Doctoral thesis

Doctoral Thesis:

Ice Lensing Mechanism during Soil Freezing

Graduate School of Bioresources, Mie University
Kunio Watanabe
December, 1999

PDF version of [chap. 1 (148KB)] [chap. 2 (116KB)] [chap. 3 (1800KB)] [chap. 4 (276KB)] [chap. 5 (48KB)] [chap. 6 (112KB)] [chap. 7 (172KB)] [chap. 8 (8KB)] [References (20KB)] [Abstract and contents (20KB)]

ABSTRACT

When soil is cooled under 0 ‹C, soil water solidifies with decreasing temperature. This is known as soil freezing. In Japan, the ground is frozen seasonally in Hokkaido, Tohoku, and some mountain areas. Short-term frozen ground is even found in the western part of Japan. Soil freezing is estimated to occur in almost 70 per cent of the Earthfs land surface. Furthermore, soil is sometimes frozen artificially for use. There are many situations in which soil freezes.

If soil expands when it freezes, the volume change is called frost heave. When heaving occurs, some of the ice layer formed near the ground surface is present as seems of pure segregated ice; this ice layer is called an ice lens. The ground surface sometimes expands dozens of centimeters due to the ice lensing. It causes much damage to paved road, drainage system in farmland, foundation, and so on. The growth of ice lenses induces soil water to flow from the unfrozen zone to the surface zone. In this case, some solutes illuviate to near the ground surface with the water flow. To overcome such frost-action damage, it is important to clarify the mechanism of ice lensing during soil freezing. Furthermore, ice lensing is not a phenomenon limited to soil. It has also been observed in various porous materials. Knowledge of the mechanism of ice lensing will be applied to fields that involve these frozen materials, including physical chemistry, biology, material science, food processing, and medicine.

Numerous studies with the intention of clarifying the mechanism of ice lensing have been reported. Presently, one of the theories most often used is the secondary frost heave theory. In this theory, presence of partially frozen region near growing surface of ice lens is assumed. The intermittent formation of ice lenses is then explained from calculating stresses in the region. However, the partially frozen region has not been confirmed experimentally. And the stress partition factor, which is important for calculating the neutral stress, has not been theoretically verified. Summarizing the historical studies, there are two problems associated with clarifying mechanism of ice lensing. One is to clarify the microstructure near the freezing front, i.e. to clarify water conditions and particle migration in the partially frozen region. The other is to explain the dynamic mechanism of ice lensing, in which the generation and growth of an ice lens is repeated to form intermittent layers. In order to solve these problems, we performed three series of freezing experiment using a nidirectional freezing apparatus. In the first experiment, ice lensing in soil and porous media consisting of fine particles were microscopically observed, then, following results were obtained. Ideal ice lenses for modeling can be made using uniform sized glass beads. The growth of ice lens is dependent on supercooling of the growth surface. The freezing rate influences the ice lens growth more than temperature gradient. It is suggested that freezing rate, supercooling degree at growth surface of ice lens and particle condition near growing ice lens were important factors for considering ice lensing model. From the second experiment, in which ice interface in water with dispersed glass particles was observed, the criteria for exclusion and encapsulation of particles during ice formation with respect to particle size and freezing rate was shown. The relationship between particle size and critical freezing rate was explained by KšberÕs theory. It is suggested that the critical freezing rate was important for the generation of ice lens. In the third experiment, microstructure in the vicinity of ice lens is observed using Raman spectroscopy. It is obtained that no ice was found in any pore warmer than the warmest ice lens in the porous media and the ice lens grew without penetrating the warmer pores.

Based on the experimental results, we then presented a model for simulating the formation of ice lenses during freezing of unconfined uniform porous media is presented. The main notions of the model are that generation and jump are dependent on the freezing rate, and growth is dependent on supercooling. The critical freezing rate is assumed to vary with changes in the number of particles near the ice lens as it grows. The model was demonstrated for the unidirectional freezing of a porous medium consisting of fine glass particles. The numerical results show that this model can represent the formation of intermittent layers of ice lenses in such a system. This model can be applied to ice formation in unconfined water-soaked porous media.

As mentioned above, we clarified the microstructure near the freezing front in water-saturated porous media experimentally and made a model, which can explain the formation of intermittent layers of ice lenses.

CONTENTS

page

1. Introduction

1.1

Soil freezing

1

1.2

The importance of studying soil freezing

2

1.3

Ice lensing and the purpose of this study

5

2. Factors and theories of ice lensing

2.1

Introduction

10

2.2

Main factors considered in soil freezing

11

2.2.1

Unfrozen water

11

2.2.2

Ice formation

14

2.2.3

Water flow during soil freezing

15

2.2.4

Heat transfer during soil freezing

15

2.3

Ice lensing theories and models

17

2.3.1

Capillary theory

18

2.3.2

Hydrodynamic model

21

2.3.3

Secondary frost heave theory

21

2.3.4

2.3.4 Osmotic model 25

25

2.3.5

Takashifs theory

26

2.3.6

Adsorption force theory

27

2.3.7

Segregation potential concept

27

2.3.8

Kinetic model

28

2.3.9

Thermomolecular pressure model

29

2.3.10

Thermodynamics approach

30

2.4

Problems for theories and models of ice lensing

31

3. Materials and apparatus

3.1

Characteristics of samples and grain-size distribution

33

3.2

Surface conditions

36

3.3

Specific surface area and distribution of pores

45

3.4

Unfrozen water content

45

3.4.1

NMR methods

45

3.4.2

Test procedure

46

3.4.3

Unfrozen water content

46

3.4.4

Effect of solute on unfrozen water content

47

3.5

Unidirectional freezing apparatus

52

4. Observation of ice lensing and frost heaving (Exp. 1. 2)

4.1

Sample and method

57

4.2

Experimental results

61

4.2.1

Freezing experiment with zero sample rate

61

4.2.2

Freezing experiment with constant sample rate

74

4.3

Discussion

85

4.4

Summary

88

5. Observation of freezing of dispersed glass beads in water (Exp. 3)

5.1

Sample and method

89

5.2

Experimental results

90

5.3

Discussion

93

5.4

Summary

94

6. Observation of the vicinity of ice lens by Raman spectroscopy (Exp. 4)

6.1

Sample and method

95

6.2

Raman spectroscopy

100

6.3

Experimental results

100

6.4

Discussion

101

6.5

Summary

102

7. A Model for the Formation of Ice Lenses in an Unconfined, Watersaturated, Porous Medium consisting of Spherical Particles

7.1

Model

103

7.1.1

System and conditions

103

7.1.2

Generation of an ice lens

103

7.1.3

Growth of the ice lens

105

7.1.4

Heat transfer

106

7.1.5

Number of particles near the growth surface

107

7.1.6

Ice lensing

109

7.2

Simulation

112

7.3

Discussion

115

7.4

Summary

118

8. Summary and conclusions

References

		page
1.	Introduction
1.1	Soil freezing	1
1.2	The importance of studying soil freezing	2
1.3	Ice lensing and the purpose of this study	5
2.	Factors and theories of ice lensing
2.1	Introduction	10
2.2	Main factors considered in soil freezing	11
2.2.1	Unfrozen water	11
2.2.2	Ice formation	14
2.2.3	Water flow during soil freezing	15
2.2.4	Heat transfer during soil freezing	15
2.3	Ice lensing theories and models	17
2.3.1	Capillary theory	18
2.3.2	Hydrodynamic model	21
2.3.3	Secondary frost heave theory	21
2.3.4	2.3.4 Osmotic model 25	25
2.3.5	Takashifs theory	26
2.3.6	Adsorption force theory	27
2.3.7	Segregation potential concept	27
2.3.8	Kinetic model	28
2.3.9	Thermomolecular pressure model	29
2.3.10	Thermodynamics approach	30
2.4	Problems for theories and models of ice lensing	31
3.	Materials and apparatus
3.1	Characteristics of samples and grain-size distribution	33
3.2	Surface conditions	36
3.3	Specific surface area and distribution of pores	45
3.4	Unfrozen water content	45
3.4.1	NMR methods	45
3.4.2	Test procedure	46
3.4.3	Unfrozen water content	46
3.4.4	Effect of solute on unfrozen water content	47
3.5	Unidirectional freezing apparatus	52
4.	Observation of ice lensing and frost heaving (Exp. 1. 2)
4.1	Sample and method	57
4.2	Experimental results	61
4.2.1	Freezing experiment with zero sample rate	61
4.2.2	Freezing experiment with constant sample rate	74
4.3	Discussion	85
4.4	Summary	88
5.	Observation of freezing of dispersed glass beads in water (Exp. 3)
5.1	Sample and method	89
5.2	Experimental results	90
5.3	Discussion	93
5.4	Summary	94
6.	Observation of the vicinity of ice lens by Raman spectroscopy (Exp. 4)
6.1	Sample and method	95
6.2	Raman spectroscopy	100
6.3	Experimental results	100
6.4	Discussion	101
6.5	Summary	102
7.	A Model for the Formation of Ice Lenses in an Unconfined, Watersaturated, Porous Medium consisting of Spherical Particles
7.1	Model	103
7.1.1	System and conditions	103
7.1.2	Generation of an ice lens	103
7.1.3	Growth of the ice lens	105
7.1.4	Heat transfer	106
7.1.5	Number of particles near the growth surface	107
7.1.6	Ice lensing	109
7.2	Simulation	112
7.3	Discussion	115
7.4	Summary	118
8.	Summary and conclusions
	References