Soil Spectral Behavior Related to Its Load-Bearing Capacity Based on Moisture Content

Abstract

Soil’s load-bearing capacity is a crucial property from which the ability of soil to resist the vertical deformation resulting from a normal load can be determined, and this property is essential for analyzing a vehicle’s performance over soil terrain in terramechanics studies. Soil’s moisture content has a significant impact on its load-bearing capacity and spectral behavior. This study aims to show the relation between the load-bearing capacity and the spectral behavior of sandy loam soil. The study presents the load-bearing capacity and color results of sandy loam soil at different moisture contents. The load-bearing capacity was measured using the Bevameter technique, and the color was measured using spectrophotometer technology that sends waves in the visible range (400–700 nm). The pressure–sinkage results of the tested soil show that with an increase in the moisture content, the bearing capacity of the soil decreases, and the color results show a decrease in the color reflectance with the increase in the moisture content. The measurements were performed in the laboratory of the vehicle technology department at Hungarian University of Agriculture and Life Sciences (MATE).

1. Introduction

When it comes to land locomotion studies, it is important to take into consideration the material that the vehicle is moving on; thus, studying the strength of this material is an important step for studying the performance of the vehicle over the material (terrain). Studying the soil’s mechanical properties is important for understanding how the soil will behave under the excitation loads resulting from the movement of the vehicle on it [1,2,3].

The most important mechanical properties reflecting the strength of the soil are the load-bearing capacity and the shear strength [4,5]. If the normal pressure resulting from the wheel on the soil exceeds the load-bearing capacity of the soil, the wheel sinks into the soil by compacting the soil [6], and if the applied pressure is of a value less than the bearing capacity of the soil, the wheel will remain on the surface and the motion of the vehicle will not be influenced. The shear strength of the soil is what keeps the material intact and prevents internal shear failure when a shearing force is applied to it; thus, the wheel will not slip due to internal shear failure in the soil material (the shear stress applied is less than the shear strength) [7]. Studying the load-bearing capacity of a soil type experimentally is an important step that will provide a basis for further studies related to that type of soil. With tangible load-bearing capacity results, the soil’s longitudinal behavior (vertical deformation) due to the normal applied pressure from the tractive element (wheel/track) can be understood. Based on the measured load-bearing capacity results, one can choose which vehicle to operate on the soil type (wheel design and the vehicle’s mass affect the applied normal pressure), or, if not choosing the suitable machine to operate, at least avoid moving on the soil zone that is of a weak bearing capacity. Terramechanics researchers have used and are still using the Bevameter technique for measuring the bearing capacity of the soil (pressure–sinkage relationship) since it emulates the normal mechanism applied from the wheel on the soil. The Bevameter technique was developed by Bekker as a solution for ending up with suitable terrain properties when it comes to the off-road case of study [8,9].

The load-bearing capacity of soil is measured by the Bevameter, which is a terramechanics measuring technique. The Bevameter technique was developed by Bekker to measure the soil’s strength and its sinkage parameters upon being subjected to a running wheel. The wheel or the running gear connected to the vehicle applies both the normal load (vertical) and the shear load (tangential stresses) on the terrain. The wheel contacts the terrain, thus transferring the power taken from the transmission to the wheels as tractive forces and ending with the vehicle’s movement. For simulating the real wheel–terrain contact case by studying the normal and shear loads applied on the terrain, the Bevameter technique is the best among the others [9,10].

The soil’s physical composition, wetness (water content), density, and initial compression condition affect the soil deformation caused by the vehicle’s passage. The soil becomes more compact as the moisture increases [11]. The change in the moisture content of the soil changes its mechanical properties, thus influencing its load-bearing capacity [6].

Soil spectral behavior refers to the way that soil absorbs and reflects different wavelengths of light, which can be used to determine its physical and chemical properties [12]. The moisture content of the soil is one factor that can affect its spectral behavior, as soil with higher moisture content typically exhibits different spectral characteristics to dry soil [13]. The change in the soil moisture content changes its color [14], which is an important soil parameter to rely on when linked to the mechanical properties of the soil.

There are different methods used for measuring soil color, and among these methods are the Munsell soil color charts 1975, which is a qualitative color estimation method [15,16], and spectrophotometry technology, which is a precise and quantitative approach to specify the color [17,18,19].

Scientists have agreed that the reflected spectrum describes the color that we see with our eyes, despite some differences between humans due to differences in human vision. Because of the differences in the observed color between people’s vision, it is better to rely on a method that measures the spectral reflectance (value) of the color [20,21,22].

It was described by Orna that an easily observed and described property is linked to an underlying theory through color [23]. The soil components such as organic matter, iron oxides, and carbonates have been assessed qualitatively using the color of the soil for many years in the remote sensing and soil science fields [24].

Generally, the soil color becomes darker with an increase in its moisture. The spectrophotometer sends electromagnetic waves in the visible band that reaches the soil. A part of the sent wavelengths will be absorbed by the soil, while the rest will be reflected; thus, based on the reflected wavelength, the color of the soil will be specified [25].

Linking the soil’s load-bearing capacity to its color and having it programmed in a remote sensor is a pioneering idea to be implemented in off-road engineering vehicles, especially in off-road autonomous vehicles moving on soil terrains. The sensor will detect the soil load-bearing capacity based on its color (since it will be linked and inserted in the sensor’s program); thus, based on its value, the driver or the autonomous vehicle computer will decide to go over the soil zone or to avoid it (in case high sinkage occurs).

As mentioned in some terramechanics references, the bearing capacity of a soil is the maximum force that the soil can resist, and it is achieved at the maximum sinkage, but what we care about is the ability of the soil to resist the applied normal load (from a vehicle tractive element) at low sinkage. When a vehicle moves on a soil terrain, what is important is to avoid the high sinkage of the tractive element, since this will be a drawback to the vehicle’s motion. The sinkage of the soil is directly proportional to the applied pressure, and this applied pressure changes from one vehicle design to another (the tractive element design and the mass of the vehicle). Relating the color of the soil to its resistance is applicable when knowing the applied pressure. So, the applied pressure (from the vehicle) will determine the sinkage, which can be deduced from the pressure–sinkage relationship curve (at a moisture value). The applied normal pressure should be given to be able to specify the resulting sinkage of the soil from its color.

2. Soil Type Used and the Measuring Techniques

2.1. Load-Bearing Capacity Measurement (Bevameter Test)

The Bevameter measurement (see Figure 1a) was made at Hungarian University of Agriculture and Life Sciences [26] using the university’s soil type (surrounding the university). The soil tested was the sandy loam soil of mechanical composition: sand (90.5%), silt (3.2%), and clay (6.3%). Since the soil infield is inhomogeneous and anisotropic, it is complicated to study the soil [27,28,29]. For simplifying this engineering case (studying the soil’s mechanical properties), the soil was sieved to study it as a homogeneous material. Having the soil initially at ambient conditions, holding moisture due to the surrounding humidity, the soil was sieved and prepared in the bin of the Bevameter, with height H₁ = 12 cm and inner diameter D₁ = 22 cm.

The moisture analyzer (HE53 230 V) was used to measure the soil’s moisture content. It works based on the principle of drying. An amount of soil is placed on the tray inside the moisture analyzer and the heating plates inside the moisture analyzer start to dry the sample until reaching unchangeable soil mass (dry mass).

The analyzer calculates the moisture percentage by subtracting the remaining mass from the initial mass and the answer is divided by the initial mass (the moisture content obtained is the wet-based result). The moisture analyzer is shown in Figure 2a,b.

The soil bin, with the soil filled to the height of H₂ = 10 cm, was centered under the pressing plate diameter D₂ = 20 cm of the Bevameter as shown in Figure 1b.

At the top of the Bevameter plate, there are two sensors; one sensor is used for measuring the vertical displacement (mm) and the other is used for measuring the applied load (Force; Newton is the unit) [26]. The results were transferred to a computer using a data logger (Spider 8) and read through the CATMAN 4.5 software (force–displacement graph as a function of time). The data logger and software used are shown in Figure 1c,d, respectively.

The first test was carried out on the soil at ambient conditions; the test was repeated on the same soil sample but with an increase in the moisture content. After every measurement, the soil was removed from the bin and placed in a bucket, and water was added to increase the moisture content. The new moisture content was measured by the moisture analyzer. The soil was tested at moisture contents of 1.05%, 3.34%, 4.82%, 6.05%, and 7.63%. These moisture content values were chosen randomly since it is difficult to set precise moisture value in the soil. What is important in the moisture value is its effect on the physical status of the soil; thus, upon adding water to the soil when moisturing it and recognizing a change in the soil behavior (even if a slight change), the moisture was measured by the analyzer and the measurement was carried out on the soil sample (the repeated measurements were on the soil with close moisture values).

The mass of the soil sample (at every moisture value) and the final height (after being compressed by the plate; H₃) were measured, knowing that the soil was initially at height H₂. These values were recorded for calculating the soil density before and after applying the load.

2.2. Spectral Behavior Measurements (Color Test)

The spectrophotometer CM-700d (see Figure 3a,b) was used to measure the soil’s color (soil spectral behavior). The color of the sandy loam soil was determined by the spectrophotometer at each moisture content mentioned in Section 2.1 (1.05%, 3.34%, 4.82%, 6.05%, and 7.63%) before the Bevameter test.

The spectrophotometer sends electromagnetic waves in the visible range (400–700 nm; light visible band), and the color of the soil was determined based on the reflected wavelength (spectral reflectance %). Three measurements were taken by the spectrophotometer for the soil at each moisture content, and the average of these was considered as the result.

3. Results

3.1. Soil Density at Different Moisture Contents

The properties and the measurement parameters of the five tested soil samples are shown in

Table 1. Five samples of sandy loam soil measured at different moisture contents.

Sample No.	H2/D2	MC (%)	Mass (kg)	Height $(\mathbf{c}\mathbf{m})$	Density (g/$\mathbf{c}{\mathbf{m}}^3)$	Density Ratio (−)
1	0.5	1.05	4.62	$H_2$ = 10, $H_3$ = 8.4	${\rho }_1$ = 1.47, ${\rho }_2$ = 1.75	0.19
2	0.5	3.34	4.52	$H_2$ = 10, $H_3$ = 6.9	${\rho }_1$ = 1.44, ${\rho }_2$ = 2.09	0.45
3	0.5	4.82	4.23	$H_2$ = 10, $H_3$ = 5.9	${\rho }_1$ = 1.35, ${\rho }_2$ = 2.28	0.69
4	0.5	6.05	4.21	$H_2$ = 10, $H_3$ = 5.8	${\rho }_1$ = 1.34, ${\rho }_2$ = 2.31	0.72
5	0.5	7.63	4.12	$H_2$ = 10, $H_3$ = 5.4	${\rho }_1$ = 1.31, ${\rho }_2$ = 2.43	0.85

. Each sample had a different moisture content (MC%). The dimensions of the soil container were height H₁ = 12 cm and inner diameter D₁ = 22 cm. The diameter size of the Bevameter pressure plate, D₂, was 20 cm, and the height of the soil before and after the measurement was H₂ = 10 cm and H₃, respectively. The soil was initially filled to the height of H₂. The soil mass m was measured to calculate the density of the soil at different moisture contents.

The initial density of the soil ${\rho }_1$ and the density of the soil ${\rho }_2$ after being compressed by the Bevameter were calculated using Equation (1). Having the soil filled in the bin with H₂/D₂ ≤ 0.5 showed that the soil when tested in a bin behaved in a similar way to when tested in infinite space [30]. Keeping the H₂/D₂ value fixed (0.5) at different moisture contents, the additional amounts of the moistured soil (remaining after filling the bin to the fixed H₂/D₂) were removed (after stirring the water and the soil since the water occupies space in the soil). The mass of the soil at different moisture contents used for filling the bin to the required height (keeping the H₂/D₂ value equal to 0.5) was recorded.

The density was calculated using the mass and the height.

$${\rho }_{1,2}=\frac{m}{\pi .r^2.H_{2,3}} ,$$

(1)

Table 1 shows that the density before pressing the soil decreases with the increase in the moisture content, showing that the water is occupying a place in the soil by filling the space and the voids between the particles; thus, upon removing an amount of the moistured soil from the total amount in order to keep the H₂/D₂ value fixed, the remaining soil mass decreases.

3.2. Bevameter Results (Pressure–Sinkage Relationship)

The pressure–sinkage curves of the sandy loam soil resulting from the Bevameter test were read through the CATMAN 4.5 software (see Table 1). The results are shown in Figure 4.

At each moisture content, the sinkage increases with the applied pressure (applied force divided by the plate area). The obtained pressure–sinkage relation at each moisture content complies with the pressure–sinkage equation suggested by Saakyan in 1959 that is based on elastic half-space Boussinesq theory (Equation (2)) [29,31,32,33,34].

$$\mathrm{p}=\mathrm{k}{\left(\frac{\mathrm{Z}}{\mathrm{D}}\right)}^{\mathrm{n}},$$

(2)

where:

• $\mathrm{p}$: the applied pressure;
• $\mathrm{D}$: the diameter of the plate;
• $\mathrm{Z}$: the soil vertical deformation (sinkage);
• $\mathrm{k}$: the sinkage modulus.

Moreover, Bernstein fitted the plate pressure–sinkage curve using Equation (3):

$$\mathrm{p}\cong {\mathrm{kz}}^{0.5},$$

(3)

where k is a modulus of inelastic deformation, and 0.5 is the exponent of sinkage [35].

The above two equations show that the sinkage is directly proportional to the pressure applied, and that is clear in the obtained curves.

At different moisture contents, with the same applied pressure, the resulting sinkage increases with the increase in moisture content. The pressure–sinkage curves follow the same trajectory (hyperbolic) at different moisture contents. At the maximum applied pressure (around 25 bar), it is recognized that the highest sinkage is at the highest moisture content and that the sinkage increases with the increase in the moisture at the fixed pressure.

3.3. Spectrophotometer Results (Color)

The soil spectral behavior results obtained using the spectrophotometer are shown in Figure 5. The spectral reflectance of the soil at each light wavelength can be obtained from the curves. The sent light wavelengths for measuring the color of the soil at each moisture content are in the range of 400–700 nm. Starting with the lowest moisture content, with the increase in the value of the sent wavelength, the reflected wavelength increases. This applies to the soil at all tested moisture contents.

When comparing the soil color behavior at different moisture contents, the obtained results show that with the increase in the moisture content with the sent wavelength fixed, the reflected wavelength decreases. At the moisture contents of 6.05 and 7.63%, the reflected wavelength starts overlapping at the (sent) wavelengths below 550 nm. The overlapping in the results shows that with the increase in the moisture content, the reflected wavelength will be the same in the range of 400–550 nm for the two moisture contents of 6.05 and 7.63%; thus, specifying the reflected wavelength (soil color) will rely on the range of 550–700 nm.

4. Discussion

4.1. The Influence of Moisture Content on the Soil’s Density

The density of the moistured soil after compaction under the same load shows an increase in its value. With the increase in the moisture content, this soil becomes more compressible, hence ending with a lower soil volume (higher sinkage after compaction at higher moisture contents). The same mass divided by the decreasing volume results in an increase in the density.

The compressibility of the soil can be determined by plotting the density ratio as a function of moisture content (see Figure 6). The density ratio ∆ρ is calculated using Equation (4). With the increase in the moisture content, the density ratio increases.

$$\Delta \rho =\frac{{\rho }_2-{\rho }_1}{{\rho }_1},$$

(4)

4.2. Sandy Loam Soil Maximum Sinkage

Figure 7 shows the sinkage at the maximum applied pressure (25 bar) as a function of moisture content. The results show that with the increase in the moisture content, the sinkage increases, which emphasizes that with the increase in the moisture content the soil becomes (for most of the soils) more compact under the same compressing load. This point is important in our study, since the moisture is a main factor leading to major weakness in the bearing capacity of the soil, so the moisture should be measured or identified from another soil property (such as color) to avoid the sinkage of the wheel into the terrain.

4.3. Color Reflectance Affected by Moisture Contents

For confirming the results shown in Figure 5, that the increase in the moisture leads to a decrease in the soil reflectance behavior, average values of the reflectance results of 400–500 nm and 400–700 nm were calculated. Equations (5) and (6) were used for calculating the mean of the reflected wavelength at the different moisture contents.

$${\mathrm{R}}_{\mathrm{m}}^{\raisebox{1ex}{$4$}\!\left/ \!\raisebox{-1ex}{$5$}\right.}=\frac{{{\displaystyle \sum }}_{400}^{700}{\mathrm{R}}_{\mathrm{i}}}{11},$$

(5)

$${\mathrm{R}}_{\mathrm{m}}^{4/7}=\frac{{{\displaystyle \sum }}_{400}^{700}{\mathrm{R}}_{\mathrm{i}}}{31},$$

(6)

where:

• ${\mathrm{R}}_{\mathrm{i}}$: the reflected wavelength at a specific spectrum, 400–700 nm;
• ${\mathrm{R}}_{\mathrm{m}}^{\raisebox{1ex}{$4$}\!\left/ \!\raisebox{-1ex}{$5$}\right.}$: the mean for wavelengths in the 400–500 nm range;
• ${\mathrm{R}}_{\mathrm{m}}^{\raisebox{1ex}{$4$}\!\left/ \!\raisebox{-1ex}{$7$}\right.}$: the mean for wavelengths in the 400–700 nm range.

When comparing the Z (10°/D65) value, a spectrophotometer color-measuring tristimulus parameter (in the spectrophotometer system), the results show that with an increase in the moisture, the Z parameter decreases, thus complying with the reflectance behavior.

Moreover, in comparing the reflected wavelength at 700 nm at the maximum sent wavelength (spectrum, 700 nm), the results show that there is a decrease in the reflected value (see

Table 2. This table shows some important spectral reflectance parameters.

Sample No.	MC (%)	Rm^(4/5)	Rm^(4/7)	X (10°/D65)	Y (10°/D65)	Z (10°/D65)	Ref. at 700 nm
1	1.05	8.13	11.58	11.56	11.71	8.84	16.25
2	3.34	6.66	9.69	9.68	9.77	7.23	13.92
3	4.82	5.09	7.44	7.40	7.41	5.52	10.92
4	6.05	4.31	6.01	5.94	5.95	4.66	8.67
5	7.63	4.37	5.83	5.74	5.79	4.72	8.04

) with an increase in moisture. Figure 8 shows the average (R_m; Equations (5) and (6)), Z (10°/D65), and reflected wavelengths at 700 nm (Ref. at 700 nm) values as a function of moisture content.

Table 2 presents the averaged results obtained from the measurements at 400–500 and 400–700 nm, also including the tristimulus values X(10°/D65), Y(10°/D65), and Z(10°/D65), as well as the reflectance at the maximum visible spectrum. The data demonstrate that all the measured values decrease with increasing moisture content. These findings provide important insights into the impact of soil moisture on the spectral properties of soil, which can be useful for developing methods to estimate soil moisture content from spectral data.

Taking the intercept of any reflected wavelength with the curve fitting the reflected plotted points and then projecting this intercept to the x axis (moisture axis) will result in the moisture content of the soil. Choosing the pressure–sinkage curve that is the same or close to the moisture value (the moisture value obtained from the color close to that measured by the analyzer), and by knowing the normal pressure resulting from tractive elements of the vehicle, the sinkage of the wheel that will result can be determined (sinkage prediction). The intercept of the applied pressure with the curve and then its projection on the x axis (sinkage axis) will be the sinkage of the tractive element that will occur upon the passage of the vehicle on the soil.

5. Conclusions

One promising approach involves using alternative soil properties, such as color, to infer moisture content and thereby avoid the unwanted sinkage of vehicles into the terrain. The findings in this article are highly relevant to the fields of geotechnical engineering, soil science, and civil engineering. By accounting for the impact of moisture content on soil behavior, we can develop more accurate and effective models for predicting the performance of off-road vehicles in different environmental conditions.

The moisture content in the soil influences its bearing capacity, thus influencing its vertical deformation resulting from an excitation load. At the same moisture content, an increase in the applied pressure leads to an increase in the sinkage of the soil. With the same consolidation load applied to the soil and an increase in the moisture content, the sinkage increases. Furthermore, the influence of the moisture content on the soil is clear in Figure 4, Figure 6, and Figure 7; thus, with the increase in the moisture, the load-bearing capacity of the soil decreases.

Additionally, the soil’s color (spectral behavior) is a property that can be used to identify the soil’s load-bearing capacity (based on laboratory experiments). By relating the load-bearing capacity of the soil to its color (spectral reflectance) through the soil moisture content, the color of the soil will be linked to its load-bearing capacity. Spectrophotometric technology is a precise color measurement method that converts the color of the object into a quantitative value.

The color results obtained from the measurements show that with the increase in the moisture content, the wavelength reflectance decreases until it reaches a level where the reflectance overlaps in the range below 550 nm. To overcome the overlapping of the results (color measurement) and for using results that are separated at a spectrum (points are not close to each other), it is suggested to use the reflected wavelength at the 700 nm spectrum.

Knowing the load applied from the tractive element on the soil, in addition to having the pressure–sinkage curves of the soil at different moisture contents, the resulting sinkage of the tractive element can be deduced from a curve at a moisture value. The moisture value will be predicted from the color measured using the spectrophotometer; thus, the response of the terrain (regarding the sinkage) is identified from the soil color and pressure–sinkage curves (from laboratory work).

Finally, using the results of the soil color and its load-bearing capacity at different moisture contents, the load-bearing capacity and the color (reflectance at 700 nm) can be linked at a moisture value. This idea is beneficial in off-road engineering projects for studying the performance of vehicles on soil terrains, especially in autonomous vehicles. The proposed method could help us predict the load-bearing capacity of the soil, serving in choosing the best design (design of the tractive element; dimensions of the wheel and the track) needed to operate on a specific soil terrain. The results presented in this article provide a foundation for future research in this area, with significant implications for a range of practical applications.