Frost heave tests with constant rate of heat extraction

(1)

aje Aa Alk w atetti tea air». 1x Aad Nds lite 3 ® (in.--) [e teysha o ).\ u 7 + 3 2 u 2 4 % : 7 t g ~ i ® " dos Ce k 5 4 % : ® # f f s * y 4 _ x A fw? = r a a Bd hic iTe 2 je [ 8 ax 6 s ;

s aA 5 P r Free R # i 3 iwor: ~f g MKE]f e # f

ma l'cn ) .: , a h It: L Is I $ S ¥ 4 " + a +F &

$ a is & :

Ar a s 4 R 2 2 f o P s 6

i h s

m 4 - s a g 3

e W " sand ® mind F + < Pa 4 f * < e j her

-a% a 3 <> e GSEs" . 5 i f es I f s RZa 3 94 g f si Te:\ # y ae : ra R . tx. * u 5 a E # a1 =z 4 R b P 2: %a A- o v ra ( i < C L "' Mox j * - o ia Fog" a b s g R - x $ E * fat E a w le a P3f ) j : "* 4 wL 3 r 3 M £ ¢ ¥ s D b a "C v A t - » t < 4 f r 2 Peis + a T r a f - Peet : ae Wie* 24 I - 4@ o % r m Poh U i 5, i; -. la i a e + a Lotts f a a 2 f f f s? 3 a ys x + th # r v i a f a & o f r i f a 4 © l % : > % i + x is ra 18 * % k ' : x 3 4_A b e _as ah 5 R _a 'C ₊ _S _A $ _a _f 7 _a - + " .v wa Ca Ua F +, $200s y as Chek k i T yy & § 7 a 5 f pd & } wowii u La all C :+ %A Ae#C h d a ( > a -Hd s s -A ix a i % % } >* f c £ f % & C apa ie 4 D : X a wed I y L x 1 9 ; e b - V # me Pogk I - - regen p » B i u &+ 2 ¥ o gs - i s s 4 S gy f r 3 ® - P &8 @ a 3 P A A P - « H ~ H & , " A y ] * C 4 & 5 4 3 g 3 L A f a P c . 5% f I ; j f < g s % c - s No 5 S e a * t k # A %, a /&. e fa P ¥ ; * s a P A 6- f i E v ( al -ta a ki a L Nok & F i- 3 - + A p f # + Pich" ¢ A * o , a a % + - * % fx 2 A " 3 ; ha g e R y C i x P f 3 ® 4 lbicten (flees.

(2)

(3)

Jr 220A - 1981

Statens vag- och trafikinstitut (Vl'l) - 581 01 linkiiping

LSN 0347-6030 National Road & Traffic Research Institute - S-581 01 linkiiping - Sweden

Frost heave tests with constant

rate of heat extraction

(4)

(5)

PREFACE

Part of this report has been presented in a paper con tributed to "The 2nd International Symposium on Ground Freezing" in Trondheim 1980.

The report is intended to give background information on and a verification of the complete set of tests performed.

The three subtitles refer to papers in swedish printed or in print.

Linkoping May 1981

(6)

(7)

CONTENT Page

ABSTRACT I

INTRODUCTION 1

DESCRIPTION OF THE APPARATUS 1

SCOPE OF INVESTIGATION 3

A. STUDY IN FROST HEAVE EFFECTS

FROM SAMPLE PREPARATION TECHNIQUE Choice of sample size

Sample preparation

Influence of saturation time on water

content

Freezing without water supply

Freezing with and without water supply 7

B. STUDY IN THE RELIABILITY AT THE APPARATUS 12 Freezing test on the soil in two

cylinders 12

Repeatability 12

C. FROST SUSCEPTIBILITY. COMPARISON BETWEEN

GRAIN SIZE DISTRIBUTION, CAPILLARITY AND

DIRECT FROST HEAVE TESTS 16

Frost susceptibility classification 16

Relations between frost heave rates

and capillarity 16

Study of frost heaving rates of sandy

tills 19

RESULTS 19

REFERENCES 22

(8)

(9)

FROST HEAVE TESTS WITH CONSTANT RATE OF HEAT EXTRAC-TION

by Lars Stenberg

National Swedish Road and Traffic Research Institute Pack

581 01 LINKUPING

ABSTRACT

An apparatus for determination of the frost suscepti bility of soils has been tested. Repeated freezing was performed with constant heat extraction. The soils

tested were frozen under various load pressures. Freeze thaw cycles on the same sample showed that the repeata-bility is good. Freeze-thaw cycles on one soil under similar conditions but in different cylinders showed that the reliability is good.

In the same equipment were several tills frozen. The results show that tills of nearly the same mechanical

composition show great differencesilltheir sensibility

to changes in load pressure. Comparisons with capil» larity and percent finer than 0.074 mm were made in an

(10)

(11)

INTRODUCTION

At the request of the National Road Administration an

apparatus for direct testing of the frost heave

be-havior of soils, has been constructed at the National Swedish Road and Traffic Research Institut (VTI). The following preurequisites were to be emphasized

A) Time of freezing reduced to 1 2 days. Standard tests

usually take 2-12 days.

B) The apparatus constructed must be easy to handle C) The sample preparation must be simple and not to

time consuming

D) Preferably one easily read parameter to settle the

frost heaving ability of the soil

DESCRIPTION OF THE APPARATUS

The apparatus, in its standard version, operates with constant rate of heat extraction and with the option of constant freezing temperature. Soil samples are pre~ pared in a cylinder of acryl plastics, figure 1. Free

zing is performed from below by Peltierelements. The

advantage of the upside down freezing is that the un frozen part of the soil may slide against the cylinder

wall and thus eliminates the counteraction of the

adfreezing forces on the frost heaving forces.

To make the operation possible at room temperature the cylinder was given a thick styrofoam insulation.

The tests performed aspires to attain the local field conditions refering to load conditions. This is at

(12)

IE:=£..

_{POROUS DISC INSIDE THE}

PERFORATED STAMP OF BAKELIZED FABRIC

<::::::_______ :::::>

FILTER PAPER

CYLINDER (ACRYL PLASTICS)

INTERNAL DIAMETER Qbi==110 MM

THICKNESS

MM

HEIGHT

200 MM

FILTER PAPER

POROUS DISC (BRONZE)

O *RING

-*";1 HEAT FLOW INDICATOR

.~....IIIIIIIIIIIIII "/ 2

(13)

A more detailed description was given in a paper presen

ted at the 2nd Ground Freezing Symposium (Fredén, S and

Stenberg, L 1980). SCOPE OF INVESTIGATION

The results reported mainly deal with C and D. Following questions were to be answered.

1) Influence of

a) soil compaction on repeated freeze thaw cycles

b) preloading not preloading of soil samples 2) Influence of

a) water content before freezing on frost heave b) saturation time

3) Relations between frost heave rate and known soil prOperty determining parameters as

a) percent fines

b) uniformity c) capillarity

4) The reliability of the tests performed

A. STUDY IN FROST HEAVE EFFECTS FROM SAMPLE PREPARATION TECHNIQUE

Choice of sample size

When performing laboratory freezing tests on soils in order to draw conclusions about their natural freezing

behavior, it is important that the composition of the

laboratory samples does not deviate to much from the natural soil. By experience it is well settled that the

(14)

fine grained part of the soil determines the frost

heave. In the tests performed particles ¢ > 20 mm were removed. It is also argued that the smaller the volume of soil tested the less representative is the test re sult, in that a larger part of the coarse particles has to be removed. The choice of ¢; = 110 mm is i.a. to

make it possible to make compaction by standard Proctor,

if needed, although the method is found to be no good

for freezing tests.

Sample preparation

The test cylinder, figure 1, with the porous disc rests

in a bowl of water. Dry soil material is placed spoon ful into the cylinder and slightly compacted by a hand~ stamp during the saturation by capillary suction. The sample height was 100 mm. Advantages with this prepara tion technique instead of using standard Proctor

compac-tion:

a) It is well known that the compaction degree is

lowered after a number of freeze thaw cycles.

b) The volume trapped air will be less than when using

Proctor compaction on moist soils.

c) No risks for unwanted ice-lenses due to soil struc

tures originating from sample preparation technique used. Fine grained tills tested has shown a tendency to ice-lens build-up at the layer interfaces in a Proctor compacted sample.

d) The deairing procedure can be eliminated.

The design also allows of drainage of the over

satura-ted soil sample after thawing and makes repeasatura-ted freeze

thaw cycles of the same sample possible. The cylinder with the still frozen soil is lifted off the cooler

(15)

and is allowed to thaw and drain under the same load conditions as during the freezing test. In the meantime, another test is performed. In all tests the heat extrac~ tion rate was % 4.7 W. With a cylinder cross section area of 95 cm2 this gives 490 W/m2.

Influence of saturation time on water content

One silt and three tills were used. Dry soil material

was portioned into the cylinder and compacted slightly

during the capillary rise of water into the soil. Satu~

ration was considered complete when water separated on

the surface under a slight extra load, for instance, by pressing a finger against the soil surface. The cylin~

der was left in the water for 1 10 days. After this period of time the water content was determined. Deter

minations of the water contentwas also made on samples saturated under a surcharge of 50 g/cm2. The results are illustrated in figure 2.

Conclusions:

a) The water content varies a lot both for loaded and unloaded samples.

b) for the soils used, the saturation time is of secondary importance.

c) Variations in water content are less for samples saturated under load pressure.

It is argued that the variation observed in the water content of the samples prepared are within those from

(16)

73 A

20< o o O a D D o o 15 i i - O o

I

l I t) A a I .A 10 i x x .A A .A .A .A 5 J

'0 FINE SILT SALEN

. II H II g/cmz)

o TILL SALEN

I "

"

(~50 g/cm2>

.A MORA TILL E:l

x H fl 6

'

' f

5 10 15 DAYS

Figure 2. Water content, W % (by weight), as function of time of saturation. Load pressure % 50 g/cm was used on the Salen soils, the other samples were unloaded

(17)

.Freezing without water supply

To find out the extent to which the variationsi11water

content affect the frost heave and frost heave rates,

freezing without water supply was performed on the silt

(Salen) and a till (VTI 94685). The results are presen-ted in table 1 and figure 3.

Concerning the silt we find that the heave amounts, AH (mm) agree well with the water content. From this we conclude that, with the heat extraction rate and

with the load pressures used in the tests, no expul

sion of water, which affects the total frost heave, takes place. The heave of the till, is somewhat less than expected. However, it is found that the variation

in frost heave amount is less for the preloaded sampu

les and that a load pressure of m 50 g/cm2 (0.5 tons/m2) will be enough for the preloaded samples. The satura

tion time is of subordinate importance. Freezing with and without water supply

In a closed saturated system frost heave is

prOpor-tional to the water content 1 e 10 % of the volume of

water of the sample. This is not the case for

unsatu-rated soils, in which the pore structure allows the free water to be pushed into the air filled voids by the expansion water + ice. Tests were performed to

compare the heave amounts with the water contents and

if the load pressures used affected the heave amounts. The influence of heat flow was not looked into. We

know however that in an open system very slow freezing may cause accumulation of ice in sand, this has been observed in the fraction O,125~O,25 mm. On the other hand higher freezing rates expells water. Although frost heave in many cases might have been observed

(18)

Table 1. Frost heave tests on saturated soil samples without water supply

Fine Silt salen

P

Atf

AtH

AH

AH/AtH

2 Saturation

(g/cm )

(kPa)

(s)

(5)

(mm)

(mm/s)

21.5 31700

23400

2.8 1.20 x 10"4 P1 10 kg (2 days)

47.9 22700 15500

2.2 1.42 x 10'4 not P1 (37 days)

47.9 32400

21600

2.1 0.97 x 10'4

P1 5 kg (2 days)

47.9 25200

16200

3.7 2.28 x 10 4 P1 5 kg (2 days)

47.9 27700

16200

3.5 2.16 x 10"4 P1 5 kg (4 days)

47.9 25200 17100

3.9 2.28 x 10'4 P1 5 kg (4 days)

47.9 27000 14400

3.5 2.43 x 10 4 P1 5 kg (6 days)

74.2 25900

19100

3.5 1.80 x 10"4 P1 7.5 kg (3 days)

100.5 27000 17100

3.0 1.75 x 10'4 not P1 (6 days)

100.5 26600

19100

2.6 1.36 x 10 4 not P1 (6 days)

100.5 24100

19800

2.9 1.46 x 10"4 not P1 (14 days)

100.5 23400

15300

3.2 2.09 x 10'4 P1 10 kg (14 days)

Till VTI 94685

21.5

2.1 25200

19800

0.74 x 10'4 not P1 (1 day)

21.5

2.1 27000

18900

0.69 x 10'4 not P1 (1 day)

21.5

2.1 21600 16200

0.93 x 10'4 not P1 (6 days)

47.9

4.7 22700

17600

2. 1.19 x 10 4

not P1 (15 days)

47.9

4.7 23000

14400

.

0.76 x 10'4

P1 5 kg (1 day)

47.9

4.7 20900

14400

.3 0.90 x 10 4

P1 5 kg (2 days)

47.9

4.7 19800

12600

.

0.79 x 10 4

P1 5 kg (7 days)

74.2

7.3 23400

15100

0.72 x 10 4

not P1 (6 days)

74.2

7.3 23000

18700

.

1.50 x 10 4 not P1 (24 days)

74.2

7.3 25200 16200

.

0.62 x 10 4

P1 10 kg (3 days)

74.2

7.3 22700 16200

.

0.86 x 10 4 P1 7.5 kg (6 days)

127.0

12.4 20500

14400

.

0.76 x 10"4

P1 12.5 kg (5 days)

Pl=Pre loaded

(19)

%¥mm/s

0 10 ?

9 ..

3 m

o

SILT SALEN

7 . PL

6 x

TILL VTI 94685

5 q 9 PL [4 '1 3 -<

2 a

i

'

O O 0

<3

X

:3

0 x

10 .

9 l

x

G

a

£73 _ ))(( 8 x Q 6 .. 0 5 J L} . 3 q 2 d 10.5 : # + % > 2 20 50 100 150 Q/cm

6

15 .

1% |<Po

Figure 3. Frost heave rate vs load pressure for a silt and a till. Freezing performed on capillary

(20)

10

clearly inidicates water expulsion.

In a finegrained soil frozen without supply of water,

slow freezing will cause water migration towards the

freezing front and thus a drying out of the still unfrozen parts. This drying out effect will decrease

with increasing freezing rates. At still higher free~

zing rates water will be expelled and at very high freezing rates water expulsion cannot keep pace with

the fase transitional volume expansion and frost heave. will occur. The more fine grained the soil, the lower can the freezing rate be to cause frost heave. In a

clayey soil the resistance to water migration is so high that no water is expelled even at low freezing

rates.

In an open system with access to water the same arguing applies with the exception for the drying out effect in the unfrozen soil is eliminated by suction of free water to the dried out zones which results in accumu

lation of water and frost heave.

The difference in frost heave behavior between freezing

with and without access to water thus also gives an

indication of the frost heaving and draining ability of the soil which is of great importance when con

sidering Spring thaw weakening on roads.

E§E§EEE§EE§l_QE§§£Y§EEQE

Some of the till samples were frozen without water

supply butsaturated in the same manner as those fro zen with free access to water. For the till VTI 94684 the heat extraction rate was lower for the lower load

than for those with higher load, which is reflected in a lower heaving rate in figure 4. However the observed

(21)

10 Figure 4. 11 VT! 94677 Nonnot VTI 94679 -I/ VT! 94682 VTI 94684 X ® 0 o A A D I ~

150 g/Cm2

3 l 50 6 1O 15 14%

Comparison between normal freezing = free

water supply and freezing of capillary satu-rated sample with no water supply, filled symbols

(22)

12

B. STUDY IN THE RELIABILITY OF THE APPARATUS

Freezing test on one soil in two cylinders.

Two cylinders, BI and B11, were prepared with salen silt to find out if the manual preparation technique used affects the frost heaving properties of the soil.

Repeated freeze thaw cycles were performed. The

re-sults are compiled in table 5. Heaving rates vs load

pressure are shown in diagram figure 5. Note that the heat extraction rate is less (m280 W/m2) in these tests than in the tests performed on the till samples (N490

2 W/m ).

The result showed that for this test the freezing rate V 8 load pressure was not different for the two samples. However the pre-loaded samples showed somewhat higher heaving rates.

Repeatability

The repeatability was studied on three tills, Mora E:1,

E:2 and G, by several freeze-thaw cycles on the same sample. Mechanical composition see Appendix 2. The sam-ples were saturated in accordance with the procedure described. Both preloaded, i.e. load pressure applied

during saturation, and not preloaded sample testings

were performed. Results are compiled in table 2 4, Appendix 1.

The observed differences in the heaving rates cannot

be ascribed to whether the sample was preloaded or not.

However, the reliability is higher for the preloaded one's the variations in heaving rates are less for the

preloaded samples. The heaving rates as a function of

(23)

II

1.3

Table 5. Frost heave tests on silt (Salen) in cylin

ders B I and B II. Heat extraction m 280 W/m2.

Day 9 2 P HO tf AH/A th AHtot Anm

g/cm kPa mm S mm/s mm 800211 74.2 7.3 109 103.103 3310'4 62.6 not Ere-leaded 800215 74.2 7.3 109 134-103 3910'4 51.4 " 800220 74.2 7.3 100 87-103 3.1 10"!+ 50.6 " 800225 74.2 7.3 100 166-103 3410 4 57.6 " 800303 74.2 7.3 100 132-103 3.7-10'4 52.0 " - 74.2 7.3 - - - capillary saturated. No water supply 800519 180.0 17.6 76 59.8103 2.5-10'4 14.2 not_prejloaded 800522 180 0 17.6 74 587-103 2210 4 12.2 "

800528

284.6 27.9 70

42.1-103

1 5-10'4

8.0 "

800314 74.2 7.3 93 161-103 4.4-10 4 70.0 p-1 5kg(1dy) 800324 74.2 7.3 86 965103 4410"4 81.2 " 5kg(3dy) 800328 74.2 7.3 86 177-103 3910'4 71.0 " 5kg(1dy) 800408 74.2 7.3 84 42.5-103 4.0-10"4 34.0 " 5kg(5dy) 74.2 7.3 - - " 800319 74.2 7.3 86 36-103 1 2-10 4 4.0 " (1dy)capillary saturated. No water supply 800213 180.0 17.6 116 3310 4 28.0 not prejloaded eller _4 2.5-10

800218

180 0 17.6 116

94-103

3 1-10'4

45.8 "

800222 180.0 17.6 116 182-103 3 1-10'Z+ 55.8 " 800227 180.0 17.6 100 165-103 1.3-10 4 51.4 " 800305 180 0 17.6 100 83.2-103 2910"4 39.0 " 800311 180.0 17.6 99 33'103 1.3'10-4 3.6 capillary saturated. No water supply 800520 74.2 7.3 89 140.4°103 4.4-10 4 60.0 not pre-loaded 800527 74.2 7.3 85 756103 -10_4 38.2 " 800529 232.0 22.7 83 59.4-103 1 910 4 11.2 " 800313 180.0 17.6 99 77.8103 3.3-10"4 25.2 p.1. 7kg(1dy) 800317 180.0 17.6 99 85-103 3.1'10 4 32.4 7kg1dy

800321

180.0 17.6 98

147-103

3.5-10'4

67.0 "

7kg1dy

800326 180.0 17.6 89 90-103 3810"4 33.8 " 12kg1dy ' 800331 180 0 17.6 89 142103 3.4~10'4 47.2 " 12kg3dy

(24)

14

F=mm/s

-3

10-

preloaded

.

H max

6 min

4 *

%

not preloaded

max

average

mm

2 q 0 O

12»

19

29 30 kPa

l l r 1 I -2

0

50

100

200 300 g/cm

Figure 5. Frost heave rate vs load pressure for two samples of silt (Salem). Heat extraction

(25)

15

mm/5

A

A Moro TiLL E:1

10_3_ _pre-looded

9: A Moro TiLl E21

,87_ not pre-Loaded 6 _ 5 - A

4-

8

3a A e

24

161' . g 9 7 50 100 150 g/cm2 r 1E) 15 kPo mmls

?

_3 5] More TiLL E:2

10- pre-Looded

C] More TiLL E:2

74 not pre tooded

6* o 5< D D 1w a El 31 E] D 2. C] E] 4. 10 ₅₀-% ₁₀₀5 it₁₅₀ _g/cm2 y

'

_S

#

₁₀

:

₁₅

:

_ch1 mm/s l OMora Titl G 10 3d pre-looded : : OMoro Till G

not pre looded 7 + 5 4 O 5 G) L -< O 3 4 O O G) 2 J 8 '1 A l L a m 50 160 1'5p g/cm2 ' é b E k%

Frost heave rate vs load pressure for

Mora tills E:1, E:2 and G. Pl means that the sample was surcharged during satu ration

(26)

16

C. FROST SUSCEPTIBILITY. COMPARISON BETWEEN GRAIN SIZE DISTRIBUTION, CAPILLARITY AND DIRECT FROST HEAVE TESTS.

Frost susceptibility classification

The methods most commonly used for classifying the

frost susceptibility of soils are basedcn1particle size distribution. In Sweden, the percentage of material finer than 0.074 mm in a cumulative curve representing particles ¢ < 16 mm is the determining factor for tills. In some cases, capillarity tests are performed on sedi mentary soils. The soils are classified as follows:

% 0.074 Cap (Sed) not frost susceptible I < 16 < 1 m

frost susceptible II 16 - 43 1.0-1.5 m highly frost susceptible III > 43 > 1.5 m

From table 6 and figure 8 it is seen how poor the agree

ment between % < 0.074 mm and frost heave rate is.

Relations between Frost Heave Rates and Capillarity At the end of the series of freeze-thaw cycles,

capil-larity tests were performed on the upper and lower parts of the till samples, Mora E:1 and E32. Capillari

ty was also determined on six other tills, see table 6.

To study whether any vertical sorting had taken place during the repeated freeze thaw cycles, a sedimentation analysis was made on material < 2 mm from three levels of the Mora E:1 and E:2 samples. (See table 6). From table 6 it appears that the difference, in % < 0.074 mm and % < 0.060 mm, is to small to indicate any vertical movement due to freeze-thaw cycles.

(27)

17

Table 6. Capillarity and soil parameters from grain

size distribution curves for soils used in

direct freezing tests.

. Cap % < % < d d Cu

5°11 cmVp 0,074 0,060 60 10 d60/d10

Mora Till E:l upper mtrl <2 mm 42.2 37.0 0.17 0.0042 40.5

" " " middle " " 45.0 40.0 0.15 0.0037 40.5 " " " lower " " 43.4 38.8 0.16 0.0036 44.4 n u E:2 upper " n 45.0 39.8 0.137 0.0037 37.0 " " " middle " " 46.0 41.4 0.133 0.0036 36.9 " " " lower " " 47.0 41.5 0.128 0.0035 36.6 " " G " " 43.2 37.8 0.16 0.005 32.0 Till VTI 94677 " " 48.6 44.5 0.142 0.0051 27.8 " " 94679 " " 43.4 38.3 0.17 0.0096 17.7 " " 94682 " " 39.0 34.0 0.22 0.0104 21.2 " " 94683 " " 46.2 39.0 0.138 0.010 13.8 " " 94684 " " 32.6 29.0 0.22 0.007 31.4 " " 94685 " " 43.0 33.5 0.13 0 008 16.3

Mora Till E:l upper tot mtrl >750 24.8 21.3 2.3 0.011 209-1

" " " middle " " 33.3 29.9 0.37 0.0074 50.0 " " " lower " " 731 28.8 25.0 0.75 0.009 83.3 " " E:2 upper " " 494 35.0 31.3 0.28 0.0066 42.4 " " " middle " " 36.8 33.0 0.25 0.0062 40.3 " " " lower " " 601 33-7 29.9 0.37 0.0068 54.4 " " G " " 524 29.6 25,3 0_67 0,011 60.9 Till VTI 94677 " " 366 43.3 39.8 0.225 0.0066 34.1 n u 94679 n u . 379 38.3 34.0 0.26 0.010 26.0 " " 94682 " " 330 32.0 27.2 0.41 0.0138 29.7 n " 94683 " " 283 40.4 35.2 0.21 0.0127 16.5 " " 94684 " " 225 27.7 25.2 0.32 0.0112 28.6 " " 94685 " " 301 40.8 32.0 0.14 0.0094 14.9

(28)

mm/s-10'l' A . 2.1 kPCl O X 7.3 "'" 6 5 12.5 *H 5 C O LIJ

(I zu

-

x

LlJ > <[ IJJ I 3. {7; o /A

E?

o

A

LL.

2 X

X

C A X 1 q A T . . . :0 100 200 300 400 500 600 700 HC CAPILLARITV

Figure 7. Frost heave rate vs capillarity of tills at

different load pressures and constant rate

(29)

19

capillarity at three different load pressures is shown

in figure 7. The expected linearity decreases with load

pressure. Thus, capillarity is still a good means in a relative classification, but it cannot yet be used in predicting frost heave behaviour.

Study of Frost Heaving Rates of Sandy Tills

Sandy tills, mechanical composition is shown in appen

dix 2, were frozen under constant heat flow but various

load pressures. The object was to compare the frost heaving rates with the general classification of frost susceptibility based on particle size composition. The

results shown in figure 8 illustrate how the heaving

rates and the sensibility to load pressure vary for soils with small differences in particle size

composi-tion.

Comparing the results in figure 8 with table 6, we will find that the tills, VTI 94677-94685, should all be classified as "frost susceptible".

However, the observed low heaving rates, by the direct

freezing test, indicates that both VTI 94682 and 94683 can be made "not frost susceptible" even when

under a small load pressure. The reasoning is based on

the fairly well established assumption that the dis tance to ground water level and the overlying soil pressure are additives.

RESULTS

Direct freezing tests by the constant heat flow appara-tus have the following advantages:

1) The method is fast. None of the freezing tests has

(30)

MM/S

4 0 WI 94677

10'3-

O " 94679

g _

8 -

A

"

94682

7 4

6 _

X

94683

5 _

I:

"

94684

a q

+

4 " 94685

3 ..

2 _

O

10'

-9 .-8 -l 7 _ 6 _ 5 .4 4 _ 3 .. 2 _

,10 5

.

e

20

50

100 150 g/cm2

2.0 3.3 4.7

7.3

9.8 12.5 13.6 kPo

Figure 3. Frost heave rate vs load pressure for tills of about similar composition in regard of 6

< 0.074 mm .

(31)

2)

3)

4)

21

By simulating the pressure conditions at the frost line in a direct freezing test, soils classified as

frost susceptible from grain size distribution can be found to be not frost susceptible.

The heat flow control makes it possible to simulate conditions in a partially insulated road.

Direct freezing of waste products. The mechanical composition of many waste products often make capil-larity tests and particle size analyses irrelevant.

To that there is to be added that their properties as road material mostly are of a chemical nature.

(32)

22

REFERENCES

Fredén, S & Stenberg, L 1980, Frost Heave Tests on Tills with an Apparatus for Constant Heat Flow. VTI

(33)

Appendix 1

Page 1 (3)

Table 2- Frost Heave Tests on More Till E:1

P 2

P

if

Atf

AtH

AH

AH/AtH

saturation

(g/cm ) (kPa) ( C) (S) (S) (m) (mm/S)

21.5

2.1

0.1 32400 29500 16.7 5.66 x 10'4 not P1 (1 day)

21.5 2.1 -0.1

42500 39600 27.4 6.92 x 10'4 not P1 (6 days)

74.2 7.3 -0.1

26600 23000

8.7 3.78 x 10"4 not P1 (2 days)

74.2 7.3 -0.2

21600 18000

5.6 3.11 x 10"4 not P1 (7 days)

74.2 7.3 -0.0

28100 24100 10.2 4.23 x 10'4 not Pl (8 days)

74.2

7.3

0.1 24100 19400

7.6 3.91 x 10 4 P1 7.5 kg (4 days)

74.2

7.3

0.1 27000 25200 12.4 4.92 x 10'4 P1 7.5 kg (4 days)

127

12.4

0.1 21200 18000

5.4 3.00 x 10'4 not P1 (5 days)

127 12.4 -0.1

22000 19400

5.8 2.98 x 10 4 not P1 (5 days)

127 12.4 -0.1

23000 18400

5.1 2.78 x 10'4 P1 12.5 kg (5 days)

180 17.6 -0.1 14400 10400 1.6 1.53 x 10"4 not P1 (5 days)

180

17.6

0.1 19100 15100

3.2 2.12 x 10"4 not Pl (7 days)

180

17.6

0.5 19400 14400

3.3 2.29 x 10'4 not Pl (8 days)

180 17.6 -0.0

21200 15800

3.2 2.02 x 10 4 P1 20 kg (1 day)

180 17.6 -0.1

20900 15800

2.6 1.64 x 10 4 P1 20 kg (3 days)

(34)

Appendix 1

Page 2

Table 3. Frost Heave Tests on Mora Till E:2

P 2 P if Atf AtH AH AH/At Saturation

(g/Cm ) (kPa) ( C) (s) (8) (mm) (mm/s)

21.5 2.1 -0.3 33800 30600 15.7 5.13 x 10 not P1 (1 day)

21.5 2.1 Extensive Ice lensing 6.04 x 10 not Pl (3 days)

74.2 7.3 -0.2 24800 18400 8.1 4.41 x 10' not P1 (7 days) 74.2 7.3 -0.1 26300 20700 8.2 3.96 x 10' P1 5 kg (1 day) 74.2 7.3 -0.2 23000 20200 7.0 3.47 x 10 P1 7.5 kg (8 days) 127 12.4 18400 4.8 2.52 x 10' not P1 (3 days) 127 12.4 -0.1 23400 20500 5.2 2.53 x 10 not P1 (7 days) 127 12.4 -0.1 20500 14800 3.8 2.58 x 10' : P1 12.5 kg (2 days) 180 17.6 -0.3 18000 11900 2.7 2.27 x 10- not P1 (4 days) 180 17.6 0.0 21600 16200 2.9 1.79 x 10' not P1 (10 days) 180 17.6 -0.2 17600 13000 205 1.66 x 10- P1 17.5 kg (4 days) Pl=Pre-loaded

(35)

Appendix 1

Page 3

Table 4. Frost Heave Tests on Mora Till G (Vines)

P 2 P 3f Atf AtH AH AH/AtH saturation

(g/cm ) (kPa) ( C) (s) (s) (mm) (mm/s)

21.5 2.1 0.0 Ice-lenses ~5.14 x 10:: not P1 (1 day) X 47.9 4.7 -o.2 32400 29700 16.5 5.56 x 10 Pl 5 kg (5 days) 74.2 7.3 -0.1 26300 23000 10.2 4.43 x 10'4 P1 7.5 kg (6 days) 100.5 9.8 -0.3 25600 22100 8.0 3.61 x 10 4 P1 10 kg (2 days) 127 12.4 -0.1 27400 22300 6.1 2.73 x 10"4 not P1 (1 day) 127 12.4 - 20200 22700 5.5 2.43 x 10'4 P1 12.5 kg (7 days) 180 17.6 -0.1 33500 17300 3.7 2.14 x 10'4 not P1 (9 days) 180 17.6 -0.3 22700 18400 4.2 2.29 x 10'4 P1 20 kg (1 day) 180 17.6 -0.5 21200 16600 4.2 2.54 x 10'4 P1 20 kg (5 days) Pl=Pre-loaded

(36)

Appendix 2 page 1 (7)

CLAY SILT SAND GRAVEL

0.002 0.000 0.02 0.00 0.2 0.0 2 6 20 60 PERC EN T FI NE R 0.001 0.000 0.000 0.01 0.02 0.074 0.12.5 0.25 0.5 1.0 4 5.0 0 11.3 16 32 50 64 PARTICLE SIZE NM

MORA MORAN Ell MIDDLE <2 MM SEDIMENTATION ANALYSIS (HYDROMETER METHOD)

0.000 0.000 0.02 0.00 0.1 0.0 2 6 10 00 PE RC EN T F IN ER 0.001 0.001 0.005 0.01 0.02 0.014 0.125 0.15 0.5 1.0 2 4 5.0 11.1 10 u n 50 64 PARTICLE SIZE NM

MORA MORAN E:2 MIDDLE <2 MM SEDIMENTATION ANALYSIS (HYDROMETER METHOD)

0.002 0.000 0.02 0.00 0.2 0.0 2 6 20 60 PE RC EN T FI NE R 0.001 0.000 0.000 0.01 0.00 0.014 0.125 0.25 0.5 1.0 4 5.0 0 11.: 1020 n 5064 PARTICLE SIZE IVM

MORA MORAN G <2 MM SEDIMENTATION ANALYSIS (HYDROMETER METHOD)

Figure . Grain size distribution curves of soil used

(37)

Appendix 2 Page 2

CLAY I SILT SAND GRAVEL

2.0 0 N O 0. ! 0. 0.02 0.00 0.2 0.6 PE RC EN T FI NE R 0.001 0.000 0.005 0.01 0.02 0.074 0.125 0.25 0.5 1.0 2 4 5.6 8 11.3 16 20 32 50 64 PARTICLE SIZE NM VTI 94677 < 2 MM, SEDIMENTATIONSANALYSIS (HYDROMETER METHOD)

0.002 0.006 0.02 0.06 0.1 0.6 2 6 20 60 _. Z

52

E 0.001 0.000 0.005 0.01 0.02 0.014 0.125 0.25 0.5 1.0 2 4 ' 5.0 O 11.3 16 20 32 50 64 PARTICLE SIZE IVM

VTI 94677 TOTAL SAMPLE

(38)

PE RC EN T FI NER 0. ! 0.002 PARTICLE SIZE PM 9. VTI 94679 Appendix 2 Page 3 SAND SILT 0. ! 0.01 0.02 0.074 0,125 0.25 < 2 MM, SEDIHENTATIONSANALYSIS (HYDROMETER METHOD)

CLAY SILT SAND

0. 0.01 0.32 0.014 0.125 0.15 '2

:5

mm 0. ! PARTICLE SIZE MM VT] 94679 TOTAL SAMPLE

Figure . Grain size distribution in frost heave tests.

GRAVEL

2 4 5.6 B 11.3 16 20 32 $064

GRAVEL

50 64 2 4 5.6 I 11.3 1620

(39)

Appendix 2

Page 4

0.002 0.006 0.01 0.05 0.2 0.6 2 6 20 60 PE RC EN T FI NE R 0, ! 0.002 0.005 0.0! 0. 0,014 0.125 0.25 0.5 1.0 2 4 5.6 O 11.! 16 20 32 50 64 PARTICLE SIZE NM VTI 94682 < 2 MM, SEDIMENTATIONSANALYSIS (HYDROHETER METHOD)

0.152 0. 0.01 0. 0.2 0.6 2 6 20 0 PE RC EN T F IN ER I J 0.001 0.002 0.005 0.01 0.02 0.074 0.125 0.25 0.5 1.0 2 4 5.0 0 11,3 16 20 32 so 64 PARTICLE SIZE MVI

VTI 94682 TOTAL SAMPLE

(40)

Appendix 2

Page 5

PE RC EN T VF IN ER 0M 0. ! 0. ! 0.0! 0.02 M4 0.125 0.15 0.5 1.0 2 4 5.6 | 11.) i6 3 32 50 64 PARTICLE SIZE NM VTI 94683 < 2 MM, SEDIMENTATIDNSANALYSIS (HYDROMETER METHOD)

CLAY I SILT SAND GRAVEL

man on an mu m: o¢~ i a 20 so PE RC EN T FI NE R QM 0,002 0.005 0.01 0.02 0.074 0.125 0.15 0.5 1.0 2 4 5.6 8 11.3 16 20 32 50 64 PA'RTICLE SIZE MM VTI 9A683 TOTAL SAMPLE

(41)

h vh -d f m W a r s ; s n ub -1 m m CLAY PE RC EN T F IN ER Appendix 2 Page 6

SILT _'L SAND GRAVEL

0.1111 0.002 0.005 0.01 0.02 0.014 0.125 0.25 0.5 1.0 2 4 5.6 8 11.3 16 20 32 50 64 PARTICLE SIZE MM

TILL VTI 94684 LASTAD ®< 2 MM

A CLAY SILT SAND l GRAVEL

LER MJKLA MO SAND GRUS STEN

T Flnmjdla g Grovmlllc Flnmo ' Gmmo {Mollcnnnd Gromnd Flngnn A Grovgrus V

om 0.6 0.62 o. 0.2 0.c i 6 20 so

0.002 0.005 0.01 0.02 0.074 0.125 0.25 0.5 4 5.6 8 11.3 16 20 32 50 64 PARTICLE SIZE MM

TILL VTI 94684 LASTAD

E Ex I? i: " 1-3E

E E

in. i i i 3 0M" Q<20 MM

Figure Grain size distribution curves of soil used in frost heave tests.

(42)

Appendix 2 Page 7

20 60 0.002 0,0 0.01 was v.1 v.0 1 6 P E R C E N T F IN ER J 0.001 0.002 0.005 0.01 0.02 0.014 0.125 0.25 0.5 1.0 2 4 5,6 8 11.3 16 20 32 50 64 PARTICLE SIZE NM VTI 94685 < 2 MM, SEDIMENTATIONSANALYSIS (HYDROMETER METHOD)

CLAY SILT SAND GRAVEL _T

0.002 0.006 0.01 0.06 0.2 0.6 2 6 20 60 PE RC EN T F IN ER 0.014 0.125 0.2! (L5 1.0 2 4 5.6 8 11.) 1610 32 5064 mm 0.002 0. 0.01 0.02 PARTICLE SIZE WT VTI 94685 TOTAL SAMPLE

(43)

(44)