Anda di halaman 1dari 81

The 3rd International Conference

on Site Characterization

April 1-4, 2008 Taipei

Short Course on Flat Dilatometer (DMT)


DMT Equipment and Testing

Authors: D. Marchetti, S. Marchetti, P. Monaco


Presented by: D. Marchetti
Presentation Outline
References and Background
Test Layout
Field Machines for Blade insertion
Equipment details (blade, control box, cables,adaptors)
Calibration: A and B
Test Procedure: A and B readings
Disassembling a blade (& membrane replacement)
Checks and Quality Control
Accuracy of DMT measurements
Dissipation Test
C-reading
Reduction Formulae
Software
References and Background (1/2)
First Publication:
Marchetti S. (1980) In Situ Tests by Flat
Dilatometer, ASCE Jnl GED, Vol 106, No. GT3,
Mar, 299-321
Standards:
ASTM D 6635-01 (2001). "Standard Test Method
for Performing the Flat Plate Dilatometer".
Eurocode 7 (1997). Geotechnical design - Part 3:
Design assisted by field testing, Section 9: Flat
dilatometer test (DMT).
References and Background (2/2)

TC16 (2001) "The Flat Dilatometer Test (DMT)


in soil investigations"
Report by the ISSMGE Committee TC16
comprehensive document incorporating all the most
important information on DMT (procedure, interpretation,
applications)

DMT on the Internet:


Key papers on the DMT can be downloaded from the
bibliographic site: www.marchetti-dmt.it
DMT Test Layout

3 4
5 6 1. Dilatometer Blade
2. Push Rods (i.e. CPT)
7
3. Push Force or dynamic penetration
4. Pneumatic electric cable
5. Control Box
2 6. Pneumatic cable
7. Gas Tank
P0 (A) Lift-off pressure
1
P1 (B) Pressure for 1.1 mm expansion

Independant from insertion method


Ways of Inserting the blade (1/2)

Most efficient method: pushing from truck

Cable coming out of the rods


Ways of Inserting the blade (2/2)

Driven by Spt tripod Pushed by drill rig Driven or pushed by


light penetrometer

Pushed by a fixed platform Driven from a barge


Driven by drill rig
Torpedo Method
To 76mm rods

TORPEDO
L3m

TORPEDO pre-assembled before test and joined to 76 mm rods


Cable exits from slot, runs through channel in collar and is taped to rods
Rods + torpedo are inserted at bottom of borehole
Upper Slotted Adaptors
Soils testable by DMT
Suitable for SANDS, SILTS, CLAY
(grains small vs membrane D=60
mm). But can cross through GRAVEL
layers 0.5 m
Very robust, can penetrate soft rocks
(safe push on blade 25 ton)
Clays: Cu = 2-4 kPa to 1 MPa (marls)
Moduli: M = 0.5 to 400 MPa
Penetrates fast and easily in hard
soils, PROVIDED sufficient pushing
capacity (e.g. 20 ton trucks)
DMT Test Connections
Ground Cable Connection

Ground cable must provide good contact with rods (blade)


DMT Complete Equipment
tank

rods

p/e cable
to DMT
DMT Readings

P0 Lift-off pressure

P1 Pressure for 1.1 mm


expansion
DMT Working Principle

Works like electric switch


(On/Off)
System must be clean (no
dirt on contacts!)
gas must be non-corrosive
(generally air or nytrogen)
Control box
Control box details
Pneumatic-electric cables 1/3

Pneumatic nylon tube


with conductive steel
wire inside

NOTE: near the joints,


teflon sleeve teflon sleeve the inner wire is
insulated from
connectors by a teflon
sleeve
Pneumatic-electric cables 2/3

start with
extendable cable
Unscrew cable leader
with cable leader
(when using both)

Three different joints:


1. Quick connector male: to
control box (not isolated)
2. Terminal male: to blade or
to extendable cable
(isolated)
3. Terminal female: to cable
1 2 3 leader or other cable
(isolated)
Pneumatic-electric cables 3/3

Metal connectors electrically insulated by from inner wire


Joints must be airtight ( 80-100 bar)
Cables and joints are not easily repairable in the field
Lower adaptor to DMT
Use of stronger rods 1/2

Most commonly CPT or drill rig


rods
Often rods "weakest element in
the chain" (20 ton truck+ high
strength 25 ton blades)
Use of STRONGER RODS may be
convenient e.g. commercially
available 45 mm rods used for
pushing 15 cm2 cones (same
steel as CPT rods)
Use of stronger rods 2/2
ADVANTAGES
Can penetrate through cemented layers / obstacles
Lateral stability against buckling in the first few
meters in soft soils or in empty borehole
Use completely the push capacity of the truck
Less risk of deviation from verticality
Risk of losing the rods 0 (wall = 22 mm)

DRAWBACKS
(Initial cost) and heavier (9 kg/m rather than 6.4
kg/m of 36 mm CPT rods + 40 %)
No big advantage in OC clay (+ skin friction)
Calibration of Membrane
The natural membrane position is between A and B.

A
free

Membrane has non-zero rigidity


Definitions:
A = external pressure which must be applied to the
membrane in free air to collapse it against its seating (i.e. A-
position)
B = internal pressure which in free air lifts the membrane
center 1.1 mm from its seating (i.e. B-position)
A and B
A & B are used to correct the A & B readings
into p0 & p1 ( TARES to be detracted)

A & B must always be measured before and


after each test

Calibration is a good indicator of equipment


condition, hence of quality of data

A large difference between before/after A & B


values should prompt a membrane change (usually
apparent)
Calibration of Membrane

Use syringe and short calibration cable for performing calibration


Determination of A and B
To obtain A
Apply vacuum by pulling back the syringe piston (vacuum
causes an inward deflection of the membrane similar to that
due to external soil pressure at the start of the test)
buzzer becomes active.
Slowly release the piston and read A on the low-range
gage when buzzer stops.
Note this negative pressure as a positive value, e.g. A = 15
kPa for a vacuum of 15 kPa (the correction formula for p0
takes into account that a positive A is a vacuum).
To obtain B
Push slowly the piston into the syringe and read B on
the low-range gage when buzzer reactivates.

REPEAT SEVERAL TIMES


CALIBRATION (with blade inaccessible)
At the beginning of a sounding, the blade is in the
hands of operator: 1st configuration OK.
Later, when the blade is inaccessible (under the
truck) a 2nd configuration is used.
The configuration is the same as during current
testing, with cables of normal length (20-30 m).
Procedure is identical. However in 2nd case, due to
the length of the DMT tubings, there is some time
lag (easily recognizable by the slow response of the
pressure gages to the syringe). Therefore, in that
configuration, A & B must be taken SLOWLY
(15 sec OK).
Dilatometer Test Sequence
STEP-by-STEP procedure: A, B (, C)
As soon as rig operator reaches test depth,
he signals go-ahead to DMT operator, who:

(Closes yellow vent valve). Slowly opens


black micrometer valve (signal on). When
sound stops, reads A.

Continues to inflate (signal off). When


sound reactivates (1.1 mm) reads B.
Immediately after B, follows 4 operations:
1. Open vent valve: depressurize membrane.
2. Close micrometer (pressure supply).
3. Gives go-ahead to rig operator to advance Step Z.
4. Write A and B.
Avoid Overinflating Membrane
TRIVIAL REASON: FORGET DEFLATING AFTER B-signal.
May happen to BEGINNER.
SERIOUS REASON: absence of B reading (due to DIRT)
CLEAN 4 POINTS BELOW DIRT, GRAINS, CLOTH ...
NO CONTACT = NO B-signal. KEEPS INFLATING ...

compressed air
with air pistol

NO NEED to clean periodically (day, week ...) but only after damage (dirt inside).
Cleaning inside
the blade
Opening the blade
Extracting the sensing disc

Sensing Disc extractor


Feeler
Sensing Disc

The feeler is spring loaded

Screw holding the spring


Cleaning the Central Hole
1. Clean the central hole
2. Place the steel spring and
cilinder in the hole
3. Place the sensing disc
4. Ensure sensing disc is
well pressed down
Disc stationary: must fit TIGHTLY
Disc must get forced inside the insulating seat, thanks to
lateral gripping force.
Extraction force > weight of blade. If sensing disc is lifted
(using extractor), blade is lifted too.
If coupling becomes loose (e.g. damage) and disc free to
move, increase gripping force.
A quick fix: insert a small piece of plastic sheet (lateral to
disc, not on bottom). Then trim.
Ideal Relative Elevations

TRIPOD & DIAL GAGE


TOLERANCES:
Sensing Disc : 0.04 to 0.07 mm above surrounding plane
Feeler : 0.04 to 0.07 mm above Sensing Disc
Electrical Sharpness of B-reading
Just before closing the blade:
Apply continuity tester to the blade:
Push up/down quartz cylinder 10 times: B-contact sharp and prompt

CONTINUITY
TESTER

Just after closing the blade:


CONTINUITY When blade is closed,
TESTER push center of
membrane. Buzzer
should sound
Exercise a new Membrane and Check Airtightness

1. Apply exercise syringe


to back of blade
2. Push piston from start
to end (5 bar)
Achieves both functions
Longitudinal wire
Note: longitudinal wire running inside blade is the "live" (+)
electrical pole.
Such + pole is insulated from body of the blade. It is the same
pole as the steel wire in pneumatic tube.

The wire cannot be disassembled from the "metal base of


sensing disc".
To clean: use compressed air inside conduit.
Use continuity tester to check contact / no contact (tubelet &
metal base must not be in contact with body).
Cables: Electrical Checks

Check electrical continuity of pole +


Check electrical isolation of pole + and connectors
Cables: Airtightness Check

PRESSURIZE at CLOSED ENDED


40-70 bar TERMINAL

SUBMERGE IN WATER
Quality Control

CHECKING TOOLS that must be available at site:


1. EXERCISE SYRINGE (5 bar) for exercising new
membranes and checking airtightness.
2. CONTINUITY TESTER
3. L-SQUARE to check blade coaxiality.
4. FEELER GAGE 0.5 mm thick and RULER 15 mm
long to check blade planarity.
5. TRIPOD with dial gage.
6. STOPWATCH for checking time to A and B, and
for the dissipation tests.
Main Checks 1/2
Checks on hardware
The main checks concern essentially the blade. Problems
in cables and control box are generally self apparent.

1. Using the tripod, verify proper elevation of sensing disc


(0.04 to 0.07 mm above surrounding plane) and feeler
(0.04 to 0.07 mm above disc).

2. In the calibration configuration: Apply 10 cycles of


push-pull to the syringe piston to verify sharpness of
signal inversion (off-to-on and viceversa).
Main Checks 2/2
3. Verify that all blades have A, B in tolerance:
A = 0.05 to 0.30 bar B = 0.05 to 0.80 bar

4. Using the 5 bar syringe verify airtightness in a


bucket of water.

5. A should be reached in 15-20 sec, B within 15-


20 sec after A.

6. The C-reading should be reached within 30 to


60 sec after starting the deflation
Acceptance Values of A and B
EUROCODE 7 (1997)
INITIAL values of A, B (before inserting the blade)
must be in the ranges:
A = 5 - 30 kPa
B = 5 - 80 kPa
If not, replace the membrane before testing.
FINAL values of A, B
The change of A or B between beginning and end
of sounding must be 25 kPa
If not, test results must be discarded.
TYPICAL values of A, B
A = 15 kPa
B = 40 kPa
Importance of Accurate A and B
Inaccurate A, B are virtually the only potential source of
DMT instrumental error
Any inaccuracy in A, B would propagate to all A, B of a
sounding
Accurate A, B are necessary in soft soils ( liquid clays or
liquefiable sands) where A, B are just a bit higher than A,
B (correction difference between similar numbers)
Small inaccuracies in A, B are negligible in medium to
stiff soils (A, B are a small part of A, B)

How A, B can go out of tolerance :


overinflating the membrane far beyond the B-position
(once overinflated a membrane requires excessive suction to
close - A > 30 kPa - even B may be a suction)
Planarity Check

Flat 0.5 mm
feeler gage

Place a 15 cm ruler against the face of the blade parallel


to its long side.
The "sag" between the ruler and blade should not exceed
0.5 mm.
Coaxiality Check
Between blade and axis of the rods

With the lower adaptor mounted on


the blade, place the inside edge of
an L-square against the side of the
adaptor.
L-SQUARE
Note the distance from the
penetration edge of the blade to
the side of the L-square.
Turn the blade 180 and repeat the
measurement.
The difference between the two
distances should not exceed 3 mm
ADAPTOR (corresponding to a coaxiality error
of 1.5 mm).
DMT data field form

Even if laptop available, it


is good practice to write
readings on field form

Every 20 cm 2 numbers
h (kPa) Dissipation Tests

IN CLAYS AND SILTS


(not feasible in sandy
Time (min) silt, sand and gravel)
Stop the blade at a given depth
Monitor decay of the total contact horizontal stress h with time
Infer coefficient of consolidation / permeability (ch , kh ) from rate
of decay of h
DMT-A Dissipation Procedure
RECOMMENDED METHOD (for other methods see TC16 2001)

Timed sequence of A-readings (no membrane expansion to B)

DMT-A PROCEDURE
Stop blade at a given depth and start a stopwatch (t = 0 when
pushing is stopped). Slowly inflate the membrane to take the
A-reading. Vent blade immediately at A. Record A-value
and stopwatch time at the instant of A-reading.
Continue to take additional A-readings e.g. by a factor 2
increase in time (0.5, 1, 2, 4, 8, 15, 30 etc. minutes after
stopping the blade).
Plot in the field a preliminary Alog t diagram (usually S-
shaped). Stop dissipation test when the Alog t curve has
flattened sufficiently to clearly identify the time at
contraflexure point tflex (used for interpretation).
Dissipation Field Forms

Draw Graph

Write Numerical Data


Dissipation with laptop 1/2
Dissipation with laptop 2/2
Advantages of DMT-A Dissipation
Similarity to the well-established "holding test" by
pressuremeter
- theory for the DMT h vs time decay curve not available yet
(but expected similar)
- fixity inherently insured for the DMT blade (solid object)
NO problems of filter smearing / clogging / loss of saturation
- DMT membrane = non-draining boundary
- what is monitored is a total contact stress
Straightforward interpretation (no need to know uo )
- Plot Alog t curve
- Identify tflex
- Calculate ch 7 cm2 / tflex
C-Readings in sands
Besides "normal" A & B readings, a third
reading C - closing pressure - can also
optionally be taken by slowly deflating the
membrane just after reading B is reached.
To perform the C-reading, there is only one
difference in the normal test sequence:
After B, open the slow vent valve instead of
the fast toggle vent valve and wait 1 min
until the pressure drops approaching the
zero of the gage. At the instant the signal
returns, take the C-reading.
Corrected C- reading Note that, in sands, the value to be
p2 = C - ZM + A expected for C is a low number (usually <
100 - 200 kPa, i.e. 10 or 20 m of water).
Frequent Mistake in C-Readings
After B, i.e. when the slow deflation starts, the
signal is ON. After some time the signal stops (from
ON to OFF). The mistake is to take the pressure at
this inversion as C, which is incorrect (at this time the
membrane is the B-position).

The correct instant for taking C is some time later


( 1 min), when, completed the deflation, the
membrane returns to the "closed" A-position, thereby
contacting the supporting pedestal and reactivating
the signal.
Graph of Frequent Mistake in C-Readings

Pressure
Incorrect C
Vent Reading !!!
B

Correct C
A
C

Time

Acoustic On
Signal Off
Frequence of C-readings
(a) SANDY SITES
In sands (B 2.5 A) C-readings may be taken
sporadically, say every 1 or 2 m, and are used to
evaluate u0 (equilibrium pore pressure) as u0 p2
Repeat the A-B-C cycle several times to insure that
all cycles provide similar C-readings (A-B taken only
the first cycle).
(b) INTERBEDDED SANDS AND CLAYS
If the interest is limited to finding the u0 profile,
then C-readings are taken in the sandy layers (B
2.5 A), say every 1 or 2 m.
When the interest, besides u0 , is to discern free-
draining layers from non free-draining layers, then
C-readings are taken at each test depth.
C-reading (pressure on membrane at
"membrane closure") in SAND = Uo

water film
between
sand and
membrane

Campanella et al. 1985 (Proc. 11th ICSMFE, San Francisco, 2, p. 852)

No contact sand/membrane = ' + Uo


In sand C Uo (in situ equilibrium pore pressure)
Uo profiles from C-READINGS in SAND

Same Uo as from
piezometer, without
problems of:
Filter clog
Smearing
Saturation

Schmertmann 1986. Some 1985-86


Development in Dilatometer Testing
and Analysis. Proc. PennDOT and
ASCE Conf. on Geotechnical
Engineering Practice, Harrisburg, PA
C-reading Results

in SANDS C Uo
measures Uo
( piezometer)

in CLAYS C > Uo
evidences u
(i.e. non free-draining)

Schmertmann 1988 (DMT Digest


No. 10, May 1988, Fig. 3)
Distinguishing free-draining / non free-draining layers

Definition of Pore Pressure Index:

UD = (p2 - u0) / (p0 - u0)

UD 0 free-draining
UD > 0 non free-draining

Benoit 1989
DMT Reduction Formulae

Po and P1

Intermediate
parameters

Interpreted
parameters
Chart for Soil Description Unit Weight
SOIL DESCRIPTION
and ESTIMATED w
2000
EQUATION OF THE LINES: SAND
(n+m log ID ) 5
ED =10
1000
m n
2.1
SILT SILTY
Dilatometer Modulus ED (bar)

SANDY
A 0.585 1.737

CLAYEY
B 0.621 2.013
C 0.657 2.289 2
2.1
500

Fig. 16 TC16 Report


D 0.694 2.564

CLAY
SILTY 5 1.9
1.9
200
2.0
5 (Marchetti & Crapps 1981)
1.8 1.8
1.9
100
D Used to construct an
1.8 1.7 1.7
50 C approximate profile of 'vo
B
1.7 1.6 (needed in the elaboration)
20
1.6
A
0.33 0.8 1.2 3.3
12
MUD 10 MUD
0.6 1.8
and/or ( ) If PI>50, reduce by 0.1
1.5
PEAT
5
0.1 0.2 0.5 1 2 5

Material Index I D
How to use A, B (p0, p1)
STEP 1 STEP 2
CALCULATE Convert Id Kd Ed to
INTERMEDIATE COMMON PARAMETERS
(OBJECTIVE) (via CORRELATIONS)

Z Po P1 Id Kd Ed Ko OCR M Cu
m bar bar bar bar bar deg

1.0 1.1 3.3 1.87 6.3 73 151 38.3


1.2 1.3 1.8 .33 6.6 15 1.4 6.5 31 .19
1.4 1.2 1.7 .37 5.7 15 1.3 5.1 29 .17
1.6 1.2 1.6 .28 5.3 11 1.2 4.6 21 .16
1.8 1.1 1.4 .21 4.6 8 1.1 3.6 13 .14

Basic philosophy: evaluate familiar parameters (users can check vs


other tests). Design via parameters.
Generally, No correlations directly to bearing capacity, foundations
etc.
M and Cu are the most useful and accurate parameters by DMT.
How to use DMT Results

ID
SOIL KD
TYPE shape similar
to OCR
sand
useful to
silt
understand
clay history of
deposit

KD = 2
M Cu OCR = 1
common use
ID Material Index (Soil Type)

Performing DMTs, immediate notice that:

CLAY SAND
p0 p p
p1 p0 p1

came natural (apart theory) define ID as a "vicinity ratio"

(p1 - p0)
ID =
(p0 - u0)
ID Material Index (Soil Type)
Experience has shown
ID very sensitive, 0.1 to 10 (2 log cycles)
0.1 0.6 1.8 10
CLAY SILT SAND

Like FR in CPT but: amplified, highly reproducible


Not primary scope, but a nice extra generally reliable
ID not result of sieve analysis, but from mechanical
response ( rigidity index)
E.g. clay + sand described by ID as silt behaves
mechanically as ... (incorrect for grain size, + relevant
mechanical behavior)
If interest in permeability (besides ID ) other index UD
DMT Software
SDMT Elab is a software for handling combinedly DMT and Seismic DMT
data (requires USB hardware key).
Create a new Site
Create a new Vertical

Pore pressure U assigned with water level Zw or by layers


Gamma calculated from DMT -chart or assigned by layers
DMT header data

** Depths always from groundlevel


Aided data input
Draft of Input Data
Dat File Specifications
Example of Dat File:
.8,5,21,.22,.65,34.7,0,1,0,1.75,.2
"GEOTEST","NEW HARBOUR"
"LIVORNO","WEST QUAY",""
"8 AUG 1998","D1"
740,2210
465,1570
345,1310
282,1070
208,880
145,545
115,408
100,215
110,265
318,1510
422,1530
440,1480
361,1140
255,1110
331,1210
145,605
210,715
408,1220
258,910
235,545
145,675
165,275
Tabular Output (export to Excel)
Automatic Report Generator (MS Word)
Thank you for your attention

END OF PRESENTATION