Published online 9 August 2007
Published in Soil Sci Soc Am J 71:1455-1459 (2007)
DOI: 10.2136/sssaj2006.0044
© 2007 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
SOIL PHYSICS
Mechanical Stresses in Soils Assessed from Bulk-Density and Penetration-Resistance Data Sets
R. Horna,*,
K. H. Hartgeb,
J. Bachmannc and
M. B. Kirkhamd
a Institute of Plant Nutrition and Soil Science, Kiel Univ. Olshausenstr. 40, Kiel, 24118 Germany
b Habichtshorst 9, Garbsen 30823, Germany
c Univ. of Hannover, Herrenhaeuserstr. 2, Hannover, 30419 Germany
d Dep. of Agronomy, 2004 Throckmorton Hall, Kansas State Univ., Manhattan, KS 66506-5501
* Corresponding author (rhorn{at}soils.uni-kiel.de).
 |
ABSTRACT
|
|---|
Measurement of soil strength with sophisticated parameters is tedious and expensive. Therefore, we developed two straightforward methods to determine this parameter down to about 80 cm, based on the classical measurements of bulk density and penetration resistance as a function of depth. They were applied to three profiles of arable Luvisols, all developed from glacial till. For each method, a procedure was worked out that allows expression of the results in terms of a normal (NC) or precompacted (PC) state. We defined the NC state as that observed in packing characteristics of virgin soils like forests and meadows, and the PC state as the packing characteristics that exist in the topsoil of agricultural soils and intensely grazed areas. Bulk density data were used to examine the packing characteristics and overburden pressures with the assumption that the horizon was in a NC state below 80 cm. For penetration resistance, we assumed a linear increase in penetration resistance with depth to represent the hydrostatic stress distribution in the NC state and deviations of measured values from this line as the PC state. The upper approximately 60 cm of all three soils were compacted, which is proofed both for the penetration resistance and for the bulk density data. For both approaches, the dimensionless coefficient of "stresses at rest," K0, was calculated following the line of thought used in engineering soil mechanics (K0 = 
x/z, where x and z are the horizontal and vertical stresses, respectively). The K0 values are highest in the precompacted soil horizons and decrease with depth.
|
INTRODUCTION
|
|---|
Compaction occurs when heavy farm machinery travels over a field. Compacted soils can make it difficult for crops to grow properly if plant roots are unable to penetrate dense soil layers except through cracks or pores. Despite the importance of soil compaction, few scientists paid attention to it in the past. Research on agricultural soils at the time of Liebig and Thaer focused more on soil chemistry than on soil physics, although Wollny (1890) was doing soil physics experiments in the 1800s. After the two world wars, chemical properties to improve plant growth and food production were of main importance. Development of methods to study soil physical properties was of minor interest and tests were not standardized. Soil deformation processes were rarely studied, probably due to the limited number of methods and the high cost of regenerating compacted soils.
Methods from civil or construction engineering have existed for a long time to measure the most important soil physical parameters, including texture, permeability, and rheological characteristics (Blanck, 1936). When the first serious damage to soil structure was recognized during the middle of the last century, specific remedial approaches for arable soils were developed. Applied methods, like the water stability test of soil aggregates, were used on individual topsoils in America and Europe. In particular, between 1955 and 1968, much research was done concerning the problem of structural deformation of soil. Special issues of handbooks that focused on methods were published, and a large number of symposia were held in Eastern and Western Europe; however, the research had no clear, scientific focus. Most activities still concentrated on the development and application of methods (DeBoodt et al., 1967; Kullmann,1968). In contrast, soil hydraulic methods were intensively investigated (e.g., Bolt and Miller, 1958), and they were focused on the newly introduced concept of soil water potential. Although development of the theoretical basis of water potential goes back to the beginning of the last century (Buckingham, 1907), apparently a lack of methods hindered further progress in experimental soil hydrology, while the theoretical state of knowledge advanced.
Now, a century after recognition of the damage done by soil compaction, the soil science community has returned to study it because of the increasing problem of compaction and soil-structural deformation. A new branch of research has been established in which the mechanical stress of soils under the temporary impact of heavy machinery is theoretically and experimentally evaluated (Drescher et al., 1988; Hartge and Sommer, 1982; Hartge, 1988; Horn, 1986, 1988; Horn et al., 2000). The increasing wetness of agricultural soils and the need to redrain large areas of agricultural land in the USA have added urgency to the studies. In the new work, earlier techniques used to loosen the subsoil or drain agricultural soil have been improved. The increased sensitivity to soil compaction under wet conditions, due to heavier machines, and the necessary increase in vibration energy to facilitate the loosening process, have resulted, however, in more weakened soils with less desirable soil structural properties throughout the whole "ameliorated" soil profile (Jayawardane and Stewart, 1994; Horn, 1994).
Soil scientists and consultants at agricultural services have not focused on deformation of the solid phase or the pore system. A reason for this lack of research might be because great effort is needed to study soil structural changes at the field scale, although similar problems have been studied by civil engineers at large construction areas or at remediation and surface-mining sites. In particular, deep chiseling or deep tillage has been commonly used. Due to destruction of aggregates and homogenization of the soil, however, physical properties often have deteriorated further with time after remediation, even when follow-up tillage has been handled carefully and expertly.
Easy-to-use methods to detect unfavorable stress situations, applicable to a wide range of agricultural soils, are essential. Therefore, we developed two straightforward methods, based on the known parameters of dry bulk density and penetration resistance, to assess soil strength. We compared the results from the two methods to derive information about the deformation of soil profiles. For the comparison, we used the "stresses at rest" coefficient (K0), which is a parameter based on early soil mechanical concepts (Terzaghi, 1943). We define the normal compression state (NC) as that observed in virgin soils like forests and meadows, and the precompacted state (PC), which exists in the topsoil of agricultural soils or intensely grazed areas. If an uncompacted state occurs in such soils, it is confined to the subsoil.
|
MATERIALS AND METHODS
|
|---|
The investigations were performed on three soil profiles in the federal state of Schleswig-Holstein (northern Germany, close to the border with Denmark). The parent material of the soils is Weichselian, calcareous, glacial till. In Profile 1, this material is buried under colluvial material from higher elevations of the site. In Profile 2, some of the Ah/p horizon had been eroded, and in Profile 3, the subsoil layers are temporarily waterlogged. All of them are Luvisols (LV) according to FAO (2002). All sites were in permanent conventional agricultural use. The crops were wheat (Triticum aestivum L.), barley (Hordeum vulgare L.), and rape (Brassica napus L. var. napus) as the usual crop rotation. The sampling was performed in early spring when the soils were at field capacity.
The first method makes use of depth-dependent changes in dry soil bulk density. This parameter was chosen because it is considered to be the most frequently determined one in soil physical field research (Walczak et al., 1997). Consequently a large number of data sets for comparative analysis of soil profiles are available worldwide. Soil core samples (200 cm3) were taken at nine depths down to 120 cm in two pits, while in the third pit only seven depth layers were cored. At each depth, three replicates were secured. Bulk density (Db) was determined after oven drying the samples at 105°C. The soil packing characteristic, i.e., the void ratio 
= [(
s/Db) – 1], where s is specific density, in relation to overburden pressure (log transformed) of the dry soil column (vertical stress, z, kPa) for sampled depths, was calculated (Hartge, 1988), and the regression lines for the NC and the PC states defined by the various slopes were obtained. Figure 1
shows a schematic plot of the soil packing characteristic. The relation of NC and PC void ratios at a given weight of the soil column, i.e., the vertical stress, is taken as a parameter of the structural state (Bachmann and Hartge, 2006).
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 1. Schematic plot of the soil packing characteristic. The linear compression line represents the greatest stress that a soil column will hold whether it has been loosened or has fresh sedimentation. It represents the normal-compacted state (NC). States diverging from this line are produced by temporary stress. They represent the precompacted state (PC).
|
|
For the second method, penetration resistance was measured at the same sites with hand-driven conical probes (Eijkelkamp Agrisearch Equipment, Giesbeek, the Netherlands; cone angle 60°; cone base area 1 cm2 for Profile 2, 2 cm2 for Profiles 1 and 3; rate of penetration 4 cm s–1). Dial readings were used. Adjustment for the different radii was not considered because evaluation was restricted to depth relations, and it is well known that a simple relation between area and resistance does not exist. These instruments are generally cheap and easy to operate. Therefore, their use is mentioned in numerous variations in most textbooks or standard-method books (Davidson, 1965; DeBoodt et al., 1967; Kullmann, 1968; Bradford, 1986). Readings were taken at the same depth layers as coring. Ten replicate probe readings down to the 100-cm depth were obtained from each profile. Stress distributions were obtained on the basis of an assumption of hydrostatic stress distribution at 120-cm soil depth, as explained by Hartge (1993).
To compare results of both methods, the term stresses at rest was chosen. It is represented by the expression
which has been proposed in the literature on soil mechanics (Bernatzik, 1947; Craig, 1978; McKyes, 1989). When using soil bulk density data, K0 was calculated from the relation between the NC and the PC cases of the packing characteristic (Bachmann and Hartge, 2006). When using penetration resistance data, the relation between an assumed hydrostatic stress and the measured probe resistance was used (Hartge and Bachmann, 2004).
|
RESULTS AND DISCUSSION
|
|---|
Results of the first method (dry bulk density vs. depth approach) are shown in Fig. 2
for the Luvisol (Profile 1). As stated, the upper part of the profile is precompacted. Table 1 shows that precompaction corresponding to a soil depth of 60 cm (which is equal to logz:2.1 [kPa]). Zones deeper than this are considered to be NC. Thus the NC regression can be obtained. The relation between the NC and the PC branch (b/a) is the analog to stresses at rest (K0) (Fig. 2).
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 2. Packing characteristic (void ratio vs. vertical gravity stress z) of Profile 1 (Luvisol). Filled symbols are measured values, showing the precompacted state for the first five values. It approximates the 60-cm depth. The line, also for Profile 1, shows the normal-compacted state, calculated by extending the linear regression of the normal-compacted state from the lowest three data points.
|
|
In Profile 2 (eroded Luvisol), precompaction is obvious down to the fifth sampling depth, which again corresponds to 60 cm (Table 1). In Profile 3 (aquic Luvisol), precompaction is indicated down to the sixth sampling depth, which also correponds to 60 cm (Table 1).
Results of the second method (penetration resistance in relation to soil depth) are shown in Fig. 3
for the same profile as in Fig. 2. To incorporate these values into the more general system, a soil profile specific reference value at the lowest point of probing was chosen, following the approach of Hartge and Bachmann (2004). At that depth, the hydrostatic state of mechanical stresses is assumed (K0 = 1), which means that stress components in all directions are similar (x = y = z). For hand-driven penetrometer equipment, the penetration resistance (PR) at that depth is usually <5 MPa. For the soils presented here, this threshold value is assumed to be lower (4 MPa; Fig. 3).
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 3. Mean penetration resistance (PR) values and their standard deviations in relation to soil depth. The straight line shows the assumed hydrostatic stress distribution, which defines the boundary conditions for the evaluation of the stresses at rest coefficient K0 and the depth of the precompression stress. Soil depths with PR values greater than this line characterize the precompacted state.
|
|
From this point on the graph, a straight line is drawn toward the origin of the coordinates at the soil surface (z = 0, PR = 0). This line is assumed to correspond to the hypothetical hydrostatic stress situation that characterizes the NC state. The measured penetrometer resistance values are now related to their corresponding values of the hydrostatic reference state, e.g., the linear regression x/z.
For both procedures, the coefficient of stresses at rest was determined by calculating either K0 from penetrometer data or the relation between for normal compaction (NC) and for precompaction (PC). The results are shown in Fig. 4
and Table 2.
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 4. Relation between the rest coefficient values obtained on the basis of bulk densities, K0(Db), (x axis) and on the basis of penetration resistance, K0(PR), (y axis) as shown in Fig. 2 and 3 for three soils derived from glacial till at multiple depths under agricultural use. The positions of the symbols reflect the different influence of moisture tension on soil strength at different bulk densities. Symbols in the normal-compaction (NC) area (K0  1) are not shown; only symbols in the precompacted (PC) area are shown.
|
|
View this table:
[in this window]
[in a new window]
|
Table 2. Linear regressions for the precompacted part of the stress-at-rest coefficient K0 originated from bulk density Db (Fig. 2) vs. K0 data originated from penetration resistance PR.
|
|
The relation between both data sets is significantly linear in all three cases (Fig. 4). The equations in Table 2 show similar slopes for Profiles 1 and 3. The lowest correlation coefficient, but still with P <0.1, is obtained for Profile 2. The differences observed for Profile 2 compared with Profiles 1 and 3 might be due to the relatively low number of depth-dependent data for dry bulk density and the different diameter of the PR cone.
From Fig. 2 and 3, it can be seen that the PC state, as represented by K0, is highest in the upper soil sections and decreases downward. The graph shows that this trend is more pronounced for the PR values. This shows an essential difference between the methods: the PR-originated values are dependent on actual in situ soil moisture content. The correlation between PR and soil moisture is, in many cases, so clear that it has been used to calibrate PR data for quick soil moisture determination in situ (Boguslavski and Lenz, 1959). But the comparison with Db–originated data shows that the moisture effect does not mask the general trend of K0 development in a profile, since the moisture content is excluded by the method—drying at 105°C. The Db–originated K0 is a parameter that is much more a structural factor, because it would be affected by soil moisture only if shrink–swell processes were considered. Such movement, however, would change the internal pore geometry more readily than the total volume of (Horn, 2004).
|
CONCLUSIONS
|
|---|
These two methods show that either soil dry bulk density or penetration resistance give information on the precompacted state of a soil profile. If we consider that these kinds of depth-dependent data sets are available worldwide, such procedures are easily applicable and can be used to define the overall strength or stress status of a site. The fact that Db and PR data sets diverge systematically is a specific effect of the approach, just as different soil pH values are obtained depending on the salt solution added to the suspension for measurement. The necessary number of depth-dependent data sets is easily obtained with a penetrometer for the PR-originated version. For the respective NC data, a linear regression can be assumed using data from a soil depth deep enough to reach a NC state. Details for this case have been published (Hartge, 2000). From the results, we conclude that simple field data sets of Db or PR might be used to characterize mechanical stress with a parameter analogous to the stresses at rest coefficient K0.
Because coefficients at rest of 1 or <1 show the NC state of stresses by definition, it is tempting to localize the depth of the PC state as brought about by agricultural trafficking. This, however, does not produce the same result with both approaches. Our assumption that it does not exceed 4 MPa was based on the limitations of normal measuring probes and it can also be assumed as the maximum value in the rooted soil volume.
|
NOTES
|
|---|
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.
Received for publication January 28, 2006.
|
REFERENCES
|
|---|
- Bachmann, J., and K.H. Hartge. 2006. Estimating soil stress distribution by using depth-dependent soil bulk density data. J. Plant Nutr. Soil Sci. 169:233–238.[CrossRef]
- Bernatzik, W. 1947. Baugrund und Physik. (In German.) Schweizer Druck- und Verlagshaus, Zürich.
- Blanck, E. 1936. Handbuch der Bodenkunde. (In German). Vol. 6. Springer Verlag, Berlin.
- Boguslavski, B.V., and K.O. Lenz. 1959. Untersuchungen über mechanische Widerstandmessungen mit einer Rammsonde auf Ackerböden. (In German with English summary.) Z. Acker- Pflanzenbau 109:33–48.
- Bolt, G.H., and R.D. Miller. 1958. Calculation of total and component potentials of water in soil. Trans. Am. Geophys. Union 39:917–928.
- Bradford, J.M. 1986. Penetrability. p. 463–478. In A. Klute (ed.) Methods of soil analysis. Part 1. Physical and mineralogical methods. 2nd ed. SSSA Book Ser. 5. SSSA, Madison, WI.
- Buckingham, E. 1907. Studies on the movement of soil moisture. Bur. of Soils Bull. 38. USDA, Washington, DC.
- Craig, R.F. 1978. Soil mechanics. 2nd ed. Van Nostrand-Reinhold, New York.
- Davidson, D.T. 1965. Penetrometer measurements. p. 472–484. In C.A. Black et al. (ed.) Methods of soil analysis. Part 1. Physical and mineralogical properties, including statistics of measurement and sampling. Agron. Monogr. 9. ASA and SSSA, Madison, WI.
- DeBoodt, M., H. Frese, A.J. Low, and P.K. Peerlkamp (ed.) 1967. West European methods for soil structure determination. Faculty of Agric. Sci., Ghent, Belgium.
- Drescher, J.R., R. Horn, and M. DeBoodt (ed.). 1988. Impact of water and external forces on soil structure. Catena Suppl. 11. Catena Verlag, Reiskirchen, Germany.
- FAO. 2002. World reference base for soil resources. Available at www.fao.org/ag/agl/agll/wrb/ (verified 16 May 2007). FAO, Rome.
- Hartge, K.H. 1988. The reference base for compaction state of soils. p. 73–77. In J.R. Drescher et al. (ed.) Impact of water and external forces on soil structure. Catena Suppl. 11. Catena Verlag, Reiskirchen, Germany.
- Hartge, K.H. 1993. Stress distribution in the solid phase of soils. Soil Technol. 6:83–87.
- Hartge, K.H. 2000. The effect of soil deformation on physical soil properties—A discourse on the common background. Adv. Geoecol. 32:32–43.
- Hartge, K.H., and J. Bachmann. 2004. Ermittlung des Spannungszustandes von Böden aus Werten des Eindringwiderstandes von Sonden. (In German with English summary.) J. Plant Nutr. Soil Sci. 167:304–308.
- Hartge, K.H., and R. Horn. 1992. Soil physical methods. (In German.) 3rd ed. Enke Publ., Stuttgart, Germany.
- Hartge, K.H., and C. Sommer. 1982. Effect of soil formation on packing and on vertical stresses in soils. J. Plant Nutr. Soil Sci. 145:25–35.[CrossRef]
- Horn, R. 1986. Auswirkungen unterschiedlicher Bodenbearbeitung auf die mechanische Belastbarkeit von Ackerböden (In German with English summary.) J. Plant Nutr. Soil Sci. 149:9–18.[CrossRef]
- Horn, R. 1988. Compressibility of arable land. p. 53–71. In J.R. Drescher et al. (ed.) Impact of water and external forces on soil structure. Catena Suppl. 11. Catena Verlag, Reiskirchen, Germany.
- Horn, R. 1994. Stress transmission and recompaction in tilled and segmently disturbed soils under trafficking. p. 53–87. In N. Jayawardane and B.A. Stewart (ed.) Subsoil management techniques. Adv. Soil Sci. CRC Press, Boca Raton, FL.
- Horn, R. 2004. The dependence of soil mechanical properties and pore functions for arable soils. Soil Sci. Soc. Am. J. 68:1131–1137.[Abstract/Free Full Text]
- Horn, R., J.J.H. van den Akker, and J. Arvidsson (ed.). 2000. Subsoil compaction: Distribution, processes and consequences. Adv. Geoecol. 32. Catena Verlag, Reiskirchen, Germany.
- Jayawardane, N., and B.A. Stewart (ed.). 1994. Subsoil management techniques. Adv. Soil Sci. CRC Press, Boca Raton, FL.
- Kullmann, A. (ed.). 1968. Untersuchungsmethoden des Bodenzustandes. (In German.) Deutsch. Landw. Verlag, Berlin.
- McKyes, A. 1989. Agricultural engineering soil mechanics. Dev. Agric. Eng. 10. Elsevier, Amsterdam.
- Terzaghi, K. 1943. Theoretical soil mechanics. John Wiley & Sons, New York.
- Walczak, R.T., B. Witkowska-Walczak, and P. Baranowski. 1997. Soil structure parameters in models of crop growth and yield production: Physical submodels. Int. Agrophys. 11:111–128.
- Wollny, E. 1890. Physik des Bodens: Untersuchungen über die Beeinflussung der Fruchtbarkeit der Ackerkrume durch die Tätigkeit der Regenwürmer. (In German.) Forsch. Geb. Agriculturphys. 13:381–394.
TOP
ABSTRACT
INTRODUCTION
