SSSAJ Curious? Visit www.highwire.org
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Institution: Danmarks || Sign In as Member/Individual (Non-Member)
Abstract of this Article
Figures Only for this Article
Reprint (PDF) Version of this Article
Similar articles found in:
SSSAJ Online
Search Medline for articles by:
Munkholm, L. J. || Kay, B. D.
Search Agricola for articles by:
Munkholm L. J. || Kay B. D.
Alert me when:
new articles cite this article
Download to Citation Manager
Collections under which this article appears:
Structure and Properties
Soil Physics
Water Content
Soil Science Society of America Journal 66:702-709 (2002)
© 2002 Soil Science Society of America

DIVISION S-1—SOIL PHYSICS

Effect of Water Regime on Aggregate-tensile Strength, Rupture Energy, and Friability

Lars J. Munkholm*,a and Bev D. Kayb

a Dep. of Crop Physiology and Soil Science, Danish Institute of Agricultural Sciences, Research Centre Foulum, P.O. Box 50, DK-8830 Tjele, Denmark
b Dep. of Land Resource Science, Univ. of Guelph, Guelph, ON, Canada N1G 2W1

* Corresponding author (Lars.Munkholm@agrsci.dk)


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
THEORY
MATERIALS AND METHODS
RESULTS AND DISCUSSION
CONCLUSIONS
REFERENCES
 
The effect of water regime on aggregate-tensile strength, rupture energy, and friability index was studied in two experiments on sandy loam soils (Glossic Phaeozem/WRB 1998). In Exp. 1, soil from a long-term, 5-yr crop rotation receiving animal manure (dairy farm and grass, DFG) was compared with a soil (continuous cereal crop, CCC), which had grown almost continuous cereals and not received animal manure. In Exp. 2, soil from compacted plots (PAC) and uncompacted reference plots (REF) was sampled. The soil was air-dried and separated into four aggregate-size classes (i.e., 1–2, 2–4, 4–8, and 8–16 mm). Aggregate-tensile strength and rupture energy were measured on air-dry aggregates and on aggregates adjusted to -10, -30, -100, -350 kPa, and -166 MPa pressure potential. Soil friability index was estimated from the tensile strength or specific rupture energy results. Aggregate density was determined on 2- to 4-, 4- to 8-, and 8- to 16-mm aggregates. The CCC soil displayed a greater increase in strength and specific rupture energy with increased dryness than the DFG counterpart. This may be related to differences in organic matter content and dispersible clay. Generally, the relationship between tensile strength or specific-rupture energy and pressure potential could be described by a power function. However, the rupture energy results could not be fitted by a power function for the Exp. 2 soils. The Exp. 2 soils differed in the stress–strain relationship, i.e., the compacted soil displayed the highest Young modulus, (Y/{epsilon}) in all cases. Maximum values of the friability index were found between -10 and -100 kPa.

Abbreviations: CCC, continuous cereal crop • DFG, dairy farm with grass • PAC, compacted plots • REF, uncompacted reference plots


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
THEORY
MATERIALS AND METHODS
RESULTS AND DISCUSSION
CONCLUSIONS
REFERENCES
 
TENSILE FAILURE is the desired mode of failure during seedbed preparation as large clods are fragmented without disturbing soil microstructure. However, the energy input should not be so high as to result in pulverization of the soil. A friable soil that crumbles to yield an optimal aggregate-size distribution is desirable. Therefore, the energy required in seedbed preparation depends on the strength of the large clods and the fragmentation pattern of the soil.

Tensile strength of soil aggregates is generally measured in a compression test (e.g., Dexter and Kroesbergen, 1985). Based on the measurements of tensile strength on different sized aggregates, a friability index has been proposed (Utomo and Dexter, 1981). However, measuring tensile strength in a compression test at soil water contents similar to those in the field at tillage is complicated. The mode of failure shifts from pure tensile to shearing and compression with increasing water content. Therefore, aggregate-tensile strength has up till now almost exclusively been determined on air-dry or even oven-dry aggregates. Moreover, the drying of soil aggregates to very low water contents may alter the soil fabric or change the nature of cementing materials (Caron et al., 1992), and there is a risk that the drying procedure may influences aggregates from different treatments differently (Caron and Kay, 1992). If so, extrapolating results based on dry aggregates to soil behavior in the field would be problematic.

The objective of this research was to study the effect of water regime on aggregate-tensile strength, rupture energy, and on the soil fragmentation pattern as determined by the friability index.


   THEORY
TOP
ABSTRACT
INTRODUCTION
THEORY
MATERIALS AND METHODS
RESULTS AND DISCUSSION
CONCLUSIONS
REFERENCES
 
Determination of aggregate-tensile strength, Y, has been widely applied to characterize ease of aggregate fracture in tillage. As it is difficult to measure Y directly in a direct tension test, most authors have used compression tests to determine Y. For a spherical aggregate, Y can be determined from the equation (Rogowski, 1964; Dexter, 1975)

[1]
where a is a constant, F (N) is the polar force required to fracture the aggregate, and d (m) is the mean aggregate diameter. The constant, a, depends on the relationship between the compressive and tensile stress in the center of the aggregate (Hiramatsu and Oka, 1966; Dexter, 1975). The most extensively used value for a (i.e., a = 0.576) is based on the assumption of spherical form and perfect elastic behavior (Poisson ratio = 0.5) (Dexter, 1975). The assumption of perfect elastic behavior until failure can most easily be fulfilled by dry soil. Farrell et al. (1967) found that the relationship between stress and strain became increasingly nonlinear as the water content increased, indicating plastic deformation during the fracture process. Because of the increased plastic deformation at increasing water content, Dexter and Kroesbergen (1985) suggest using oven-dry aggregates in the compression test. Also, effects on pore-water pressure during the test may affect the strength measurement of moist aggregates. In a direct-tension test, the pore water pressure gets increasingly negative during the test (Junge et al., 2000), whereas it gets increasingly positive in an indirect tension test (compression) (Dexter and Watts, 2000). Utomo and Dexter (1981) proposed the scale factor between aggregate size and aggregate strength as an index of friability, k

[2]
where A (kPa) is the predicted log strength of 1 m3 soil and V (m3) the aggregate volume.

When applying conventional Weibull analysis (Braunack et al., 1979; Perfect and Kay, 1995), k is estimated from the spread of strength. For a specific size of aggregates, i, the cumulative relative frequency distribution of Y can be expressed as:

[3]
where c1 is a fitting parameter.

Solving Eq. [3] for the mean aggregate strength yields (Braunack et al., 1979)

[4]
where c2 is a fitting parameter.

Equation [2] may be obtained by taking the logarithm of Eq. [4], thus A = log c2.

Perfect and Kay (1994) recommended using rupture energy, E, for the statistical characterization of aggregate strength in tillage studies. They concluded that rupture energy gave a more appropriate estimate of dry aggregate strength than tensile strength since no assumptions on mode of failure have to be taken. This conclusion would appear to be particularly relevant to the fragmentation of wet aggregates. The rupture energy was obtained by integrating the stress-strain curve until the point of failure (Vomocil and Chancellor, 1969).

[5]
where F(s) is the compressive force at a specific strain between 0 and the strain at rupture, r.

The specific rupture energy may be defined on a gravimetric basis, Esp whereby no measure of the aggregate diameter is needed.

[6]
where m is the mass of the individual aggregate.

As for tensile-strength data, a friability index can be estimated from Eq. [3] by replacing Y by Esp as input variable. For air-dry aggregates, they found a linear correlation between k calculated on the basis of Y and Esp (Perfect and Kay, 1994) (R2 = 0.54). The k value estimated using Esp, kE, was 1.45 times higher than those estimated using Y, kY.


   MATERIALS AND METHODS
TOP
ABSTRACT
INTRODUCTION
THEORY
MATERIALS AND METHODS
RESULTS AND DISCUSSION
CONCLUSIONS
REFERENCES
 
Soils
In this study, soil was sampled in two experiments: Exp. 1 cropping systems, Exp. 2 traffic experiment. All soils were sandy loams developed on till plains from the Weichselian glacial stage and classified as Oxyaquic Agriudolls/Glossic Phaeozems according to the USDA/WRB98 classification systems. Some basic soil characteristics are given in Table 1.


View this table:
[in this window]
[in a new window]
 
Table 1. Basic soil characteristics for all soils in the investigation (6 to 13-cm depth).

 
Experiment 1.

A soil from a dairy farm with a 5-yr crop rotation including a 2-yr grass (DFG) was compared with a neighboring soil managed for at least 20 yr with no animal manure application and almost continuous small grain cereals production (CCC). Both soils had been managed according to conventional tillage practices that included moldboard plowing. Seedbed preparation was carried out using a rotary harrow in all soils. Further, all soils were grown with an annual crop in the year preceding sampling. This also meant that for all soils, a time period of at least ~1.5 yr had passed since incorporation of a grass ley. The management characteristics of the soils are described in detail by Schjønning et al. (2002).

Experiment 2.

Experiment 2 was a traffic experiment located on the organically managed Rugballegård Experimental Station, Denmark. The experiment was concerned with the effect of topsoil compaction on soil physical properties. A compaction treatment, labeled PAC, was compared with a reference treatment, labeled REF. The experiment was established in 1997 and the treatments were applied to plots in a randomized block design with three blocks. In each experimental year (1997–1999), the PAC treatment was carried out in early spring on soil at field capacity or wetter than field capacity. The treatment was performed after moldboard plowing to a 20-cm depth. The PAC treatment was obtained by creating adjacent wheel tracks using a 6- to 8-Mg tractor (inflation pressure 125 kPa). Secondary tillage and sowing were carried out later in the spring by an S-tine harrow and a traditional drill at moisture content around field capacity. Small grain cereals were grown in all experimental years.

Sampling
In Exp. 1, a representative and uniform area of each field was selected to fit texturally the counterpart. For each field, nine sampling points were identified at the intersection of a 10 by 10 m grid. At each grid point ~1 m2 of land was used for sampling and field measurements. To reduce effects of temporal variation in the parameters studied (e.g., Mulla et al., 1992), sampling and field measurements was carried out in early spring when the soil first reached water content of field capacity in an autumn-sown winter small grain cereal crop. Bulk soil was sampled from the 6- to 13-cm depth at all nine grid intersections and stored in plastic containers.

For the Exp. 2, soil sampling was performed in May 1999 shortly after plant emergence at a water content slightly drier than field water capacity (Table 2). For each plot a subsample was taken at the 7- to 12-cm depth in three 0.5 m2 areas. Subsequently, the three subsamples from each plot were bulked.


View this table:
[in this window]
[in a new window]
 
Table 2. Water content, porosity and density of soils (6 to 13-cm depth). Values followed by the same letter in rows for the specified experiment are not significantly different at the P < 0.05 level.

 
Soil Separation and Wetting Procedure
A gentle soil separation procedure was followed that involved air-drying, gently crushing of large air-dry clods and separation by sieving. The procedure allowed a separation of the soil into small aggregates without kneading and with limited energy input. However, some disturbance of the physical characteristics of the soil because of air-drying and handling cannot be excluded.

The samples were air-dried by spreading the soil in a dry ventilated room at ~25°C immediately following arrival at the laboratory. The air-dry soil was separated into four aggregate size-classes 8 to 16, 4 to 8, 2 to 4, and 1 to 2 mm. First the air-dry soil was passed through a nest of sieves with the openings of 16, 8, 4, and 2 mm. An amount (130 g) of 8- to 16-mm aggregates was taken out randomly and placed in a box and stored. The rest of the material on the 8-mm sieve and the material on the 16-mm sieve were crushed to <8 mm, using the roller method suggested by Hartge (1971). The soil was then again placed on the nest of sieves (except for 16-mm sieve). The soil material on the 4-, 2-, and 1-mm sieves was placed into boxes and stored at a constant temperature of 20°C.

An amount of the air-dried aggregates of each size class was transferred to metal cylinders (2.2-cm diam., 3.4-cm height, 50-cm3 volume) and slowly capillary wetted to -1.5 kPa on tension tables to minimize slaking. Thereafter, the aggregates were adjusted to -10 kPa pressure potential. A number of aggregates were taken out randomly and used in the crushing test. The remaining aggregates in the metal cylinder were subsequently dried to -30, -100, and -350 kPa using pressure plates. For each pressure potential, a similar number of aggregates was taken out and subjected to the crushing test. Aggregate strength was also determined on air-dry aggregates. The pressure potential in air-dry soil was calculated to be -166 MPa assuming the aggregates were in equilibrium with the air in the laboratory, which had a constant temperature of 20°C and a relative humidity of ~30%.

Aggregate Density
Dry aggregates were weighed and coated with Saran resin (Brasher et al., 1966). The coated aggregates were then weighed in air and then reweighed immersed in water at 20°C. The weight loss equals the volume of water displaced, which is equal to the volume of coated aggregate. The measurements were adjusted for the weight and volume of the Saran resin according to Schafer and Singer (1976):

[7]
where md denotes the weight of an oven-dried aggregate, mw is the aggregate weight when immersed into water, {rho}w the density of water at 20°C, ms is the weight of Saran coating and {rho}s is the density of Saran coating. The latter was determined to be 1.36 g cm-3. The factor, {alpha}, accounts for Saran resin sucked into the aggregates. Schafer and Singer used {alpha} = 0.3 for large clods (>50 cm3). In this study, a value of 0.3 was used for 8- to 16-mm aggregates. To take account of the higher surface area in relation to aggregate volume for smaller aggregates, arbitrary values of 0.6 and 1.0 were used for 4- to 8- and 2- to 4-mm aggregates, respectively.

Aggregate bulk density was determined on fifty-four 4- to 8-mm aggregates for the Exp. 1 soils (i.e., six from each field grid point). For the Exp. 2 soils aggregate density was determined on 39 aggregates per treatment (i.e., 13 for each plot) from the 2- to 4-, 4- to 8-, and 8- to16-mm size fractions.

Crushing Test
The aggregates were crushed individually between two flat parallel plates using the indirect tension test described by Dexter and Watts (2000). All of the tests were performed at constant rate of displacement of 2 mm min-1. The compressive force was measured 30 times per second by a load cell (0–100 ± 0.03 N) and recorded automatically on an adapted computer. A larger load cell (0–500 ± 0.15 N) was used for the largest air-dried aggregates from the Exp. 1 soils. The strain rate was calculated from the rate of displacement.

The aggregate tensile strength, Y, was calculated from Eq. [1] with a = 0.576 in all cases. Generally, a linear stress–strain relationship until failure was found at all the tested pressure potentials and the soil aggregates broke into two halves along the vertical axis. This indicates that tensile failure was the dominant mode of failure. No corrections were made of flattening of the moist aggregates before failure. According to Frydman (1964), a high degree of flattening is needed to substantially give errors in using Eq. [1]. In addition, the smallest size fractions that showed a tendency to compress rather than to fail in tension mode at high pressure potentials were excluded. Tensile strength was not measured on the 1- to 2-mm aggregates at -10 and -30 kPa and on the 2- to 4-mm aggregates at -10 kPa for the Exp. 1 soils. The aggregate diameter, d, was estimated from "Method 4", described by Dexter and Watts (2000).

[8]
where xi is the mean of the openings of the upper and lower sieves for the ith size class; m is the mass of the individual aggregate and i is the mean mass of the aggregates in the ith size class.

The rupture energy, E, was derived by calculating the area under the stress-strain curve:

[9]
where F(si) is the mean force at the ith subinterval and {Delta}si is the displacement length of the ith subinterval. The specific rupture energy, Esp, was determined according to Eq. [6]. The weight of the two largest size classes of aggregates (i.e., 4–8 and 8–16 mm) was determined directly by measuring the weight of the aggregates, individually. The recorded weights were adjusted to oven-dried weight and these were used as the input variable, m, in Eq. [6] and [8]. The 2- to 4-mm aggregates were classified as small, medium, or large when performing the test. The weight of the individual 2- to 4-mm aggregates was then calculated from the estimated-size and the measured-aggregate density for the specific soil and treatment. For the 1- to 2-mm aggregates, m was estimated from the mean-aggregate size and the aggregate density. In both experiments, aggregate density for the 4- to 8-mm aggregates were used to estimate m for the 1- to 2-mm aggregates. The mean-aggregate size of the 1- to 2-mm aggregates was calculated assuming spherical form with a diameter of 1.5 mm (i.e., the mean of the upper and lower sieve opening). The friability index, kY, was estimated from Eq. [2] and kE was estimated by replacing Y by Esp in Eq. [2].

For the Exp. 1 soils, 45 aggregates (i.e., five from each soil grid point) were crushed for each combination of soil, pressure potential, and size class with the exceptions mentioned above for the smallest aggregates at the two highest pressure potentials. For the Exp. 2 soils, 45 aggregates were used for each combination of treatment, pressure potential and size class with the exceptions for the highest pressure potentials as mentioned above. In all, ~3150 aggregates were subjected to the crushing test.

Statistical Analysis
The Y and Esp parameters were log-transformed to yield normality. For the Exp. 1 soils, the variation between soils was compared with inter-grid variation (F-tests, using the PROC GLM, SAS-Institute, 1996) for testing significant (P = 0.05) difference between soils. For Exp. 2 the variation between treatments was compared with the intra-plot variation for testing significance between treatments.


   RESULTS AND DISCUSSION
TOP
ABSTRACT
INTRODUCTION
THEORY
MATERIALS AND METHODS
RESULTS AND DISCUSSION
CONCLUSIONS
REFERENCES
 
Soil Density
In Exp. 1, the forage cropped DFG soil had significantly lower density of bulk soil and of 4- to 8-mm aggregates than the CCC counterpart (e.g., 1.74 and 1.77 Mg m-3, respectively, for the 4- to 8-mm aggregates) (Table 2). In Exp. 2, the compacted soil had significantly higher bulk density and aggregate density than the reference treated soil (e.g., 1.71 and 1.63 Mg m-3 for 4- to 8-mm aggregates, respectively). The aggregates had a higher density than the bulk soil indicating exclusion of macroporosity with decreasing sample size.

The values of bulk density and aggregate density reflected differences in soil management practices. In Exp. 2, the difference was a direct consequence of soil compaction. The difference in density between the soils used in Exp. 1 may be because of lower aggregate stability in the CCC soil. It has a lower content of organic matter (Table 1) and biological-aggregate binding and bonding material and a higher content of mechanical dispersible clay (Schjønning et al., 2002).

Tensile Strength and Specific-Rupture Energy
In the air-dry state (~-166 MPa), the strength of the 8- to 16-mm aggregates and the mean strength of CCC soil were significantly greater than those of the DFG soil (Table 3). The aggregate-tensile strength decreased with increasing water content. At -10 kPa the 4- to 8-mm aggregates of the CCC soil were apparently weaker than the corresponding aggregates for DFG. The specific rupture energy results correspond to the tensile strength results. In air-dry soil, the CCC soil had significantly higher Esp, whereas at the highest pressure potential (-10 kPa) the CCC soil had significantly lower mean Esp.


View this table:
[in this window]
[in a new window]
 
Table 3. Geometric mean values of tensile strength, Y, specific rupture energy, Esp, in Exp. 1. Values followed by the same letter for a given aggregate size and pressure potential are not significantly different at the P < 0.05 level.

 
The PAC-treated soil had significantly higher tensile strength for one or more aggregate sizes at all pressure potentials, except at -30 kPa (Table 4). In general, all the aggregate-size fractions were affected by the PAC treatment when tested at pressure potentials <=-100 kPa. Contrary to the Exp. 1 soils, the specific-rupture energy results from Exp. 2 did not show a good agreement with the tensile-strength results. No significant difference in Esp was found between the treatments at any combination of aggregate size and pressure potential.


View this table:
[in this window]
[in a new window]
 
Table 4. Geometric mean values of tensile strength, Y, specific rupture energy, Esp, in Exp. 2. Values followed by the same letter for a given aggregate size and pressure potential are not significantly different at the P < 0.05 level.

 
Influence of Pressure Potential
The tensile strength of the aggregates increased with decreasing pressure potential and could be modeled by a power function of the type:

[10]
where q and n are regression coefficients and {Psi} is the pressure potential. Equation [10] may be linearized by taking the logarithm form,

[11]
where the pressure potential is given in 0.1 kPa units, and the coefficient, q, is an estimate of the tensile strength of the soil aggregates at pressure potential equal to -0.1 kPa, i.e., close to saturation. The relationship between Y and {Psi} is illustrated in Fig. 1a . For clarity, only data for the 8- to 16-mm aggregates are shown. The correlation coefficient varied between 0.92 and 0.94 for the Exp. 1 soils and between 0.67 and 0.81 for the Exp. 2 soils (Tables 5 and 6). The relationship between Esp and pressure potential could also be fitted by a power function for the Exp. 1 soils, although the correlation coefficients were lower than for the Y data (r2 = 0.70–0.87) (Table 5, Fig. 1b). For the Exp. 2 soils, Esp also increased with decreasing pressure potential but the relationship between Esp and pressure potential could not be fitted by a power function as shown for the 8- to 16-mm aggregates in Fig. 1b.



View larger version (23K):
[in this window]
[in a new window]
 
Fig. 1. (a) Aggregate tensile strength, Y, and (b) specific rupture energy, Esp, for 8- to 16-mm aggregates as a function of pressure potential for the Exp. 1 and 2 soils. Lines are linear regression lines except for the Esp data from Exp. 2.

 

View this table:
[in this window]
[in a new window]
 
Table 5. Linear regression parameters for the relation between log Y and log (-{psi}) or log Esp and log (-{psi}) in Exp. 1. Values followed by the same letter for a given aggregate size are not significantly different at the P < 0.05 level.

 

View this table:
[in this window]
[in a new window]
 
Table 6. Linear regression parameters for the relation between log Y and log (-{psi}) for the Exp. 2 soils. Values followed by the same letter for a given aggregate size are not significantly different at the P < 0.05 level.

 
An increase in aggregate strength upon drying agrees with other studies (Lipiec and Tarkiewicz, 1986; Guérif, 1988; Causarano, 1993). The increase in strength with decreasing water content can be ascribed to an increase in the cohesive forces of capillary-bound water by decreased pore water pressure as described by the effective stress theory (Bishop, 1961; Snyder and Miller, 1989) and to increased effectiveness of cementing materials (Caron et al., 1992).

Snyder and Miller (1989) highlighted the importance of pore characteristics in relation to tensile failure. A strong correlation between structural porosity and soil fragmentation and tensile strength of aggregates has been found in a number of studies. Hallett et al. (1995) found that dry natural soil blocks fragmented mainly along preexisting crack surfaces. Guérif (1990) showed a strong negative correlation between macroporosity and tensile strength of dry soil. Moreover, soil water influences other soil bonding and binding agents (e.g., inter-clay forces, clay-organic matter, polysaccharides, fungal hyphaes, and roots). Cementation of dispersed clay contributes to an increase in aggregate strength as the soil dries (Caron et al., 1992).

In Exp. 1, the DFG soil displayed a smaller increase in aggregate strength with decreasing pressure potential than the CCC, although, significant differences were only found for the 4- to 8- and 8- to 16-mm aggregates (Table 5). The DFG soil had significantly lower slope values, n, for both size classes and had a higher intercept value, q, for the 8- to 16-mm aggregates. The Y and Esp data showed a very similar trend. The stronger aggregates found in the DFG soil at high water contents may be ascribed to a higher organic matter content (Table 1) as well as a higher content of biological bonding agents (e.g., polysaccharides and fungal hyphaes) (Schjønning et al., 2002). The DFG soil also had a lower amount of dispersible clay (Schjønning et al., 2002), which probably can be related to the higher content of organic matter and biological bonding agents. Cementation by dispersed clay can serve as an explanation for the larger increase in Y and Esp with decreasing pressure potential in the CCC soil than in the DFG soil. A number of workers have reported a similar marked effect of organic matter and dispersible clay on the tensile strength of dry aggregates (e.g., Kay and Dexter, 1992; Watts et al., 1996; Watts and Dexter, 1997). The study of Watts et al. (1996) supports the finding that monocultural growing of cereals in comparison with crop rotations may result in stronger dry aggregates. The weaker aggregates found in the DFG soil in dry state may also be related to the significantly higher macroporosity found for this soil in comparison with the CCC soil. For further discussion on the effect of pore characteristics on dry-aggregate strength and friability index see Munkholm et al. (2001).

Aggregates from the PAC treatment in Exp. 2 were stronger than those of the REF treatment at all pressure potentials except at -30 kPa. The marked influence of soil compaction on all size classes agrees with the aggregate density results. A strong influence of wheel traffic on tensile strength is in accordance with a number of other studies (e.g., Arvidsson and Håkansson, 1996; Watts et al., 1996; Watts and Dexter, 1998). Interestingly, no significant interaction between soil and pressure potential was found for any of the four size classes, i.e., no significant difference in n (Table 6). The difference between the soils may be related to the difference in pore characteristics. The compacted soil (PAC) exhibited very different soil pore-size distribution than the uncompacted soil (REF) as indicated by the bulk density and aggregate density results. A major effect of soil compaction was a decrease in macroporosity (Table 2).

In conclusion, an interaction between pressure potential and soil was found for the Exp. 1 soils but not for the Exp. 2 soils. The results indicate that care must be taken in evaluating soil management effects on aggregate strength properties only from measurements on air-dried or oven-dried aggregates. In Exp. 1, the contrasting long-term soil managements had affected a number soil properties influencing soil strength properties (e.g., organic matter, biological activity, clay dispersibility, bulk density, pore characteristics, and cation-exchange capacity, CEC). In contrast, the differences between the Exp. 2 soils were because of short-term effects of soil compaction with main influence on bulk density and soil-pore characteristics. This means that the effect of the contrasting soil managements on soil were much more complex in Exp. 1 than in Exp. 2.

Correlation between Aggregate-tensile Strength and Specific-Rupture Energy
The tensile strength and Esp data showed very similar trends in Exp. 1. This was, however, not the case for the Exp. 2 soils. Clear differences between the treatments were found in tensile strength but not in specific-rupture energy. When looking at the stress-strain curves, the Exp. 2 soils displayed distinctly different stress–strain relationships at all pressure potentials and for all aggregate-size classes. A significantly sharper increase in stress was found for the compacted soil. An estimate of Young's modulus can be made by measuring the gradient of the stress-strain curve at the point of failure (Y/{epsilon}) assuming linearity up to that point. The relative strain at rupture ({epsilon}) was calculated as the strain at rupture (mm) divided by the estimated aggregate diameter (mm).

As illustrated in Fig. 2 the Young's modulus, (Y/{epsilon}), was higher for the dense compacted soil at all pressure potentials. For clarity, only data for the smallest and largest aggregate-size classes are shown. The intermediary aggregate-size classes showed similar trends. This indicates that the applied stress is less effectively used for fracturing the aggregates for the REF than for the PAC soil. This may be related to the supposedly more complex pore structure for the REF soil, i.e., crack branching and interaction between the larger numbers of macropores may slow down crack propagation and also result in fracture at more points within the soil mass. Rogowski et al. (1968) also found that Y/{epsilon} was strongly positively correlated to bulk density in a study where they tested the stress–strain relationship for 2- to 8-mm aggregates from a wide range of soils. A strong increase in Y/{epsilon} with decreasing pressure potential (i.e., stronger bonding forces and less plastic deformation at low water contents), agrees with findings by Panayiotopoulos (1996).



View larger version (18K):
[in this window]
[in a new window]
 
Fig. 2. The relationship between stress and strain (Young's modulus), Y/{epsilon}, as a function of pressure potential for the Exp. 2 soils. Bars indicate ± 1 standard error of mean.

 
In conclusion, fruitful information was gained when using both Y and Esp to characterize aggregate strength. It was especially surprising that the Y and Esp data showed different trends in Exp. 2. The lack of significant differences between the differently treated soils in Esp was contrary not only to the Y data but also to the soil fragmentation data obtained by Munkholm et al. (2002) (i.e., REF fragmented more readily than PAC in a drop shatter using undisturbed bulk soil). This implies that modeling fragmentation of bulk soil from the fracture energy of single aggregates may be difficult.

Friability Indices
The DFG treated soil tended to have higher friability index than the CCC treated soil (Fig. 3a) . This was the case for calculations based on both Y (kY) and Esp (kE). However, the difference between the treatments was not significant in any case when tested in an F-test. For the Exp. 2 soils, there were no marked trends (Fig. 3b). For both experiments, kY had a maximum value between -30 and -100 kPa pressure potential. The kE values from the Exp. 1, showed a similar trend. In contrast, kE showed an increase with increasing pressure potential for the Exp. 2 soils. The increase was most marked from -30 to -100 kPa.



View larger version (25K):
[in this window]
[in a new window]
 
Fig. 3. Friability indices, kY and kE as a function of pressure potential for the Exp. 1 soils (a) and the Exp. 2 soils (b). Bars indicate ± 1 standard error of mean.

 
The finding of a maximum value between -30 and -100 kPa for kY agrees quite well with results by Utomo and Dexter (1981). They found a maximum friability for two sandy loam soils at ~ -100 kPa pressure potential, which was close to the plastic limit. Shanmuganathan and Oades (1982) also found a maximum friability at water content around the plastic limit for a remolded soil. However, in neither of these studies was the friability index compared with values determined on air-dried or oven-dried soil. Causarano (1993) determined friability index at three very different pressure potentials including air-dry soil. He found a higher friability of moist soil (-10 kPa) than of dry soil (-1.5 MPa) and air-dried soil (-82.5 MPa).

Values of kY and kE were rather poorly correlated (r2 = 0.41) (Fig. 4) . In most cases kE was larger than kY but the difference decreased with increasing values. Larger values of kE than of kY are in accordance with Perfect and Kay (1994). However, they found that kE increased linearly with kY. Contrary, to this study they determined the friability index from the spread of strength on aggregates from a specific size fraction according to Eq. [3].



View larger version (12K):
[in this window]
[in a new window]
 
Fig. 4. The relationship between the friability indices, kY and kE (all data included).

 

   CONCLUSIONS
TOP
ABSTRACT
INTRODUCTION
THEORY
MATERIALS AND METHODS
RESULTS AND DISCUSSION
CONCLUSIONS
REFERENCES
 
  1. The increase in tensile strength with increased drying was described by a power function. The specific-rupture energy also increased with increased drying and the relationship could be described by a power function in Exp. 1 soils but not in Exp. 2.
  2. In Exp. 1, a CCC soil displayed higher aggregate strength and specific-rupture energy in air-dried state and a tendency for weaker aggregates in a wet state than the diversely cropped DFG soil receiving animal manure. This may be related to a lower organic matter content and higher amount of dispersible clay for the CCC soil.
  3. In Exp. 2, a compacted soil showed higher aggregate strength than the reference treated soil, whereas no difference was found in specific-rupture energy. The compacted soil displayed a higher Young modulus, at all pressure potentials and for all the size classes tested.
  4. The interaction between aggregate-strength properties and pressure potential found in Exp. 1 indicate that care must be taken in evaluating soil-management effects on aggregate-strength properties solely from measurements on air-dried or oven-dried aggregates.
  5. No significant differences between the soils were found between the treatments in friability index. However, the friability indices scaled with pressure potential showing maximum values between -10 and -100 kPa.


   ACKNOWLEDGMENTS
 
The technical assistance of Bodil B. Christensen is gratefully acknowledged. This work was financed by the Danish Environmental Research Program and was performed under the Danish Research Centre for Organic Farming.

Received for publication October 12, 2000.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
THEORY
MATERIALS AND METHODS
RESULTS AND DISCUSSION
CONCLUSIONS
REFERENCES
 


Abstract of this Article
Figures Only for this Article
Reprint (PDF) Version of this Article
Similar articles found in:
SSSAJ Online
Search Medline for articles by:
Munkholm, L. J. || Kay, B. D.
Search Agricola for articles by:
Munkholm L. J. || Kay B. D.
Alert me when:
new articles cite this article
Download to Citation Manager
Collections under which this article appears:
Structure and Properties
Soil Physics
Water Content


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Vadose Zone Journal Journal of Environmental Quality