Normal Shrinkage Research Articles

Soil physical properties related to soil structure

The aim of this paper is to clarify the effect of soil aggregation on soil physical and chemical properties of structured soils both on a bulk soil scale, for single aggregates, as well as for homogenized material. Aggregate formation and aggregate strength depend on swelling and shrinkage processes and on biological activity and kinds of organic exudates as well as on the intensity, number and time of swelling and drying events. Such aggregates are, most of all, more dense than the aggregated bulk soil. The intra-aggregate pore distribution consists not only of finer pores but these are also more tortuous. Thus, water fluxes in aggregated soils are mostly multidimensional and the corresponding water fluxes in the intra-aggregate pore system are much smaller. Furthermore, ion transport by mass flow as well as by diffusion are delayed, whereby the length of the flow path in such tortuous finer pores further retards chemical exchange processes. The chemical composition of the percolating soil solution differs even more from that of the corresponding homogenized material the stronger and denser the aggregates are. The rearrangement of particles by aggregate formation also induces an increased apparent thermal diffusivity as compared with the homogenized material. The aggregate formation also affects the aeration and the gaseous composition of the intra-aggregate pore space. Depending on the kind and intensity of aggregation, the intra-aggregate pores can be completely anoxic, while the inter-aggregate pores are already completely aerated. The higher the amount of dissolved organic carbon in the percolating soil solution, the more pronounced is the difference between the gaseous composition in the inter- and in the intra-aggregate pore system. From the mechanical point of view, the strength single aggregates, determined as the angle of internal friction and cohesion, depends on the number of contact points or the forces, which can be transmitted at each single contact point. The more structured soils are, the higher the proportion of the effective stress on the total stress is, but even in single aggregates positive pore water pressure values can be revealed. Dynamic forces e.g. due to wheeling and/or slip processes can affect the pore system as well as the composition of the soil by: (1) a rearrangement of single aggregates in the existing inter-aggregate pore system resulting in an increased bulk density and a less aerated and less rootable soil volume, (2) a complete homogenization, i.e. aggregate deterioration due to shearing. Thus, the smaller texture dependent soil strength coincides with a more intensive soil compaction due to loading. (3) Aggregate deterioration due to shearing results in a complete homogenization, if excess soil water is available owing to kneading as soon as the octahedral shear stresses and the mean normal stresses exceed the stress state defined by the Mohr-Coulomb failure line. Consequently, normal shrinkage processes start again. Thus, the rearrangement of particles and the formation of well defined single aggregates even at the same bulk density of the bulk soil both affect, to a great extent, various ecological parameters. Environmental aspects can also be correlated, or at least explained with the processes in soils, as a major compartment of terrestial ecosystems, if the physical and chemical properties of the structure elements and their composition in the bulk soil are understood.

Shrinkage of some New Zealand soils and its implications for soil physics

Aggregates from three New Zealand soils were used to produce shrinkage curves. Each soil had a clay mineralogy dominated by a different mineral, namely allophane, halloysite and kaolinite.The curves showed marked differences. Only the allophanic soil showed structural shrinkage, and only the halloysitic soil showed residual shrinkage. When the slope of the normal shrinkage line is about unity, this indicates the soil is liable to crack- this was only shown by the allophanic soil. The implication for soil physics is that, since all soils with at least a moderate clay content shrink, bulk densities change with moisture and this must be allowed for in such calculations as soil moisture on a volume basis.

Effect of development and cultivation on physical properties of peat soils in New Zealand

The effect of development and cultivation on the physical properties of two peat soils formed from restiad/sedge and sphagnum peat and developed for agriculture in the Hauraki Plains, New Zealand is reported by comparing physical properties of the plough layer (0–15 cm) and the relatively unaffected underlying peat (15–40 cm). Drainage and cultivation reduced soil moisture content, increased bulk density and reduced organic matter content. Bulk density was correlated with organic matter content only for the plough layer and indicates that bulk density alone is not a good indicator of decomposition in weakly decomposed peat soils in this region. Moisture retention curves indicated that a large amount of the water in the soil could not be removed at high suctions (up to 65% of the water was retained at suction oof 1500 kPa) and may cause moisture stress for plants, particulary in the plough layer of sphagnum peat, despite the high water content of the soil. This has important implications for maintaining a water table which maintains an adequate available water content for plants in the unsaturated zone immediately above the water table. Hydraulic conductivity differed between peat types in both the plough layer and the underlying peat. The presence of a root layer at 35–40 cm in the restiad/sedge peat was considered to substantially influence the readings and has important implications for drainage design in the soil. A simple method for measuring shrinkage in peat soils which involves coating natural clods in a saran resin is presented. A shrinkage curve indicated that shrinkage in peat is directly related to the gravimetric water content and does not contain zones of residual or structural shrinkage which are often associated with mineral soils. It is suggested, therefore, that a simple linear model based upon normal shrinkage will successfully predict volume changes associated with drainage of peat soils for agriculture.

Water movement associated with overburden potential in a shrinking marine clay soil

In unripe marine clay soils, water flow occurs in relation to shrinkage and swelling of the soil material. In the initial stage of physical ripening, when the soil retains its higher moisture ratio, dewatering is mainly accompanied by normal and unidimensional shrinkage. Soilwater flow at this stage, in the absence of significant mateic potential differences, can be explained only by overburden potential. In a one-dimensional vertical system, numerical solutions of non-steady flow during normal shrinkage have been validated with experimental data. For the condition of a closed boundary at the top and a constant pressure head at the bottom, there are significant differences in downward flux density with and without an overburden component to total potential. Further, for the condition of a closed bottom boundary, the generation of an upward flux can only be attributed to the overburden potential. This implies that the overburden potential plays an important role in the Darcy flow, especially for the unripe marine clay soil.

Seasonal variation in field‐saturated hydraulic conductivity in two swelling clay soils in Sweden

SUMMARYA simple method (the inversed auger hole method) for measuring field‐saturated hydraulic conductivity (Kfs) was investigated. Measurements were carried out in the spring, summer and autumn at three depths in two Swedish clay soils (Ultuna and Limsta, with clay contents of 45–60%0 and 65–80%, under barley and grass/clover ley respectively). Seasonal fluctuations in Kfs at Limsta were more pronounced, and were observed deeper in the profile. This was attributed primarily to larger structural changes due to a higher capacity for swell/shrink (normal shrinkage over the available water range) and an earlier drying up of the soil under grass/clover ley.It was shown that the measured Kfs values were strongly correlated with the total inter aggregate (macro‐) porosity (et), estimated from a simple model of soil shrinkage. Combining the data from both soils, a single power‐law relation was adequate (r= 0.73) to describe the variations in Kfs with et

Shrinkage of Dutch clay soil aggregates.

Shrinkage characteristics and COLE and PLE values of undisturbed natural aggregates of clay soils from the Netherlands were measured. The course of the shrinkage process upon drying varied strongly between soils, and very often the measured characteristics diverged from the theoretical curve. Some Dutch clay soils are amongst the strongest swelling and shrinking soils in the world, with volume decreases of aggregates up to 49% between saturation and oven-dryness, and 42% between saturation and a pressure head of -16 000 cm. Potential subsidence of a Dutch field soil due to shrinkage is up to 15 cm. In some Netherlands clay soil, as a result of normal shrinkage, the aggregates remain saturated throughout the whole year; only inter-aggregate pores such as shrinkage cracks, contain air. (Abstract retrieved from CAB Abstracts by CABI’s permission)

Shrinkage Curve Indices to Quantify Cultivation Effects on Soil Structure of a Vertisol

AbstractDifferences in selected parameters and derived variables of models fitted to soil shrinkage curves were used to investigate soil structural degradation of an irrigated Pellustert in northern New South Wales, Australia. Significant differences between clods taken from the same depths in land prepared for cotton (Gossypium hirsutum L.) when “dry” and “wet” were found in specific clod volume at zero water content, the slope of the line in the normal shrinkage zone, the airfilled specific pore volume throughout normal shrinkage, and the range in water content in the structural shrinkage zone. These indicators showed that this soil prepared wet exhibited increased bulk density, decreased porosity, loss of structural pores, and tended to have weak structural development.

Collapse, volumetric shrinkage, specific gravity and extractives in Eucalyptus and other species

The influence of extractive content on the shrinkage behaviour in eucalypts was shown by “mathematical extraction” after making certain assumptions with respect to available data. The slope of the volumetric shrinkage/specific gravity (S/ϱ) ratio changed from negative to positive after the theoretical removal of alcohol/benzene/water soluble or NaOH soluble extractives. Statistical analysis of available data suggested that collapse was positively related to the encrusting and extraneous materials and negatively related to the polysaccharide cell wall component. Conversely, normal volumetric shrinkage was positively related to the polysaccharides and negatively related to lignin and extractive content. Mechanisms by which extractive content might influence collapse development were discussed. R-values (change in external volume during shrinkage or swelling per change in weight of equivalent volume of water) appeared to be negatively correlated with extractive content in eucalypts and calculations for other species showed that R progressively increased with cold and hot water extraction. Further analysis suggested that the apparent change in lumen dimensions suggested by particular R-values was mitigated by the bulking effect of extractives in the cell wall. A possible role for R as an indicator of dimensional stability was postulated.

Effects of flooding and subsequent drying on consolidation of beds of aggregates of a gray clay from Narrabri, NSW

Changes in porosity of beds of aggragates of a grey clay, from Narrabri, NSW, on drying after flooding were studied in a laboratory experiment. For both surface soil (0-0.2m) and subsoil (0.8-1.0m) as a result of quick wetting, air-filled porosities sufficiently low (<0.1m3m-3) to be adversely affecting plant growth were recorded in the aggregate beds for much of the plant-available moisture range. Resultsindicated that, for the subsoil, the structural shrinkage phase was completely absent. Normal shrinkage prevailed and air-filled prosity remained at<0.1m3m-3 until the commencement of the residual shrinkage phase, which occurred at moisture content drier than the permanent wilting point (-1.5kj kg-1). Calcium chloride (0.01M) did not have an effect on the surface soil but significantly increased air-filled porosity of the subsoil by creating a structural shrinkage phase and hence reduced the extent of normal shrinkage (consolidation), Relevance of these laboratory results to field situations under furrow irrigation was discussed.

Shrinkage and Structure of Allophane Soil

The shrinkage of allophane soil was studied in relation to structure. The structure was described in terms of two fabric units called S -unit and L -unit which were observed at different magnification levels with a scanning electron microscope. An S -unit is 1-2 μ m in diameter and an L -unit is 100-200 μ m in diameter. The characteristics of shrinkage and structure which were found to correspond to each other are as follows: (1) the commencement of residual shrinkage at pF 4. 1 and the large amount of this shrinkage corresponds to the skeleton which is framed by a series of S -units and wherein allophane is continuously distributed. (2) the increase in normal shrinkage range after remolding corresponds to the breakdown of the structure with L -units and the subsequent release of the void water in the structure by remolding. (3) the reappearance of pF 4.1 as the value at which residual shrinkage commences after remolding corresponds to the fact that S -units remain after remolding. As for the effect of the shrinkage on the structure, it was concluded that the S -units became compacted and then combined, causing the L -units to become rigid. The structure information presented here was applied also to the description of such behavior as the decrease of liquid limit with decreasing water content and the sensitivity of allophane soil.

Using bulk density to calculate soil water changes in soils with shrinkage cracks

Apparent bulk �a measured between shrinkage cracks is used to assess vertical shrinkage on the assumption that soil cores shrink isotropically from a fully swollen reference bulk density pr. A fully swollen depth increment �X corresponds to an increment �x = (�t/�a)1/3 �X at any other water content, field bulk density (which includes cracks) is given by and the fractional area of shrinkage cracks is 1 - (�t/�a) 2/3. When gravimetric water contents are known, volumetric water contents can be calculated in terms of the fully swollen reference depth scale and changes in water contents of soil layers obtained. The procedure applies to any soil, from rigid through to one which undergoes normal shrinkage, provided the above assumption is satisfied. A BASIC program to perform the calculations is given.

GASEOUS DIFFUSION IN SHRINKING SOILS

To assess the effect of shrinkage of two zero-tilled clay soils on gaseous diffusion, we measured relative diffusivity on large samples (15.8 cm diameter, 10 cm high) that showed shrinkage cracks and on small samples (7.5 cm diameter, 5 cm high) that had no visible shrinkage cracks. The relationship between air-filled porosity and relative diffusivity was similar for both soils. Upon drying, both clay soils showed structural shrinkage followed by normal shrinkage. During structural shrinkage, relative diffusivity (Ds/D0) and air-filled pore space (α) were related by the expression Ds/D0 = α2]; with the onset of normal shrinkage and the appearance of shrinkage cracks, further increases in Ds/D0 were nearly equal to increases in α. At the same gravimetric water content, relative diffusivity decreased from the large to the small cores. This was more pronounced for the soil with the weak blocky structure than for the soil with the strongly developed block structure. Measurements of relative diffusivity in vertical and horizontal cores were similar when the cores were near saturation. As the air-filled pore space increased, however, vertical diffusivity of these zero-tilled soils was about twice as large as horizontal diffusivity. This anisotropic behavior is related to the larger vertical than horizontal shrinkage in the large cores. Both observations point to a difference in pore orientation or tortuosity in the vertical and horizontal direction. In plowed plots, horizontal relative diffusivity appeared larger than vertical diffusivity, but in the undisturbed subsoils the directional dependence of relative diffusivity was similar to that in the zero-tilled soils.

Shrinkage characteristics of soil clods from a grey clay under intensive cultivation

Structural degradation in a grey clay under intensive cropping was investigated by studying the shrinkage characteristics of soil clods collected from the field. Soil clods with different cropping histories differed in the amount of structural shrinkage relative to normal shrinkage. Soil clods (c. 50 cm3) from the irrigated site were found to have a smaller degree of structural shrinkage than those from the dryland wheat site, resulting in higher bulk density and lower air-filled porosity. The 'degraded' soil clods remained almost saturated with water over the available moisture range.

ROTHAMSTED STUDIES OF SOIL STRUCTURE III

SummaryThe response of three pairs of soils having contrasting management behaviour to changes in water content was investigated by measuring moisture characteristics, shrinkage, and pore size distributions by mercury porosimetry. Sample preparation for mercury porosimetry was by direct drying, or water replacement by methanol and liquid CO2 followed by evaporation above the critical temperature. In heavy textured horizons, water release on shrinkage between ‐0.05 and ‐15 bar matric potentials occurs when pores of 10–200 nm equivalent plate separation contract. Structure development is dependent on the stability and regeneration of pores in the size range 200 nm‐30 μm during cycles of wetting and drying. Soils with little stable porosity in this size range show nearly normal shrinkage and have coarse structural units, whereas soils with stable or regenerating porosity depart more from normal shrinkage and structure in the profile tends towards a finer per unit. The application of mercury porosimetry shows promise for determining possible porosity limitations in soil profiles.

Engineering problems associated with expansive clays from Romania

Engineering problems associated with expansive clay formations, which have a large swelling and shrinking potential, exist in different regions of Romania. In the general framework of the study of the behaviour of expansive clays, the present paper describes some particular aspects connected with the swelling and shrinkage processes and their engineering effects in Romania. The differences between swelling potential relationships, obtained by two different methods are presented and discussed. Swelling pressure tests were accomplished using the available standard techniques, as well as a specially developed cell pressure equipment. Attention is drawn to the stages of the shrinkage process as related to soil structure. It was found that, in a general case, there are three distinctṡtages: initial, normal and residual shrinkage. Laboratory investigations have shown that in soils with an anisotropic structure, the directional variation of shrinkage strains is elliptical in shape, with the maximum shrinkage normal to particle orientation. The engineering aspects of the results are discussed.

Effect of restraint on the thermal transitions of tendon

Sheep tendon samples that had been given a heat treatment in 0·9% sodium chloride solution at temperatures above their normal shrinkage temperature (ST) were examined by differential scanning calorimetry. If the tendon was not restrained from shrinkage during the heat treatment, no endothermic transition in the region of 60°C was subsequently detected by calorimetry. If it was restrained an endothermic transition was then detected, the magnitude of which was inversely related to the shrinkage allowed to occur during heat treatment. Values for transition heat (ΔH) and transition temperature (T 1,2) of tendon samples restrained at their original taut length and heat treated 10°–15°C above the ST were similar to those found for unheated control samples. However, increase in the temperature of the heat treatment to 20°–30°C above the ST resulted in a decrease in both ΔH and T 1,2 and the transition became broader.

The nature of changes in bulk density with water content in a cracking clay

The effect of changing water content on the bulk density of undisturbed cores of a cracking clay was examined in laboratory experiments. The results were compared with the relationship between bulk density and water content established by core sampling the same soil in the field. Over the water content range measured in the field soil, the laboratory cores shrank three-dimensionally and normally. Small departures from normal shrinkage were attributed to the formation of cracks within the cores, and to the occurrence of some structural shrinkage in cores previously wet to high water contents. Swelling of cores was approximately three-dimensional, except for some unconfined swelling which occurred in the core surface. Unidimensional swelling was induced by confining dry cores to reduce the void ratio before wetting. Subsequent shrinkage was three-dimensional, indicating that the soil particles were reoriented during the unidimensional swelling phase. While the laboratory measurements showed that the soil volume changes were essentially three-dimensional and normal, the field data indicated that unidimensional shrinkage occurred at water contents greater than 0.47 g g-1. These field results were attributed to sampling inaccuracies associated with the use of a small-diameter core sampler, the actual field bulk density relationship being considered three-dimensional.

The effect of height of sample and confinement on the moisture characteristic of an aggregated swelling clay soil

The pore pressure/moisture content relationship of 0.5-1.0 mm aggregates of Narrabri soil, initially 1.66 cm high, was examined at 1/3 cm height intervals by means of dual source gamma ray scanning, for confined and unconfined conditions. In the unconfined sample, absorption and desorption of water to or from the interaggregate pore space occurred at less negative pore water pressures than in the confined sample. In the latter, the swelling aggregates deformed into the inter-aggregate pore space during absorption and reduced the size of the pores. In the unconfined sample, absorption into the inter-aggregate pore space occurred at less negative water pressures with increasing height in the sample. During absorption, bulk density decreased with height in the sample, and bulk density at all points within the sample decreased until pore water pressures were between -20 to - 10 cmH2O. At pore water pressures closer to zero, bulk density increased as the sample compacted under its own weight. During desorption, bulk density increased at all heights as the pore water pressures decreased from zero to - 8 cmH2O, because of isotropic stress imparted by the hanging water column. During this phase normal shrinkage occurred, therefore overburden potentials, and hence moisture potentials, could be calculated. Between - 10 and - 50 cmH2O, the inter-aggregate pore space drained rapidly, and no further increase in bulk density was detectable and hence moisture potential equates to manometric pressure. It was found that the derived moisture characteristics (� versus �) are different for different heights in the aggregate mass. This result is of consequence to the theory of hydrostatics in swelling media. In the unconfined sample, no differences in the pore water pressure/moisture content relationships occurred with height during absorption or desorption, and bulk density remained constant throughout. The application of these results to the behaviour of swelling soils in the field is briefly discussed. The use of dual source gamma ray scanning seems an essential tool for further investigations.

Kinetics of sorption and volume change in three-component systems

The paper analyses one-dimensional absorption (swelling) and desorption (shrinkage) in media subject to volume change (such as soils of high colloid content) and containing both water and air. The void ratio, the hydraulic conductivity, and the moisture potential are taken to be arbitrary known functions of the volumetric moisture content �. Then the combination of Darcy's law, applied to water flow relative to the soil particles, and the continuity requirement yields the general flow equation for systems of this type. In terms of material coordinates, the equation takes a nonlinear heat-conduction form. For media exhibiting (a) zero and (b) normal volume change, the equation reduces, as it should, to two known equations: (a) the nonlinear diffusion equation in physical coordinates, and (b) the nonlinear diffusion equation in material coordinates. The specific problem is that of sorption consequent on a step-function change in moisture potential at the column surface. The column is taken to be effectively semi-infinite (and constrained at infinity). The relevant solution of the general flow equation is of the similarity formm(�,t) = �(�)t1/2 where m is the material coordinate, t is time, and 4(8) is found by the solution of a nonlinear, ordinary, integrodifferential equation. A rapid and accurate numerical method of evaluating �(�) from any given set of soil characteristics is provided by a minor variant of an established technique for solving the nonlinear diffusion equation subject to similar conditions. With �(�) known, it is an elementary matter to restate the solution in physical coordinates, and to deduce a great variety of properties of the sorption process : profiles of moisture content, moisture potential, and void ratio; cumulative total volume change and the instantaneous rate of total volume change; cumulative total sorption and the instantaneous sorption rate; profiles of soil-particle velocity; profiles of volume flux density of water (i) relative to the soil particles and (ii) due to mass flow; profiles of absolute volume flux density of water; and the displacement history of soil particles initially at various positions in the column. Two examples are worked out fully, one for absorption and one for the converse desorption process. Graphs are presented illustrating the various aspects of the solution mentioned above. It is found, in keeping with previous work on two-component systems exhibiting normal volume change, that swelling is propagated more rapidly than is shrinkage but that it is associated with a slower rate of exchange of water. As in the earlier study (Philip 1968), this seeming paradox arises from the influence of the mass flow of water on the phenomena. For the example of absorption, the total column swelling amounts to only 43% of normal swelling; on the other hand, the converse desorption process produces over 99% of normal shrinkage. The moisture profiles computed for absorption exhibit a shallow surface layer of large moisture gradient, reminiscent of the transition zone sometimes observed in infiltration and absorption experiments. The present analysis applies to the drying of a cracking soil, so long as the (properly defined) void ratio function exists.

A Precision Weighing Lysimeter for Row Crop Water Use Studies1

AbstractA relatively simple and economical weighing lysimeter specifically designed for evaluating hourly row crop water use under dryland conditions has been developed and tested. The lysimeter soil container is 1.83 m (6.0 ft) square by 1.22 m (4.0 ft) deep. This allows a representative rooting profile for two rows of crop spaced 0.92 m (3.0 ft) apart and also allows normal soil shrinkage cracks to form within the soil mass. The thin soil container and retaining tank walls are separated by a 10‐mm air gap, resulting in a surface disturbance of only 4.5% of the lysimeter area. Soil water suction at the bottom of the soil container is controlled. A commercially available floor stand tank scale lever assembly is used to transfer up to 600 kg of the total lysimeter weight to a precision strain gauge load cell, the remainder being counter‐balanced under the lysimeter. The sensitivity of the system (± 0.025 mm equivalent water weight change) is adequate for short‐period evaporative flux measurements. The electrical output signal from the load cell can be measured remotely, minimizing site disturbance. The weighing mechanism has suitable utility for other weighing lysimeter applications.