Soil Science Society of America Journal 65:1400-1403 (2001)
© 2001 Soil Science Society of America
DIVISION S-1 - NOTES
Measurement of soil aggregate density by volume displacement in two non-mixing liquids
Guillermo O. Sarlia,
Roberto R. Filgueiraa and
Daniel Giménez*,b
a Facultad de Ciencias Agrarias, Univ. Nacional de La Plata, Calles 60 y 119, CC 31, 1900 La Plata, Argentina
b Dep. of Environmental Sciences, Rutgers, The State Univ. of New Jersey, 14 College Farm Road, New Brunswick, NJ 08901-8551
* Corresponding author (gimenez{at}envsci.rutgers.edu)
 |
ABSTRACT
|
|---|
Soil aggregate density is an important property influencing soil biological, chemical, and physical processes. Current methods used to estimate soil aggregate density are based on more or less restrictive assumptions or require specialized equipment. This study was conducted to develop an alternative method for measuring aggregate density of soil and other porous objects that is both simple and theoretically sound. The proposed method solves the balance of forces resulting when a kerosene-saturated aggregate is immersed in a mix of water and glycerin. We tested the accuracy of the method by measuring the volume of 40 aggregates with diameters ranging from 4 to 20 mm and with variable densities. Soil aggregates were saturated in kerosene, drained at a tension of -30 mm, and their volume estimated using a pycnometer. Upon resaturation, aggregates were suspended from a thread and successively weighed in air, in kerosene, and in a mix of water and glycerin. Aggregate volumes and densities varied between 1.20 x 10-7 and 3.85 x 10-6 m3 and between 1.05 to 1.86 Mg m-3, respectively. On average, aggregate volumes estimated with the pycnometric method were 2.6% smaller than the volumes obtained with the proposed technique, suggesting that the new method is less aggressive in evacuating pores open to the surface than a tension of -30 mm. The new method was easier to use, did not require previous preparation of the sample, and was less time-consuming than older methods.
|
INTRODUCTION
|
|---|
SOIL BULK DENSITY is defined as the ratio of the mass of dry solids to the bulk volume of the soil (Blake and Hartge, 1986). The bulk volume includes both the volumes of the solid phase and the pore space. A common technique to determine bulk density is to measure the mass of soil contained in a core of known volume (core method). A bulk density determined with the core method includes the effect of both the intra- and inter-aggregate pore spaces.
Soil aggregate properties are usually different from bulk soil properties (Horn, 1990; Santos et al., 1997). Swelling and shrinking and wetting and drying processes result in a concentration of clay, microorganisms, and nutrients in the outer layers of aggregates. Aggregate size and density are identified as two of the most important properties of soil aggregates. Aggregate density is used to estimate inter-aggregate porosity, and, together with aggregate size, influence plant growth (De Freitas et al., 1996). During the last decade, fractal models of soil structure have been developed that use variations in aggregate density with aggregate size as a fundamental scaling property related to soil water retention and other soil properties (Rieu and Sposito, 1991). There is an interest in methods to determine density of soil aggregates, and several methods have been proposed.
Some methods use a group of aggregates to determine an average aggregate density. In this group is the method proposed by Chepil (1950), which was successfully used to determine the density of aggregate size fractions separated by sieving dry soil (Eghball et al., 1993; Filgueira et al., 1999a, 1999b). The method, however, assumes that aggregates and sand grains of similar size pack in the same way. This assumption is likely to introduce error in the determination of the density of irregularly-shaped aggregates. Voorhees et al. (1966) circumvented this problem by using a mixture of glass beads and aggregates packed to a known volume to determine aggregate density. The same principle was applied by Olson and Zobeck (1989), who filled a chamber of known volume with soil aggregates and applied mercury to fill the inter-aggregate pore space and obtain aggregate volume as the difference between the chamber volume and that of the applied mercury. Sometimes it is important to determine aggregate density on individual aggregates.
Typically, techniques suitable to measure density of individual aggregates somehow seal the aggregate and determine its volume by submerging it in a fluid and either measuring volume displacement or change in weight with respect to its weight in air. An exception to this general procedure is the use of gamma-ray attenuation to determine aggregate density (Benjamin and Cruse, 1985). Brasher et al. (1966) proposed the use of saran resin dissolved in methyl ethyl ketone to coat soil clods prior to submersion in water. Clods with large pores are sometimes difficult to seal and several coats of a more viscous saran solution are needed. Measurement errors with the method increase when clods of 
<40 g are used (Blake and Hartge, 1986). Organic fluids are commonly used to saturate an aggregate. McIntyre and Stirk (1954) measured the displacement of kerosene-saturated aggregates submerged in kerosene. Prior to submersion, aggregates were subjected to -30 mm tension to remove the kerosene from the aggregate surface. Ross and Prebble (1989) submerged dry aggregates in hexane (a highly volatile and extremely dangerous organic fluid) and measured their weight. After a short drainage period, in which hexane is removed from the surface by evaporation, the weight in air of hexane-saturated aggregates is recorded. Hallett et al. (1998) modified the method of Ross and Prebble (1989) by replacing hexane by xylene. The advantage of the method proposed by Ross and Prebble (1989) is that measurements of weight are simpler and potentially more accurate than volume measurements. The method, however, implicitly assumes that the organic fluid fills completely the pores. Errors are introduced when this is not the case.
The objective of this work was to develop an alternative displacement method, using two non-mixing liquids, to estimate the bulk density of porous objects. The new method was tested with soil aggregates covering a range of soil volumes and bulk densities.
|
Materials and Methods
|
|---|
Theory
The principle of the method proposed in this paper is based on solving the balance of forces on an aggregate when immersed in two liquids (Fig. 1a, 1b)
. In general, any combination of two non-mixing liquids with different densities can be used. We worked with kerosene and distilled water, and kerosene and water mixed with glycerin in different proportions. Finally, kerosene and a mixture of 50% water and 50% (v/v) glycerin proved to be the best for the procedure.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 1. Force diagram applied on a porous object submerged in either kerosene (a) or in a water-glycerin mixture (b). The symbols are F = force, W = weight, and B = buoyancy. The object contains variable amounts of solid (-S), pore-filling liquid (kerosene, -K) and trapped air (-A). See text for explanation.
|
|
From the force diagrams for a dry porous object submerged in kerosene (Fig. 1a) and in water-glycerin mixture (Fig. 1b) we obtained:
 | (1) |
 | (2) |
where F is the force exerted by the electronic balance on a kerosene-saturated aggregate submerged in either kerosene (FK, Fig. 1a) or in the water-glycerin mixture (FMix, Fig. 1b); W represents weight of either the porous solid in air (WS), the pore-filling kerosene (WK), or the entrapped air (WA); and B is the buoyancy of the solid volume (BS), the pore-filling kerosene (BK), and of the entrapped air (BA), in both kerosene (BSK, BKK, BAK, respectively) and in the water-glycerin mixture (BSMix, BKMix, and BAMix, respectively). Subtracting Eq. [2] from Eq. [1] and expressing a buoyancy force by the combination of the volume and density of a constituent and the gravity acceleration, g, we obtain:
 | (3) |
where 
K and Mix are the densities of kerosene and the water-glycerin mixture, respectively, and V is volume of the solid phase (VS), of the pore-filling kerosene (VK), and of the entrapped air in the pores (VA). Finally, we have that the bulk volume of a porous object, Vbulk, is
 | (4) |
and its aggregate density, bulk, can be expressed as:
 | (5) |
where mS represents the mass of the solid.
Experimental Procedure
The measurement procedure involves suspending an aggregate from the bottom of a balance (Fig. 2)
, by means of a thin synthetic thread, and recording its dry weight in air, the weight when submerged in kerosene and when submerged in the water-glycerin mixture, respectively. The aggregate density is then estimated by Eq. [5] using the appropriate liquid densities. The densities of the kerosene and the water-glycerin mixture, measured with a Mohr-Westphal balance, were K = 0.81 Mg m-3 and Mix = 1.14 Mg m-3, respectively. It is important to note that the proposed technique is independent of the relative proportion of the components of a bulk volume (i.e., solid, air, and water). In other words, it does not assume that an aggregate is completely saturated with kerosene. When placed in the same container, kerosene and the water-glycerin mix formed a two-layer system because both liquids are immiscible and have different densities (Fig. 2). Typically, soil aggregates were weighed in air (WS) and after being submerged in kerosene (FK). The container with non-mixing liquids was then slowly raised until the aggregate was completely submerged in the water-glycerin mixture. Some excess of kerosene could be detached from the surface of the clod and went up by buoyancy. At this point, we assumed that a kerosene film coated completely the surface of the aggregate, and the apparent new weight (FMix) was read. We also applied this technique by vacuum-saturating aggregates with kerosene, and also by saturating them at atmospheric pressure, prior to submerging them in the two non-mixing liquids, with similar results.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 2. Diagram of the experimental setup to measure bulk volume of a porous object using two non-mixing liquids, which in this work were kerosene and a mixture of water and glycerin.
|
|
Soil Aggregates
Soil cores (0.1 m in diameter and 0.40 m long) were sampled from the Experimental Farm of the School of Agronomy and Forestry Engineering (34°54' S and 57°57' W) of the National University of La Plata, Argentina. The soil is a Typic Argiudoll. Soil cores were air-dried at room temperature (25°C). Soil aggregates were obtained from both ends of an air-dried soil core (corresponding to depths of 0–0.1 m and 0.3–0.4 m, respectively) to maximize the probability of having differences in soil aggregate density. The textural composition of the surface soil (0–0.1 m) was 23% clay, 50% silt, 27% sand, and changing in depth (0.3–0.4 m) to 31% clay, 48% silt, and 19% sand. Dry soil samples were sieved mechanically in a nest of sieves having mesh openings of 16, 8, and 4 mm. A total of forty aggregates with diameters ranging between 4 and 20 mm were used in the experiments.
Prior to measuring aggregate density with the proposed technique, soil aggregate density was measured using a modified kerosene method (McIntyre and Stirk, 1954). We used a 25-mL pycnometer with a wide neck that enabled aggregates to pass through it. The weights of a clean and empty pycnometer (WP) and of the same pycnometer filled with kerosene (WP+K) were recorded. Aggregates were saturated with kerosene, subsequently drained on a sand table to a tension of -30 mm for 15 min. Kerosene-saturated aggregates were introduced in the pycnometer, and weighed again (WP+SK). The weight of the pycnometer plus an aggregate and kerosene (WP+SK+K) was obtained by filling the pycnometer, inserting the stopper and carefully removing any kerosene excess with tissue paper. The bulk volume of an aggregate was calculated using the following equation:
 | (6) |
|
Results and Discussion
|
|---|
Volume of soil aggregates measured with the two non-mixing liquids and the modified pycnometer methods varied between 1.2 x 10-7 m3 and 3.85 x 10-6 m3. Volumes obtained with both methods were highly correlated (Fig. 3a)
. The residuals of the linear fit showed no trend with aggregate volumes (Fig. 3b). The average soil aggregate volume estimated with the two non-mixing liquids method (1.51 x 10-6 m3) was, however, significantly higher (P > 0.001) than the average soil aggregate volume estimated with the modified kerosene method (1.47 x 10-6 m3), according to a paired comparison analysis with n = 40 (Steel and Torrie, 1980). Ross and Prebble (1989) also used a tension of -30 mm to measure the volume of soil aggregates using the kerosene method. Their results from the kerosene method were slightly (0.6%), but systematically, higher than those of the hexane displacement method. In our measurements, the modified (pycnometric) kerosene method resulted in volumes that were on an average 2.6% smaller that the obtained with the two non-mixing liquids measurement.
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 3. A comparison between bulk volumes of soil aggregates measured with a modified pycnometric method with kerosene, and with the proposed technique using two non-mixing liquids (a), and a plot of the residuals of the fitting procedure (b). The symbols represent three ranges of soil aggregate densities covered. Values on axes equal reported values times the indicated factor.
|
|
The discrepancy in values of soil aggregate volumes obtained with different methods is a function of how extensively the saturating liquid is displaced from an aggregate surface and of the amount of large pores open to the surface. The former effect is a function of the procedure used to displace the liquid (e.g., suction vs. evaporation), whereas the latter depends on soil properties. Olson and Zobeck (1989) showed that when mercury was applied at a pressure of 110 kPa to a container with aggregates, it partially filled intraaggregate pore space resulting in higher aggregate density than when the same aggregates were measured with a mercury pressure of 14 kPa. Thus, the physical limit of an aggregate cannot be unambiguously defined. McIntyre and Stirk (1954) applied suction to kerosene-saturated aggregates and estimated bulk volume by kerosene displacement. They concluded that, for the soils they worked with, tensions of -30 mm or lower resulted in a constant volume and hence a constant aggregate density. There is no physical reason to generalize such finding to other soils or to soils subjected to different soil management. In fact, one would expect that larger suctions would empty increasingly smaller pores resulting in a continuous variation of soil volume, which should be a function of an aggregate pore distribution.
|
Conclusions
|
|---|
The technique proposed in this paper, using Archimedes's principle with two non-mixing liquids, provided similar volumes of soil aggregates of various sizes and bulk densities, when compared with volumes obtained with a modified kerosene method. Our results showed that, when used with 50% water and 50% glycerin, the technique removed less kerosene from the surface of soil aggregates than a -30 mm tension. This result, however, may be different for other soils or different proportion of the two components in the water and glycerin mixture. The proposed method can be used irrespective of the fluid filling the pores (i.e., air or water). It is easy to implement, can be adapted to measure several aggregates supported by a mesh, and requires no special care regarding toxicity or flammability.
|
ACKNOWLEDGMENTS
|
|---|
The authors want to acknowledge the laboratory assistance provided by Janice Karmon (Rutgers University), and the economical support from the Universidad Nacional de La Plata, the Fundación Campodónico (La Plata, Argentina), and the New Jersey Agricultural Experiment Station.
|
NOTES
|
|---|
Roberto R. Filgueira, Researcher of Consejo Nacional de Investigaciones Científicas y Técnicas, Argentina.
Received for publication September 28, 2000.
|
REFERENCES
|
|---|
- Benjamin, J.G., and R.M. Cruse. 1985. Measurement of shear strength and bulk density of soil aggregates. Soil Sci. Soc. Am. J. 49:1248–1251.[Abstract/Free Full Text]
- Blake, C.R., and K.H. Hartge. 1986. Bulk density. p. 363–375. In A. Klute (ed.) Methods of soil analysis. Part 1, 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison. WI.
- Brasher, B.R., D.P. Franzmeier, V. Valassis, and S.E. Davison. 1966. Use of saran resin to coat natural soil clods for bulk-density and water-retention measurements. Soil Sci. 101:108.
- Chepil, W.S. 1950. Methods of estimating apparent density of discrete soil grains and aggregates. Soil Sci. 70:351–362.
- De Freitas, P.L., R.W. Zobel, and V.A. Snyder. 1996. A method for studying the effects of soil aggregate and density. Soil Sci. Soc. Am. J. 60:288–290.[Abstract/Free Full Text]
- Eghball, B., L.N. Mielke, G.A. Calvo, and W.W. Wilhelm. 1993. Fractal description of soil fragmentation for various tillage methods and crop sequences. Soil. Sci. Soc. Am. J. 57:1337–1341.[Abstract/Free Full Text]
- Filgueira, R.R., L.L. Fournier, G.O. Sarli, A. Aragón, and W.J. Rawls. 1999a. Sensitivity of fractal parameters of soil aggregates to differences in management practices. Soil Tillage Res. 52:217–222.
- Filgueira, R.R., Ya.A. Pachepsky, L.L. Fournier, G.O. Sarli, and A. Aragón. 1999b. Comparison of fractal dimensions estimated from aggregate mass-size distribution and water retention scaling. Soil Sci. 164:217–223.
- Hallett, P.D., N.R.A. Bird, A.R. Dexter, and J.P.G. Seville. 1998. Investigation into the fractal scaling of the structure and strength of soil aggregates. Eur. J. Soil Sci. 49:203–211.
- Horn, R. 1990. Aggregate characterization as compared to soil bulk properties. Soil Tillage Res. 17:265–289.
- McIntyre, D.S., and G.B. Stirk. 1954. A method for determination of apparent density of soil aggregates. Aust. J. Agric. Res. 5:291–296.
- Olson, K.R., and T.M. Zobeck. 1989. Improved mercury-displacement method to measure the density of soil aggregates. Soil Sci. 147:71–75.
- Rieu, M., and G. Sposito. 1991. Fractal fragmentation, soil porosity, and soil water properties: I. Theory. Soil Sci. Soc. Am. J. 55: 1231–1238.[Abstract/Free Full Text]
- Ross, P.J., and R.E. Prebble. 1989. Non-destructive measurement of soil clod volume using hexane displacement. Aust. J. Soil Res. 37: 39–44.
- Santos, D., S.L.S. Murphy, H. Taubner, A.J.M. Smuker, and R. Horn. 1997. Uniform separation of concentric surface layers from soil aggregates. Soil Sci. Soc. Am. J. 61:720–724.[Abstract/Free Full Text]
- Steel, R.G.D., and J.H. Torrie. 1980. Principles and procedures of statistics: A biometrical approach. 2nd ed. McGraw-Hill, New York.
- Voorhees, W.B., R.R. Allmaras, and W.E. Larson. 1966. Porosity of surface soil aggregates at various moisture contents. Soil Sci. Soc. Am. Proc. 30:163–167.
TOP
ABSTRACT
INTRODUCTION
