E M 1110-2-1906

Change 1
1 May 80
APPENDIX XB:

DETERMINATION OF CRITICAL VOID RATIO

i. DEFINITION OF CRITICAL VOID RATIO.

Critical void ratio is

that void ratio at which a cohesionless soil can undergo deformation

or actual flow without volume change. In general, it is a function of
confining pressure, i.e., an increase in confining pressure yields a
lower value of critical void ratio.

Theoretically, a soil with a void

ratio above the critical value for the confining pressure is subject to
flow failure or liquefaction if it undergoes sufficient undrained stress,
whether that stress is cyclic or monotonic (steadily increasing). However, if the void ratio is lower than the critical value for the given
confining pressure, the soil will not liquefy under monotonic stress
increase.

Liquefaction resulting from monotonic stress increase and

initial liquefaction occurring under cyclic stresses are not the

same phenomenon.

For the purposes of this appendix, liquefaction

refers to the behavior of a mass of cohesionless soil during flow

For monotonic loading conditions, failure is caused by a sub,
stantial reduction in shear strength due to large increases in pore
slides.

pressure and the consequent great reduction of effective stress.

2. PRINCIPLES OF MONOTONIC TRIAXIAL TESTING. The monotonic stress-controlled triaxial compression test is used to determine
the approximate values of critical void ratio over an applicable range
of confining pressure.

The test is comparable to an R test with pore

pressure measurements (see Appendix X, TRIAXIAL COMPRESSION

TESTS) with the exception that load is applied to the specimen in
increments.

The increments are arbitrary but should be chosen to

produce several points on the stress-strain curve prior to peak deviator stress or the tendency to dilate, depending on the response of the
XB-4

E M 1110-2-1906
Appendix XB
Change 1
1 May 80
specimen as outlined below.

The ability to execute an efficient test

program to establish critical void ratio over a range of confining

pressures requires recognition of the nature of specimen behavior.
In general, three types of response from specimens can be anticipated, as illustrated in Figure 4.

I I

C-DILATANT FA,,?E

B-,M,TED LIOUEFACTION FA,LRE

, . /_A~,OEFACT!ON FA,LRE
------____---

Figure 1. Typical stress-strain and pore

pressure response, monotonic triaxial E
tests on sands
Liquefaction. If the void ratio after consolidation
-a. Complete
is sufficiently higher than the critical value for the given confining
p r e s s u r e , the specimen will bear several load increments with minor
axial strain (usually less than 2 percent) while exhibiting steadily
XB-2
-

E M 1410-2-1906
Appendix XB
Change 1
1 May 80
accelerating pore pressure increases until failure by liquefaction occurs catastrophically at peak deviator stress.

During the collapse of

the specimen, which occurs over a fraction of a second, the deviator

stress actually borne by the specimen declines rapidly while the pore
pressure approaches, but not necessarily equals, the confining pressure.
-b. Partial or Limited Liquefaction. If the void ratio after consolidation is slightly higher than or very nearly equal to the critical
value for the given confining pressure, the specimen will tolerate
several load increments with small axial strain until at peak deviator
stress sustained deformation will occur.

The deviator stress will

essentially remain constant or suffer minor decline during the

deformation. In some cases, deformation will cease, but the addition
of small increments of load that counter the effects of area increase
due to strain will maintain deformation. During the deformation, the
pore pressure will rise to a value somewhat less than the confining
pressure.

In the latter stages of the test, the specimen response may

become dilative, i.e., it will bear additional deviator stress, and the
pore pressure will begin to decrease.
c. Dilatant Failure. If the void ratio after consolidation is lower
than the critical value for the given confining pressure, the specimen
will tend to dilate, i.e. continue to accept loading, and the pore pressure
decreases.
3. APPARATUS. All the equipment is listed in Appendix

X, T R I A X I A L

COMPRESSION TESTS, for performing triaxial compression R tests

with pore pressure measurements plus (a) electronic transducers and
high-speed recorders for data acquisition, and (b) stress-controlled
loading capacity.
a. Loading Devices. The equipment may be either (1) a pneumatic system, (2) a closed-loop electrohydraulic system, or (3) a
deadweight system.

A deadweight system is practical only for

in.-diameter specimens.

i.4-

Such small specimens may be used, but

XB-3

E M 1110-2-1906
Appendix XB

Change 1
1May 80
they are undesirable from the standpoint of specimen preparation
(see Appendi.x X, TRIAXIAL COMPRESSION TESTS, pp X-35 and
X-16). Figure 2 shows a satisfactory arrangement for deadweight
loading, which consists of a horizontal loading bar bearing on the load
rod (piston) and connected to a weight hanger suspended beneath the
triaxial chamber.

A stopblock must be provided beneath the weight

hanger that will permit deformation of the specimen equivalent to not

less than 30 percent axial strain but will prevent contact of the
horizontal lo,ading bar with the top of the triaxial chamber. The
loading assembly should be designed to support 300 lb and constructed
of lightweight material.

Pneumatic or closed-loop electrohydraulic

systems must be capable of maintaining the axial load, i.e., supply

sufficient air or oil to allow the piston to follow the sudden and rapid
deformation during liquefaction.
b. e:cimen Cap. The specimen cap should be of a lightweight
noncorrosive material equipped with porous .metal or porous stone
inserts and drainage connections.

The cap can be similar to those in

Figure 4 of Appendix X, TRIAXIAL COMPRESSION TESTS, but

lubricated end platens similar to those used by
desirable .

Castrot are more

Preferable specimen diameter is 2.8 in. or larger.

c . Tr i a x i a l C o m p r e s s i o n C h a m b e r . The chamber is the same

as that in paragraph 3b
- of Appendix X, TRIAXIAL COMPRESSION
TESTS, or the newer design shown in paragraph

3c
- of Appendix XA,
CYCLIC TRIAXIAL TESTS, except that the rigid connection of the

loading piston to specimen cap is not required.

Piston to Loading Equipment Connection. See
-d. Loading
paragraph 3d
- of Appendix XA, CYCLIC TRIAXIAL TESTS.
Equipment. See paragraph 3e - of Appendix XA,
-e. Recording
CYCLIC TRIAXIAL TESTS.
t G . C a s t r o , Liquefaction of Sands, Harvard Soil Mechanics Series
No. 81, Jan 1969, Cambridge, Mass.
XB-4
-

E M 1110-2-1906
Appendix X B
Change 1
1 WY 80
J LOADING BAR

7
i
-kJ-

TRIAXIAL C H A M B E R

NARROWWORKBENCH
OR CANTILEVERED
PLATFORM

I
A-

LOADHANGER

NOTE

a
f

C L E A R A N C E B MST B E
EQUIVALENT TO NOT LESS
T H A N 30 P E R C E N T A X I A L
S T R A I N A N D L E S S THAN
C L E A R A N C E A.

STOP BLOCK

Figure 2.

Deadweight loading system

XB- 5

E M 1110-2-1906
Appendix XB
Change 1
J_s May 80
- f. M e-a s u r e m e n t T r a n s d u c e r s .

See paragraph 3f- of Appendix

XA, CYCLIC TRIAXIAL TESTS.

g. Back-Pressure Saturation Equipment.

See paragraph 3g

of Appendz<A, C Y C L I C T R I A X I A L T E S T S .
Rod for Moist Tamping Specimen Preparation.
-h. Tamping
(Optional) See paragraph 3& of Appendix XA, CYCLIC TRIAXIAL
TESTS.
4. TESTING PROCEDURES. Specimen preparation, back-pressure
saturation, and consolidation procedures specified in paragraphs 4band 7b(l)
through 7b(7)
of Appendix X, TRIAXIAL COMPRESSION
TESTS, are adequate and pertinent. However, preferred techniques
for handling undisturbed samples of cohesionless soils, preparing
remolded specimens, and anisotropic consolidation are as follows:
2. Reconstituted Specimen Preparation (Moist Tamping).
The procedure is as outlined in paragraph 4a- of Appendix XA,
CYCLIC TRIAXIAL TESTS, except that the initial water content of
the material should be near the bulking water content, i.e.,
percent.

5- 10

This water content permits molding of specimens in the

vicinity of the critical void ratio, which is typically a relatively high

value, i.e ., usually between 10 and 40 percent relative density,
depending on the confining pressure.
b. Specimen Preparation (Undisturbeg). See paragraph 4b- of
Appendix XA, CYCLIC TRIAXIAL TESTS.
2. Specimen
Measurement. Inasmuch as void ratio is a
principle of the test results, it is imperative that accurate initial
specimen dimension measurements and volume change determinations
during saturation and consolidation be made. Diameter measurements
should be made to the nearest 0.001 in. using a circumferential

tape7

t Commercially available from PI Tape, Box 398, Lemon Grove,

Calif. 9 2 0 4 5 .
XB-6

E M 1110-2-3906
Appendix X B
Change 1

.l May 80
at the top, midheight, and bottom of the specimen or dial-gage
calipers at the same locations, except two readings should be taken
at each position by rotating the dial-gage calipers 90 deg to produce
a total of six diameter readings.

Care must be taken not to deform

or disturb the relatively loose specimen during diameter measurements.

Height measurements to the nearest 0.001 in. at four locations

and weights to the nearest 0.1 g are recommended for 2.8-in.-diameter

specimens.

Record the data on the data sheet. (Plate

XB-1).

d. Preliminary Seepage Saturation. See paragraph 4d-of

Appendix XA, CYCLIC TRIAXIAL TESTS.
2. C o n s o l i d a t i o n .

See paragraph 45 of Appendix XA, CYCLIC

TRIAXIAL TESTS.
-f. Monotonic Loading. If it does not already exist, a large air
pocket should be formed at the top of the triaxial chamber by draining water-t from the chamber but leaving enough water to cover the
top of the specimen.

The air pocket is required, so that the large

rapid piston movements into the chamber at the onset of failure do

not create chamber pressure fluctuations.

The procedure is as

follows:
(1) Record test number and specimen identification on
recorder trace.
(2) Zero the recorder and transducer outputs, and record
calibration steps and scale factors.
(3) Close valve F (D should have already been closed) as
shown in Figure 16 of Appendix X, TRIAXIAL COMPRESSION TESTS.
(4) Initiate incremental axial loading using an increment

t Compressed air instead of water may be used as the confining

medium provided saturation and consolidation procedures do not
exceed 8 hr.
XB-7

EM 1410-2-4 906
Appendix XB
Change 1
1 May 80
determined as described in (5) below.

During loading, the chamber

pressure is :maintained constant, and the axial load, axial deformation,

and change in pore water pressure are recorded with time. Slow
recorder speeds can be used as long as the specimen is deforming
slowly; manual recording of load, pore water pressure, and deformation can be made as long as specimen behavior permits (this practice
can save timte in data reduction in that recorder traces are troublesome to read).

Loading is continued on approximately l-min intervals

until either (a) the specimen liquefies or partially liquefies, or

specimen tends to dilate.

(b) the

A complete stress-strain record, such as

shown for liquefaction or limited liquefaction in Figure

obtain and requires experience-

1, is difficult to

However, it is achieved by increasing

the recorder speed to 10 to 20 in. of record per second just as the specimen indicates impending failure.

The clues to approaching rapid fail-

ure are an obvious increase in deformation rate and steadily accelerating increases in pore water pressure. If the specimen deforms
steadily but not catastrophically, additional load increments may be
applied as the straining tends to cease until the pore pressure begins
to decline.
(5) The magnitude of the monotonic load increments is
arbitrary but should be selected to provide several points on the
stress-strain curve prior to peak deviator stress.

For a confining

p r e s s u r e , T3c, of 0.5 tsf, a reasonable increment would be equivalent

to about 4.0 :psi of deviator stress based on the initial area of the
specimen. If the confining pressure is 5.0 tsf, an increment equivalent to about 5.0 psi is more practical.

In any case, it is necessary to

reduce the load increment as peak deviator stress is approached in

order to prevent the application of the last increment from significantly
exceeding the true peak value.

Confident use of such a procedure must

be learned through experience with initial tests on a given soil.

&. Z&ecimen R e m o v a l .

Following testing, the specimen should

XB-8

E M 1110-2-i906
Appendix XB
-

Change 1
I May 80
or limited liquefaction,

is taken as the difference between the

3f
consolidation confining pressure, T
and the maximum pore water
3c
pressure generated during the test (see Figs. 1 and 3). Figure 4 illustrates construction of the critical void ratio, which is the approximate
relationship between e

and ; 3f

for the range of u3c

The most

desirable test series includes two to three tests at each of several

pertinent confining pressures,

u3c ; e.g., 0.5, 4.0, 2.0, 4.0, and 8.0

tsf. The objective is to build the specimens near the critical void ratio,
but slightly above it, to obtain complete and limited liquefaction in failure.

This can be accomplished by trial. A trial procedure is not

wasteful since all tests are of value in establishing the critical void
ratio relatifonship.

It is suggested that a pilot test be performed with

a specimen built to about 20 percent relative density and by applying a

moderate consolidation stress (about 2 tsf). The behavior of this specimen should indicate the approximate position of critical void ratio.
Smaller confining pressures would produce higher critical void ratios;
and larger confining pressures, lower critical void ratios.

Those

specimens that tend to dilate upon loading serve to help fix a lower
bound to the position of the e
versus the s-3f r e l a t i o n s h i p .
7. POSSIBLE ERRORS. In addition to those described in paragraph

9.

of Appendix X, TRIAXIAL COMPRESSION TESTS, the following are

possible errors that would cause inaccuracies in the determination
of critical void ratio.
a. A
- p p a r a t u s . (I) L o a d i n g s y s t e m . Insufficient air or hydraulic
fluid at failure conditions will cause unacceptable load reduction. Misalignment between the loading rod piston and load actuator or air piston
may also cause unacceptable loading conditions.
(2) Electronic transducers and recording equipment. Improper calibration or sensitivity of the electronic transducers, incorrect
balancing of amplifiers or zero settings, or improper range settings
will result in inaccurate recording of actual loads, deformations, and
XB-40

E M ~110-2-~I.906
Appendix XB
-

Change 1
1 May 80
C O N S O L I D A T I O N P R E S S U R E m3c, TSF
1
I

.o

ljllll

10.0

5.0

pIIll

ll I

\,ASXJMED c~fnc,4L vo1~3 R A T I O

-\

0
6
>

STRESS
4

0.77

3c
af
-6-7
0

RANGE IN INDICATED POSITION

OF CRITICAL VOID RATIO

I
LIQUEFACTION
LIMITED LIQUEFACTION

1 1

0.01

DILATIVE

0.75 I

FAILURE RESPONSE

0.05

0.1

IllIll
0.5

IllIll

1 .o

EFFECTIVE CONFINING PRESSURE AT FAILURE

5.0

10.0

E3f, T S F

Figure 4. Example of the determination of approximate

critical void ratio
pressures occurring during the test.

It is also essential that the re-

corder response be rapid enough to follow all changes in the

trans-

ducer output.
b_. Specimen Preparation and Testing.

(I) S p e c i m e n d i m e n s i o n s

not measured precisely or density improperly calculated. A circumferential tape for measuring specimen diameter is recommended for
obtaining precise measurements.

Twice the thickness of the membrane

must be subtracted for measurements of single membrane-encased

specimens.

An improperly calibrated burette will lead to incorrect

volume cha.nge measurements during consolidation with resulting errors

in specimen void ratio computations.
(2) Percent undercompaction in lower specimen layers improper for achieving desired average density.
XB-12

EM gdjO-2-1906
Appendix XB
Change 1
1 May 80
(Uniform density in a reconstituted specimen is essential).
(3) Incomplete saturation resulting in low B values.

The

following problems can cause low B values: (a) use of insufficiently

de-aired water may prevent dissolving of air in the specimen without
resorting to extremely high back pressures; (b) incomplete de-airing
or saturation of pore pressure transducer and drainage lines (can be
avoided by applying a vacuum); and (c) system leaks due to punctured
membrane, poor membrane sealing to cap and base, loose fittings, or
improperly designed O-ring grooves (can be detected by using a bubble
chamber while applying vacuum to the system).
(4) Assuming that the loading piston is not connected to the
specimen cap, the static uplift load equal to the area of the loading
piston multiplied by the cha.mber pressure must be accounted for when
applying axial stresses.
(5) Scale factor for recorder traces not in agreement with
actual data measurements.

Reduction of data from recorder traces

where scale factor, i.e., inches on recorder trace per pound (load),
per inch (deformation), or per pounds per square inch (pressure), will
result in data reduction error.

The calibration steps should always

be recorded on the recorder trace prior to and after testing.

XB-13