Published in Soil Sci. Soc. Am. J. 68:7-16 (2004).
© 2004 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
DIVISION S-1—SOIL PHYSICS
Effects of Compaction on the Acoustic Velocity in Soils
Zhiqu Lu*,
Craig J. Hickey and
James M. Sabatier
National Center for Physical Acoustics, Univ. of Mississippi, 1 Coliseum Drive, University, MS 38677
* Corresponding author (zhiqulu{at}olemiss.edu).
 |
ABSTRACT
|
|---|
Soil compaction induced by the use of agricultural machinery changes the mechanical and structural properties of soils. Assessment of its effects may be made by acoustic techniques. In this study, a triaxial cell modified to measure the acoustic velocity was used to simulate the compaction process. Unconsolidated-undrained triaxial tests were performed on two air-dry remolded soils and one undisturbed field soil taken from sites in Sharkey, Neshoba, and Marshall Counties, Mississippi. Both the deformation and the acoustic behaviors of the soils were studied during a compaction process. It was found that the acoustic velocity and the deviator stress behaved similarly. Both the acoustic velocity and the deviator stress increased linearly at the early stage of compaction and they changed nonlinearly with intermediate compaction. At the extreme case where the soil was compacted to failure, the acoustic velocity and the deviator stress changed only slightly with further compaction. During the unload-reload cycle, the acoustic velocity and the deviator stress varied steeply and presented hysteretic and load-history-dependent properties. Significant influence of water content on both acoustic and deformation behaviors was found. The similarity between the acoustic behavior and the deformation characteristics make the acoustic velocity a promising parameter for monitoring the ongoing compaction process in situ and for long-term soil survey.
Abbreviations: LVDT, linear variable differential transformer
|
INTRODUCTION
|
|---|
SOIL COMPACTION problems induced by intensive use of powerful machines in agricultural land have adverse effects on crop growth and yield (Voorhees, 1991). Compaction increases the bulk density and reduces the porosity of the soil. Increased mechanical impedance (or penetrometer resistance) inhibits seed germination and hampers root growth (Bengough and Mullins, 1991). Low porosity gives rise to insufficient aeration, reduction of water intake, and poor nutrient transport (Grable and Siemer, 1968). Heavy load decreases the permeability and capability of drainage and results in increased possibilities of surface runoff and erosion (Germann, 2002). Compaction may deteriorate the self-remediation ability of soil and also affect the subsurface soils where a plow pan may develop (Laird, 1998). An assessment of the influence of compaction on soil physical properties is necessary in agricultural research and it usually involves the monitoring of variations in soil structure, bulk density, porosity, water content, air permeability, pore size distribution, and so on under loading and unloading.
Many conventional techniques for measuring soil physical properties in situ and in the lab exist. Such techniques include the gravimetric technique (Hillel, 1982), porous plate, sand table (Van der Haarst and Stakman, 1965), pressure membrane cell (Avery and Bascombe, 1974), air pycnometer (Pidgeon, 1974), porosimeter (Janse, 1969), permeameter (Grover, 1955), and penetrometer (Carlos and Hopmans, 2001). These methods have been routinely used in soil science to measure soil physical properties such as water content, degree of saturation, density, porosity, permeability. However, these techniques are generally destructive, labor intensive, and slow.
In recent years, acoustic techniques for in situ soil characterization have been researched (Don and Cramond, 1985; Hunter et al., 1989; Hess et al., 1990; Miller et al., 1990; Sabatier et al., 1990, 1993, 1996; Frederickson et al., 1996; Baker et al., 1999; Jarvis and Knight, 2000; Flammer et al., 2001; Oelze et al., 2001). These methods consist of either reflection or transmission measurements with acoustic-to-seismic coupling or seismic techniques. The quantitative information of surface and near surface of soil regarding impedance, porosity, air permeability, tortuosity, surface roughness, layering, and water flow can be determined. Compared with the conventional techniques mentioned above, the acoustic methods are rapid, nondestructive, can be applied to a large area of soil, and might be explored for in situ, real-time, and long-term monitoring of the variations of soil physical properties due to compaction.
Clearly, a basic understanding of the acoustic wave propagation in a soil would be helpful before any practical application of these acoustic techniques could be used to monitor the soil compaction process. For a porous material, Biot's theory (Biot, 1956a, b) predicted the existence of three types of waves: two compressional (or longitudinal) waves, the fast P-wave and the slow P-wave; and one shear (or transverse) wave, the S-wave. The slow P-wave is generally difficult to observe because of its high attenuation nature and is hence neglected in an ordinary study.
If the soil is treated as an effective solid continuum, the compressional wave velocity Vp and the shear wave velocity Vs can be expressed in the form as follows:
 | [1] |
and
where B and G denote the effective bulk and shear moduli, and 
b the bulk density of soil.
A soil consists of a granular skeleton and pore fluids (usually water and air). The above expressions can be rewritten as (source from Santamarina et al., 2001):
 | [2] |
and
where Bsk and Gsk are the bulk and shear moduli of the skeleton, and Bsus denotes the bulk modulus of a suspension that is a mixture of the fluids and soil particles without contacting each other.
The bulk and the shear moduli of the skeleton are complex functions that depend on the conditions of interparticle forces and local particle contact interactions (Santamarina et al., 2001). In the case of a real soil, where partial saturation is expected, the capillary forces between the contact (their macroscopic effect can be assessed by the water content) and the external force transmitted through the skeleton are believed to be the major factors that govern the changes of the elastic moduli of solid skeleton.
The contribution to the bulk modulus of the suspension is given by (source, Santamarina et al., 2001):
 | [3] |
where Bw, Ba, and Bg denote the bulk moduli of water, air, and the material that makes the grains, respectively; n denotes the porosity, and S stands for the degree of saturation.
The density of a fluid-filled soil is:
 | [4] |
where w and g are the densities of water and the material that makes the grains; the density of air is neglected.
Compaction induced by passing wheeled and tracked vehicles is a physical process that exerts pressure on the soil surface and reduces the volume of the soil. Because soil particle and water within the soil are relatively incompressible, this process causes reorientation of soil particles and soil fabric change. The variations of the porosity and the degree of saturation (also water content) affect the bulk modulus of the suspension and the density of the soil through Eq. [3] and Eq. [4]. The variation in water content also affects the capillary forces between contacts and results in changes in the elastic moduli of the granular skeleton. Moreover, surface pressure can propagate through the skeleton into the soil 
60 cm deep below the surface (Söhne, 1958). The increased pressure causes a significant increment in the elastic moduli of the skeleton. Consequently, compaction significantly affects the acoustic velocity through Eq. [2], and on the other hand, the variation of the acoustic velocity will reflect the ongoing process of compaction.
On the basis of the above knowledge, a study of the effects of compaction on the acoustic velocity in soils was conducted with a conventional triaxial cell apparatus. In this study, the device was modified to measure the velocity of a compressional wave propagating through a soil sample during triaxial testing. Three soil samples taken from sites in Sharkey, Neshoba, and Marshall Counties, Mississippi, were chosen for triaxial tests. Soils were compacted vertically to simulate a compaction process. The compressional wave velocity in the axial direction was measured along with the measurement of the stress-strain response during both loading and unloading. A comparison between the acoustic velocity-strain behavior and the deviator stress-strain response was made. A discussion of the relation between the acoustic velocity and the compaction was addressed.
|
SOIL DESCRIPTIONS AND PREPARATIONS
|
|---|
Three kinds of soils were used for testing. They were Sharkey silty clay (very-fine, smectitic, thermic Chromic Epiaquerts), Neshoba soil (fine, mixed, thermic Rhodic Paleudults), and Marshall soil (fine, mixed, active, thermic Oxyaquic Fraglossudalfs), and were obtained from sites in Sharkey, Neshoba, and Marshall Counties, Mississippi. The Sharkey clay is brown in color and contains 64% clay, 34% silt, and 2% sand. The Neshoba soil is red in color and is a sandy loam to sandy clay loam texture (Ruston Series) commonly found in the southeastern USA. Soil size analysis indicated that it is composed of 20% clay, 2.9% silt, and 77.1% sand. The Marshall soil is brown in color and has components of 22% clay, 75.1% silt, and 2.9% sand.
Both Sharkey clay and Neshoba soil were prepared as remolded soils. They were crumbled, sieved, and put into an oven at 110°C for 12 h to remove the water content within the soils. Since negligible moisture could be absorbed during the specimen processing, the water content was believed to be near zero, an air-dry condition. According to the standard procedure of the ASTM International (1999), a certain amount of soil was divided into six parts. With the aid of a funnel, each part of the soil was rained successively into the pedestal of the triaxial cell, which was encompassed by a rubber membrane. The soil specimen was compacted in six layers by tamping each layer uniformly with each layer having approximately equal thickness.
The Marshall soils were taken in the field 10 cm below the ground surface by an AMS (Art's Manufacturing and Supply Inc.) split core sampler (Forestry Suppliers Inc., Jackson, MS). The cores were transported to the lab where they were extruded from the sampler and trimmed to form a cylindrical shape with the ends perpendicular to the longitudinal axis of the specimen. During processing, the specimens were handled with caution to minimize disturbance. The undisturbed core was then placed on the base pedestal of the triaxial cell, sealed with the rubber membrane, and subjected to triaxial testing.
The height and the diameter of the soil specimens were measured before the triaxial test. For remolded soils, the mass of the specimen was measured before the triaxial test. For the undisturbed Marshall soil, the soil sample was removed from the test chamber immediately after the triaxial test and weighed. It was then placed in an oven at 110°C overnight to remove water content. The dried soil was weighed again and the difference between the two weights was used to determine the water content of the specimen.
The initial physical properties of soil specimens were calculated and tabulated in Table 1. A value of 2.65 g cm–3 for the density of the materials that form the solid fraction was used for the calculation. As can be seen from Table 1, variations in properties of the same kind of remolded soil do exist among separate preparations, although efforts have been made to obtain identical specimens. Significant differences in physical properties were found for the field soil samples. However, it should be noted that the sample with a water content of 29.0% was taken from the field one month before a second sampling. The other three samples taken from the second sampling were tested in a period of 9 d. Some moisture losses were expected to occur during this period of time. It will be shown later that the variation in water content could significantly affect the acoustic behavior as well as the deformation characteristics of the soil.
|
MATERIALS AND METHODS
|
|---|
The triaxial cell is a principal laboratory shear device mainly used in geotechnical engineering, civil engineering, and soil science to determine the shear strength and stress-deformation behaviors of a soil. Recently, there are growing interests in using the triaxial cell to study the dynamic soil properties by acoustic velocity measurements (Bates, 1989; Viggiani and Atkinson, 1995; Nakagawa et al., 1996, 1997; Gajo et al., 1997; Lu et al., 2002). The triaxial cell provides versatile methods of drainage (UU test, CU test, and CD test; see below), loadings (compression, extension, K0, and cyclic tests), and pressure controls (cell pressure, pore pressure, and axial stress). These capabilities make the triaxial cell appropriate for simulating a soil compaction process. Since a routine triaxial test is not the purpose of the study, a detailed description of its principle and operation is beyond the scope of the paper. Interested readers can refer to the relevant books (Bishop and Henkel, 1962; Donaghe et al., 1988).
The triaxial cell was a Bishop and Wesley hydraulic cell which is shown in the schematic diagram of Fig. 1
. The cylindrical soil specimen shown in Fig. 2
was approximately 5.0 cm in diameter, 6.5 cm in height, and was placed on a pedestal inside the plastic confining cylinder cell that was filled with degassed water. The initial 38-mm-diam. pedestal was modified to a 50-mm-diam. pedestal to match the size of the field soil core. A watertight rubber membrane surrounded the test specimen and was sealed against the base pedestal and the top cap with four O-rings. The axial load was applied by increasing the pressure in the ram pressure chamber. This pressure pushed the loading ram upwards, thereby vertically compacting the soil specimen. A linear variable differential transformer (LVDT) served as a displacement transducer by which the axial deformation, and hence, the axial strain were measured. An internal channel in the base plate that connected to the inside of the cell served to measure the cell pressure with a cell pressure transducer. A load transducer that was mounted on the top plate measured the force transmitted through the sample. An inlet located in the center of the pedestal passed pore pressure through a 1-mm i.d. polythene tube to a pore pressure transducer. A filter paper and two pieces of porous discs that enclosed the soil sample were inserted in place to facilitate the pore pressure measurement. For the field soil core testing, a thin layer of silicon grease was applied between the acoustic transducer and the tested soil to increase the acoustic coupling.
The test cell was modified to measure the compressional wave velocity during the triaxial test. Research is under way to add the capability of measuring shear wave velocity. As shown in Fig. 2, the acoustic cell contained three piezoelectric discs (Morgan Matroc, Inc., Bedford, OH; 20-mm diam., 0.4-mm thickness, and working in a bender mode): Disc 1 served as a transmitter, Disc 2 was used as a reference, and Disc 3 was used to receive the signal once it had traveled through the sample. In a traditional time-of-flight method for measuring the acoustic velocity, one measures the time delay between two subsequent signals, that is, the direct arrival pulse and the first echo. However, because of the high absorption, only the direct signal can be received in this study. The travel time was determined as the difference between the direct arrival and the reference. To obtain an adequate reference, a specially designed acoustic transducer was made. Disc 1 and Disc 2 were glued with a conductive epoxy onto both sides of a brass plate. A brass cap was machined to form a geometry that allowed the discs to be accommodated into the cap. The brass cap had enough space for the inner disc (Disc 1) to vibrate freely, and it was flushed with epoxy at the bottom. The transducer in this configuration was found to have a final resonant frequency of 6.5 kHz when in contact with soil. The receiver had a similar configuration to the transmitter.
Acoustic travel times were measured as follows. Disc 1 was excited by a function generator (Agilent 33120A, Function/Arbitrary Waveform Generator, Palo Alto, CA). It generated a 1-cycle 6.5-kHz tone burst signal with initial phase of 90° (to get the typical waveform as shown in Fig. 3)
and repetitive frequency of 20 Hz. A compressional wave generated by Disc 1 vibrated the brass plate plane perpendicularly and propagated through the soil sample along the axial direction. The vibration was detected immediately by Disc 2 and used as a reference signal. The acoustic wave was received by the receiver (Disc 3). The electrical, the reference, and the received signals were amplified by three preamplifiers (SRS 560 low noise amplifier, Standford Research System, Sunnyvale, CA) and displayed by an oscilloscope (Infiniium, Agilent) with average number of 256. The typical signals are shown in Fig. 3, which was a hardcopy from the oscilloscope screen.
The travel time of the acoustic wave (
t) was obtained by measuring the time delay between the maxima of the first peaks of the reference and the received signals (as shown in Fig. 3). The propagation distance was the initial height of the sample L corrected by the deformation L that was measured by LVDT. The compressional velocity was determined by:
 | [5] |
Multiple trials were conducted to determine the accuracy of the acoustic velocity measurement. For each remolded soil, three specimens were prepared separately and isotropically loaded by increasing the cell pressure from 0.0 to 34.5, 69.0, and 103.4 kPa, respectively. For one kind of soil, three curves of acoustic measurements with overlapping pressure ranges were obtained. It was found that these curves fell closely into one line with the standard deviations of ±5.70 m s–1 for Sharkey soil and ±5.93 m s–1 for Neshoba soil. The differences were believed to be associated with the variations in the soil properties between samples as shown in Table 1.
A computer communicated with the oscilloscope and the function generator through a GPIB board. A program written by LabView (National Instrument, Inc., Austin, TX) software was used for measurement control, data acquisition, signal processing, and computation.
As mentioned previously, there are three types of triaxial tests: unconsolidated-undrained test (UU test), consolidated-undrained test (CU test), and drained test (CD test). In an UU test, a soil specimen is first isotropically loaded by increasing the cell pressure to a desired value, which is then kept constant during the remaining test. After that, the specimen is sheared in compression at a constant rate of axial deformation. No drainage of the specimen is permitted during the test. The pore pressure is established during the isotropic loading stage. In a CU test, the specimen is isotropically loaded and consolidated. The pore pressure is allowed to dissipate to zero during the isotropic loading stage. After that the specimen is axially compressed to failure without permitting any further drainage. In a CD test, the drainage is permitted for all the testing stages and the pore pressure is kept constantly zero.
Among the three procedures, the UU test takes the least time to carry out and is more likely to represent the conditions of real compaction in the field, where pore pressure could be established deep in the soil due to the overburden of the upper soil layers. Furthermore, the compaction event occurs so quickly that pore fluids do not have enough time to be dissipated or drained. Therefore, the UU test was chosen for simulating the compaction process.
In the study, the UU tests were performed on each soil under different cell pressures to take into account the effect of lateral pressures that increase with depth. The soil specimen was compressed vertically to simulate the compaction process. Two parameters were used to observe and evaluate the compaction response: the axial strain and the deviator stress. These correspond to the field compaction response: volume reduction and excessive pressure induced by applied load. The axial strain 
is the ratio of axial deformation L to the initial height of soil sample L. The deviator stress 
is the difference between the axial stress a and the cell pressure Pc and is given by (ASTM International, 1999):
 | [6] |
where F denotes the force measured by the load cell and A0 is the initial average cross-sectional area of the specimen corrected by the term of (1 – ) to reflect the change of volume of the specimen under pressure.
The soil specimen was gradually compacted up to a high stress level where the specimen was brought to failure to take into account some extreme cases of very heavy vehicles. One unload-reload cycle in the test was conducted to simulate and examine the unloading behavior in the stage after the passage of a vehicle and the effects of recompaction due to repeated agricultural tillage.
|
RESULTS AND DISCUSSION
|
|---|
The experimental results of the UU tests for air-dry remolded Sharkey and Neshoba soils are shown in Fig. 4a,b
and Fig. 5a,b . The tests were performed under cell pressures of 34.5, 69.0, and 103.4 kPa (5, 10, and 15 psi; here psi stands for pounds per square inch). The soil samples were vertically compacted and subjected to axial strains up to 8% for Sharkey soils and 14% for Neshoba soils.
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 4. The unconsolidated undrained test for air-dry remolded Sharkey soils. (a) The deviator stress vs. the axial strain. (b) The acoustic velocity vs. the axial strain. Pc, cell pressure.
|
|
View larger version (32K):
[in this window]
[in a new window]
|
Fig. 5. The unconsolidated undrained test for air-dry remolded Neshoba soils. (a) The deviator stress vs. the axial strain. (b) The acoustic velocity vs. the axial strain. Pc, cell pressure.
|
|
Figures 4a and 5a represent the load-deformation behaviors of Sharkey and Neshoba soils. The deviator stresses increase almost linearly at the early stage of compaction where the axial strain is <5%. With further increase in the axial strain, the rate of increment of the deviator stress decreases, which reveals the nonlinear behavior of soil under loading. The deviator stresses reached their maximums where the axial strains ranged from 5 to 7% for Sharkey soil and 10 to 14% for Neshoba soil. The maximum deviator stress is generally regarded as an indicator of soil failure in which significant fabric change in the soil occurs. Continued compression of the soil after failure leads to larger axial strains, but the deviator stress shows a slowly decreasing trend.
Soils display hysteretic behaviors when subjected to an unload-reload cycle. The deviator stress does not follow its original loading path when the axial pressure is released. Instead, it decreases sharply with a small decrease in the axial strain in the unload stage and increases steeply with the axial strain. The unload-reload deformation curve forms a clockwise loop in a small strain range. Further reloading the soil after the current deviator stress exceeds the original deviator stress from which the unloading started, the soil resumes its original deformation curve. This represents the load-history dependent behavior of soil. The compression curve without the unload-reload cycle is usually called the normally consolidated line and the unload-reload stress cycle is called the overconsolidated line or the preconsolidated line in geophysics and soil mechanics (Terzaghi et al., 1996).
The compressional wave velocities under compaction for Sharkey clay and Neshoba soil are shown in Fig. 4b and 5b. As compared with Fig. 4a and 5a, the acoustic behaviors are very similar to those of the load-deformation behaviors. The acoustic velocities increase linearly with the axial strain in the early compaction stage. The nonlinear behavior is also observed for the acoustic velocity, which is indicated by a progressively slower increase in the acoustic velocity with the axial strain. The curves also exhibit maxima when the soil specimens were further compressed. It appears that these acoustic velocity maxima occur within the same axial strain range as the deformation peaks associated with failure. This demonstrates that the maximum acoustic velocity, like the maximum deviator stress, can be used as an indicator of failure. Across the whole axial strain range, the acoustic velocities increase with the cell pressure. After reaching the maximum, the acoustic velocities change very little, which resembles the slow change in the deviator stresses in this stage of compaction.
The acoustic velocity during the unload tests, as in the case of the load-deformation curves, does not recover its original loading path and decreases sharply with the change in axial strain. The reload path follows approximately, in an opposite direction, the unload path. The cycle forms a clockwise loop representing a hysteresis for the acoustic velocity in this small strain region. The soil displays the load-history dependent properties for the acoustic velocity. The acoustic behavior resumes its normally consolidated line after passing the point where the unload-reload cycle starts.
The initial acoustic velocities before triaxial testing and their maxima for the Sharkey and Neshoba soils are tabulated in Table 2. Also listed in Table 2 are the slopes of the acoustic velocity vs. the axial strain during the early stage of normal consolidation and the unload-reload cycles. The relative increment ratios of the acoustic velocity are also presented. As can be inferred from Table 2, all these quantities increase with an increase in the cell pressure or the confining pressure. The values of Slope 2 are consistently higher than those of Slope 1 in every test, which indicates that the acoustic velocity changes faster during the unload-reload stress cycle than it does during the normally consolidated line.
The deformation and acoustic velocity behaviors of the Marshall field soils are plotted in Fig. 6a,b
. Four undisturbed cores were tested under cell pressures of 34.5, 41.4, 69.0, and 103.4 kPa (5, 6, 10, and 15 psi), respectively. The specimens were subjected to axial strains up to 10%. As can be seen from Fig. 6a, the deviator stress increases linearly in the early stage of compaction and rises nonlinearly in the intermediate axial strain range. The stress-strain behavior once again shows the hysteretic and load-history dependent properties, like in the cases of air-dry remolded soil. However, with further compaction, in most cases, except for the soil tested under cell pressure of 34.5 kPa, the deviator stress rises slightly and no peak is found. This continuing increment of the deviator stress is an indication of a plastic deformation. For the soil sample tested under cell pressure of 34.5 kPa, the deviator stress shows a distinct peak at the axial strain of 3.5% that indicates a failure and it drops quickly after failure.
View larger version (41K):
[in this window]
[in a new window]
|
Fig. 6. The unconsolidated undrained test for partly saturated Marshall field soils. (a) The deviator stress vs. the axial strain. (b) The acoustic velocity vs. the axial strain. Pc, cell pressure.
|
|
Figure 6a shows the significant influence of water content on the load-deformation behaviors of soils. The deviator stresses of tests under cell pressures of 34.5 and 69.0 kPa have overall higher values than those of the tests under cell pressures of 41.4 and 103.4 kPa, respectively, unlike in the cases of air-dry remolded soil, where the deviator stresses show a constant increasing tendency with cell pressure. This inverted variation in the deviator stress with cell pressure could be attributed to the influence of water, in which high water content decreases the maximum deviator stress or the undrained shear strength of a soil (Bishop and Henkel, 1962; Terzaghi et al., 1996). The soil specimens tested under cell pressures of 41.4 and 103.4 kPa, as listed in Table 1, have higher water contents than those under cell pressures of 34.5 and 69.0 kPa, respectively.
The load-acoustic velocity behaviors of Marshall field soils are plotted in Fig. 6b. Once again, the acoustic velocity behaviors of the field soils resemble their deformation behaviors in all stages of compaction. The slopes of the acoustic velocity vs. the axial strain are calculated and listed in Table 2. Also listed in Table 2 are the initial acoustic velocities, the maximum acoustic velocities, and their relative increment ratios, respectively.
The influence of water content on the acoustic velocities of the undisturbed Marshall soils was evident as shown in Fig. 6b and Table 2. The values of the acoustic velocity for tests under cell pressures of 34.5 and 69.0 kPa have overall higher values than those under cell pressures of 41.4 and 103.4 kPa in the large strain region. The variation of water content was believed to be the main reason that caused the scattering of the values in the slope of the acoustic velocity vs. the axial strain for undisturbed field soils during the early stages of compaction. No tendency of slopes with respect to the cell pressure was found, whereas in the cases of dry remolded soils, the slopes increase slightly with the cell pressures. The scattering in the slopes can also be found for unload-reload cycles for Marshall soils. Water content within the field soils significantly decreased the values of the maximum acoustic velocity attainable during the tests. As compared with the data of dry-remolded soils, although the initial acoustic velocities under corresponding cell pressures are roughly the same, the maximum acoustic velocities for field soils are much lower than those of dry-remolded soils. As a result, the corresponding relative increment ratios in the acoustic velocity for field soils have smaller values than those of dry-remolded soils and do not show the increasing dependence on cell pressure. It is clear that the water content plays an important role in governing the acoustic velocity behaviors of soils. A systematic study of the influence of water content on the acoustic velocity of soil is needed to address this issue.
For a routine triaxial test, it is desirable to get three or more identical soil specimens tested under different cell pressures to determine the effective shear strength of the soil represented by the effective friction angle and the cohesion (Bishop and Henkel, 1962). However, it is not the primary purpose of this study. The major purpose of this investigation is to compare the compaction-deformation characteristics with the compaction-acoustic velocity behaviors of soils under different soil conditions. It can be seen from Table 1 that the soil samples are far from identical, especially for the four field soil samples. The different physical properties of the soil samples originate from the different preparations, heterogeneity, and the variation of water content of soil samples as mentioned in the section of SOIL DESCRIPTIONS AND PREPARATIONS. The diversities have been represented by different strain-stress responses as shown in Fig. 6a, particularly for the test under cell pressure of 34.5 kPa where its shape is distinctively different from the others. The interesting thing is that, as demonstrated by the experimental results, the acoustic velocity resembles fairly well the behavior of the compaction-deformation of the soils in all stages of compaction in spite of the diversities of the soils. Therefore, the acoustic velocity could be explored for monitoring the ongoing compaction process in situ, real-time, and for long-term soil survey.
According to some investigations (Söhne, 1958; Vanden Berg and Gill, 1962), the magnitudes of pressures exerted on the soil surface by wheeled and tracked vehicles range from 104 to 207 kPa (15 to 30 psi), which depend in a combined way on characteristics of the soil and the tire pressures of the vehicles involved. In this pressure range, the vertical deformation (or volume reduction) increases almost linearly with the vertical pressure for air-dry remolded soils as shown in Fig. 4a and 5a. For a partially saturated soil, as in the case of field Marshall soils in Fig. 6a, the surface pressure could cause linear or nonlinear increments in the vertical deformation and, in the worst case, bring the soil into failure. After passage of the vehicle, the applied surface pressure returns to zero; however, the vertical deformation will not correspondingly recover to its original state and the soil will remain in some compacted state due to the hysteresis of the soil. It implies that repeated tillage after the first passage of the vehicle does not necessarily lead to further compaction unless the repeated passage of vehicle exerts more surface pressure than the previous ones.
With regard to the effects of field compaction on the acoustic velocity, several qualitative conjectures can be made. The initial acoustic velocity before compaction increases with the depth due to increased confining pressure or lateral pressure. The compressional wave velocities of deep soils are more sensitive to the excessive pressure induced by passing of an agricultural vehicle than those of shallow soils, provided their water contents are the same. For most agricultural vehicles where the surface pressure exerted by various tractors is 104 kPa (15 psi, Söhne, 1958), first compaction causes a linear rise in acoustic velocity. Postpassage of the vehicle that causes a pressure release dramatically decreases the acoustic velocity. After first compaction, repeated tillage will cause a pseudoelastic behavior of the acoustic velocity and the rate of change in the acoustic velocity with pressure is greater than that of first compaction. The same values of the acoustic velocity, as in the case of the deviator stress, could be found in different compaction stages. Therefore, single measurement of the acoustic velocity cannot be used to assess the soil state of compaction. It is the evolution of the acoustic velocity that reflects the ongoing process of compaction.
It must be pointed out that a real compaction process is much more complicated than the one simulated by the triaxial compression. The compaction event in the field occurs generally much quicker than the one in the UU test in which a strain rate of 0.1% min–1 was used. In case of moist natural soils, the acoustic velocity depends on the speed of deformation. Field soils are confined under increasing horizontal and vertical pressures with depth. The excessive pressures induced by the agricultural vehicles decrease with depth. These nonuniform stress distributions plus the heterogeneity, anisotropy, and different moisture content of field soil make the lab simulation far from reality. In situ measurement is necessary to validate the present acoustic technique for assessment of the soil compaction problem. To achieve this goal, a specially designed acoustic probe has recently been developed at the National Center for Physical Acoustics at the University of Mississippi. The probe consists of five separate transducers embedded longitudinally along a probe frame. It can be inserted into the ground 36 cm below the surface with minimum disturbance of the soil. One pair of the probes can be used to measure the compressional wave velocity horizontally at different depths. With these probes, an in situ compaction study will be conducted in the future.
|
CONCLUSIONS
|
|---|
A laboratory study of the effects of compaction on the acoustic velocities as well as the deformation behaviors of soils was presented. The triaxial compression tests were designed to simulate the compaction process induced by agricultural machinery driving over a field soil. Experimental results showed a significant influence of compaction on the acoustic velocity due to compaction-induced pressure, variations of porosity, and water content. It was found that the acoustic velocity and the deviator stress behaved similarly in all stages of compaction in spite of the conditions of the soils. The excellent similarity of the acoustic velocity and load-deformation behaviors of soils under compaction clearly demonstrated that the acoustic velocity could be used as a promising candidate instead of the deviator stress, which can only be determined in the laboratory, to monitor the ongoing compaction process in situ and for long-term soil survey. A field test with the acoustic probe is needed to further validate the present acoustic technique.
|
ACKNOWLEDGMENTS
|
|---|
This work is supported by the U.S. Department of Agriculture. The first author is indebted to F.D. Shields for many helpful discussions and suggestions, to M.J.M. Römkens for providing the triaxial cell, and to D. Dicarlo for soil sample analysis.
Received for publication November 18, 2002.
|
REFERENCES
|
|---|
- ASTM International. 1999. Standard test method for unconsolidated-undrained triaxial compression test on cohesive soils. Historical Standard D2850–95. Am. Soc. for Testing and Materials Int., West Conshohocken, PA.
- Avery, B.W., and C.L. Bascombe. 1974. Soil survey laboratory methods. Vol. 6. Soil Survey of Great Britain, Harpenden, UK.
- Baker, G.S., C. Schmeisneer, D.W. Steeples, and R.G. Plumb. 1999. Seismic reflection from depths of less than two meters. Geophys. Res. Lett. 26:279–282.[ISI]
- Bates, C.R. 1989. Dynamic soil property measurements during triaxial testing. Géotechnique 39:721–726.
- Bengough, A.G., and C.E. Mullins. 1991. Penetrometer resistance, root penetration resistance and root elongation rate in two sandy loam soils. Plant Soil 131:59–66.
- Biot, M.A. 1956a. Theory of propagation of elastic waves in a fluid-saturated porous solid. I. Low frequency range. J. Acoust. Soc. Am. 28:168–178.
- Biot, M.A. 1956b. Theory of propagation of elastic waves in a fluid-saturated porous solid. II. High frequency range. J. Acoust. Soc. Am. 28:179–1191.
- Bishop, A.W., and D.J. Henkel. 1962. The measurement of soil properties in the triaxial test. 2nd ed. Edward Arnold, London.
- Carlos, M.P.V., and J.W. Hopmans. 2001. Simultaneous measurement of soil penetration and water content with a combined penetrometer-TDR moisture probe. Soil Sci. Soc. Am. J. 65:4–12.[Abstract/Free Full Text]
- Don, C.G., and A.J. Cramond. 1985. Soil impedance measurements by an acoustic pulse technique. J. Acoust. Soc. Am. 77:1601–1609.
- Donaghe, R.T., R.C. Chaney, and M.L. Silver (ed.) 1988. Advanced triaxial testing of soil and rock. Publ. STP 977. ASTM Int., West Conshohocken, PA.
- Flammer, I., A.B. Blum, A. Leiser, and P. Germann. 2001. Acoustic assessment of flow patterns in unsaturated soil. J. Appl. Geophys. 46:115–128.
- Frederickson, C.K., J.M. Sabatier, and R. Raspet. 1996. Acoustic characterization of rigid-frame air filled porous media using both reflection and transmission measurements. J. Acoust. Soc. Am. 99:1326–1332.
- Gajo, A., A. Fedel, and L. Mongioci. 1997. Experimental analysis of the effects of fluid-solid coupling on the velocity of elastic waves in saturated porous media. Géotechnique 47:993–1008.
- Germann, P.F. 2002. Relations between acoustical and conventional mechanical soil properties. p. 38–42. In H.E. Bass and J.M. Sabatier (ed.) Proc. Bouyocous Conf. Agro-acoustics 4th Symp., University, MS. 6–9 May 2002. National Center for Physical Acoustics, University, MS.
- Grable, A.R., and E.G. Siemer. 1968. Effects of bulk density, aggregate size and soil water suction on oxygen diffusion, redox potentials and elongation of corn roots. Soil Sci. Soc. Am. Proc. 32:180–186.
- Grover, R.L. 1955. Simplified air permeameters for soil in place. Soil Sci. Soc. Am. Proc. 19:414–418.
- Hess, H.M., K. Attenborough, and N.W. Heap. 1990. Ground characterization by short–range propagation measurements. J. Acoust. Soc. Am. 87:1975–1986.
- Hillel, D. 1982. Introduction to soil physics. Academic Press, London.
- Hunter, J.A., S.E. Pullan, R.A. Burnes, R.M. Gagne, and R.L. Good. 1989. Applications of shallow seismic reflection methods to groundwater and engineering studies. Ont. Geol. Surv. Spec. Vol. 3:704–715.
- Janse, A.R.P. 1969. Sound absorption at the soil surface. Cent. Agric. Publ. Doc., Wageningen, the Netherlands.
- Jarvis, K.D., and R. Knight. 2000. Near-surface VSP surveys using the seismic cone penetrometer. Geophysics 65:1048–1056.
- Laird, D. 1998. Soil structure: Scales of observation. p. 91–104. In H.E. Bass and J.M. Sabatier (ed.) Proc. Bouyocous Conf. Agro-acoustics 3rd Symp., Tishomingo, MS. 4–5 Nov. 1998. National Center for Physical Acoustics, University, MS.
- Lu, Z., C.J. Hickey, and J.M. Sabatier. 2002. Load-deformation and load-acoustic velocity behaviors of soils over high axial strain range. p. 135–151. In H.E. Bass and J.M. Sabatier (ed.) Proc. Bouyocous Conf. Agro-acoustics 4th Symp., University, MS. 6–9 May 2002. National Center for Physical Acoustics, University, MS.
- Miller, R.D., D.W. Steeples, and M. Brannan. 1990. Identifying intra fluvial and bedrock structures shallower than 30 meters using seismic reflection techniques. p. 75–88. In S.H. Ward (ed.) Geotechnical and environmental geophysics. Vol. II. Environmental and groundwater. Soc. of Exploration Geophysicists, OK.
- Nakagawa, K., K. Soga, and J.K. Mitchell. 1996. Pulse transmission system for measuring wave propagation in soils. J. Geotech. Eng. April:302–308.
- Nakagawa, K., K. Soga, and J.K. Mitchell. 1997. Observation of Biot compressional wave of the second kind in granular soils. Géotechnique 47:133–147.
- Oelze, M.L., J.M. Sabatier, and R. Raspet. 2001. Roughness characterization of porous soil with acoustic backscatter. J. Acoust. Soc. Am. 109:1826–1832.[Medline]
- Pidgeon, J.D. 1974. A portable air pycnometer for rapid field measurement of air-filled porosity. Dep. Note DNSSN/169. Natl. Inst. Agric. Eng., Penicuik, Scotland.
- Sabatier, J.M., H. Hess, W.P. Arnott, K. Attenborough, M.J.M. Römkens, and E.H. Grissinger. 1990. In situ measurements of soil physical properties by acoustical techniques. Soil Sci. Soc. Am. J. 54:658–672.[Abstract/Free Full Text]
- Sabatier, J.M., R. Raspet, and C.K. Frederickson. 1993. An improved procedure for the determinations of ground parameters using level difference measurements. J. Acoust. Soc. Am. 94:396–399.
- Sabatier, J.M., D.C. Sokol, C.K. Frederickson, M.J.M. Römkens, E.H. Grissinger, and J.C. Shipps. 1996. Probe microphone instrumentation for determining soil physical properties: Testing in model porous materials. Soil Technol. 8:259–274.
- Santamarina, J.C., K.A. Klein, and M.A. Fam. 2001. Soils and waves—Particulate materials behavior, characterization and process monitoring. John Wiley and Sons, Chichester, UK.
- Söhne, W.H. 1958. Fundamentals of pressure distribution and soil compaction under tractor tires. J. Agric. Eng. 39:276–281, 290.
- Terzaghi, K., R.B. Peck, and G. Mesri. 1996. Soil mechanics in engineering practice. 3rd ed. John Wiley and Sons, New York.
- Vanden Berg, G.E., and W.R. Gill. 1962. Pressure distribution between a smooth tire and soil. Trans. ASAE 5:105–107.
- Van der Haarst, G.G., and W.P. Stakman. 1965. Soil moisture retention curves II. Directions for use of the sand box apparatus, range pf 0 to 2.7. Inst. Land Water Mange. Res., Wageningen, the Netherlands.
- Viggiani, G., and J.H. Atkinson. 1995. Stiffness of fine-grained soil at very small strains. Géotechnique 45:249–265.
- Voorhees, W.B. 1991. Compaction effects on yield—Are they significant? Trans. ASAE 34:1667–1672.
TOP
ABSTRACT
INTRODUCTION
