Cambridge 73mm High Pressure Dilatometer - Working Instructions

WORKING INSTRUCTIONS FOR
CAMBRIDGE 73MM
HIGH PRESSURE DILATOMETER
Serial Number 060110
CAMBRIDGE INSITU
Old Church School
High Street
Little Eversden
Cambridge
ENGLAND
CB3 7HE
Telephone +44 (0)1223 262361

Fax +44 (0)1223 263947
Email caminsitu@aol.com
HTTP www.cambridge-insitu.com
WORKING INSTRUCTIONS FOR 73MM HIGH PRESSURE DILATOMETER
INDEX
1. Introduction
2. Description of the Instrument
3. Dismantling the Instrument
4. Description of the Test
5. Analysing a test in soil
6. A simple model for a test in rock
7. Calibrating the Instrument - theory and practice
8. From Raw Data to Engineering Units
9. Preparing the Instrument
10. Lowering the Instrument
11. Raising the Instrument
12. Getting ready for the next test
13. Working in a dry hole
14. Servicing details for probe and interface unit
15. Details of the Oil Pump
16. Inflating with nitrogen gas
17. Nulling a strain gauge transducer
18. Miscellaneous
Appendices
A Specification
B Manufacturer's leaflet
C Detailed Parts List
D Cable and Plug Connections
E 'O' Ring List
F Bolt List
G Packing List for Site
H References
I Glossary of Terminology
Separate
A fold-out plan of the instrument
HPD Working Instructions.doc Version 060110 Page 2 of 113

PART 1. INTRODUCTION
The High Pressure Dilatometer (HPD) is a pre-bored hole pressuremeter for testing a 76mm
diameter (NX) pocket. It carries out a loading test on the ground. The central part of the
instrument is covered by a tough rubber membrane; pressure is applied to the inside of the
instrument and the membrane expands, pressing against the borehole wall. The radial
displacement of the inside boundary of the membrane is measured at six points equally
distributed around the centre of the expanding section.
This displacement, and the pressure necessary to cause the movement, is continously
monitored by strain gauged transducers contained within the instrument. Also within the
instrument is the analogue and digital electronic circuitry necessary to condition the signals
from the transducers. Every ten seconds a set of readings from all the measuring circuits are
transmitted to the surface as an RS232 data stream which may be connected directly to the
serial port of a microcomputer. Plotting these readings of displacement against pressure
produces a loading curve for the material being tested; a number of mathematical analyses
can be applied to this loading curve in order to derive fundamental strength parameters for
the ground.
The HPD can apply up to 20MPa of pressure to the ground, and can expand from an initial
diameter of 73mm to nearly 100mm. It will resolve movements of less than 1 micron and
pressure changes less than 1kPa. Hence although it was developed to test weak rock it can
make a test at two extremes of ground conditions - stiff soils, which yield at pressures below
1MPa, and soft to weak rocks with a shear modulus greater than 5GPa.
The instrument has had a long and successful history of site work and has been used
worldwide. It is a development of an instrument invented by Dr J.M.O. Hughes in 1978.
Although internally complex by the standards normally applied to instrumentation of this
kind, it is reliable and robust, and the routine maintenance is straightforward. The HPD can
be gas or oil inflated. There are advantages and disadvantages to both systems. For reasons of
safety and to reduce the amount of ancillary equipment needed to carry out a test, oil is
preferred. It is strongly recommended that all pressure calibration and system checking is
carried out using oil.
For a guide to an understanding of the test itself we recommend the Ciria report on
Pressuremeter Testing by Mair and Wood. See Appendix H for the full reference. There is
additional information on our web site http:\\www.cambridge-insitu.com
Although much has been written about the pressuremeter test, some of what follows in these
instructions is new material. Like all expansion pressuremeters, the HPD has one
fundamental limitation- the loading curve produced is derived from following the movement
of the inside boundary of an elastic membrane. This is different from the movements of the
outside boundary of the membrane, and hence the movement of the material itself. For the
majority of the tests for which the HPD is used, this uncertainty is not significant. However
for a small number of tests it is critical; for this reason the calibration procedure described in
Part 7 is necessarily complex in order to reduce the margin of uncertainty and set limits to it.

PART 2. DESCRIPTION OF THE HPD
Figure 2.1 The basic HPD-MPX kit
THE INSTRUMENT
The instrument described here is a digital version of an analogue only instrument that has
been in production for many years. It is known as the Multiplexed High Pressure Dilatometer
(HPD-MPX). The significant advantage over the earlier design is that digital signals require
only a pair of wires. This has allowed the relatively fragile cable of the analogue machine to
be replaced with a steel logging cable of great strength, removing a potential source of
weakness. Placing electronics in the instrument inevitably makes it more complex. However
this is offset by a corresponding reduction in the amount of instrumentation required at the
surface. Experience has shown that conditions down the borehole are normally more benign
than those exhibited at ground level.
The HPD-MPX is a pre-bored hole pressuremeter. When a test is required it is lowered into a
pocket approximately 76mm in diameter. On completion of a test it is removed from the
ground and the borehole is extended by conventional drilling techniques. The pockets for the
tests are normally formed by coring with an N size barrel, although rock-rollering has been
used with some success. The instrument is 73 mm in diameter and has an overall length of
about 1.8m of which approximately 0.5m can be expanded. The expanding part of the
instrument is a nitrile rubber membrane with a wall thickness of 5mm.
The ends of the membrane are stiffened with rings of stainless steel fingers to prevent the
membrane extruding up and down the borehole, and are known because of their appearance
as 'Christmas Trees'. The expanding length of rubber membrane is covered with a sheath of

FIGURE 2.2 The Arms eighteen stainless steel strips which are axially stiff but free to
expand radially. This sheath protects the membrane from sharp
edges, and is known as a 'Chinese Lantern'. The individual strips
do not overlap in the closed position. The instrument measures
expansion by means of six strain gauged leaf springs which
follow the movement of the inside of the membrane. The springs
are equi-spaced around the circumference of the instrument. The
maximum expansion which the instrument is able to measure is
an increase of its diameter by 35%. The minimum movement
which it can resolve is normally one micron, and under certain
circumstances which will be described later (section 5) it can
resolve approximately 0.3 microns. The internal pressure is
measured by a transducer within the body of the HPD. The
maximum pressure capability of the instrument is 20MPa
(although it can be taken higher under certain circumstances),
and the minimum pressure change which it can resolve is about
1kPa.
If the ground conditions are such that the full pressure capability
of the probe is not necessary then a slightly different membrane
can be used, with advantage. The ends of the membrane have
kevlar strings embedded in them, making the membrane much
stiffer at this point than over the central part. The effect is similar
to that achieved with stainless steel fingers, albeit not quite so
strong. Fitting a reinforced membrane is more straightforward
than the standard version.
The instrument is inflated by oil or gas. A strong hose connects

the instrument to the pressure source, either a manually operated
hydraulic pump or a pneumatic control system. The passage
down the centre of the hose is large enough to incorporate a steel
logging cable with four electrical conductors. Three of these
conductors are used; one carries the digital signals output by the instrument, and two carry
power to the instrument from a conventional 12 volt vehicle battery. Current versions of the
HPD use a cable coupler system to link the probe to the hose. A short length of hose (less
than 1 metre) comes out of the instrument and is terminated with an in-line connector. The
reel of hose ( typically over 100 metres in length) is itself terminated with a matching
connector so that the probe and hose can be separated for transport. It also allows longer
lengths of hose to be added for deeper testing.
The power consumption of the pressuremeter is small; up to 500 metres of hose and cable can
be connected to the instrument with only minor modification. Because the instrument has six
strain arms there is some redundancy in the measurement of strain, and this enables the user
to carry out a successful test even if one of the arms are defective. In order to give a similar
level of reliability to the pressure measuring system a second pressure cell is included in the
HPD-MPX, and its readings provide a check of the performance of the first transducer.
The instrument is fitted with some internal electronics that provide power to the strain
gauged transducers and amplification of the resulting signals; in addition a detachable
electronics module is fitted to the foot of the instrument. This contains electronics that

multiplex, digitise and transmit the signals as an RS232 data stream to the surface. A cycle of
readings is sent every 10 seconds. One significant advantage of having complex electronics in
the downhole instrument is that every variable associated with converting the strain gauged
signals to a final value is contained within the instrument. As a result the decoded signals
prove to be exceptionally free of noise degradation.
An additional feature of this pressuremeter is a special electronic compass module fitted to

the foot of the instrument. This gives a continuous reading of the orientation of a fixed
reference on the instrument with respect to magnetic North. The compass consists of two flux
gate magnetometer sensors at right angles to each other. The output of the compass therefore
is two signals which are the sine and cosine of the angle made with the Earth's magnetic field.
The quotient of these gives an unambiguous direction.
Software to log the output of the instrument is provided with the instrument - there are
separate working instructions to describe the use of this software.
FIGURE 2.3 The HPD-MPX Disassembled

PART 3 DISMANTLING THE INSTRUMENT
The details below describe how to completely dismantle and reassemble the HPD. The parts
of the instrument described can be identified from the large pull-out assembly drawing.
Tools required: C Spanners, 30/32mm O/E spanner, nylon headed hammer, 5/3/2.5 mm hex
wrenches, drip trays - and another pair of hands.
If repair work other than replacing a membrane is going to take place then a full kit of tools
will be required.
Fig 3.1 Probe on handling frame 1. Start with the complete instrument
placed on the instrument handling frame,
allowing the probe to be firmly gripped
without damage using the plumbers vice.
2. With the 5mm hex wrench take out the

4 bolts holding the two halves of the
Chinese lantern (CHL) together.
3. Remove the CHL end float ring, which

is held together with two M5 bolts and is
located just inside the top end of the CHL.
4. Separate the electronics cover from the

lower end thread adaptor - this may require
the use of two C spanners. Pull the cover
away enough to expose the multiway connector which joins the electronics can to the
HPD lower end.
5. Break this connection - it is a bayonet fitting, and takes only a half turn to undo. Lay the
electronics cover together with the electronics can to one side.
6. Unscrew the lower end thread adaptor and put to one side.
7. Now release the ends of the membrane. This requires a 32mm spanner and a C spanner.
The 32mm spanner locates on flats machined on the stainless steel upper and lower ends
of the instrument; the C spanner locates in slots provided in the brass ring nuts. Undo the
ring nuts, moving the upper ring nut as far back as it can go and removing the lower one
from the instrument.
8. What is presented now is two stainless steel tapered clamp rings which have been tightly
pushed over the xmas trees/membrane. To release the clamp rings bash them with the
nylon headed hammer, and they will spring free. Never use anything other than a soft
headed hammer.
9. Remove the lower membrane clamp ring altogether, and push the upper ring as far up the
instrument as it can go. Slide off the lower Xmas tree, and unwrap the upper Xmas tree

enough to ease it off the membrane. The ends of the membrane are now exposed, and
probably there is a small dribble of oil appearing.
10. Procedure for an oil inflated instrument:

Now attempt to pump up the instrument. When oil starts to pour out of the membrane
ends, grasp the bottom of the membrane with both hands and pull off the membrane with
one swift movement. Be generous with pumping the oil and if this procedure is done fast
enough the membrane slips off extremely easily. Oil will pour out from the slots of the
exposed arms and can be caught by drip trays but there is certain to be some mess.
Discard the oil.
11. If the operator were too timid the membrane will be stuck half on the HPD. However it
can be banged back on to the instrument with the flat of the hand or a block of wood, far
enough to cover the holes in centre section of the instrument and then step 9 can be
repeated. It is possible to arrive at a state where the membrane cannot be moved in either
direction. In this event the procedure for gas inflated instruments must be used.
12. Procedure for gas inflated instruments:

To remove the membrane from a gas inflated probe the blow-on adaptor must be used.
This is pushed onto the HPD and just under the end of the membrane. It is then secured in
this position with a Multi-Clip. By connecting up the blow-on adaptor to a gas supply the
membrane can be 'floated' off the body of the instrument. If there is oil in the instrument
then oil will go everywhere.
13. Of course if the instrument is being dismantled because the membrane was punctured
then it is easier and quicker to cut off the membrane.
At this point the instrument is probably dismantled enough for the majority of problems. The
arms are available for calibrating and the membrane could be refitted or replaced. To do this
proceed to step 32.
Most repair operations will be easier with the cable and oil line removed:-
13. To remove the hose/cable assembly is straightforward, and can be done at any stage -
before step one if the cable alone is being replaced or removed for transport. The probe
has approximately one metre of hose connecting it to the larger reel of hose. The
connection is made with a purpose built coupler.
14. The main part of this coupler consists of a cover protecting a small coiled cable (see fig
3.2).
15. The cable coupler assembly has a left hand and right hand thread at either end. To break
the connection turn the cover (possibly using a small hook spanner) whilst the ends of the
connector stay stationary. It is useful to have a second pair of hands when doing this.
Once the connection is broken the internal cable may be unplugged. Cover the exposed
ends of the couplers with the protective black caps.
This next instruction is not a necessary step but needs to be mentioned if the user wishes to
remove the rod adapter, perhaps for repairing the hose/cable.

16. Undo the locking screws that fix the brass left hand/right hand coupler to the stainless
steel rod adaptor and the brass cable cover. Now undo the coupler, keeping the rod
adapter and the cable cover stationary. The rod adapter has a left hand thread, the cable
cover has a right hand thread. The presence of O rings under each end of the coupler
means that there is some frictional resistance to the coupler and both the instrument and
the rod adapter will want to turn. Two people here makes life easy. Note that the flat part
of the rod adapter fits a 30mm spanner, and there are C spanner notches in the coupler.
Fig 3.2 Parts of the cable coupler
17. Once the O rings are uncovered the coupler can easily be rotated by hand. Pull back the
rod adapter. What is revealed is the upper end MKIV plug on the coiled cable. Undo the
plug, and allow the coiled cable to spring back. Now remove the coupler, and fit the
special push-on cap to the cable cover. Use the locking screw to fix it in place. This will
stop oil leaking from the instrument, and is how the instrument should be prepared for
transport.
18. Screw the LH/RH coupler back onto the rod adapter. The hose assembly can now be put
to one side.
19. Remove the top brass cover sleeves, which are each secured with a single countersunk
M4 screw . Note that the cover sleeves have a legend etched on them to show how they
must be refitted. In two of the compartments rows of terminals are exposed which are the
points where analogue signals are brought before being distributed to the computer. The
terminals make a convenient place for checking that the analogue circuitry is behaving
correctly. The other four compartments contain a 5 volt power supply board, a 2.5 volt
power supply board, and the amplifier boards for the two pressure cells.
20. Removing the two lower brass cover sleeves exposes a further six compartments which
contain the amplifier boards for the strain arms.
21. In order to see the arm springs, slacken one of the M4 screws that hold down the thin
bands of stainless steel which stop the followers falling out. All six followers can be
removed, exposing the springs. Note that the followers are etched with their respective
number.
22. The pressure cells are concealed under the upper stainless steel end piece. In order to
expose them, first the cable cover must be removed - note that it is fixed with a locking
screw. Undo the MKIV plug and remove the coiled cable.
23. Now undo the six M3 screws that hold the aluminium plate with the MKIV socket to the
upper end thread adapter. Pull the plate away; there are a number of wires soldered to the

socket mounted on the plate, the wires being protected by two PTFE tubes. There is
sufficient slack in the wire loom to allow the socket plate to be pulled free by some way.
Slacken the locking bolts that fix the stainless steel upper end to the brass centre section,
and unscrew the upper end. If it is tight then spanners will have to be used on the flats
provided on the upper and lower stainless steel ends. Lay the upper end (complete with
upper thread adapter, ring nut and membrane clamping ring) down on the bench, being
careful not to stretch the wire loom. The pressure cells are now available for inspection.
24. The only significant part of the instrument that has not yet been revealed is the glass-to-
metal seal. This is recessed into the lower end of the brass centre section, and is held in
place by the stainless steel lower end. 12 wires run from the seal through the lower end to
the multiway connector. These wires must not be rotated.
25. In order to prevent the wires from twisting, the multiway connector securing screws are
removed, and the connector on its wires is pulled away from the lower end.
26. The lower end locking screw is slackened, and the lower end is unscrewed from the
centre section. It is certain to be tight to start , and will require spanners applied to the
machined flats. Once uncoupled from the centre section it can be laid far enough from the
instrument to permit access to the glass-to-metal seal.
27. The seal itself has two small through holes bored across its diameter. To remove the seal
bend a piece of stiff wire into a shaped loop; insert the short ends into the holes in the
seal and then pull the seal out from its recess.
This completes the dismantling of the HPD itself. Stripping down the hose and the electronics
module is treated separately.

Re-assembly
26. Replace all four terminal covers. Note that they are marked A - D or similar and should
be replaced in the correct position.
27. Replace the arm followers and the restraining band. The 'V' form at the ends of the band
push down into the slope of the countersunk screw head. This is easier to do if the
followers are held down with a multiclip. Make sure that the followers go back into the
slots from which they came out.
28. Replacing the upper and lower ends is a reversal of the dismantling instructions. Wires
must not be rotated; note that parts which use a locking screw must be tightened
sufficiently to align the locking screw with its recess in the mating part.
29. Refit the upper end thread adapter, coiled cable and the cable cover, again reversing the
instructions for dismantling. Note that at the cable cover is stamped 'A' at one end, where
it screws to the upper end thread adapter. Secure the locking screw.
30. Now replace the hose assembly. Take off the protective caps from the cable cover and the
rod adapter. Unscrew the LH/RH coupler from the rod adapter. Pass the coiled cable
through the LH/RH coupler, and screw the coupler onto the cable cover one turn to hold it
in place. Obtain the end of the coiled cable and screw it to the rod adapter. Now unscrew
the LH/RH coupler, hold it, the rod adapter and the cable cover in line, and connect
together by rotating the LH/RH coupler. Only the coupler must turn.
31. There are a pair of marks 'B' either side of the join with the cable cover. When these
marks are approximately aligned the locking screws should be in the right position for
being tightened.
32. Now fit the membrane.
33. Procedure for oil:

If it is a new membrane lubricate the inside wall with Diala B oil. Cover the LMH socket
on the lower end of the instrument with the protection tube. Bang the membrane onto the
instrument using the flat of the hand or possibly a block of wood. In order to refit it by
pumping oil, the membrane will have to be pushed on far enough to cover the upper
terminal covers. This is not always easy. The membrane will reasonably easily push on
enough to cover the arms, but not the upper brass cover sleeves. There are two options.
The blow-on adapter can be used to float the membrane on to the instrument; this is easy
but requires access to a gas supply and usually two people. Alternatively, before starting
to put the membrane on, a single layer of mylar tape can be wound around the upper
terminal covers, sealing the joins. This means that the membrane need not be pushed on
as far as before to cover the passages where oil escapes.
34. Repeat the pumping up procedure used to remove the membrane. When the oil starts to
flow out of the membrane it is straightforward to slide it into its proper position.

35. Procedure for a gas inflated instrument:

In this arrangement the membrane is partly pushed onto the instrument as far as it will
easily go. Now fit the blow-on adapter to the free end of the membrane and secure with a
Multi-Clip. Apply gas in a controlled manner to the blow-on adapter, and float the
membrane onto the probe. Note that the gas will in the first instance be resisting the
efforts of the person holding the adapter.
36. The membrane must be equally distributed on the instrument and must not overlap the
solid rings which terminate the tapered ends. Any excess can be trimmed.
37. Inspect the Xmas trees. If the roots of the fingers are cracked or stretched then discard
and fit new ones.
38. A new one is fitted by forming it roughly to a cone shape, sliding it on the membrane and
trapping the end with the tapered stainless steel clamp ring. The fingers must be even, the
Xmas tree should overlap itself neatly, and 10-15 mm of material below the fingers must
be visible. Note that the Xmas trees lie only on the membrane, they do not overlap at the
ends.
39. Force the clamp rings onto the membrane by tightening the brass ring nuts, by hand. The
interface between the ring nuts and the clamp rings should be lubricated with a little
copper loaded grease. Ensure that the Xmas trees look OK - the fingers should lay evenly,
and about 10mm of plain steel should lie between the end of the clamping ring and the
start of the fingers. If not, undo the ring nuts and try again. When satisfied tighten the ring
nuts further using a C spanner and 32mm spanner together.
40. If gas inflation is being employed, the brass ring nut must be very tight indeed.
41. The lower ring nut must be on far enough to expose some threads onto which the lower
end thread adapter can be screwed. Tighten the ring nut as hard as possible.
42. The instrument must now be properly tightened and the air bled out. Place the special
calibration cylinder with the close-fitting ends over the membrane and Xmas trees.
43. Pressurise the instrument to 5MPa. Hold for a short time and release the pressure. Of
course, check for leaks.
44. If oil is being used any air must be bled from the system. Fit the special bleed tube to the
fitting on the rod adapter, extracting some oil and air.
45. Re-tighten the ring nuts as far as possible.
46. Repeat steps 43 to 45 for 10MPa and 20MPa respectively.
47. Tighten the ring nuts for the last time. Refit the lower end thread adapter.
48. Now fit the Chinese lantern, and its associated spacer. This is a strip of steel which limits
the amount that the lantern can move longitudinally. Without the spacer, the Chinese
lantern would be severely damaged on removing the pressuremeter from the ground.

49. Place half of the Chinese lantern under the instrument with the lower end of it hard
against the lower ring nut. Note that the lantern will only fit one way on the instrument.
There is a visible gap between the end of the upper ring nut and the point where it would
touch the brass ends of the lantern. Select the CHL spacer ring that best fills the gap.
50. Screw together the Chinese lantern.
The instrument is now ready for calibrating the membrane.

4. DESCRIPTION OF THE TESTING PROCEDURE
Coring the Pocket
The HPD test is carried out in an NX size pocket that has been formed by coring or more
rarely by rock rollering. Coring is the conventional method because the recovered core can
give some information about the pocket before the instrument is placed. Samples of the core
are normally of sufficient quality to permit standard laboratory testing, which can then be
compared with the pressuremeter results.
The NX pocket itself should be at least 1.8 metres long. This allows the user some choice
about the exact point in the pocket in which to place the HPD. Note the following crucial
measurements:
(a) From the foot of the instrument to the centre of the measuring section is 0.8 metres.
(b) The expanding part of the instrument is about 0.6 metres.
(c) The instrument is 2.05 metres long - if the EW extension rod is added to this then the
effective length from the foot of the pressuremeter to the start of the rotary drill string is 3
metres. If the pocket is longer than 3 metres then the diameter of the drill rod used to place
the pressuremeter must not exceed 50mm; this together with two thicknesses of hose, is the
maximum that will fit into an NX pocket.
From these dimensions it is apparent that the instrument must be at least 1.3 metres into the
pocket, and for safety we would suggest 1.5 metres into the pocket as a minimum. There is
evidence that the very lowest part of the pocket should be avoided as this is often the area
most spoiled by the coring, but as the test centre is some distance from the foot of the
instrument this should not present a problem.
Soft patches in the core suggest a point that ought to be in the test section. However soft
patches at the ends of the expanding section may result in the membrane bursting before the
pocket has been properly loaded.
Heavily fractured material around the test centre, although not likely to burst the membrane,
will present problems of adifferent kind. The analyses that currently exist for the
pressuremeter test do assume that the material is intact - if this is not the case, then although
the data may be good,deciding what it means may prove too complex.
It is a truism that the quality of the test is very much connected to the quality of the initial
coring. However many good tests have been made in pockets where no core at all was
recovered. Standard coring practice is designed to preserve the core without regard for the
borehole wall. This is the opposite of what the priorities need to be for the pressuremeter test.
All details about the core must be noted on the Test Record Sheet as they can be vital in
interpreting the test results.
Description of the Test
The test which the HPD makes is a loading test. Pressure is applied to the instrument and
readings are taken of this pressure and the displacements of the cavity which the pressure

produces. Plotting pressure against displacement (or strain) results in a test curve with a
characteristic 'S' shape.
The tests which the HPD makes can for convenience be considered to be of two kinds:
(a) Tests where the material is failed before the maximum pressure capability of the
equipment is reached.
(b) Tests where the maximum pressure capability of the HPD is the limiting factor.
Tests in material which can be failed - Type A

It is the loading curve of these tests which tend to take the full 'S' shaped form. At the start of
the test the membrane is lying on the body of the HPD. There are then several distinct parts to
the test:
(1) Pressure is applied, the membrane lifts off the body of the HPD and expands out to touch
the sides of the borehole. Hence there are large displacements for very little pressure.
(2) Further increments of pressure define the first curve of the 'S' shape; this curve is the
instrument taking up the contour of the pocket and the cuttings left behind by the coring
being squeezed out.
(3) Following this is a linear part of the loading curve. The pressure being applied is
remoulding the material adjacent to the HPD which has been failed by the coring process, but
is not yet large enough to extend the zone of failure into fresh material. In this part of the test
the pressure is increasing but is having little effect on the strain.
(4) Once the pressure is sufficient to extend the zone of failed material the upper curve of the
'S' starts to be defined, eventually reaching an almost linear condition once more where strain
is increasing rapidly for very little increase in pressure.
(5) At this stage the cavity can be unloaded and the full unloading curve be drawn. The first
part of this will be linear, curving away as the pressure and hence the strain are reduced to
zero.There are several ways of using the data contained in the pressure/strain curve but by far
the most common is to use it to derive fundamental parameters about the strength of the
material being tested. Therefore the precise manner in which the test is carried out should be
chosen to make it easier to assess these parameters.
The slope of the linear part can give a value for the initial shear modulus, G(i), but because
this part of the curve is sensitive to the disturbance produced by the coring the derived value
is of little significance in itself.

FIGURE 4.1 Example of HPD test in a material showing large deformation

1000
TEST TYPE A
900
800
Corrected Total Pressure (kPa)
700
ONSET of PLASTICITY
Slope of final
600
unloading, similar
to loops
Initial
500 Slope
LOOP 3
400
LOOP 2
300
LOOP 1
200
HPD Test in Coal

100 Original radius of hole cut by core barrel February 1994
15.0 Metres
0
0 1 2 3 4 5 6 7 8 9
Average Radial Displacement (mm)
A better estimate of G can be made by taking unload/reload loops at intervals along the
loading path. To do this a little of the pressure that has been applied is released and then
reapplied in a controlled manner, taking readings of the changing strain. This produces a
characteristic loop. The slope of the best fit straight line through the long axis of the loop can
be used to derive G(r).
The value of G produced in this way is relatively insensitive to the initial drilling disturbance.
Although an assessment of Po is required in the conversion from slope to a value for Gr the
assessment can be grossly in error without serious consequence. The shear modulus is
probably the single most useful parameter that the HPD test can produce, and it is used
extensively in design calculations. Because of this significance it is usual to take at least two
loops at suitable points on the test curve (even if the engineer supervising the test specifies
only one). Suitable points would be on the linear part of the curve and as soon as there are
indications of failure (item 4, above). Loops can also be taken on the unloading curve.
The ratio of Gi to Gr allows some assessment to be made about the extent of the disturbance
created by the coring of the pocket.
Deductions about the undrained shear strength, Cu, are made from the plastic loading or
unloading data using the assessment of the insitu lateral stress in the calculation to decide a
strain origin.
Creep Readings:
From the linear part of the curve, (3 above), can be deduced the insitu lateral stress, Po. There
are several methods. We use a modified version of the Marsland & Randolph argument (see
references, Appendix H). Another aid to identifying the part of the curve where Po may lie is

to use creep strain, and this is frequently specified in contracts. To derive creep strain each
pressure increment applied to the test cavity is held for a fixed period of time. The change of
strain over this time period is recorded. Plots of this creep strain can in certain circumstances
allow an estimate to be made of Po.
Current thinking is that for an 'A' type test creep readings should only be taken during the
linear phase, where the lateral stress must lie. After this point, when the material is failing,
holding the pressure for extended periods of time will allow any excess pore water pressure
generated by the expansion to dissipate, seriously affecting the shear strength analysis which
assumes that the expansion is undrained.
Creep readings should never be taken during an unload/reload loop and there is no point in
taking them when unloading the test cavity.
For contracts where no creep readings are specified we would nevertheless take creep
readings as described above; the time period we would use would be 30 seconds.
The exploration of creep effects is important for other reasons not directly related to the insitu
lateral stress.Before taking a reload loop all residual creep in the material should be allowed
to slow to an apparent stop, otherwise it is possible that the loop will be spoiled. As a guide
the average loop will take two minutes - therefore the creep strain must be demonstrated to be
insignificant over a similar time span.
Tests in material where all deformation is elastic - Type B Test:

Tests where the material does not fail before reaching the maximum pressure capability of the
HPD need to be treated with great care. Such a test will have only the parts 1 - 3 of the 'S'
shape curve described above. This means that for a large change in pressure there is little
apparent strain. Any errors in the displacement sensing system will have important
consequences for the results.

FIGURE 4.2 Example of an HPD test where all rock response is elastic
25000
EXAMPLE OF TYPE B TEST
20000
Corrected Total Pressure (kPa)
15000
SECOND LOOP
10000 FIRST LOOP

THIRD LOOP, from 22MPa
This line denotes stress at down to 5MPa and back
which sheath deforms to the
shape of the container
5000
HPD Test in Andersite

71.2 metres
Feb 1991
0
1.15 1.25 1.35 1.45 1.55 1.65
Average Radial Displacement (mm)
Frequently in material of this type the consulting engineer will only be looking for a value for
modulus, because making assessments of Po or Cu is dependent on the material at least
starting to fail.
It is important therefore that there should be absolutely no possibility of creep in the material
spoiling any unload/reload loop, and we would recommend holding the pressure for a
significant length of time before taking a loop. Note also that the value of modulus is certain
to be high, probably in excess of 1GPa. This indicates that the displacements seen during the
loop will be very small, and therefore the pressure drop should be substantial to obtain the
maximum possible definition of strain. This in turn implies the unload /reload loop be started
from a point on the loading curve near the maximum pressure limit.
In a 'Type B' test, if unusually large creep behaviour is seen at any point then we would
advise exploring this creep behaviour by making pressure holds at increasing stress levels and
for longer periods of time. Creep, if present, has important consequences for the final design.
As will become clear from reading the section on calibrating the HPD, the instrument itself is
deforming under the influence of the large internal pressures. Removing this system error
from the test data is not always straightforward and the engineer carrying out the test must
ensure that sufficient information is available about the test and the characteristics of the
instrument in order to facilitate the correction procedure.
Immediately before the actual test is carried out there should be a corresponding calibration
test in a special cylinder supplied with the HPD. This is always good practice but is
especially important for a 'Type B' test.

Another technique that has proved rewarding is to make two loadings of the material. After
reaching maximum pressure (20MPa) the cavity is unloaded to about 5MPa. It is then
reloaded once more to maximum pressure before unloading to zero, taking unload/reload
loops at appropriate stress levels.
A comparison of the two curves can be very revealing, particularly if there is any uncertainty
as to whether the material has begun to fail.
Concluding Remarks
This is a brief introduction only to the test; the intent is to ensure that the test is carried out in
a manner that will recover the maximum amount of good quality data.
From remarks made above it is clear that we expect to gather more data than may have been
specified.
It is difficult to separate a description of the test from a description of the analysis, which is
not the concern of these working instructions. An appreciation of what will be done with the
data contributes significantly towards the success of the test.

5. ANALYSING A TEST IN SOIL
Index to this section:
1. Introduction
1.1. The pressuremeter test in soil - initially elastic response followed by failure in
shear
1.2. Defining strain
1.3. Average displacements versus the output of the separate axes
1.4. The Analysis program
2. Analyses for expansion
2.1. Overview
2.2. Marsland & Randolph (1977) Analysis
2.3. Undrained shear strength (CU)
2.3.1. Gibson & Anderson (1961)
2.3.2. Palmer
3. Shear Modulus
3.1. The Initial Shear Modulus
3.2. Cycles of elastic unloading and reloading
3.3. Non-linear stiffness/strain response

Deriving parameters from pre-bored pressuremeter tests in soil.
1. Introduction
There are two approaches to the interpretation of pressuremeter test data. The first, that
developed by Menard, uses empirical correlations to allow measured co-ordinates of pressure
and displacement to be inserted directly into design equations. This approach depends on a
standardised test procedure and a large data bank of pressuremeter tests correlated with
observations of the response of finished structures.
The second approach, which will be described briefly here and is the usual way of
interpreting the pressuremeter test in the UK, relies on solving the boundary problem posed
by the pressuremeter test.
The aim of the pressuremeter test is to expand a long cylindrical cavity within an undisturbed
mass of soil. Fundamental strength properties of the material can be deduced from
measurements made of cavity pressure and displacement.
In practice no instrument can be placed into the ground without affecting in some way the
surrounding soil. In the case of a pre-bored pressuremeter test the disturbance is significant
because the cavity wall is affected by the process of making the hole and is then completely
unloaded prior to the placing of the instrument. The measured pressure/displacement curve
does not represent the insitu response and must be redrawn to compensate for insertion
disturbance prior to attempting to derive strength and stiffness parameters. The problem is
two-fold:
Expand the pressuremeter sufficiently far so that the effects of disturbance are erased by a
new stress regime where the pressure applied by the instrument exceeds any previous
stress experienced by the surrounding soil mass.
Identify a point on the loading path that can be used as an origin for converting pressures
and displacements to stresses and strains.
1.1 The pressuremeter test in soil - initially elastic response followed by failure in shear.
Consider that the soil is homogeneous, and shows elastic behaviour before failing in shear.
The stress path followed by an element of soil adjacent to the cavity is given in figure 1.1 and
the corresponding pressure /strain curve is shown alongside.
The radial stress, initially zero for a pressuremeter placed in a pre-bored hole, increases at the
same rate as the circumferential stress decreases, regardless of whether the material is
deforming under plane strain or plane stress conditions. The line 0 - 0 represents stress
equality, so that in the ideal case considered here the point po is the insitu lateral stress..
Once the radial stress increases above the insitu stress then the shear stress in the soil at the
cavity wall will increase. If the insitu lateral stress is low, then it is possible that the
circumferential stress would go into tension. However in this example the insitu stress is high
enough to ensure that the shear stress limit is reached before tensile stresses can be generated.
The pressure necessary to initiate shear failure is denoted pf in figure 1.1. After this pressure
the strain rate shows a substantial increase, and the form of this part of the pressure/strain
curve

FIGURE 1.1 - Elastic Response followed by failure in shear
will be a function of the shear strength of the material. Radial stress and circumferential stress
now increase together. If the shear stress limit is constant, and is not influenced by pressure,
and if the material deforms at constant volume, then the failure shear strength can be
determined by the analytical solution developed by Gibson & Anderson.
Before the shear stress limit is reached the pressuremeter response is elastic, both in loading
and unloading. Were the installation to be perfect then the slope of the initial loading path
would give the shear modulus of the material, using the classic procedure of Bishop, Hill &
Mott. In practice, in a pre-bored test where significant unloading of the cavity wall has taken
place prior to the pressuremeter being in position, the initial loading is pseudo elastic - it will
under-estimate the true stiffness of the material. No such reservation applies to the initial
unloading. In addition, small cycles of unloading and reloading taken after the shear stress
limit is exceeded give plausible and repeatable estimates of the shear modulus (Hughes
1982).
As figure 1.1 implies, the complete unloading of the pressuremeter can also be analysed to
give strength and stiffness parameters comparable with those obtained from the loading path.
1.2 Defining strain

For a pre-bored pressuremeter measuring the radius of the expanding cavity the conversion
from displacement to strain is [R-R0]/R0, where R is the current radius of the cavity and R0 is
the original radius of the cavity in the insitu state. This is simple strain and when
displacements are measured at the borehole wall is termed cavity strain, c.
R0 can be roughly approximated by the at rest radius of the instrument. Usually, in a pre-
bored test, the initial part of the loading curve shows the instrument taking up the shape of the
test cavity. The radius when this process is complete (indicated by a point of inflexion) is
often taken to be R0 . In addition there are procedures for identifying when the applied
pressure has reached the insitu lateral stress, and interpolating from this the corresponding
radius which then becomes R0.

Note that although the pressuremeter measures the radius of the cavity wall, c is actually
circumferential strain. It is usually expressed as a percentage.
Figure 1.2 shows how pressures and strains in the expanding borehole are defined.
FIGURE 1.2 Pressures and strains around the expanding cavity
The other strain commonly used is the constant area ratio, which is shear strain. As figure 1.2
indicates it can be expressed in terms of simple strain.
1.3 Average displacements versus the output of the separate axes

Although there are a number of displacement sensors in the probe recommended practice is to
quote parameters from the average displacement curve. This is for two reasons:
1. The reference for the measured displacements is the body of the instrument itself - trying
to separate the individual axes means assuming that the body of the instrument remains
fixed at all times, which is not realistic.
2. All available analyses assume isotropic properties in the surrounding soil, and only the
average pressure/strain curve represents this condition.
The High Pressure Dilatometer (HPD) has six equi-spaced arm followers in the same plane so
that the effects of lateral translations of the instrument are cancelled by taking the average
radius from a pair of opposing sensors. However because of the fairly poor fit to the borehole
the probe moves in other directions than the horizontal plane and it is wise to use the average
whenever possible to minimise these effects.
These remarks assume that the instrument is in full working order throughout the test - failure
of a displacement follower means that alternative strategies must be adopted.
1.4 The analysis program

We use (and supply to others) software for analysing a pressuremeter test. The program is
called INSITU, it has been in use for a number of years and is well proven.

To use the program the user must first read in a text file of test data in engineering units. The
program needs to know the type of instrument being used, and the user may choose to enter
additional background information about the test.
The next task is to identify for the program the nature of the individual data points. Broadly,
speaking the options are these:
a point can be part of the expansion curve

or part of a reload loop
or part of the contraction curve
or part of a creep waiting period
or none of the above. This might mean a rogue data point, but it is more likely to be true
of parts of the loading where the expansion was slowed prior to taking an unload/reload
cycle. Data points recorded at this time are neither part of the expansion nor part of a
cycle, and should be identified as such.
There is a quick on-screen routine for marking the points. Once marked, they appear in
different colours and have different shapes (so that the distinction can be made clear on a
black and white printout). Most of the analyses use a limited set of the available data - for
example the Gibson & Anderson analysis for undrained shear strength uses only points on the
expansion curve. The smoothest data are obtained when end creep readings are plotted.
These are the strains recorded at the end of a creep waiting period. Providing the creep
interval is constant throughout the loading and providing there are sufficient steps of stress
these are the ideal data to plot.
The program implements all the standard analyses mainly in a graphical form. As figure 1.1
implies, there are significant changes of gradient in the pressure/strain curve which denote
critical soil parameters. The user of the program is provided with on-screen tools to mark
these breakpoints or to obtain the slope of the loading curve. The tools can be visualised as
rulers, and the chosen position of any ruler is stored by the program in the file of test data.
The evidence for any derived parameter is a screen dump of the appropriate analysis showing
the position of any rulers set by the user and quoting the parameter obtained.
Even when the user declines to make a choice it is good practice to provide the screen dump
as evidence of why a choice is difficult.
The results for a test appear as a summary sheet of derived parameters followed by a number
of plots showing the application of the various procedures.
Sometimes analyses are required which are not included in the INSITU program. In such
instances commonly available spreadsheet software is used to implement the new analysis.
Inevitably in such circumstances there is some risk of human error affecting the conversion of
data in engineering units to the form required for analysis. INSITU has comprehensive export
facilities and wherever possible is used as the data source for the spreadsheet.

2. ANALYSES FOR EXPANSION
2.1 Overview
The pressuremeter test is a sequence of measured co-ordinates of pressure and displacement
of the cavity wall (once suitable corrections have been made to compensate for the response
of the elastic membrane).
In order to solve the boundary problem, an origin for the expansion has to be determined. For
insertion methods which imply stress relief, the origin is taken to be the point where insitu
conditions are restored to the cavity. This means that an estimate of the insitu lateral stress
has to be made, and the measured radius of the cavity at the point where the insitu lateral
stress is restored is used to convert subsequent displacements to strain.
For a pre-bored pressuremeter test it is not normally possible to recognise the insitu lateral
stress by inspection. However it is possible to recognise by inspection the shear stress limit
(the point marked pf in figure 1.1) as this is signified by the onset of a markedly non-linear
response. An iterative procedure first suggested by Marsland & Randolph (1977) allows the
insitu lateral stress to be inferred.
Once the origin is known, the expansion phase of the test can be used to determine the
material shear strength. For an undrained expansion the classic procedure is that developed
by Gibson & Anderson (1961) where the slope of the pressure /strain curve plotted on semi-
log axes gives the shear strength directly and an estimate of the ultimate limit pressure. In
most circumstances the assumption of failure at a constant shear strength is reasonable, but
the complete shear stress:shear strain response of a material deforming at constant volume
can be described by applying the analysis due to Palmer (1972).
For materials where the expansion is drained, so that there are volumetric as well as shear
strains to take into account, the analysis due to Hughes et al (1977) is used to derive the peak
angle of internal friction and dilation. Manassero (1989) is a more complex analysis that
offers a complete description of the shear strain versus volumetric strain response of a
material deforming under drained conditions.
Estimates of shear modulus are obtained either from the slope of the pressure/strain curve at
places where elastic response can be assumed, or (preferably) from the slope of the chord
which bisects small rebound cycles. The values for shear modulus obtained in this second
manner are repeatable and seem to be largely independent of any disturbance caused to the
material by the placing of the pressuremeter. It is important, however, to correlate the
measured stiffness with the strain range covered by the rebound cycle.
2.2 Marsland & Randolph (1977) Analysis

Marsland & Randolph analysis relies on being able to identify the onset of plastic behaviour,
the yield stress pf. The argument is as follows:
that in the vicinity of the insitu lateral stress the soil behaves elastically and therefore the
pressure/strain plot will be linear
that this elastic behaviour will cease when the undrained shear strength of the soil is
reached in the wall of the cavity, and hence the pressure /strain plot will begin to curve
(see Figure 1.1).

This can be expressed as:

pf = po + cu .....[2.1]
From this it follows that po can be deduced by a process of iteration. Initially a guess is made
of a value for po; using this guess to define a temporary strain origin a total pressure:log
volumetric strain plot is then generated in order to derive a value for cu. The sum of these two
parameters is compared with the selected value of pf. The choice of po is then suitably
adjusted and the process repeated until a match is found. It is a straightforward matter to
carry out this procedure on the computer.
The modified method in current use is a response to the difficulty that the Gibson &
Anderson model is too simple for use in most materials and yield may occur at a shear stress
that is different from the large strain shear strength. Hawkins et al (1990) suggested that the
most appropriate choice was that value of shear stress pertaining at the apparent onset of
plasticity, so that equation [2.1] above now becomes:
pf = po + f .....[2.2]
f can be obtained from a total pressure:log volumetric strain plot by selecting the slope at the
pressure and strain corresponding to the choice of pf (in practice, using the Palmer (1972)
argument to identify the mobilised shear stress at failure).
The analysis is implemented graphically, using a number of rulers to identify significant

points on the curve.
FIGURE 2.1. AN EXAMPLE OF THE MARSLAND & RANDOLPH ANALYSIS

The main problems associated with the analysis are these:
The assumption of simple elastic response - in practice most soils exhibit marked non-
linear elastic characteristics, so that the pressure at which the material appears to go fully
plastic is more than one increment of shear stress above p0.
The assumption that the unloading of the cavity prior to the placing of the instrument is
elastic and hence recoverable. It is obvious that in most tests this assumption is unjustified.
This does not by itself affect the relationship between pf and p0 but it does mean that the
associated yield strain (iterated from the two stresses) is too large, and hence the strain
origin will be miscalculated.
2.3 Undrained shear strength (cu)

There are two analyses for deriving estimates of undrained shear strength :
Gibson & Anderson (1961)
Palmer (1972)
Usually only parameters derived from the Gibson & Anderson analysis are quoted, although
plots of the Palmer analysis are routinely supplied.
2.3.1 Gibson & Anderson (1961) - Prior to yield, the assumption of linear elastic
behaviour means that pressures and shear strains measured at the borehole wall are related by
pc- Vho = ('J)G ... [Equ. 2.3]
where pc is the total pressure applied to the cavity wall. For pressuremeters measuring
changes in volume such as the Mnard probe for which the Gibson & Anderson solution was
developed, shear strain J is the change in volume divided by the current volume, usually
expressed 'V/V. For pressuremeters measuring the radius of the cavity wall, J is the change
in area divided by the current area, 'A/A. Because plane strain expansion is assumed both
expressions are identical.
Shear Stress
SIMPLE ELASTIC/PERFECTLY PLASTIC MODEL
Stress to failure = Cu
Slope of elastic response = G

Strain to failure = 1/ Ir
Shear Strain
FIGURE 2.2.

For pressures exceeding the yield stress of the soil, the simple elastic/perfectly plastic
solution is usually expressed in terms similar to the following:
pc - Vho = cu{1+Ln[(G/ cu)('V/V)]} ... [Equ. 2.4]
Strictly, equation 2.4 is a simplified version of the Gibson & Anderson solution developed by
Windle & Wroth (1977).
Equation [2.4] can also be written:

pc - Vho = cu[1-Ln(cu /G)+Ln('V/V)] ... [Equ.2.5]
From this it is clear that the change from simple elastic to perfectly plastic response occurs at
a shear strain cu /G at which point the log terms in equation [2.5] cancel and
pc - VHO equals the undrained shear strength cu.
The limiting pressure pLimit at which indefinite expansion of the borehole occurs is given by
pLimit- Vho = cu[1-Ln(cu /G)] ... [Equ. 2.6]
Equations [2.4] and [2.6] can be combined to give the result

pc = pLimit + cu Ln['V/V] ... [Equ. 2.7]
This is a particularly convenient form of the solution with the undrained shear strength and
limit pressure being the gradient and intercept respectively of a plot of total pressure against
the natural log of the shear strain at the cavity wall.
The analysis assumes constant shear strength, so the slope chosen should be that which best
fits this assumption. There is no theoretical justification for selecting a slope at a specific
strain and claiming this to be cu. If the plot indicates that the gradient of the plastic phase is
not more or less constant then the assumptions underpinning the analysis have been violated,
and other strategies must be adopted.

FIGURE 2.3 THE GIBSON & ANDERSON ANALYSIS
Equation [2.6] shows that if pLimit , p0 , and cu are known then the rigidity index Ir and the
shear modulus at yield strain Gmin can be determined without going outside the analysis for
additional data. This has been done for the example plotted in Figure 2.3. In this example the
rigidity index and the modulus value seem plausible, suggesting that the origin for strain is
about right and that the material response is approximately linear elastic. The slope of the
semi-log plot is relatively insensitive to the choice of the initial conditions so that the value
for shear strength is better conditioned than that for all other parameters.
2.3.2 Palmer
The Palmer analysis (1972) shows that more information can be obtained from the
pressuremeter test if fewer assumptions are made. The analysis shows that the pressure:strain
graph is the integrated shear stress:shear strain curve. Taking the slope of the pressure:strain
graph at any point gives the mobilised shear stress directly, and allows the complete shear
stress:strain curve to be plotted. In terms of cavity strain the shear stress is:
= c(1+c)(2+c)dP/dc .....[Equ. 2.8]
More conveniently, perhaps, equation [2.8] can also be written in terms of shear strain as:
= dP/d[ln(V/V)] .....[Equ. 2.9]

This implies that the gradient at any strain of the semilog plot used for the Gibson &
Anderson analysis gives the mobilised shear stress directly. The example below is the same
test as Figure 2.3:
FIGURE 2.4 ` AN EXAMPLE OF THE PALMER (1972) ANALYSIS
The analysis is awkward to implement on the computer because the differentiation process
highlights any irregularities in the data. This is especially irritating because the stress strain
response must be a smooth curve. Possible strategies involve curve fitting the measured data
prior to applying the solution, but this may be a mistake. Minor changes of gradient on the
loading path are usually not random, but a response to some event such as the taking of an
unload/reload cycle.
If the material does indeed deform in the manner required by Gibson & Anderson then the
two analyses give identical answers. Included on the plot is a horizontal ruler marking the
value of shear strength obtained from applying the previous Gibson & Anderson analysis. If
there are clear indications of peak and residual shear strength then additional horizontal rulers
are available to mark these values. However the true value of the plot is that it is a 'map' of
the shear stress, and it is the form of the complete curve which is of interest.
The analysis is very sensitive to insertion disturbance - in particular insufficient allowance for
stress relief will give an apparent peak in the stress/strain response. Figure 2.5 is an example.
A high peak followed by a fall to a much lower value may be real, but it may indicate the
effects of disturbance. It is likely that the initial expansion is not occurring at constant
volume, and as a consequence the analysis fails. The residual value of about 1900kPa may be
a better choice, and there are indications here that if the expansion is taken far enough the
effects of disturbance can be overcome. Note that it may take a substantial expansion for this
to be achieved.

FIGURE 2.5 A shear stress/cavity strain plot indicating possible disturbance
3 Shear Modulus
There are four parts of the pressuremeter curve capable of providing information concerning
shear modulus:
From the slope of the initial elastic loading phase

From the slope of the chord bisecting small rebound cycles
From the analysis for shear strength
From the slope of the first part of the contraction curve
Whatever method is used to determine shear modulus, the value obtained is strain dependent.
3.1 The Initial Shear Modulus

Shear modulus derived from the slope of the initial part of the loading curve is quoted as part
of the Marsland & Randolph analysis (see figure 2.1). In a pre-bored pressuremeter test,
unless the probe is in good rock, this invariably underestimates the true elastic properties of
the material because the initial part of the test is affected by the process of making a pocket.
As figure 1.1 shows, the calculation for shear modulus G is:
G = dP/2dc .....[3.1]
Strictly, because the cavity strain c is simple strain it is only the initial modulus for which
this calculation is true. Once the expansion is underway modified calculations should be used
that take account of the change in the reference condition.
3.2 Cycles of elastic unloading and reloading

Graphical plots of the reload loops are the preferred method of obtaining a value for shear
modulus. The plots provided show the position of a cursor which has been placed by eye to
bisect the loop. The slope of the cursor is the gradient of the reload loop and the program uses
this slope to derive a value for shear modulus. This value is quoted in the top left hand corner
of the plot together with an indication of the size of the loop expressed as the change of
pressure and strain, and the co-ordinate of the centre of the loop. The equation used is:
G = [1+c][dP/2dc] .....[3.2]

In addition, the program carries out a regression analysis of the data points that are part of the
reload loop. If the loop is good, that is symmetrical and without indications of scatter, then
the two values of modulus obtained will be the same. However the regression analysis is
sensitive to erroneous data points, which the visual technique can ignore. The value obtained
by regression is quoted in the bottom right hand corner of the plot.
FIGURE 3.1 EXPANDED VIEW OF AN UNLOAD/RELOAD CYCLE
It is important that the effects of creep (for whatever cause) be minimised before starting the
cycle, and in Figure 3.1 deleted points before the start of the unloading show where the
pressure in the probe was held for a period of time.
3.3 Non-linear stiffness/strain response

In recent times it has become widely acknowledged that the stiffness/strain relationship is not
linear. The unload/reload cycle can be made to give a comprehensive description of this non-
linear relationship by looking at smaller steps of pressure/strain other than the points at the
extreme ends of the cycle.
For reasons explained in Whittle et al (1992) it is preferable to examine one half of the
rebound cycle only, that following the reversal of stress in a loop. The lowest recorded value
of stress and strain then becomes the origin for subsequent data points until the original
loading path is rejoined.

FIGURE 3.2
F
E
Measured
Loading data point
curve D
PRESSURE
C
B Conventional choice of
slope for defining modulus
New origin
for pressure New
origin for TYPICAL
strain UNLOAD/RELOAD
STRAIN LOOP
The reloading data can be plotted on axes of log 'pc versus log 'V/V. Figure 3.3 is an
example, using the same data as that in figure 3.2. The gradient of the best fit straight line to
the data points gives the non-linear elastic exponent, where 1 is a linear elastic response.

FIGURE 3.3 The non linear elastic response
The linear relationship between pressure and shear strain on log scales expands to a power
law of the form
pc = KJE ... [3.3]
where pc is the change in radial stressat the cavity wall, J is the corresponding shear strain
and K and E are the intercept and gradient of the log log relationship.
Palmer (1972) shows for undrained plane strain loading the shear stress at any point
on the pressure versus strain plot is given by
dP
W =J ... [3.4]
dJ
Substituting [3.3] into [3.4] gives

d ( KJ E )
W =J ... [3.5]
dJ
The differential equation can now be solved
W = J (KEJ E 1 ) = KEJ E ... [3.6]
Hence the shear stress is related to the radial stress measured at the cavity wall by
W = E pc ... [3.7]
It is convenient at this point to replace the combined coefficient KE in [3.6] with a single term
D, where
D=.KE ... [3.8]

Note that [3.8] can be turned into a general expression for secant shear modulus Gs by
dividing both sides by the shear strain J:
Gs = KEJE-1 = DJE-1 ... [3.9]
and because the tangential modulus Gt is related to the secant modulus by the following
relationship (Muir Wood 1990, Jardine 1992)
dG S
Gt = Gs + J ... [3.10]
dJ
it follows from [3.9] that the solution to [3.10] is
Gt = KE2JE-1 = DEJE-1 ... [3.11]
Figure 3.3 shows an almost linear elastic response and for most tests in rock this is typical.
Where the material is friable and behaves almost like a soil a more marked non-linear elastic
response is apparent. Often loops carried out later in the loading when the applied stress is
higher show the influence of grain crushing, revealed as a tendency for the exponent to
become more non-linear. If the test is drained, meaning the mean effective stress increases
throughout the loading, then successive loops will have a higher intercept.
Our practice is to give the exponent and intercept of the power law, and for comparative
purposes to quote secant shear modulus parameters at three levels of plane shear strain, 10-2,
10-3 and 10-4. It is unwise to use the power law to predict modulus for strains smaller than
10-4.

6. A SIMPLE MODEL FOR A TEST IN ROCK
INTRODUCTION
There is little in the way of published text on techniques for analysing pressuremeter tests in
rock. The methods commonly applied are essentially analytical techniques for deriving
strength parameters from soil and the assumptions implicit in these techniques may invalidate
their use on rock data.
Almost all the analytical models for deriving strength parameters assume that the material
being tested has reached large strains. They are closed end solutions for soil where failure in
shear is a measured boundary.
Because of the strength of rocks it is often not possible to fail the material before reaching the
pressure limit of the equipment used. Consequently there is little scope for deriving the
maximum shear strength. In most cases all that can be quoted is a value that the shear
strength must exceed.
One of the virtues of the Cambridge High Pressure Dilatometer is that relative displacements
of less than one micron can be measured reliably. It is possible to interpret with confidence
small details of loading paths where there are little or no indications of plastic behaviour.
The pressuremeter test in rock can be considered from several aspects; two will be considered
here:-
1. It can be considered as a gross "proof load test" of the rock to enable relative stiffness and
strength of strata to be determined.
2. The data can give indications of the material properties of the rock.
PROOF LOAD TEST

In this approach the pressuremeter can be treated as a large jack that forces the walls of the
borehole apart. The pressuremeter can be imagined to consist of two half cylindrical plates
100mm wide and 600mm long pushed against the rock as shown in figure A. If viewed in this
way the instrument can apply a large force of about 120 tonnes on the borehole wall.
When testing rock with the HPD a typical maximum deformation or separation of these
'plates' would be about one millimetre and is normally much less. Hence the pressuremeter
tests can be interpreted as providing an indication of the gross behaviour of the rock mass.
These gross displacements can be translated into a 'modulus' by assuming that the rock
behaves as an elastic continuum.
This procedure is analogous to the empirical procedure developed by Menard for determining
the deformation modulus. It is only useful for comparative purposes. The actual 'elastic
modulus' could well be several times greater.

FIGURE A THE PROOF LOAD TEST
The Determination of Material Properties from Pressuremeter Tests

In view of the complex nature of the rock a simple material model of the rock behaviour is
proposed. This model attempts to take into account the major material properties of the rock.
Some limitations of this model are considered, and it is used to give some understanding of
the behaviour revealed by the tests. There are six variations of the model:
TYPE (1) MODEL. HOMOGENEOUS/ELASTIC

Consider that the rock is dry, homogeneous and behaves in an elastic manner. The stress path
followed by an element of the rock adjacent to the cavity is given in figure B and the
pressure/ strain curve for this assumed material is shown for increasing and decreasing
pressure alongside.
The original insitu lateral stress is represented by point po. The radial stress, initially zero,
increases at the same rate as the circumferential stress decreases, regardless of whether the
rock is deforming under plane strain or plane stress conditions.
As the pressure in the pressuremeter increases the shear stress in the rock at the boundary
wall will change, reducing at first up to the line O - O (the line containing the insitu lateral
stress), then increasing. If the lateral stress is low, as represented by point po', then the
circumferential stress will go into tension; as long as the rock does not fail in tension the
loading and unloading pressure/expansion curves will be straight lines.
The shear modulus of the rock can be determined from the slope of the pressure/expansion
curve using the classic procedure of Bishop, Hill & Mott (1945). However not much else can
be derived.

This simple model is of limited use unless the pressure/expansion curve is linear.
Figure B Homogeneous/Elastic
Type (2) Model. Elastic behaviour before failing in Shear

In this next modification the effect of material strength is introduced. In this model it is
assumed that there is a limit to the shear stress the rock can sustain. This limit is shown
diagramatically in figure C. The rock is assumed to behave as an elastic material prior to
failing in shear.
FIGURE C ELASTIC RESPONSE BEFORE FAILING IN SHEAR ONLY
As the pressure exceeds pf, the pressure required to initiate shear failure, the strain rate will
show a substantial increase. The pressure/expansion curve that results from this model is
shown alongside.

The form of the non-linear portion of the pressure/expansion curve will be a function of the
shear strength of the rock. If the shear stress limit is constant and not influenced by pressure,
and the rock deforms at constant volume, then the failure shear strength can be determined by
the analytical procedure developed by Gibson & Anderson (1961).
FIGURE D THE LINEAR ELASTIC/PERFECTLY PLASTIC MODEL
On unloading the material will behave elastically until the failure strength is reached. At this
point the circumferential stress becomes the major principal stress.
In the ideal test shown in Figure C, the elastic shear modulus can be evaluated from both the
slope of the loading curve (up to the onset of yield) and the unloading curve. In an actual test
it is unlikely that the shear modulus can be determined from the loading phase. The true
behaviour of the rock is often masked by the disturbance caused by the forming of the test
cavity or by soft cuttings being lodged between the pressuremeter and the cavity wall.
The defect of this model is the assumption that the rock only fails in shear and not in tension.
If the tensile strength of the rock is low then the lateral stress must be high enough to ensure
that the rock does not fail in tension and crack, otherwise the analytical procedures developed
for this model will be inappropriate.
Type 3,4,5,and 6 models. Elastic, Shear and Tension

Most rocks do not have isotropic strength properties and they are usually weak in tension and
much stronger in compression or shear. If the insitu lateral stress is low and the shear strength
high then tensile stresses can develop during the test. The stress path followed during such a
test is given in figure E.

As the pressure increases the stress path will follow from X to A' to B'. At B' the
circumferential stress will start to go into tension. At some stress the rock will crack,
theoretically in a radial direction (as illustrated in figure F). With further increase in pressure
Figure E Elastic response followed by tensile failure
the stress path will shoot from B to C as soon as the radial cracks form and as the pressure
increases the cracks will grow and the stress path will continue along C to D.
Figure F The onset of radial cracking The pressure/expansion curve

will show a distinct step at the
initiation of the cracks. From
then on, as the pressure
increases and the cracks spread,
the slope of the
pressure/expansion curve
flattens. On unloading, the
stress path will follow down the
same curve until the cracks
close completely; then the
curve will follow the initial
'elastic path' as shown in the
pressure/strain plot of Figure E.
If the pressure continues to

increase the mechanism of
failure at the boundary will change to one of a shear failure mode as the stress path intersects
the shear failure boundary at D-E in figure E. Hence in the material around the pressuremeter
there will be three distinct zones, illustrated in figure G:-
An outer zone, zone 1, in which the rock is uncracked and behaves elastically.

Zone 2, where the rock will have radial cracks and a low circumferential stress. In this
zone the rock will probably not be deforming at constant volume because it is likely to
have open cracks.
An inner zone, zone 3, immediately adjacent to the borehole wall. Here the rock will be
failing in shear, and the voids will be closed.
FIGURE G THE SPREAD OF RADIAL CRACKING
The pressure/expansion curve will have the form shown in figure H.
FIGURE H. TENSILE FAILURE FOLLOWED BY If prior to the test there are

FAILURE IN SHEAR existing radial cracks then no
tension can develop. The
resulting stress path is shown
in figure I. If the shear
strength of the rock is too
high to fail under the
maximum pressure of the
instrument then the
pressure/expansion curve will
have the form shown
alongside. There is a distinct
change of curvature when the
crack starts to open. If the
pressure is reduced the curve
will retrace itself because the
material is behaving
elastically at all times.
However the conventional
method for determining the shear modulus will only be appropriate for the initial part of the
curve, not the upper section.

FIGURE I THE INFLUENCE OF EXISTING RADIAL CRACKING
If existing cracking does not dominate then as soon as a crack forms, the stress path path will
shoot from B to C in figure E. The usually appears on the pressure/strain plot as a sudden
increase in expansion. This behaviour is frequently observed in hard rocks near the surface.
In view of the erratic nature of the tensile strength several abrupt steps are often seen in the
pressure/expansion curve as the pressure increases. It is reasonable to suppose that the lowest
limit of tensile stress is denoted by the onset of 'cracking'.
Note: Knowing the lowest limit of tensile stress gives a maximum value for the insitu lateral
stress. As the symmetry of the stress diagram in figure E indicates, the insitu lateral stress at
point Ac can be no more than half of the stress necessary to make the material fail in tension.
FIGURE J TENSILE FAILURE IN A MATERIAL OF VARIABLE TENSILE

STRENGTH

The pressure/expansion curve will have the form shown in figure J. In general it will be a
straight line containing small steps but with no significant increase in strain (unless or until
the rock begins to fail in shear). On unloading the curve will follow a similar path to the
loading except that it will be smooth, without steps.
However if on loading the cracks do not
FIGURE K CRACK FILLING remain open but become filled with rock
cuttings then the unloading curve will be
much stiffer than the loading curve. One
could expect a pressure/expansion curve
of the form given in figure K. In addition
both tension and shear failure could occur
with the cracks being filled before
unloading.
INFLUENCE OF FLUID IN THE

HOLE
Throughout the above description the hole
is assumed to be dry during the test.
However water can exist either as a
consequence of the drilling process or
from the hydrostatic water pressure.
If the hydrostatic water table is near the

surface then pressure will exist throughout the formation. In this situation the previous
discussions still apply with the provision that the stresses considered must be effective rather
than total stresses. Figure D is shown redrawn in terms of effective stress in figure L.
FIGURE L THE EFFECTIVE STRESS If the rock is of low permeability

PATH (in terms of drilling time) and the
natural water table is below the
bottom of the hole a very different
situation can occur. If in the test
the hydrostatic pressure is below
the pressure of the column of fluid
in the borehole, and if the rock is
fractured in tension by the
instrument, then displacement of
the borehole wall is augmented by
fluid pressure acting on the whole
crack. If these fluid pressures are
sufficiently high then the crack
could extend itself without any
additional pressure increase from
the instrument.
This will be a time dependent

phenomenon. If the pressures are held constant then the displacements that result are likely to
be a combination of creep, consolidation and crack growth.

CONCLUSION
A simple model has been proposed constructed from three material parameters; shear
strength, tensile strength, and insitu lateral stress. It is apparent that the simple model can
have at least six distinct forms. These are shown together in figure M and are summarized
below:
(1) Elastic only

(2) Elastic and shear failure only (which implies high lateral stress)
(3) Elastic and no tension allowed (low lateral stress)
(4) Elastic and tension failure (low lateral stress)
(5) Elastic and tension failure with crack filling (low lateral stress)
(6) Elastic, tension and shear failure, with crack filling (low lateral stress).
These characteristic pressure:strain responses provide a context from which to consider

pressuremeter tests in rock. The principle use of these diagrams is to aid in judging when a
particular analysis is appropriate, and when it is not.

FIGURE M
The six basic forms of pressure:strain curve for a pressuremeter test in rock

PART 7. CALIBRATING THE INSTRUMENT - THEORY AND PRACTICE
INTRODUCTION
The instrument determines displacements and pressures by the use of strain-gauged bridge
circuits. There are eight in the High Pressure Dilatometer (HPD) Six of these are used to
determine displacements, two are used to determine stress.
The means by which a physical event ends up as a number on the printed page or computer
screen is as follows:-
Displacements and pressures are measured using strain gauges

The strain signals are amplified
A set of signals is captured at regular intervals and is multiplexed
The multiplexed analogue signal is converted to digital format
The digital signal is then made available as ASCII numbers which in turn are transmitted
via an RS232 interface to a computer
All these functions are carried out in the instrument itself.
The displacement followers of the HPD consist of strain gauged springs which in concept
bear directly on the inside wall of the rubber membrane covering the probe. In practice the
springs push on an aluminium follower which in turn presses on the membrane. The
aluminium follower is an averaging device, to minimise the risk that small discontinuities in
the material would have an exaggerated influence on the form of the test loading curve.
The hydraulic or gas pressure being delivered to the instrument is monitored by two total
stress transducers; these transducers consist of a thin strain gauged diaphragm, with a
vacuum on one side and the pressure in the probe on the other.
SENSITIVITY
Each strain gauge bridge circuit produces an output dependent on the voltage being applied to
it and the stress that is deflecting the strain gauges.
The instruments contain electronic devices that provide a regulated voltage to the strain
gauge circuits. Because this voltage is a fixed part of the system it is not mentioned when
presenting calibrations. The electrical output of the strain gauge bridge, in volts, is quoted
only as a function of the stress that is deflecting it. This is termed 'sensitivity'.
The signals from the strain gauges are small, a few millivolts only. Before they can be read
by the data recording system they must be amplified.The electronics for this amplification are
contained in the pressuremeter itself, in a water tight compartment close to the strain gauged
bridge circuits. Each circuit has an independent bridge amplifier of high quality with good
noise and temperature characteristics and the components around the amplifiers have a
similarly high specification.
The output of each amplifier is a signal changing between typically minus and plus one volt.
These analogue outputs are taken to a multiplexer which makes a cycle of readings every 10

seconds. The multiplexer output connects to the input of a high resolution analogue to digital
converter which contains an additional x2 stage of amplification.
The output of the system is numbers in ASCII format transmitted as RS232 signals via the
logging cable to the surface. The multiplexer, A to D converter and serial transmitter are all
controlled by a microprocessor within the instrument. Once power is supplied to the
instrument the system runs automatically.
All variables associated with producing the final digital output from the strain gauge signals
are a function of the pressuremeter itself, and are independent of external changes such as
replacing the cable.
SENSITIVITY
The displacement measuring system is normally referred to as 'the arms'. The arms are
calibrated by mounting a micrometer assembly above each arm and recording the output for a
given deflection. When calibrating the instrument it is necessary to plot these readings for
both an increasing and reducing deflection. The difference at a given point between
increasing readings and reducing readings is a measure of the hysteresis. The worst case
figure is noted, and steps are taken to reduce the friction in the system if the hysteresis is
outside an acceptable limit - normally 0.5% of the range being measured.
The slope of the best fit straight line through all the points is used to quote the arm sensitivity
- as an output for a given deflection in units of millivolts per millimetre.In general the
sensitivity should be set to be the maximum possible output that keeps within the limits of the
amplification circuit and the operating limits of the probe. The only requirement is that the
sensitivity be known and be linear and stable.
The sensitivity of the internal pressure cell is determined by placing a large metal cylinder
over the membrane, and applying a known pressure to the inside of the instrument. The gas
or oil pressure being applied is measured by a standard test gauge.
As with the arms, the readings are plotted, the hysteresis noted, and the best fit straight line
drawn through the plotted points. Pressure sensitivities are quoted in units of millivolts per
MegaPascal.
When using the sensitivity calibrations to convert readings from volts into engineering units
we make two important assumptions about this output; that it is linear and that the hysteresis
is small enough to be ignored.

ZERO
The other parameter that the strain gauge circuits have is a known output for an 'at rest'
position. This is the value of the outputs produced by the circuits with atmospheric pressure
on the inside of the instrument, and the displacement measuring arms at the initial radius
position. This is called Zero. The absolute value of this figure is normally unimportant - it is
not necessary, for example, that the figure be zero volts for the zero strain position, just that it
be known. For practical purposes, as the circuits are linear over the range plus or minus 3.0
volts, it is best that the 'zero' readings be about minus 1.0 volts.
It is normal to take zero readings both at ground level and also immediately prior to carrying
out a test. A significant change between zero readings must be investigated. 'Significant'
would mean a change of 0.03 volts from the last set of zero readings. It is, however, not
unusual for shifts of a few millivolts to occur from day to day. It is much more important that
the zero readings be stable when viewed over a period of a few minutes.
ACCURACY
Although it is important regularly to check the sensitivities of the strain gauge circuits, it is
unusual for them to alter significantly. Indeed it is common for the hysteresis to improve with
use. 90% of the performance of a strain gauge bridge application can be predicted from its
design; the calibration removes the uncertainty due to manufacturing tolerances, and can give
early warning of impending problems in a particular circuit.
Pressuremeter tests make relative measurements, not absolute measurements. The

displacement measuring system will resolve movements of less than 0.5 microns over a range
of 13 millimetres; the pressure measuring system will resolve changes of 0.5 kPa over a range
of 20MPa. This resolution is considerably higher than can be seen with a standard
micrometer or test gauge. To put it into context, 0.5 microns is approximately the wavelength
of ultraviolet light. Clearly there is no practical possibility of checking by measurement a
movement so small.
Hence the term 'calibrating' is inappropriate. What is done in practice is to check that the
various sensors are linear over a number of relatively coarse steps or intervals; the size of
these steps being decided by the limits of the pressuremeter (10 being the usual number of
steps). We assume that this linear behaviour will be true for very much smaller changes.
For this reason alone, without considering additional sources of error such as the skill of the
operator carrying out the calibration, the accuracy of the standard used to derive this linearity
is of secondary importance. We would expect successive calibrations on the same sensor to
be within 2%.
We also ignore secondary sources of error in this assumption of linearity, such as temperature
change. When critical measurements are being made during a test, for example when taking
a reload loop, it is reasonable to assume that the temperature remains constant.

Using spreadsheet software to present the results of the calibrations for sensitivity has
become common practice. One benefit of this is that slopes can be calculated by linear
regression routines; this ensures that different operators given the same set of data will derive
identical calibration factors.
MEMBRANE CORRECTION
The membrane that is expanded by the instrument has its own initial tension requiring a finite
pressure to move it. The readings measured by the stress cells need to be reduced by this
pressure in order to determine the net stress being applied to the ground.
The term 'membrane' is used here to mean both the sealed rubber sleeve over the instrument
that contains the pressure, and the stainless steel protective sheath that covers this. The sheath
is known as the 'Chinese Lantern'.
The membrane correction has two components - the pressure to move the membrane from its
position at rest on the instrument, and a second component that depends on the radial
expansion.
In addition because the membrane cannot be assumed to act isotropically, it is possible to

apply these corrections separately at the six points on the instrument where the strain arms
are positioned. However an average figure can sensibly be used in stiffer materials.
The technique for obtaining the correction data is to inflate the vertical instrument in free air,
using the same rate of expansion as would be applied in a real test. The slope and the
intercept on the pressure axis of the graph produced by this test give the membrane correction
information for each arm.
The membrane correction data is quoted as a pressure in kPa to move the membrane from its
rest position together with a second pressure in units of kPa,/mm representing the pressure
increase necessary to maintain the inflation.
Typical correction figures might be 20 kPa and 7.0 kPa/mm.
MEMBRANE DISPLACEMENT WITH PRESSURE
The pressuremeters measure the movement of the inside wall of a substantial membrane. For
most purposes the membrane is assumed to be incompressible, and hence the movement of
the outside of the membrane is taken to be the same as the movement of the inside of the
membrane.
When testing in very stiff materials and measuring modulus values in excess of 0.5GPa, then
small instrument movements can seriously upset these assumptions. There are a number of
effects to consider but they are collectively determined using a single procedure. The
correction figure which results is known somewhat inappropriately as 'membrane
compression'.

The procedure which is normally suggested to obtain correction data for 'membrane
compression' is to inflate the pressuremeter inside a number of cylinders of different bores;
by comparing these known bores with the displacements actually obtained from the
pressuremeter then a correction curve can be obtained. Because the correction has been
assumed to be a function of membrane thickness, then it is expected that the effect reduces as
the membrane thins. In other words, it is treated as a strain dependent variable, and a change
in membrane means a new correction curve must be derived.
In our experience real membrane compression, that is the membrane changing in thickness as
a direct result of the pressure differential across it, is so small as to be almost unmeasurable.
There are a number of other factors to consider of significantly greater magnitude than
membrane compression.
One component of the correction is the strain gauged springs. In addition to the output related
to the stress which is deflecting them, they can have a small output which is proportional to
the pressure under which they are operating. This output can be positive or negative.
However it is usually linear.
The other significant source of error is the deformation of the instrument as a result of the
pressure being internally applied. Put simply, the instrument stretches. Because the
displacement measuring system uses the body of the instrument as a reference, movements of
the body are seen as apparent displacements of the membrane; some ingenuity is needed to
immunise the displacement measuring system from this problem.
Inflating the instrument inside a steel cylinder will in theory provide data on the magnitude of
these effects. However a third factor, which is a function of the calibration procedure itself,
then becomes apparent. The membrane is able to expand axially by a small amount, and as a
result experiences a change in thickness which would not occur in the ground. Although steps
can be taken to keep this axial movement to a minimum, it cannot be easily eliminated.
Applying correction factors from expansion in a cylinder may result in large errors, possibly
greater than would be obtained from ignoring the correction altogether.
As a consequence of the poor fit of a calibration cylinder, and also of the relatively low
coefficient of friction between the membrane and the steel by comparison with the membrane
and the ground, the instrument will move about in the cylinder - its centre will not be the
same as the centre of the cylinder. Only average radial movement can be derived from this
calibration process, and it is not possible to obtain good data for each arm.
One approach is to take the membrane out of the correction loop by removing it altogether. A
special cylinder is then fitted which seals to the body of the instrument, which is then
pressurised. The displacement data which this test produces is used to determine the purely
instrument related factors. Typically the data is reduced to a slope correction, on the order of
3 - 4 millimetres per GPa, and is a constant, being a function of the physical properties of the
instrument.
The membrane is then fitted, and the instrument is expanded in the cylinder. The slope of the
unloading path of the average radial displacement in this cylinder is used to obtain a value - it
has been noted that the unloading path is relatively unaffected by instrument movements.

The slopes obtained from the two methods are then compared. Typically they are the same
within 1mm/GPa. This is to be expected. The bulk modulus of rubber is about 1GPa, and
hence a membrane that is about 2mm in thickness will have a slope of 1mm/GPa. Further
expansions inside other cylinders will not improve the quality of the correction so obtained.
To put the correction in context, a slope of 5mm/GPa is equivalent to a modulus greater than
4GPa.
Note that before the correction data is quoted the expansion of the metal cylinders themselves
must be removed from the data.
The correction data can be used in two ways. Applied as 'mm per GPa' it can be used to
correct individual data points before analysis; quoted as a modulus it can be applied
subsequently to modulus parameters determined from analysing the uncorrected test data.
MEMBRANE THINNING WITH STRAIN
Deriving the Equation
As the membrane expands, its circumference increases. To keep the cross-section area
constant the thickness must decrease. Conversely the thickness must increase as the
membrane collapses. This change can be calculated and the calculation is incorporated into
the computer program that converts raw data into engineering units.
Note that the term 'membrane' includes the stainless steel protective sheath, and that the
measurement made by the arms is the radial distance to the inside of the membrane.
Definition of Terms
2a is the I.D. of the membrane at rest

2b is the O.D. over the membrane at rest
2c is the I.D. of the membrane expanded
2r is the O.D. over the membrane expanded
t is the thickness of the stainless steel sheath strips
d is the measured movement of the strain arm
E is the actual expansion of the membrane
Calculation
At rest the cross-section area of rubber = (b t ) a
The expanded cross-section area of rubber = ( r t ) c
Because the rubber is incompressible, these must be equal:-
therefore (b t ) a = ( r t ) c
Now:- c= a+d

and:- r =b+ E
therefore (b t ) a = [(b + E ) t ] (a + d )
[(b t ) + E ] = (b t ) a + ( a + d )
= (b t ) + d ( 2a + d )
= (b t ) + E = [(b t ) + d ( 2a + d )]
E = [(b t ) + d (2 a + d ) ] (b t )
This is the form in which the calculation is commonly applied to the data, with 2a, 2b and t
being known from the manufacturer's data, and d being the measurement made by the
displacement sensors during the test. For an HPD fitted with a nitrile rubber membrane and
Chinese lantern:-
2a = 61.4 mm
2b = 72mm
t = 0.5334 mm
ORIENTATION
The electronics module fitted at the lower end of the instrument contains an electronic
compass which is used to identify the orientation of the probe with respect to magnetic
north.The compass consists of two sensors whose output is proportional to the Earths
magnetic field. The sensors are fitted at right angles to each other, each giving a maximum
output when that sensor is in line with magnetic north. The consequence is that that at any
time the sensors give the sine and cosine of the angle made with magnetic north, permitting
an unambiguous direction to be inferred.
The calculation of direction is implemented in the logging software package.
To calibrate the sensors, the instrument is rotated slowly through 360 degrees whilst the
output of the sensors are logged. From this, the maximum and minimum output of each
sensor is derived and is stored. Thereafter, selecting an option Heading in the logging
software uses the derived maximum and minimum values and the current data line to
determine a direction.
Because the magnitude of the output is affected by the dip angle made with magnetic north, it
is best if such a procedure is carried out in the ground immediately above the test pocket.
The sensors are hidden inside the electronics module container. A brass stud on the outside
indicates the position of the Cos sensor, and the electronics module fits to the instrument so
that this stud is in line with arm 1. The direction that the compass produces, therefore, is the
angle of arm 1 with respect to magnetic north.

8. FROM RAW DATA TO ENGINEERING UNITS
What is described in some detail in this part are the steps necessary to convert the raw data
output from the pressuremeter into engineering units.
In order to convert pressuremeter signals into calibrated data the following steps are taken:
A. The raw data is in units of volts, and needs to be corrected for zero offsets and scaled using
the sensitivities quoted in the calibration data.
The calibrations for this particular test are presented as follows:-
INSTRUMENT CALIBRATIONS: B1T1 DEPTH: 11.2M DATE: 22/9/95
ZERO SLOPE CORRECTION & COMPRESSION
ARM 1 -1313.9mV & 251.4mV/mm 30kPa & 10 kPa/mm 2.0mm/GPa

TPC A -600.6mV & 99.5mV/MPa
TPC B -631.0mV & 100.4mV/MPa
The line of raw data reads from left to right as follows. Note that the units are volts:-
LINE TPC A ARM 1 ARM 2 ARM 3 ARM 4 ARM 5 ARM 6 TPC B SIN COS
214 0.6570 -0.6102 -0.6158 -0.2087 -0.4579 -0.3515 -0.3979 0.6473 -0.2502 -0.6144
The first operation is to deduct the zero offsets. These are the figures found in the first
column of the calibration information. They are quoted here in volts. Note that the columns
for Sin and Cos disappear, as they are not transferred to the calibrated data file:
TPC A ARM 1 ARM 2 ARM 3 ARM 4 ARM 5 ARM 6 TPC B

Output 0.6570 -0.6102 -0.6158 -0.2087 -0.4579 -0.3515 -0.3979 0.6473
Zero -0.6006 -1.3139 -1.2767 -1.1803 -1.3801 -1.1781 -1.0025 -0.6310
Result 1.2576 0.7037 0.6609 0.9716 0.9222 0.8266 0.6046 1.2783 ......[1]
This result [1] can now be scaled. The information for this is found in the second column of
calibration data, and is expressed as millivolts per millimetre to calculate displacement, and
as millivolts per megaPascal to calculate pressure. At this point in the procedure, a choice
has to be made about which total pressure cell or combination of cells to use in producing the
calibrated data. In this example cell A was used, and the data for cell B are no longer
required.

As before, the results of the calculations are quoted in volts: They are written below as volts:-
ARM 1 ARM 2 ARM 3 ARM 4 ARM 5 ARM 6 TPC

From [1] 0.7037 0.6609 0.9716 0.9222 0.8266 0.6046 1.2576
Slope 0.2514 0.2197 0.2761 0.2420 0.2480 0.2436 0.0995
Result 2.7991 3.0082 3.5190 3.8107 3.3331 2.4819 12.6392 ....[2]
(mm) (mm) (mm) (mm) (mm) (mm) (MPa)
B. The data is now in engineering units which reflect what is taking place inside the membrane.
The remaining corrections are introduced to give a better representation of what is taking
place at the point where the membrane bears on the borehole wall.
The displacement data is adjusted for the instrument displacements due to the pressure being
applied to it. This is expressed as a linear movement in millimetres per gigaPascal of
pressure being applied, and is found in the 5th column of the calibration details:
ARM 1 ARM 2 ARM 3 ARM 4 ARM 5 ARM 6

Correction Factor 2.0 2.0 2.0 2.0 2.0 2.0 column 5
(mm/GPa)
Internal Pressure (MPa) 12.6392 12.6392 12.6392 12.6392 12.6392 12.6392 ...... [2]
Adjustment ((5)*[2])/1000 0.0253 0.0253 0.0253 0.0253 0.0253 0.0253 .......[3]
Internal Displacement 2.7991 3.0082 3.5190 3.8107 3.3331 2.4819 ...... [2]
(mm)
Corrected Displacement 2.7738 2.9829 3.4937 3.7854 3.3078 2.4566 .......[4]
(mm)
C. The strain data calculated so far is the movement measured by the strain arms to the inside of
the membrane. The figures quoted in the calibrated data listings are the movement of the
outside of the protective sheath. This is derived from the internal strain movement by
assuming that the cross-section area of the membrane is a constant. A full explanation of this
and the derivation of the equation used is discussed in the appendix on calibration technique.
The equation is E = [( R t ) + D (2 r + D ) ] ( R t ) ............. [a]
where E is the actual expansion of the pressuremeter

2R is the O.D of the pressuremeter at rest
2r is the I.D of the membrane at rest
D is the movement measured by the strain arm
t is the thickness of the Chinese lantern steel
For the pressuremeter used to produce this example:-
2R = 94 mm
2r = 82 mm
t = 0.51 mm

Because the membrane can be assumed to have the same thickness at all points on the cross-
section the technique employed is to calculate a scale factor from the average strain.

Corrected Displacements 2.7738 2.9829 3.4937 3.7854 3.3078 2.4566 ...[4]
Average Displacement 3.1334 3.1334 3.1334 3.1334 3.1334 3.1334 .....[5]
Result of equation [a] using D 2.7855 2.7855 2.7855 2.7855 2.7855 2.7855 .....[6]
= [5]
Scale Factor [6]/[5] 0.8890 0.8890 0.8890 0.8890 0.8890 0.8890 .....[7]
Apply [7] to [4] 2.4659 2.6518 3.1059 3.3652 2.9406 2.1839 .....[8]
D. The result, using displacements from [8] and Total Pressure from [2] :
LINE ARM 1 ARM 2 ARM 3 ARM 4 ARM 5 ARM 6 TPC
183 2.4659 2.6518 3.1059 3.3652 2.9406 2.1839 12639.2 .....[9]
In practice the errors introduced by rounding-off calculations may result in small differences
in the final figure.
This is the line of data seen in the calibrated data printout prior to analysis being carried out.
E. However the conversion to data ready for analysis is not yet complete. The column for
pressure is the pressure inside the membrane. What is required is the pressure on the outside
of the membrane where it bears against the borehole wall. Before using the calibrated data
file, therefore, the analysis program corrects the pressure data for the influence of the
membrane, using the data in the calibrations for membrane correction. It is separately
calculated for each arm position, although in practice an average correction value tends to be
used. The correction figure is the sum of the zero figure (column 3 in the calibrations) plus
the increased stiffness with strain (column 4):-
From Result [8] 2.4659 2.6518 3.1059 3.3652 2.9406 2.1839

Average 2.7856 2.7856 2.7856 2.7856 2.7856 2.7856 ....... [10]
Displacement
kPa/mm (column 4) 10 10 10 10 10 10 ....... [11]
Result [10]*[11] 27.9 27.9 27.9 27.9 27.9 27.9 ....... [12]
(kPa)
Correction zero 30.0 30.0 30.0 30.0 30.0 30.0 (column 3)
(kPa)
Add zeroes to result 57.9 57.9 57.9 57.9 57.9 57.9 ....... [13]
[12]
This is the total membrane correction at each arm position and is now deducted from the total
pressure cell readings to give six files of corrected pressure corresponding to six columns of
corrected displacement:

TPC 1 TPC 2 TPC 3 TPC 4 TPC 5 TPC 6

Uncorrected pressure in 12639.2 12639.2 12639.2 12639.2 12639.2 12639.2 (from result
kPa [2])
Membrane correction in 57.9 57.9 57.9 57.9 57.9 57.9 (from result
RESULT 12581.3 12581.3 12581.3 12581.3 12581.3 12581.3 .............[14]

9. PREPARING THE INSTRUMENT (prior to entering the borehole)
This procedure should be carried out about one hour before a test is scheduled to take place.
It is assumed that the HPD is working:
1. Check that the instrument zero readings are as expected. It is unusual for them to change
by more than about 50 mV.
2. Dress the Christmas trees (extrusion limiters). Check that they are not severely cracked at
the base of the fingers.
3. Fit the Chinese lantern, and make sure that a spacer of the correct length is fitted.
4. Now ensure that the plain calibration cylinder will slide over the instrument - this
guarantees that the HPD is the correct diameter. If the cylinder won't fit then check the
zero readings - if the arms are all a little positive this would suggest that there is too much
oil in the HPD. Connect up the bleed tube and release some of the surplus.
5. Alternatively the Christmas trees made need a little more dressing. Beat lightly with the
nylon headed hammer.
6. With the calibration cylinder in place run a quick test. If the material is very hard this is
especially important, because the way the instrument is set up has a bearing on the test
data. See the section on calibrations for a full explanation of why and how the test should
be carried out. Ensure that the calibration is stored on disc with a reference that can be
linked to the test which will follow.
7. Add a short length of EW rod together with the necessary adaptor for the drill string to
the top of the instrument. Use wrenches to make sure that this joint is tight.
8. Remove the calibration cylinder, and bleed off any excess oil.
9. Tape down the HPD. There are two things involved here; one, bind the first one metre of
hose tightly to the rod adaptor and EW rod extension piece at the top of the instrument.
Two, close up the strips of the Chinese lantern with tape. The way to do this is to place a
single tight ring of tape around the fattest part of the Chinese lantern where the Christmas
trees fit. Now run a spiral of tape down the Chinese lantern to link the two rings of tape.
Don't overdo it because an excess of tape will interfere with the strain arms. Note that
there should be no sign of the Christmas trees protruding through the Chinese lantern
when this taping process is complete.

10. LOWERING THE INSTRUMENT
Before beginning to lower the HPD to a test location the core from the proposed test pocket
should be examined. From this examination a decision can be made about where in the
pocket the instrument is to be placed.
Now decide what drill rods the instrument is going to be lowered on. Make sure that the
length of each rod is known. Measure the instrument and calculate:
1. How many rods are going to be used, and
2. What the stick-up once the instrument is in place will be.
Connect the HPD and any associated rod adapters to the drill string, and position above the
borehole.
Now check that the instrument is producing sensible readings.
The strain readings should be close to the known zero readings - a difference of more than
0.050 volts must be explained before commencing a test, What often happens is that there is
some residual oil in the probe that is tending to push out the membrane, and this may result in
the displacement followers reading slightly positive.
Tape the hose/cable assembly tightly to the rods, particularly at the point where it bends
around the adapters and cable couplers. Note that the hole down which the hose is to pass
must never be smaller than the rod diameter plus two hose diameters, Expressed another way
the size of rod used, particularly if a long pocket is made, must be less than
76 - (2 x 13) = 50mm.
It is often a requirement that the orientation of the instrument be known, that is, that the
position of the strain arms can be referred to a compass bearing. The orientation sensor is set
so that the cos sensor is in line with arm 1, and hence when the cosine output is at a
maximum (and the sine output in its zero position), arm 1 on the instrument will be pointing
due north.
It helps if there is a rod clamp to grip the drill string whenever the lowering (and raising)
operation is interrupted in order to add or remove a length of drill rod. Normally the best that
is available is a bitch that engages with flats provided on the drill rods. Make sure that the
driller has allowed sufficient space in his set-up for the taping operation to be efficiently
performed.
Note that whatever system of supporting the rods is employed, the use of wrenches alone is
strictly for emergencies only. Should the instrument be dropped down the borehole it will
survive, but the cable within the hose may well break. Do not stint on the taping of the hose -
looking after it is by far the most important part of the lowering procedure. Be careful
therefore that it does not catch on any part of the rig, and that it is lowered smoothly and
without kinks.

Ty-wraps have been used as an alternative to tape, with mixed results. They apply a load to
the hose, which is not a rigid object. It therefore tends to 'give', resulting in the join becoming
slack. In addition, if the ty-wrap catches on any part of the drilling equipment it tends to snap.
Although easily fitted and removed, inevitably some fall down the borehole and may interfere
with subsequent coring. Taping is tedious but safer. Use duct tape in deep holes.
Particularly when lowering the instrument to an appreciable depth it pays to be monitoring

the outputs, Power should be applied to the instrument at all times and the outputs monitored,
perhaps on a printer. Glance at this printout occasionally to reassure oneself that all is well.
Note: Decide what zero to use for the Total Pressure Cell at some point during the testing
program. We use the zero at the surface immediately before lowering commences.
Alternatively the zero once at test depth could be used - this second figure will be different
from the first by the head of oil. Whatever choice is made, be consistent.
Stop lowering just before the instrument is to enter the test pocket. Now slowly and carefully
place the instrument at the desired depth. The pocket may be tight, or it may be blocked by
debris, Do not force the instrument into the pocket with any more than the weight of the drill
string. If it will not go in it is a problem for the driller - the only thing to be checked by the
HPD operator is that the displacement sensors have not come out for some reason; this is
normally only a problem in a dry borehole, when the increasing head of oil can reach a
pressure that is sufficient to start the instrument expanding.

11. RAISING THE INSTRUMENT
There are two problems associated with getting the instrument back. One, after the test the oil
can take some time to return from the instrument, meaning that the membrane remains
pressed against the wall of the test pocket. Two, and more serious, treating the instrument as
if it were a drill rod results in it being pulled out too fast without enough care being paid to
the cable.
Waiting for the Oil
At the end of the test there is a natural delay whilst the operator checks that the data are
properly saved and backed up. During this time the oil is being forced back up the hose and
cable feed line into the oil reservoir. Two things are contributing to the speed at which it
returns; the membrane stiffness and any contribution from external water pressure.
Depending on how high is the head of water it is possible that left to itself the oil would take
many days to return.
The technique is to take a note of the arm readings, and to attempt to lift the instrument by no
more than 100mm. The arm readings will go out, and then return to something like the value
before the pull. Repeat the procedure for as many times as necessary until either the
instrument measuring section is out of the pocket or the driller is certain that the HPD is
moving freely.
In deep dry boreholes there is no alternative to fitting a second hose to the instrument, and
forcing the oil to return by applying gas pressure from the surface. Handling a second hose
however is not always convenient, and we try to avoid using one.
Pulling Up
Ensure that the hose comes back in equal amounts to the drill rods. This sounds obvious but it
is not always clear that the hose has stopped moving. The risk is that the instrument can be
pulled up past the hose, probably destroying it in the process.

12. GETTING READY FOR THE NEXT TEST
The damage that the instrument; experiences in normal use can be minimised by adopting the
following procedure; it should be carried out as soon as the instrument is removed from the
borehole. It is assumed that the instrument is apparently OK, and that there were no
indications during the test of any problems:
1. Wash down the instrument, removing all the surface soil. A scrubbing brush is an
essential tool.
2. Make sure that the membrane is completely collapsed. If two oil lines are being employed
then this may well be the case. If a single oil line is being used then probably the
instrument will still contain an excess of oil. Using the bleed hose, drain off this excess.
3. Take off the Chinese lantern, and remove as much of the soil under the extrusion limiters
(Xmas trees) as can be managed. Any soil left here will tend to cut the membrane the next
time the probe is pressurised. The Chinese lantern must be checked for excessive damage,
and replaced if necessary.
4. Inspect each of the Xmas trees. If the roots of the 'fingers' are showing signs of distress,
such as cracking or becoming severely distended, then replace the Xmas tree. Any defect
in this area will not only result in damage to the membrane, but will mean that the
membrane will be expanding axially in an unpredictable manner; this may well affect the
readings taken during a test.
5. Using the plastic headed hammer, beat the Xmas trees until they are restored to their
proper shape and are lying as flat as possible on the instrument.
6. Check the tightness of the membrane clamping nuts.
7. Refit the Chinese lantern; ensure that the end float shim is in good condition - if the
clamp nuts have been moved then a new shim will have to be cut to keep the end float at
1mm.
8. Ensure that the Xmas tree fingers do not protrude through the strips of the lantern.
Probably a little more attention from the plastic headed hammer will be required. Fit the
76mm calibration cylinder to the instrument.
9. The instrument could now be pressurised, either as part of a calibration procedure or

merely to ensure that it is operating correctly. The main reason why the calibration
cylinder is fitted is to ensure that the diameter of the instrument is as it should be for
fitting an NX borehole.
10. Remove the calibration cylinder. If the instrument has been pressurised then the excess oil
must be drained off before this can be achieved.

11. Now tape down the Chinese lantern. The tape should be applied in a spiral along the
length of the lantern; the purpose is to ensure that the lantern strips are neatly butted up
against each other, and do not overlap. Note that rings of tape should not be used for this
purpose, as too thick a layer of tape might interfere with subsequent readings of
displacements in the ground.
12. Finally, take a set of zero readings for the instrument, and assure yourself that these are as
they should be.

13. WORKING IN A DRY HOLE
The membrane of the HPD requires only 100kPa more pressure on its inside boundary than
on its outside to change it from the deflated state to its maximum expansion. The head of oil
alone may be sufficient to keep the instrument expanded if the water table is more than 10
metres below ground level.
This is a problem when going down the borehole, because the instrument begins to expand.
However the volume of oil required means that the problem only truly reveals itself in a deep
(below thirty metres) dry hole. In this case, after completing the test, the membrane will not
collapse back, converting the instrument into an effective ground anchor.
There are three things that can be done:
Fill the borehole with water - usually not possible.
Use gas rather than oil to pressurise the instrument.This requires the use of some special
equipment to supply the amount of gas required, but is perhaps the best solution for very
deep testing (below 200 metres).
Fit a second oil line to the equipment.
In this arrangement the bleed nozzle on the rod adapter is removed and a hose (without an
electrical conductor) is connected to this point.
At the surface this second hose is connected to the hydraulic pump in place of the first hose.
Now this hose is used as the oil supply line.
Often this in itself is enough to allow the membrane to collapse. Without an electrical cable
running up the inside of it the increase in bore means that the oil in the instrument sees a
lower hydraulic resistance.
If the oil still is unable to return then it can be forced back by applying gas pressure to the oil
to electricity separating union. The oil will return up the second hose. In these circumstances
the volume of oil in the instrument together with that in the two hoses will exceed the
capacity of the reservoir, so it may be necessary to arrange for the returning oil to run into a
larger container.
Additional advice on this can be provided. The penalty for using a second hose is the
difficulty of handling two supply lines to the instrument, which can be awkward.

14. SERVICING DETAILS FOR PROBE AND INTERFACE UNIT
The following section contains details of the circuits and layouts for the High Pressure
Dilatometer.
The instrument is laid out as follows:
Refer to figure 2.1 of these instructions. What is shown is the view under the membrane
of the probe. Around the middle of the instrument is six long compartments containing
the arm followers and behind the followers, the strain gauged springs.
Above and below the arm followers are two rings of six short compartments. Ten of the
compartments contain printed circuit boards - the remaining two consist of terminals from
which all the outputs of the instrument (with the exception of the compass) can be
measured.
The ring of compartments below the arms contain the analogue amplifiers for the arm
followers immediately above.
Two of the compartments above the followers contain amplifier boards for the two
pressure cells, A and B. The other two circuits to be found here are power supply
regulators. These supply +5 volts and +2.5 volts to the circuits. Note that the 2.5 volt line
is the reference level for the entire instrument, and is also the power source for the strain
gauges.
All the outputs of the instrument are collected on the terminals described above. From
here a wiring loom runs to a socket above and a similar socket below the expanding part
of the environment,. The socket below is placed outside the pressurised part of the
instrument - the pressure and electronic signals are separated via a ring of glass to metal
seals.
An electronics module contained in a brass can plugs onto the lower socket (see figure
2.3). This module contains the compass and digital electronics. Note, however, that the
instrument does not require the module in order to work - as long as it is connected to a
power source then the strain gauged transducers and associated amplifiers will be
providing an analogue output. This means that finding faults can be achieved with as little
servicing equipment as a digital multimeter.
The connection between the analogue signals and what is read from the digital output is
not immediately obvious. In the analogue state the instrument appears to work backwards,
meaning that increasing displacements result in a negative going signal. Furthermore the
digital stage includes a times two amplification, so that the analogue signals are only half
the magnitude of the digital outputs.
Typically, when measuring between the reference and analogue output of a strain arm, the
readings will be 0.5 volts in the zero state, and -0.8 volts at 13 millimetres. The
corresponding digital output will be -1 volt at zero and 1.6 volts at 13 millimetres.

By far the bulk of problems experienced with the instrument will be either failure of a
transducer or a fault in the wiring. The only option available if a transducer fails is to
replace it. Obtaining access to the strain arms is straightforward, the pressure cells are not
quite so easy. They are placed within the upper stainless steel part of the instrument,
under the membrane. To reach them the top end of the instrument must be completely
dismantled until the socket at the top end (described above) is reached. This socket is
bolted to an aluminium plate, Unbolt the plate, and pull the socket away together with a
substantial length of wiring harness. Now unscrew the upper stainless steel end from the
brass centre section and the pressure cells can be exposed.
It will be noted that the wiring used is PTFE, which is difficult to strip. It is important to
use this wire, both for its high conductivity and in particular for its resistance to the oil
used to expand the pressuremeter. A suitable wire stripper is supplied with the instrument,
together with some spare lengths of PTFE wire.
If the digital can is being dismantled, then note the following: the top, with the plug, is
secured with a ring nut; remove this nut. The lower end is secured with four countersunk
M3 screws; remove these. Essentially the internals are removed by pushing the plug end
and pulling the white plastic end - do not rotate. It will appear to be tight because both the
top and bottom are fitted with O rings. Once removed the internals will appear as they do
in figure 2.3.
On replacing the works, note that the cos sensor must be in line with Arm 1 of the
instrument. The sensor can be seen from the outside of the can because it is identified by
a brass stud on the base. The final act of assembly is to fit the module to the instrument
with the ring nut slack - turn the module until the brass stud and Arm 1 are aligned, and
tighten the nut, '1 is stamped at several positions on the instrument itself, most obviously
on the upper brass thread adapter.
Repairing the hose and cable
If the hose has become damaged relatively close to either the upper or lower end, then it can
be repaired.
Assuming that the hose has become damaged at the lower end, where it connects to the rod
adapter, proceed as follows:
1. The steel logging cable runs inside the hose. At either end, when making a connection,
the steel cable must protrude about 0.3 metres beyond the end of the hose. It is not
possible to recover much of the cable inside the rod adapter itself, so the hose will have to
be cut back in such a way that the 0.3 metres of cable is made available.
2. Sever the combined hose and cable as close to the rod adapter as possible, using a chisel.

3. Recovering parts
Undo the alencoflex coupler at the top of the instrument. The coupler parts, still attached
to the hose, can be pulled over the steel cable. The alencoflex coupler is in two parts, and
the inner needs to be unscrewed from the outer. The outer is then detached from the hose
- note that the outer has a left-hand thread.
4. To strip down the rod adapter, start from the MKIV socket. Undo the long screws holding
down the aluminium plate. Pull the aluminium plate away, and cut the cable, leaving
enough wire attached to the MKIV socket so that the colours and connections can be
identified.
5. With the aluminium plate removed, the socket spacer ring is now free. Remove this, and
undo the ring of cap head bolts that lie under the castellations of the spacer piece.
6. What happens now is that the cable clamp inner and outer will come away from the rod
adapter complete with the tail of steel cable. The steel wires that make up the cover of the
cable are clamped on a taper between the inner and outer. Hold the inner in a vice and tap
the outer with a hammer to break the taper.
7. Discard all pieces of steel, clean the parts, and put on one side for re-assembly.
8. Preparing the hose

Identify the damaged part of the hose, and cut the hose beyong this point far enough to
give the necessary 0.3 metres of protruding cable. To cut the hose, use a hacksaw with a
new blade. Do not attempt to saw straight through the hose at one point, but cut gradually
and. neatly around the circumference.
9. Inspect the cut end for stray pieces of steel braid. The end must be neat, so any rough
pieces of braid must be trimmed with steel cutters.
10. Fit the alencoflex coupler. Notes on doing this are provided elsewhere in these
instructions.
11. Assemble the Rod adapter to the alencoflex coupler using a little Hydraulic Loctite (542)
to seal the join.
12. Hold the Rod adapter in a vice, so that the EW thread is pointing away. Slide the Cable
clamp outer over the steel cable.
13. What happens now is that the steel braid of the cable is unwrapped and is laid in a regular
manner down the taper.. There are two layers of braiding, the outer layer being made of
thicker wires than the inner. Use all the wires, and do not rush the process. Trim the ends
of the braid so that the wires do not extend beyond the taper.
14. After this process the inner cores of the cable are left. Slide the Cable clamp outer over
the remaining cable, replace the cap head screws and tighten down evenly.
15. Replace the Socket spacer ring. Wire the inner cores to the MKIV socket and replace the
aluminium plate.

16. If the cable clamp outer has not been screwed down evenly, then the Rod adapter will not
fit inside the mating piece. Apart from this cautionary note, the job is finished.
17. Should it be necessary to remake the connection at the upper end of the cable, the
procedure is very similar. However in this instance the braid wires are not gripped on a
taper, but are held in successive layers between clamping washers. See figure 14.1 below.
FIGURE 14.1 Oil/Gas Electricity Separating Union.

FIGURE 14.2 Computer board FIGURE 14.3 Compass Board

FIGURE 14.4 CIRCUIT OF COMPUTER BOARD

FIGURE 14.5 CIRCUIT OF COMPASS BOARD (original version)

FIGURE 14.6 LAYOUT AND CIRCUIT OF REGULATOR BOARDS

FIGURE 14.7 LAYOUT AND CIRCUIT OF AMPLIFIER BOARD

FIGURE 14.8 LAYOUT OF INTERFACE BOARD

FIGURE 14.9 CIRCUIT OF INTERFACE BOARD

FIGURE 14.10 CIRCUIT OF DISPLAY BOARD

FIGURE 14.11 LAYOUT OF DISPLAY BOARD
FIGURE 14.12 THE PRESSURE CELLS

FIGURE 4.13 INSIDE SOLID-STATE COMPASS CAN

PART 15. DETAILS OF THE OIL PUMP
The pages which follow are extracts supplied by the manufacturer of the pump.
MAINTENANCE OF HP2 HYDRAULIC HANDPUMPGENERAL

Cleanliness is of prime importance when carrying out handpump maintenance. Dirt in the oil
causes wear and premature oil seal failure.
If the pump is to be dismantled, it is recommended that all 'O' rings are renewed as a matter
of course. Before dismantling pump, open pressure release valve and drain off all surplus oil
from the pump assembly.
DlSMANTLlNG
1. Remove screw (15) and slide out lever (6).
2. Remove pin (16) and unscrew relief valve housing (5).
3. Pierce core plugs (23) and prise out.
4. Turn shaft (4) fully clockwise and prise out core plug (22) through bore.
5. Remove screw (13) and extract crosspin (3) with shaft (4) in midstroke position.
6. Withdraw plungers (2) after sliding shaft (4) from body.
7. Remove screws (13), springs (12) and balls (19).
8. Prise out balls (20) and star clips (21) using small allen key through pump suction ports in
back face.
9. Relief valve housing (5) is simply dismantled by unscrewing screw (14) and removing
spring (11), socket (7) and ball (18).
lNSPECTlON
Remove all seals and thoroughly clean all parts in a suitable solvent. Carefully examine
moving parts for wear and damage, and check that valve seats are sharp and unbroken.
RE-ASSEMBLY
Ensure that all parts are perfectly clean. Fit new seals using SAElO oil or thin grease to avoid
damage.
1. Fit balls (20) and star washers (21) to cylinder pockets.

2. Fit balls (19), springs (12) and screws (13) using suitable thread sealer.
3. Slide plungers (2) into cylinders until cross holes are in line with shaft centreline.
4. Slide shaft (4) into body until all crossholes align, and fit crosspin (3) to shaft (4) and
centralise.
5. Fit screw (13) to shaft to clamp crosspin (j).
6. Ensure shaft (4) turns freely and fit core plugs (22 8 23) using suitable sealer-.
7. Reassemble relief valve using suitable sealer on screw (14) and fit to pump.
8. fit lever (6) and retain with screw

(15).
SETTING RELIEF VALVE

1. Fill pump with oil and connect a pressure gauge to pump outlet port.
2. Operate pump several times and close release valve.

3. Adjust screw (14) until desired pressure is obtained within the range of lOOOpsi (70 bar)
to 3000psi (207 bar), do not exceed these values. Lock screw with thread sealer and replace
pin (17) to provide approximately 270 of rotation from closed to open positions.
Refit pump in accordance with product leaflet.




16. INFLATING WITH GAS
Hydraulic inflation is the preferred choice when using the High Pressure Dilatometer because
it is safer. If the membrane or connection hose breaks the stored energy is low and the risk of
someone becoming hurt is small. It is also easy to reach high pressures with oil, and the oil is
recovered so that costs are kept down. In addition if there is water in the borehole, puncturing
the membrane when the probe is full of oil prevents water incursion, a very real benefit
particularly if the water is briny.
However there are disadvantages to the use of oil, and certain situations where gas is a more
convenient option. These are:
In locations where enviromental considerations mean that leakage of oil into the soil
cannot be risked.
In very deep holes, where it would be difficult to empty the probe of oil after inflation
Where relatively low maximum pressures will be reached during a test - 50 to 70 bars
being the expected maximum.
In these cases gas inflation has much to offer and dry nitrogen or breathing quality air is the
suggested inflation medium.
If the pressures to be reached are fairly modest, then nitrogen is readily available in bottles
pressurised to about 250 bars which will do several inflations. If high pressures are required
then steps will have to be taken to obtain special bottles which can store nitrogen at much
higher pressures, 350 to 400 bars.
There are no special steps that need to be taken when switching between oil or gas inflation
as far as the instrument itself is concerned. However the following safety points must be
noted:
20MPa is the safe maximum on the gas fittings that are supplied and more importantly, on
the calibration cylinders that are provided with the instrument.
When inflating with oil, because it is inherently safe, exceeding the 20MPa pressure limit
quoted in the HPD specification is not a problem. When inflating with gas, under no
circumstances should the limit be exceeded.
Do not substitute other pieces of tube for the calibration tubes we supply without
checking the material properties to ensure that they are adequately safe.
Gas inflation is particularly hazardous when calibrating the instrument at ground level.
Do not stand any closer to the probe than is absolutely necessary.
Check the integrity of the fittings before commencing inflation. It is quite easy to appear
to couple the male and female parts, but in fact to have a mismatch which results in a gas
line becoming pressurised without any easy means of venting.
Note that the pressure line we supply for gas inflation has the outer cover pricked to
prevent explosions should the gas leak past the fittings. Ensure that replacement hose is
similarly pricked.

FIGURE 16.1 The High Pressure Control Panel
IN PU T FR O M O U TP U T 4 O U TP U T 3 O U TP U T 2 O U TP U T 1 M E TE R E D VE N T
G A S B O T TLE
D irect from Via ven t
300 B A R S regulator T hese outputs com e throu gh the outpu t m e tering va lve m ete ring valve
M A X IM U M
ON ON ON ON
ON
OFF OFF OFF OFF
OFF
VENT
E N A BLE
B O TT LE O U TP U T
C aution! use w ith com pressed P R E S SU R E P R E S SU R E
air or dry nitrogen only.
D O NO T E X C E ED M A X IM UM
IN P U T P R E S S U RE
REDUCE
R E G U LATO R
(0~ 200 B A R S )
ON
OPEN
OPEN
OFF
PRESSURE
C O N TR O L VENT
IN C R E A S E C LO SE
U N IT O U T PU T M E TE R IN G C LO SE
O U TP U T M E TE R IN G VA LV E E N A BLE VA LV E
C A M B R ID G E IN S IT U , Little E versde n, C am brid ge, E N G LA N D , C B 3 7H E Teleph one: +4 4 (0) 1223 262 361 E m a il: cam ins itu@ aol.com
Parts of the Pressure Control Panel:
1. Gas Input - The adjacent male connector is taken directly to the pressure source. The
output of the red On/Off valve feeds the input of the regulator and the Bottle Pressure
gauge.. The regulator output feeds the Output Metering Valve and Output 4 (via its
On/Off valve).
2. Output Metering Valve - This is a needle valve. Its input is the output of the regulator.
Outputs 1 to 3 and the Metered Vent are downstream of this control. The Output Pressure
gauge reflects the current pressure being delivered to the outputs. It is used to bleed
pressure at a steady rate into the pressuremeter . To stop the flow without disturbing the
setting, the Output Enable valve can be used.
3. Output 3 This usually has a male fitting in it, for use as an auxiliary vent.
4. Output 1 This is the usual output for feeding the gas-to-electricity-separating union on
the cable drum.
5. Vent Metering Valve This is a needle valve used to release pressure in a controlled
fashion via the Metered Vent. There is an associated value to allow a bleed setting to be
preserved.
There are two modes of adjusting the flow of gas during a test, via the regulator or controlled
leakage.

Testing via the regulator:

The Hale Hamilton regulator is a precise tool for setting an exact value of pressure into the
system. If the Output metering Valve is fully open and the valves downstream of this are also
open then inflation and deflation can be controlled by manually turning the regulator. The
advantages are that minor leaks anywhere in the system can be accommodated by the
regulator. The disadvantages are the manual nature of the inflation and possibly a slightly
irregular shape to the measured loading curve in materials that are rate sensitive. However
where an exact pressure needs to be maintained, for example a pressure hold in a test or when
calibrating the instrument pressure transducers, then control via the regulator would be the
preferred option.
Testing using controlled leakage:

In this mode of testing the maximum pressure required to carry out the test is estimated.
Output 1 is closed off, the Output Enable valve is open and the Output Metering Valve is
fully unwound. This expected maximum plus a bit is then set on the test gauge using the hand
regulator. Leave the regulator at this setting. The Output Metering Valve is then closed, and
the panel is vented.
To carry out the test, the valve controlling Output 1 (to the probe) is opened, and the vent
path via the Vent Metering Valve is closed. To raise the pressure in the instrument the
Output Metering Valve is cracked open and the Output Enable valve is opened. Gas is
allowed to bleed steadily into the probe. Set the Output Metering Valve to the desired
flow with a little experimentation
To lower the pressure, for example to carry out an unload/reload loop, close the Output
Enable Valve. Open the Metered Vent valve and crack open the Vent Metering needle
valve to reduce pressure via the vent. A little experimentation will be necessary to
achieve the desired release rate.
Once the Vent Metering needle valve has been set, and enough gas has been released,
close the Metred Vent valve to prevent further venting. Open the Output Enable valve
once more to continue inflation.
One disadvantage to inflation with gas rather than oil is that there is a noticeable time lag
between initiating action at the panel and the instrument responding. Take care in the first
stages of the test to allow time for the probe to respond.
There are methods available for recovering some of the gas used in the test. Advice on this
can be obtained from us if sourcing high pressure gas bottles is a problem.

17. NOTES ON THE NULLING OF STRAIN GAUGE BRIDGES
There are a number of strain gauged transducers in the probe. Occasionally the output of a
transducer needs to be nulled, that is, set to a particular output for a given state. This may be
because a transducer has been replaced, or because a gradual change in output over time has
taken the circuit beyond an acceptable limit.
The analogue to digital (A-to-D) converter in the electronics of the probe can only output a
number between 3.2767 volts. To make the most of this restriction while still leaving
something in reserve, the displacement followers (the arms) of the 95mm HPD have an
output that lies between 2 volts over their working range of 0 to 25 mm. Therefore the
nulled or zero position should be approximately -2 volts - anywhere between -1.3 and -2.3 is
acceptable.
The condition for the pressure transducers is different. For TPC A, the nulled or zero output
is set to be close to 0 volts for ambient pressure - this results in the panel display in the
interface unit giving a pressure reading in MPa that is reasonably close to system pressure.
For TPC B, however, the nulled position can be zero volts or less - a common value is -1 volt.
There are no variable potentiometer devices anywhere in the probe for nulling the
transducers. This is because no variable device can be made with the necessary electrical
performance and mechanical reliability. In order to null the transducer fixed resistors must be
soldered into place.
The figure below shows the circuit of a typical transducer. This might be a strain arm. The
two shaded resistors must be
chosen by the user. Terminal posts
are provided in the instrument
where these resistors must be
soldered - see below.
A supply of suitable resistors must

be available. These must be metal
film resistors (MF25 series) and a
comprehensive range is required of
preferred values between 100 ohms
and 10 meg ohms.
To null a circuit, arrange that the

transducer is in a known condition
(ambient pressure for a pressure
cell, arm fixed to the body of the
instrument for a strain arm).
Remove any existing resistors.
Note the output of the transducer. It

will be either positive or negative
from the desired value. If positive,
the first resistor to be used will go

into the negative pair of soldering posts, and vice versa.
With practice, a user can hold a resistor between the soldering posts long enough and reliably
enough to determine if it is the optimum value. Initially, for someone not used to doing this, it
might be easier to lightly solder a resistor into place to see what effect it has.The optimum
value is either:
The resistor that brings the reading sufficiently close to the desired output or
The resistor that is just large enough to take the ouput past the desired condition. That is, if
the output was initially too positive, the optimum resistor is that which is just sufficient to
take it negative.
Having found the first resistor and soldered it in place, another resistor can now be chosen to
go into the other side of the circuit. This can be visualised as the first resistor having too great
an effect, and something is now required to swing the output back the other way.
Provided that the user is operating with a complete set of resistors using the standard intervals
the second resistor will be about 10 times the value of the first.
The usual cautions should be taken:

Turn the power off when soldering in the new resistor.
The resistor will be affected by the heat of the soldering procedure, so allow it to cool
before taking a reading.
Make sure that the joint is sound and no trailing wires can touch other parts of the system.
It is advisable to leave a small tail on the resistor to aid in its removal at a subsequent date.
Some users unused to handling electrical components use a resistor decade box to find out
what value of resistor to install. Such a box has a pair of wires that can be connected between
the soldering posts. Switches on the box are then configured until the optimum value of
resistance is discovered. The box is then removed and the resistor is then obtained and
soldered in place.

PART 18. MISCELLANEOUS
A - Fitting Re-usable Hydraulic and Pneumatic Couplings
The alencoflex couplings on the instrument hose and gas lines have a re-usable design. To
replace an end follow the steps below:
Make sure that the end of the hose is cut square, with no ragged ends of steel braid
protruding.
Lubricate the ends of the hose with a little silicon grease and screw on the outer fitting -
note that it is a LEFT HAND thread. Usually the fitting can be screwed on most of the
way by hand, with the hose needing to be held in a vice for a final turn or two.
The outer fitting normally butts up to a shoulder.
Hold the outer fitting in a vice, and screw on the inner coupling. This is normal right hand
thread
Make sure that the end of the coupling is smeared with a little silicon grease
Make sure that it is not folding over or shaving the inner wall of the hose.
Pressure check the hose.
B - Drying out Desiccators
The electronics units are sealed boxes fitted with desiccators. These absorb the small amount
of moisture that, despite the sealing, eventually penetrates the units.
If the windows of the desiccators show pink then they must be changed. Spares are supplied,
so fit the spare and dry the old desiccator.
To do this, place it in an oven at 120- 150C for approximately 1 hour, after which time the
window should show blue. Place it in a sealed bag, and store it in a sealed canister for future
use. The metal cans in which the desiccators are supplied are suitable containers, with sealing
tape applied to the join between the lid and can.
Desiccators can be dried many times before their effectiveness is impaired.

C. Pneumatic fittings
The pnematic fittings come in two kinds. For pressures up to 20MPa (in practice every
connection apart from those on or coming from the gas bottle itself) we use pneumatic
connectors manufactured by a Swedish firm called CEJN. The correct part numbers are: 10 -
322 - 1202 for the female connector with 1/4 BSP female thread 10 - 322 - 5202 for the male
(or nipple) with 1/4 BSP female thread alternatively 10 - 320 - 5202 for the male (or nipple)
with 1/4 BSP female thread
The female couplings have the full part number and pressure rating marked, but the males
only quote 322-5202 or 320 5202. They are made from nickel plated steel , hardened and
have a dullish surface.
For pressures above 20MPa we use modified hydraulic fittings. These are the same Hanson
fittings used elsewhere with oil but have had the self-sealing parts removed. This means the
modified Hanson couplings cannot be mated with the standard items.

APPENDIX A
SPECIFICATION: CAMBRIDGE INSITU HIGH PRESSURE DILATOMETER
The instrument is a high pressure, direct strain measuring dilatometer consisting of six strain
arms giving three axes of measurement at 120 degree spacing. Inflation can be pneumatic or
hydraulic, and pressure is measured by transducers within the instrument. Orientation is
measured by an electronic compass fitted to the foot of the instrument.
The instrument includes analogue and digital electronics, and the output is a serially encoded
data stream that is RS232 compatible.
The instrument comes with an electrical interface unit (EIU), powered by a 12 volt vehicle
battery, which provides a protected DC supply to the downhole instrument and buffers the
upcoming RS232 signal. Two opto-isolated serial outputs are available for direct connection
to the serial port of a computer. A panel meter on the EIU shows either the current input volts
or a continuous reading of the total pressure being supplied to the instrument (depending on
the setting of a front panel switch). Fuse protected DC outputs are provided for powering
other equipment.
Diameter of instrument 73 mm
Overall length 2.05 M
Expanding Length 455 mm
Length to Diameter Ratio 6.2:1
Maximum Pressure Rating 20 MPa, 200 bars
Coupling Thread EW pin to DCDMA and BS 4019
Hydraulic Inflation medium Hydraulic oil (Diala B

Transformer Oil)
Hydraulic inflation method Hand Pumping
Pneumatic inflation medium Dry nitrogen gas
Pneumatic inflation method via Pressure Control

Panel
Recovery from a dry hole when inflating with oil Gas Purge
Strain Measuring System six strain gauged leaf springs

bearing on aluminium followers

Strain Range To 96 mm OD (26% radial strain in NX pocket)
Smallest detectable movement less than 1 micron
Pressure Measuring System strain gauged cell; there are two

in the probe to increase
reliability
Smallest detectable pressure change 1 kPa
Membrane: Nitrile rubber sleeve, 566mm long by

5mm wall, with a 57.2mm bore.
(Note:- these membranes are capable of going to the maximum pressure capability of the
instrument when fitted with extrusion limiters as described below).
Re-inforced membranes:- Nitrile rubber sleeves with Kevlar

cuffs, dimensions as above.
(Note:- these membranes can be used up to 10MPa without extrusion limiters).
Membrane extrusion prevention Cone of steel fingers known as

Christmas trees (Xmas trees)
Membrane Protection 18 strip s/steel (Chinese lantern)
Connections to the surface a 12mm high pressure hose with internal 4 core electrical
cable. The cable has a breaking stress of one tonne. Hose Length 75 metres standard, up
to 500 metres is possible.
Power Requirements for the system 12 volts DC at 25mA.


HIGH PRESSURE DILATOMETER
The High Pressure Dilatometer is a pre-bored or MPM type

pressuremeter. It has a radial strain range of 35% and a pressure capability of 30 MPa. It can
resolve deflections smaller than one micron and resolve changes in pressure of 1 kPa. It can
make good tests at the two extremes of stiff clays where the maximum pressure reached
during the test may be only 2MPa and of soft rocks where the maximum movement of the
rock is less than 0.5 mm. It has been used at depths from 10 metres below ground to 600
metres below ground.
The instrument works by measuring the increase in the internal radius of the membrane
directly using strain sensing feelers. Six feelers are fitted each being designed to average the
borehole deflection over an axial length of 100 mm and a width of 12 mm located in the
centre of the expanding length. The instrument contains its own amplifiers, multiplexer and
serial data interface and transmits its signals up the borehole on a four wire cable contained in
the pressure hose. At the top of the borehole a small Interface Unit provides power from a 12
volt car battery to the downhole instrument and passes the returning signals to the on-site
computer.
The complete tally of equipment required is:
The downhole pressuremeter with interface unit 12 volt car battery

IBM compatible portable computer
Hydraulic hand pump and pressure gauge
Software to log the data on to floppy disc
Software to analyse the results
Printer
Spares
For deep tests, cable handling equipment
The expanding length of the instrument is 450 mm and it is designed for use in N size
boreholes which are 76.2 mm nominal diameter. The normal maximum pressure is 20MPa
(200 bars) but the instrument, correctly assembled, has been used at pressures up to 28MPa
without harm or leakage. The maximum length of hose we can supply without joints is 700
metres.
The strain sensing system is essentially the same as that used in the Cambridge Self-boring
pressuremeters we have been making for well over a decade. An aluminium alloy strain
follower is pressed lightly against the inside of the membrane by a long beryllium copper leaf
spring. This spring carries, towards the fixed end, four strain gauges connected in a full
bridge circuit. The output of these six strain gauges gives an accurate, sensitive and linear
measure of the changes in the internal radius of the membrane. The sensitivity of the strain
measuring system is such that deflections of less than one micron can be clearly displayed on
the computer screen and printed. A typical print shows one micron as 7 mm on the strain
axis.
It is important to note here that, unlike a self-boring pressuremeter, the centre of the
instrument is not located in a fixed position in the centre of the borehole. As the expansion
proceeds the instrument may move in the borehole. To take account of this the readings of

opposite pairs of strain sensors are averaged to give a value of radial strain that is actually
half the diametral strain.
The results of a pressuremeter test are well suited to mathematical analysis as the boundary
conditions are very simple. Provided the instrument is long enough in relation to its diameter
and the material is a continuum the expansion can be treated as that of an infinite cylinder. In
this case the stresses and strains are only radial or circumferential and very well established
theories of cavity expansion can be used. The most common analysis is that due to Gibson
and Anderson (1961) though at least four others are readily available. Further details on this
subject are available in other leaflets of ours.
Use of six strain sensors allows full analysis of the anisotropy of the material under test,
provided that the material can be assumed to behave as a continuum. This assumption
however, in the materials in which these instruments are commonly used, is often not
justified either because the material cracks at some point in the expansion or because the rock
is already discontinuous. Block theory may be required to make sense of the results after the
first crack has appeared.
Calibration can be complex. Not only are the stresses applied to the soil large but the
deflections measured are small and the influence of small errors is greater.
Carrying out a test is straightforward. The procedure is for the drill rig to core an NX size
pocket for the instrument 1.5 metres long. Recovery and emptying of the core barrel allows
the user to observe the core following which the instrument is lowered into the pocket,
normally on rods. Oil pressure is then supplied to the instrument by a hand pump whilst
observing the resulting pressure:strain graph displayed on the computer. Each increment of
pressure is, usually, followed by a specified waiting period during which the strain changes
are observed and recorded as the creep strain. Towards the end of the linear portion of the
pressure:strain graph but before the onset of 'failure' at least one unload/reload loop is carried
out. Loading then continues to the end point and ray be followed by a descent during which
further reload/unload loops can be taken.
In wet holes full collapse of the instrument back to its original diameter and its recovery
from the hole presents no problems. In dry holes where there is no head of water to drive the
oil filling back into the reservoir a second hose is used to purge the instrument using gas
pressure .
In very deep dry holes we use ourselves, and supply, a gas pressurising system working
from high pressure gas in bottles at 350 bars. The bottles may be filled either from a multi-
stage engine driven gas compressor such as is used for filling Aqualung bottles or by using an
Intensifier working from a low

pressure gas compressor at about 7 to 10 bars. However appropriate and continuous attention
has to be paid to the safety aspects of high pressure gas. Substantial amounts of energy are
stored and must not be released uncontrolled.
The pressure:strain graph resulting from the test described above provides the following
parameters:
An estimate of the insitu lateral stress using the Marsland and Randolph iterative
procedure. The procedure for doing this, which is particularly easy on a computer, has
been described in Hawkins et al (1990)
The Initial Modulus, evaluated as half the slope of the straight portion of the
pressure:strain graph
Reload moduli, half the slope of a reload loop
The Undrained Shear Strength by the Gibson and Anderson analysis
Results from some typical tests are shown on this page. To get them on the page these plots
are much reduced from the A4 size at which they are printed by our software.
Computers are prone to crash unexpectedly. Although it is usually easy to reinstate them data
is often lost whilst this is done. To solve this problem we supply a Memory Data Capture
Unit which performs the following functions:
Storage of the returning digital data on a battery backed solid state memory
Feeding the data, on a Centronics parallel link, to a printer so providing hard copy on site
Feeding the data stream on to the computer.
The memory is large enough to store more than a thousand lines of data, enough for about
three normal tests. Even if the computer crashes the complete test can be recovered from the
MDCU without loss of any data. The security this provides we find very reassuring.

ORIENTATION
To special order magnetic direction sensing can be fitted giving a reading of the direction in
which strain sensing arm number 1 lies. The accuracy is 5 though caution should always be
exercised as a magnetometer may be affected by stray magnetic fields.
The instrument described is a development of that devised by Hughes and Ervin. Some
features of it are protected by patents.
REFERENCES
Baguelin, F., Jezequel, J. F. and Shields, D. H. (1978)

The Pressuremeter and Foundation Engineering" Trans Tech Publications
Ervin, M. C., Burman, B. C. and Hughes, J. M. O. (1980)

'The Use of a High Capacity Pressuremeter for Design of Foundations in Medium Strength
Rock Proc Int Conf on Structural Foundations on Rock, Sydney, N.S.W. May 1980
Gibson, R. E: and Anderson, W. F. (1961)

Insitu Measurements of Soil Properties with the Pressuremeter Civil Engineering and
Public Works Review, V. 56, No. 658, PP 615-618
Hughes, J. M. 0. and Ervin, M. C. (1980)

"Development of a High Pressure Pressuremeter for Determining the Engineering Properties
of Soft to Medium Strength Rocks" Proc 3rd A. N.Z. Conf on Geomechanics, Wellington N.Z.
Hughes, J. M, O., Wroth, C. P. and Windle, D. (1977)

"Pressuremeter Tests in Sands" Geotechnique, V. 27, No. 4, pp 455~477
Marsland, A. and Randolph, M. F. (1977)

"Comparison of the Results from Pressuremeter Tests and Large Plate Tests in London
Clay" Geotechnique, V.27. No.2 pp 217-243.

APPENDIX C - DETAILED PARTS LIST
Parts List (for 73mm High Pressure Dilatometer, machined parts)

ITEM PART NUMBER
Rod Adaptor:
Rod adaptor 701/3
Cable clamp outer 776
Cable clamp inner 777
HPD socket spacer ring 778
Rod adaptor protective cap 1143 & 1144
73mm High Pressure Dilatometer, main

assembly:
End plug 1145
LH/RH coupler sleeve 1141
Cable cover sleeve 1142
Upper thread adaptor 1105/2
Body end (upper) 1106/3
Membrane clamp ring 707/1
Membrane ring nut 705
Locking ring 1132
Centre section 1107/3
Split cover sleeve 713
Arm follower 709/2
Arm retaining strip 759
Body end (lower) 1108/4
HPD glass to metal seal holder 1146
Lower thread adaptor 1109
Compass cover 1110/1
End cap 1111
Sundries:
Arm calibration jig 755/2 & 756
Plug protector 1147
Blow-on adaptor 770
Arm springs, complete 1125, 1126 & 350R spring
Total Pressure Cell 722/1 & 723
Chinese lantern ring 730 & 731
Chinese lantern strip 732
Nitrile rubber membrane 1060
Nitrile rubber membrane, reinforced 1062
Christmas trees 729/3
Calibration cylinder (plain) 742/2
Calibration cylinder with ends 783
Windlass

Fluid to Electricity Separating Union:

Body 762
Glass to metal seal holder 726/4
OESU cap
HPD OESU Coupler 764/1
Logging cable clamp rings 773 & 774
Locking nut
Alencoflex coupler:
HPD Alencoflex coupler 760
HPD Alencoflex Union Body 761
Electronics Assembly:
Compass sensor mounting 1118/3
Compass can (plastic version)
Plug retainer
Retainer ring nut 1135
PCB support
PCB spacer

APPENDIX D CABLE & PLUG CONNECTIONS
SERIAL LEAD from interface box to computer:

4-way MKIV free socket Colour Description 9 way D-type free socket
A Brown RS232 2
B Blue 0 Volts 5
Note: the RTS, DSR, and DCD lines in the D-type socket, pins 1, 6, & 7 are linked.
Electronics Interface Unit to Separating Union lead:

6 way MKIV Description Colour 6 way MKIV
free plug free socket
A RS232 Red A
B Not used Green B
C Not used Yellow C
D + Volts Blue D
E 0 Screen E
Note: This lead can also be used to connect directly from the interface
unit to the internal Mk4 socket inside the probe for bench testing, by-
passing the FESU, logging cable and Rod Adaptor.
Fluid to Electricity separating union:

6 way Mk4 Colour Description Lemo socket
(fixed plug)
A Red RS232 1
B Green Not used 2
C White (clear) Not used 3
D Black + Volts 4
E Green/yellow 0 Volts Metal work of FESU
Logging cable:- This has a left hand cable coupler at the top (FESU) end and
a right hand cable coupler at the bottom (probe) end. The cable couplers are
terminated by a 4 way miniature Lemo socket:
Top end 4 way Lemo Colour Description Bottom end 4 way
socket Lemo socket
1 Red RS232 1
2 Green Not used 2
3 Clear Not used 3
4 Black + Volts 4
Metal work Steel braid 0 Volts Metal work

Miniature Coiled Cable - inside the cable couplers

4 way Lemo free plug Description Colour 4 way Lemo free plug
1 RS232 Red 1
2 Not used Green 2
3 Not used Yellow 3
4 + Volts Blue 4
Note: A connection to 0 volts is made through the metal work of the assembled coupler
Rod Adaptor this has a short length of hose with a left hand cable coupler at the
FESU or logging cable end and a Mk4 6 way socket at the probe end:
4 way Lemo free plug Description Colour 6 way Mk4 fixed socket
1 RS232 Red A
2 Not used Green B
3 Not used Yellow C
4 + Volts Blue D
Metal work 0 volts Green/yellow E
Large Coiled Cable - inside the left/right hand coupler at top of probe:
6 way MKIV free socket Description 6 way MKIV free socket
A RS232 A
B Not used B
C Not used C
D + Volts D
E 0 E
Note: Colours vary depending on the manufacture of the coiled cable.
Probe Wiring - Internal PTFE cable harness

LMH PT06 Description Colour 6 way MKIV fixed
(lower end) plug (top end)
A Arm 1 Brown
B Arm 2 Red
C Arm 3 Orange
D Arm 4 Yellow
E Arm 5 Green
F Arm 6 Blue
G TPC A Violet
H TPC B Grey
J Reference Violet (repeated colour)
K + Volts supply Pink D
L 0 Volts Black E and metal work
N RS232 Grey (repeated colour) A

Additional Wiring Notes
All parts of the instrument that could conceivably see pressure and hence could come into
contact with oil are wired with PTFE wires. There is only a limited number of colours
available and so the following convention has been applied throughout:
Each strain gauge transducer is connected to its electronics by four PTFE wires:
Colour Description
Red -ve signal output
Green +ve signal output
Black 0v power supply
Violet +ve power supply (2v5 - this is reference line
for the system but is also the power connection
for the transducers).
In addition:
The 0 volts line is always black
The reference line is always violet (but so is the output of pressure cell A)
The +ve raw input voltage supply is always pink
The regulated 5 volt supply is always white.

After going to the printed circuit board; the output signals follow this colour convention:
Colour Description
Brown Arm 1
Red Arm 2
Orange Arm 3
Yellow Arm 4
Green Arm 5
Blue Arm 6
Violet Pressure Cell A
Grey Pressure Cell B
Grey RS232 output
The implications of this arrangement are as follows. If a red wire is seen it can be one of two
things:-
It can be the output of Arm 2.
It can be the negative input line from any of the eight transducers.
A similar logic applies to the green wire.

APPENDIX E
List of O rings used in Cambridge 73mm High Pressure Dilatometer
Size Qty Type

Oil to Electricity Separating Union:
Glass to metal seal holder 223 1 4490
Cable Coupler (with backing ring) 223 1 4490
Face seal 029 1 4470
Instrument:
Rod Adaptor 228 1 4490
Upper End Adaptor to LH-RH Coupler 228 1 4490
Upper Thread Adaptor 228 1 4490
Face Seal 215 1 4470
Inner Seal Ring 120 1 4490
Lower End to Centre Section 125 1 4490
Lower End to Lower Thread Adaptor 218 1 4470
Lower Thread Adaptor 228 1 4470
Electronics Cover Cap 228 1 4470
Blanking Cap 228 1 4470
Electronics Unit:
Electronics Plug Holder 123 1 4470
Sensor Assembly (internal) 032 1 4470
Plug face seal 016 1 4470
Sensor Assembly (external) 032 1 4470
Cable coupler
Shaft (item 1451) 116 2 4490
Outer, RH 116 1 4490
Outer, LH 116 1 4490
Gas Bottle 111 1 4470
Note: A 4490 type O ring can be used for a 4470 type O ring - the reverse does not apply.
The harder grades are used where high pressures are likely.
The sizes quoted are British and American Standards, all rings are commercial grade for use
with mineral based oils.

APPENDIX F
List of Bolts used in 73mm High Pressure Dilatometer
LOCATION SIZE TYPE QTY

Oil to Electricity Separating M6 by 30mm A 16
Union
Chinese Lantern Rings M5 by 20mm A 4
Upper Thread Adaptor M10 by 16mm B 1
Chinese Lantern Spacer ring M5 by 20mm A 2
Rod Adaptor Cable Clamp M5 by 50mm A 6

Rod Adaptor Socket plate M5 by 40mm C 6
Upper body locking M6 by 10mm B 4

Lower body locking M6 by 10mm B 4
Centre section cover sleeves M4 by 8mm D 4
Follower retaining strap M4 by 8mm D 2
Spring clamp plates M3 by 20mm E 12
Upper socket plate M3 by 6mm A 6
Lower socket plate M3 by 6mm A 6
Compass can sensor assembly M3 by 6mm E 4
Calibration Jig M3 by 6mm special set 2

screw
Notes:
Type A is a socket head cap screw, stainless steel, ISO metric coarse thread
Type B is a socket head grub screw, stainless steel, ISO metric coarse thread
Type C is a pan head slotted screw, brass plated, ISO metric coarse thread
Type D is a countersunk socket head screw, stainless steel, ISO metric coarse thread
Type E is a countersunk slotted head screw, stainless steel, ISO metric coarse thread.

APPENDIX G. Packing list of parts supplied
SEPARATE ITEMS:-
Weight Item description Qty Remarks
Yellow box containing 73mm HPD parts 1 Contents see
below
Yellow Tool box, with spares 1 Contents see
below
Pallet containing: Contents see below
a) Cased standard test gauge
b) High Pressure Control Unit
c) Reel 8mm Hose and cable 150metres long With couplers
d) Reel 6mm Hose and cable 110metres long Oil return + fittings
e) Reel 8mm Hose and cable 110metres long Spare, no fittings
f) Large Windlass
g) Oil Hand Pump c/w 250 bar Gauge In 2 pieces
h) Air Compressor Boxed
i) Compressed air bottle, 12 litre, 300 bar 2
j) Instrument Handling Frame In three pieces
k) Pressure hoses 5
Items in yellow case for MPX-HPD :-

Item description
Qty Remarks
Multiplexed High Pressure Dilatometer MK IV 1
Serial Number:040409
Compass and Multiplexer can 1 Serial Number:040407
and fitted to probe
CHL 1 fitted to probe
CHL spacing ring 1 fitted to probe
Nitrile rubber membrane (5mm thick) 1 fitted to probe - reinforced
Black protective cover LH (for cable) 1 fitted to cable + coiled
cable link
Fluid to Electricity Separating Union. 1 With blanking plug for gas
connection.
LH/RH cable coupler 2 One fitted to separating
union, one to probe.
Calibration Cylinder 1 fitted to probe
Special Calibration Cylinder (screw in ends) 1
73 HPD Electronics Interface Unit (EIU) 1 Serial Number: 040908
Power lead (12 volt battery to interface unit) 1
RS232 lead (from OESU to EIU) 1
RS232 lead (EIU to IBM PC 9-pin serial input) 1
I metre EW rod c/w coupler 1 Drill string adaptor
0.5m EW Rod c/w coupler 1 Drill string adaptor
Cross bar for instrument handling frame 1
Pressure line Tee piece 1 For oil pump outlet
MkIV 2-way free plug 1 Spare, for EIU

Tools and Equipment in Yellow Tool Box:-

Item description
Qty Remarks
Reinforced membranes 5
Standard membranes 10
Extrusion preventers, pair 16 (Xmas trees)
Chinese lanterns complete 2
Set of 18 CHL strips 5 c/w rivets
Strain follower assembly 3
Total pressure cell 1
Software CD containing Logger 1 MSDos and Windows 95
Arm calibration device 1 In its own special box
Brass tube to protect probe lower plug 1
Membrane blow-on tube 1
Bleed hose 1
Pack of plastic funnels 1 For oil handling
Screw pack 1 Complete set for probe
O ring pack 1 Complete set for probe
CHL spacing rings 2 Remainder of spacing
pack
Pop rivet gun 1
Hand drill 1
Pack of drill bits 1 ( 1/8 dia )
Soft hammer 1
Tin snips 1 ( 8 length )
5mm Hex wrench 1
4mm Hex wrench 1
3mm Hex wrench 1
2.5 mm Hex wrench 1
5mm Ball driver 1
4mm Ball driver 1
L form Allen key, 3/16 inch 1 Used for oil pump handle
L form Allen key, 3/32 inch 1 Used for Swagelock
handles
Pack of Swagelock tools 1 Spacing piece, packing
tightening piece
Penknife of Allen Keys
Multiclip 2
30/32 OE spanner 2
Large hook spanner 2 For probe
Small hook spanner 2 For cable couplers
l8/l9mm OE spanner 2
7mm nut driver 1 (for clips securing OESU )
20mm combination spanner 2
Yellow Tape (reel) 6
Silver duct tape (reel) 1
Short Test Lead 1

Oil pump release valve handle 1

Pair of hydraulic couplings 1
Pair of pneumatic couplings 1
Large flat ended screwdriver 1
Medium Screwdriver 1
Pozidrive screwdriver 1 size 2
Set of Mk IV Plug and Socket Tools 1 shell size 1
Large Adjustable Spanner 1
Chain Wrench 1
Tube of Copper loaded grease 1
Pack of Araldite (Rapid) adhesive 1
Tube of Silicone Grease 1
Pack of assorted sealing washers and unions 1
PTFE Wire Stripper 1 (adjustable)
Pack of spare PTFE wire 1
Stillson Wrench 1 14
Desiccator 1 Spare, in sealed pack
Resistor Pack 1 MR25 series
12v soldering iron 1
Reel of 60/40 solder 1
Handling frame securing pin 3
V block for handling frame 1
Adapter, gas bottle to Hanson 1
Instruction Manuals 2

Pallet:
Item description Qty Remarks
250mm STG % accuracy, 0 to 250 bars 1 In aluminium case,
Serial No: 041052/1
High Pressure Control Box:- 1 Serial No: 4870229
Air Compressor 1 Serial No: 5204 0608 3
Gas Cylinder (12 litre capacity, 300 bar WP) 2
150 metre 8mm bore hose/cable on reel 1 With couplers and black
caps + coiled cable link
110 metre 6mm bore hose, on reel 1 Oil return line
110 metre 8mm bore hose/cable on reel 1 Spare, no couplers
Large windlass 1
Instrument Handling Frame 1 In 3 pieces two off A
frames, one off pipe vice
Oil Connection hoses 1/4" bore as follows: 3
a) Pump to gauge 3 metres
b) Pump to Interface Unit (optional) 5 metres
c) Gauge to Separating Union 5 metres
Gas Connection hoses 1/4" bore as follows: 2
a) Bottle to Control Panel Modified Hanson fittings
b) Control Panel to Separating Union Std CEJN fittings
Oil pump 1 With handle and gauge
loose.

APPENDIX H REFERENCES
ARNOLD, M. (1981)
An Empirical Evaluation of Pressuremeter Test Data, Civil Engineering Department,
University of Adelaide, South Australia.
BAGUELIN, F., JEZEQUEL, J.F. and SHIELDS, D.M. (1978)

The Pressuremeter and Foundation Engineering, Transtech Publications, Clausthal, Germany
ISBN O-87849-019-l
BELLOTT1, R., GHIONNA, V., JAMIOLKOWSKI, M., ROBERTSON, P. & PETERSON,

R. (1989)
"Interpretation of moduli from self-boring pressuremeter tests in sand". Geotechnique
Vol.XXXIX, no.2, pp.269-292.
BOLTON, M.D and WHITTLE, R.W (1996)

A non-linear elastic/perfectly plastic analysis for the pressuremeter test. Paper as yet
unpublished, available from Cambridge Insitu.
DENBY,G.M., COSTA,C.A., CLOUGH,G.W. and DAVIDSON,R.R.(1981)

"Laboratory and Pressuremeter Tests in Stiff Clay". Proceedings of the Tenth International
Conference on Soils Mechanics and Foundation Engineering, Stockholm, Sweden.
DENBY, G.M. and HUGHES, J.M.O. (1982)

Horizontal stress interpretation of pressuremeter tests. Proc. ASCE Conf. on Updating
Subsurface Sampling of Soils and Rocks and their In-Situ Testing, Santa Barbara, California.
ERVIN, M.C., BURMAN, B.C. and HUGHES, J.M.O. (1980)

The use of a high capacity pressuremeter for design of foundations in medium strength rock.
International Conference on Structural Foundations on Rock, Sydney, 1980
FAHEY, M. (1981)
A Study of the Pressuremeter Test in Dense Sand. PhD Thesis University of Cambridge.
FAHEY, M. and JEWELL, R.J. (1990)

Effect of Pressuremeter Compliance on Measurement of Shear Modulus. Paper submitted to
The Third International Symposium on Pressuremeters, Oxford University 2 - 6 April.
GHIONNA, V., JAMIOLKOWSKI, M., LANCELLOTTA, R. & MANASSERO, M (1989)

Limit Pressure of Pressuremeter Tests. Proc. of 12th ICSMFE, Rio De Janeiro.
GIBSON, R.E. and ANDERSON, W. F. (1961)

In situ measurement of soil properties with the pressuremeter, Civil Engineering and Public
Works Review, Vol. 56, No. 658 May pp 615-618.
HABERFIELD C. M. and JOHNSTON I. W (1990)

The interpretation of pressuremeter tests in weak rock - theoretical analysis. Proc. ISP.3
Oxford, April.

HAWKINS, P. G., MAIR, R. J., MATHIESON, W.G. & MUIR WOOD, D. (1990)
Pressuremeter measurement of total horizontal stress in stiff clay, Proc. ISP.3 Oxford, April.
HOULSBY, G., CLARKE, B. and WROTH, C.P. (1986)

"Analysis of the unloading of a pressuremeter in sand". Proc. of the 2nd International
Symposium on The Pressuremeter and its Marine Applications pp 245-262.
HOULSBY, G & WITHERS, N.J (1988)

Analysis of the Cone Pressuremeter Test in Clay. Geotechnique, Vol 38, No. 4, pages 573-
587.
HUGHES,J.M.O., ERVIN, M.C. (1980)

Development of a High Pressure Pressuremeter for determining the engineering properties of
soft to medium strength rocks.Proc. 3rd Aus.-NZ Conf. Geomechanics, Brisbane, pp.292-
296.
HUGHES, J.M.O., WROTH, C.P. and PENDER, M.J. (1975)

A comparison of the results of special pressuremeter tests with conventional tests on a deposit
of soft clay at Canvey Island, Internal Report No. CUED/C - SOILS TR 20, Cambridge
University Engineering Department.
HUGHES, J.M.O., WROTH, C.P. and WINDLE, D.(1977)

Pressuremeter tests in sands, Geotechnique No. 4, pp 455-477
JEFFERIES, M.G. (1988)

Determination of horizontal geostatic stress in clay with self-bored pressuremeter.
Can. Geotech. 25 (3), pp 559-573
JEWELL, R.J., FAHEY, M. and WROTH, C.P. (1980)

Laboratory Studies of the Pressuremeter Test in Sand. Geotechnique 30, No. 4 pp 507-531.
MARSLAND, A. and RANDOLPH, M. F.(1977)

Comparison of the Results from Pressuremeter Tests and Large Insitu Plate Tests in London
Clay. Geotechnique Vol. 27 No. 2 pp 2l7-243.
MAIR, R.J. and WOOD, D.M. (1987)

Pressuremeter Testing. Methods and Interpretation. Construction Industry Research and
Information Association Project 335. Publ. Butterworths, London. ISBN 0-408-02434-8
NEWMAN, R.L., CHAPMAN, T.J.P. & SIMPSON, B. (1991)

"Evaluation of pile behaviour from pressuremeter tests". Proc. Xth European Conference on
Soil Mechanics and Foundation Engineering, Florence, May 1991
PALMER, A.C.(1972)
Undrained plane-strain expansion of a cylindrical cavity in clay: a simple interpretation of the
pressuremeter test, Geotechnique Vol. 22 No. 3 pp 45l-457.
RANDOLPH, M.F. and WROTH, C.P. (1978)

An analytical solution for the consolidation around a driven pile. Int. J. Numer. Anal.
Methods in Geomech.

ROWE, P.W. (1962)

"The Stress Dilatancy Relation for Static Equilibrium of an Assembly of Particles in
Contact". Proceedings of the Royal Society. Vol. 269, Series A, pp 500-527.
WHITTLE, R.W. and DALTON, J.C.P. (1990)

Discussing 'Experience with the self boring rock pressuremeter'. Ground Engineering,
Jan/Feb pp 30-32.
WHITTLE R. W, DALTON J. C. P and HAWKINS P. G. (1992)

Shear Modulus and Strain Excursion in the Pressuremeter Test.. Proc. Wroth Memorial
Symposium, Oxford, July.
WOOD, D.M. and WROTH, C.P. (1977)

Some laboratory experiments related to the results of pressuremeter tests. Internal Report No.
CUED/C Soils TR 32, Cambridge University Engineering Department, published in
Geotechnique 27, pp 181-201 (1977).
WROTH, C. P. (1984)
"The Interpretation of In Situ Soil Tests". Twenty Fourth Rankine Lecture, Geotechnique 34,
No. 4, pp 449-489

APPENDIX I GLOSSARY OF TERMINOLOGY
ASCII - American Standard Code for Information Interchange. It is the most universal
character coding set using 7-bits.
EW - A drilling rod size
CHL - Chinese Lantern.
CI - Cambridge Insitu.
DMM - Digital Multi-Meter.
DVM - Digital Volt-Meter.
EIU - Electronic Interface Unit.
FESU - Fluid to Electricity Separating Union
GESU - Gas-to-Electricity Separating Union.
GMS - Glass-to-Metal-Seal
GPa - GigaPascal
HPD - High Pressure Dilatometer
kPa - KiloPascal
MEM - Membrane
MPa - MegaPascal
MPX - Multiplexed
NX - A drilling/coring size, approximately 76mm/3 inches.
OESU - Oil to Electricity Separating Union
PCP - Pressure Control Panel.
RR - Rock Roller bit.
RS232 - A convention for transmitting and receiving digitally encoded data in serial form
TPC - Total Pressure Cell.
Xmas - An abbreviation for Christmas, used to refer to the extrusion limiters.

Cambridge 73mm High Pressure Dilatometer - Working Instructions

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cambridge 73mm High Pressure Dilatometer - Working Instructions

Uploaded by

Copyright:

Available Formats

WORKING INSTRUCTIONS FOR

Telephone +44 (0)1223 262361

HPD Working Instructions.doc Version 060110 Page 2 of 113

HPD Working Instructions.doc Version 060110 Page 3 of 113

PART 2. DESCRIPTION OF THE HPD

Figure 2.1 The basic HPD-MPX kit

HPD Working Instructions.doc Version 060110 Page 4 of 113

The instrument is inflated by oil or gas. A strong hose connects

HPD Working Instructions.doc Version 060110 Page 5 of 113

An additional feature of this pressuremeter is a special electronic compass module fitted to

FIGURE 2.3 The HPD-MPX Disassembled

HPD Working Instructions.doc Version 060110 Page 6 of 113

PART 3 DISMANTLING THE INSTRUMENT

2. With the 5mm hex wrench take out the

3. Remove the CHL end float ring, which

4. Separate the electronics cover from the

HPD Working Instructions.doc Version 060110 Page 7 of 113

10. Procedure for an oil inflated instrument:

12. Procedure for gas inflated instruments:

HPD Working Instructions.doc Version 060110 Page 8 of 113

Fig 3.2 Parts of the cable coupler

HPD Working Instructions.doc Version 060110 Page 9 of 113

HPD Working Instructions.doc Version 060110 Page 10 of 113

32. Now fit the membrane.

33. Procedure for oil:

HPD Working Instructions.doc Version 060110 Page 11 of 113

35. Procedure for a gas inflated instrument:

45. Re-tighten the ring nuts as far as possible.

46. Repeat steps 43 to 45 for 10MPa and 20MPa respectively.

HPD Working Instructions.doc Version 060110 Page 12 of 113

50. Screw together the Chinese lantern.

The instrument is now ready for calibrating the membrane.

HPD Working Instructions.doc Version 060110 Page 13 of 113

4. DESCRIPTION OF THE TESTING PROCEDURE

Coring the Pocket

Description of the Test

HPD Working Instructions.doc Version 060110 Page 14 of 113

Tests in material which can be failed - Type A

HPD Working Instructions.doc Version 060110 Page 15 of 113

FIGURE 4.1 Example of HPD test in a material showing large deformation

HPD Test in Coal

HPD Working Instructions.doc Version 060110 Page 16 of 113

Tests in material where all deformation is elastic - Type B Test:

HPD Working Instructions.doc Version 060110 Page 17 of 113

10000 FIRST LOOP

HPD Test in Andersite

HPD Working Instructions.doc Version 060110 Page 18 of 113

HPD Working Instructions.doc Version 060110 Page 19 of 113

5. ANALYSING A TEST IN SOIL

Index to this section:

HPD Working Instructions.doc Version 060110 Page 20 of 113

Deriving parameters from pre-bored pressuremeter tests in soil.

HPD Working Instructions.doc Version 060110 Page 21 of 113

FIGURE 1.1 - Elastic Response followed by failure in shear

1.2 Defining strain

HPD Working Instructions.doc Version 060110 Page 22 of 113

FIGURE 1.2 Pressures and strains around the expanding cavity

1.3 Average displacements versus the output of the separate axes

1.4 The analysis program

HPD Working Instructions.doc Version 060110 Page 23 of 113

a point can be part of the expansion curve

HPD Working Instructions.doc Version 060110 Page 24 of 113

2. ANALYSES FOR EXPANSION

2.2 Marsland & Randolph (1977) Analysis