(b) Constant head test on a compacted granular soil, with vertical pressure added to simulate overburden. The apparatus was originally used for investigations for the South Para Reservoir.
(c) Falling head test for measuring permeability of fine-grained soils, either natural or compacted.
Students are required to carry out measurements of coefficient of permeability by method 2(a) from Procedure for falling head test, for fine-grained soil.
Procedure for constant head test on granular soil:
1. Weigh permeameter cell. Measure internal diameter of Perspex cylinder and length of sample, and compute sample volume. Measure distance L between manometer points.
2. Weigh out the quantity of soil required to give the desired dry density. Pour loose into permeameter, then gently jar or vibrate until the sample settles to the required overall length. Screw down the top of cell firmly. Weigh. (This is done for you).
3. Attach inlet, outlet and manometer tubes. Connect inlet tube to water tank at level above top of cell and allow sample to saturate from below. De-air top of cell by releasing thumbscrew. This procedure results in a minimum of trapped air. When water rises in the outlet tube to the level of the reservoir, the sample may be assumed to be saturated. Close outlet tube.
4. Allow seepage to commence by opening outlet tube and connecting discharge in measuring cylinder. Ensure water level in reservoir remains constant. Record quantity Q in 3 or 4 equal time periods = (t). The final 2 runs should agree reasonably well. Note manometer readings for each run. Record water temperature.
5. Repeat for three different ‘h’ values.
6. Dismantle apparatus.
Since q = Q/t = kiA, then
Figure 1: Schematic for Constant Head Test. Procedure for falling head test, for fine-grained soils. (Figure 1)
1. Measure and weigh permeameter cell.
2. Prepare sample, using one of the following methods:
a) For an undisturbed cohesive sample. Trim so that the sample cylinder just slides over soil, the final parings being taken off by the cutting edge. Trim ends.
b) For a disturbed cohesive sample. Compact soil at the pre-determined moisture content to the required density. Either cut specimen from the compact soil as in (a), or test in standard compaction mould.
c) For cohesionless materials. Prepare sample either by compaction in standard mould, or by dry preparation as in 2 (b).
3. Measure sample diameter and calculate its cross-sectional area, A. Weigh both the sample and cylindrical mould. Assemble cell with porous stones above and below the specimen.
4. Place sample in cell with porous stones, top cap, base plate and drainage lines in place. Non-saturated soils should then be saturated. Place cell in soaking tank with water well above the cell. Apply a vacuum of approximately 100 kPa to the underside of the sample and maintain until no more air is observed to rise. If a vacuum is not applied, much time is required to ensure a void volume of water (equivalent to the volume of the voids in the soil) has passed through the sample, ensuring saturation.
5. Remove from soaking tank and connect the cell to the standpipe (of inner area, a), taking care to exclude all air. With the standpipe full, allow seepage to commence. Note times to pass successive graduations, and the corresponding total heads.
6. If time allows, refill standpipe and repeat 2 or 3 more times.
Derivation of Formula for calculating k for falling head test.
Task No.2 Determination of flow pattern using a seepage tank
An experimental method of studying a pattern of two-dimensional seepage is to model the appropriate region of soil and the structure in a seepage tank.
The pattern of flow lines may be observed by introducing dye at the upstream boundary. The distribution of total head may be determined using piezometers arranged in a grid pattern in the region of seepage. Measurement of the quantity of water emerging from the mode enables a check to be made on the theoretical prediction of seepage rate.
1. Establish steady-state seepage in the seepage tank, below or through the model provided.
2. Place dye on the upstream face and maintain steady seepage flow until a pattern of flow lines is formed.
3. During steady-state flow, measure the quantity of seepage.
4. Sketch the tank and model to scale (see Note on last page).
5. Trace off the pattern of flow lines given by the dye.
(a) On a cross-sectional drawing of the tank and model, obtain the flow-net by sketching.
Calculate the seepage rate from:
Q = k.h Nf = No. of flow channels
Nd = No. of equipotential drops
(b) Compare the sketched and experimental flow patterns and the measured and predicted seepage rates.