Slot-type pore pressure CPT-u filters - Behaviour of different filling media.

Synopsis

In the CPT-u cone, there is a so called filter section, the purpose of which is to distinguish the generated pore water pressure from the mechanical earth pressure existing between the soil particles.
The measurement of generated pore pressure outside CPT probes has been practised since 1974. The filter section used has often been constructed using some porous material. Another type of filter- is constructed from steel only, with a tiny slot acting as filter. (Vlaasblom 1975) and (Sidey 1984). This type of filter offers simplicity in the field, and some technical advantages.
As with the porous filters, the absence of air is crucial but can be easier controlled, however it is obvious that the filling media in the slot and in the interior of the cone is of great importance.
This paper describes some laboratory tests on different filling media in slot filters.

Introduction

The function of a slot filter while performing the task of distinguishing the pore water pressure from the total pressure is to break the structure of the soil so that only individual particles floating in the liquid can enter the slot. A typical slot filter appears in picture 1.
The geometry of the slot filter comprises two separate parts, the cone and the ring, and two O-rings. The material is normally hardened stainless steel. Prior to filling, the cone and the ring are pressed together. The size of the slot will then be 0.3 mm. Note that the thread will be cut off by the two O-rings. If any bubbles are present in the thread, they cannot influence the pore pressure readings. The cavities A and B can be filled with the same or with different types of medium. Until now users have filled it with water or silicon oil. I have used water with approx. 1 % Gelatin, Petroleum grease and water or Petroleum grease and Hydraulic oil.
Since Gelatin has to be prepared in advance by boiling and cooling the filters, the grease fillings have become more popular by the users.
Experience from the field has however indicated that grease filling would introduce some hysteresis in the pore pressure measurement. A series of tests were performed in order to study the behaviour of these different combinations of filling media.

Pressure chamber

The pressure chamber is constructed from steel, and capable of 10 Bar over pressure. The
chamber is oil filled, but around the piezocone there is a rubber bag containing water. A reference pressure transducer is placed at the bottom. A dead weight pressure balance of type WIKA having an accuracy of better than 0.1% is used as pressure reference, and both the piezocone and the reference transducer are calibrated against this balance.

The piezocone

A piezocone of type Memocone has been used for the tests. The pressure sensing element consists of a stainless steel membrane with a strain gauge bridge inside, and a chopper stabilized amplifier that Will amplify approximately 100 times. This type of amplifiers are very stable, but unfortunately they produce some noise on the signal. In order to reduce the
noise, a low-pass filter is placed after the amplifier. The time constant is 0.05 sec.

This will cause a delayed reaction for pressure changes. However considering that the piezocone, penetrating at the normal rate of 2 cm per second, Will travel only 1 mm during this time it will not affect the measurements in the soil significantly. Note the difference between a reaction delay and hysteresis. In order to reduce influence from the delay, the low-pass filter has been removed in the tested piezocone.

Static measurements

The static measurements are made by pumping oil into the chamber until the weights of the balance are raised. The weights are rotated to overcome friction, and a stable pressure is achieved. The static hysteresis has been measured by altering the pressure in a sequence of 0-5-10-5-0 Bars. The different readings at 5 Bar, while increasing or decreasing the pressure, divided by 2 is then expressed as hysteresis +/-.

Dynamic measurements

The dynamic measurements are made by operating the piston by hand. This will cause pressure variations inside the chamber. These variations are picked up by the reference pressure transducer and by the piezocone. Data collection is made in a D-Mon. This instrument can be programmed to make synchronised measurements on two channels exactly at the same instant. The sampling rate was 10 readings per seconds, and the resolution of measurements 1/1024. Data is collected in the internal memory, and the files can be downloaded into a diskette afterwards.
Before measurement, zero readings and scales are checked against the pressure balance. The datalogger is started, and a pressure wave is generated by pressing the piston. The results are saved on a diskette and processed in a PC for presentation.

Filling media

The different filling media investigated here are:

1. 1. GREASE Type STATOIL LK 62
2. 2. GREASE Type BP LT 2
3. 3. GREASE Type ARIENS moly No 1
4. 4. Gelatin MIXTURE 10 gram/ litre
5. 5. OIL hydraulic oil

Type 1 grease was used earlier by Envi and is relatively stiff. Type 2 grease is temperature
stable, and is currently used by Envi. Type 3 grease is very soft, almost liquid, and is currently used by NGI.

Results from static tests

Referring to figure 1

A cavity	B cavity	Hysteresis kPa
Water	Grease 1	+/- 12
Oil	Grease 1	+/- 1
Water	Grease 2	+/- 7
Oil	Grease 2	less than +/- 1
Water	Grease 3	+/- 1-2
Oil	Grease 3	less than +/- 1
Water	Gelatin	less than +/- 1

Results from dynamic tests

For swift changes of pressure, the hysteresis is greater. Fig 6 is an example where grease in combination with water causes a clearly visible hysteresis in the order of +/- 25 kPa. However, with the same grease in combination with hydraulic oil, the hysteresis is reduced to +/- 2 - 3 kPa. (Fig. 7 ).
Grease 3 shows a better response. In combination with water +/- 12 kPa and with oil +/- 2 kPa. Gelatin and water shows excellent response also in the dynamic tests. See Fig.10 The hysteresis is less than 1 kPa which is the accuracy of the test equipment.

Conclusions

At any normal use of a piezocone, it should, in view of full scales of 1 - 2 MPa, be acceptable to use a grease/oil filling in slot filters. When ultimate accuracy is needed, there is the possibility to use Gelatin filling for these cases.

References

Vlaasblom, A. (1975). Soil probes for measuring various soil parameters.
U.S Patent No 3,906,781

Sidey, R. (1984). Pressure sensor and soil stress; isolation filter arrangement in a pore pressure probe. U.S. Patent No 4,453,401

Larsson, R. (1995). Use of a thin slot as filter in piezocone tests. International Symposium on Cone Penetration Testing - CPT95, Linköping