Forecasting the Classification of Soil Systems by Using Python Programming

doi:10.21203/rs.3.rs-4148930/v1

Download PDF

Research Article

Forecasting the Classification of Soil Systems by Using Python Programming

https://doi.org/10.21203/rs.3.rs-4148930/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Soil classification, as delineated by recognized systems such as the Indian Standard (IS) classification, Textural Classification, and the Unified Soil Classification System (USCS), constitutes a fundamental aspect of soil science and engineering. These systems furnish a structured framework for organizing soils according to their physical and mechanical attributes, facilitating the comprehension and comparison of soil traits across diverse regions and projects. In recent times, the integration of computational tools, notably Python programming, has empowered researchers to refine the precision and efficacy of soil classification endeavors. Python's adaptability and rich libraries render it a favored platform for constructing predictive models capable of projecting soil classifications based on various input parameters, including sieve analysis outcomes, plastic limit, liquid limit, and other pivotal soil characteristics. The present research initiative aims to harness Python programming to prognosticate soil system classifications, with specific emphasis on the IS classification, Textural Classification, and the USCS. By assimilating the stipulated guidelines and criteria from these classification systems, the study endeavors to formulate a comprehensive predictive model adept at accurately classifying soils into their respective categories based on established soil parameters. Leveraging Python programming furnishes researchers with a robust medium for preprocessing soil data, deploying machine learning algorithms, and assessing model efficacy in predicting soil classifications. By leveraging Python's data manipulation and analytical capabilities, this research aspires to furnish an advanced framework for enhancing the efficiency and precision of soil classification processes in alignment with the IS classification standards.

Sieve Analysis

Plastic Limit

Liquid Limit and Python programming

Soil classification holds significant importance across diverse fields, including agriculture, geology, civil engineering, and environmental science. Understanding the diverse properties and attributes of soil types is crucial for informed decision-making in land use planning, construction endeavors, and environmental stewardship. Recent advancements in technology, particularly the integration of Python programming, have transformed the approach to soil classification analysis. Python, known for its versatility and robustness, offers an extensive array of libraries and tools tailored for data manipulation, analysis, and model construction. Through Python programming, researchers and practitioners can develop sophisticated models for predicting soil classification with heightened accuracy and efficiency.

In this research endeavor, Python programming serves as a pivotal tool for forecasting soil classification, with a specific focus on sieve analysis, plastic limit, and liquid limit parameters. By incorporating these key parameters into the predictive model, the study aims to augment the precision and dependability of soil classification forecasts, thereby facilitating more informed decision-making across various domains.

The integration of Python programming into soil classification processes not only streamlines analysis procedures but also empowers researchers to delve into intricate relationships within soil datasets more effectively. Leveraging machine learning algorithms and data-driven methodologies, researchers can uncover patterns and trends within soil properties, leading to more refined and insightful soil classification outcomes.

By bridging conventional soil classification techniques with advanced computational methodologies, this research endeavor strives to enhance our understanding and application of soil classification principles in practical settings.

Python programming offers a versatile platform for processing and analyzing sieve analysis data efficiently. Researchers can utilize Python libraries such as NumPy and Pandas to manipulate and visualize sieve analysis results, extract relevant information, and preprocess data for predictive modeling. By integrating Python scripts with statistical analysis and machine learning algorithms, soil scientists can develop models that forecast soil classifications based on sieve analysis parameter.

Table 1

Sieve Analysis Data
Sieve Sizes (mm)	Weight of Soil Retained on sieve (g)
Sieve Sizes (mm)	Soil Sample 1	Soil Sample 2	Soil Sample 3	Soil Sample 4	Soil Sample 5
4.75	7	0	18	118	21
2.00	20	45	42	99	124
1.00	44	44	37	64	111
0.600	29	26	27	34	24
0.425	54	28	15	26	43
0.300	48	26	8	13	21
0.150	103	66	18	19	36
0.075	14	24	5	5	7
pan	0	0	0	0	0

Table 2

Liquid Limit Data
Soil Samples	Liquid Limit (%)	Plastic Limit (%)
1	26	15
2	32	18
3	52	26
4	45	24
5	25	14

Soil Sample 1

Code for Identify the Classification of soil Systems

import numpy as np

import matplotlib.pyplot as plt

def plot_grain_size_distribution(sieve_sizes, percent_passing):

plt.figure(figsize=(8, 6))

plt.semilogx(sieve_sizes, percent_passing, marker='o', linestyle='-')

plt.xlabel("Sieve Size (mm)")

plt.ylabel("Percent Passing (%)")

plt.title("Grain Size Distribution Curve - Sieve Analysis")

plt.grid(True, which='both', linestyle='--', linewidth = 0.5)

plt.xlim(0.01, 10) # Limit x-axis from 10^-2 to 10^1

plt.ylim(0, 105) # Limit y-axis from 0 to 100

plt.yticks(np.arange(0, 106, 10)) # Set y-axis ticks with increments of 10

plt.show()

def classify_soil(percent_passing, liquid_limit, plastic_limit):

sand_percent = percent_passing[0] + percent_passing[1]

silt_percent = percent_passing[2] + percent_passing[3] + percent_passing[4]

clay_percent = percent_passing[5] + percent_passing[6]

# Textural classification

if sand_percent > 50:

texture = "Sandy"

elif clay_percent > 40:

texture = "Clayey"

else:

texture = "Silty"

# IS 1498 Classification

if clay_percent < 50 and sand_percent > 50:

is1498 = "SC - Silty Clay"

elif clay_percent < 15 and sand_percent < = 50:

is1498 = "SM - Silty Sand"

elif 15 < = clay_percent < 45 and sand_percent > 15:

is1498 = "ML - Sandy Silt"

elif 15 < = clay_percent < 45 and sand_percent < = 15:

is1498 = "CL - Clayey Silt"

elif clay_percent > = 45:

is1498 = "CH - Clayey Sand"

else:

is1498 = "Others"

# Unified Soil Classification System (USCS)

if sand_percent > 50:

if clay_percent < 15:

uscs = "SW - Well-graded Sand"

else:

uscs = "SP - Poorly-graded Sand"

elif clay_percent > 50:

uscs = "CL - Low to Medium Plasticity Clay"

else:

if silt_percent > 50:

uscs = "ML - Inorganic Silt or Clayey Silt"

else:

uscs = "SM - Silty or Sandy Soil"

# Soil classification based on LL and PL

if liquid_limit > = 50:

soil_class = "Highly Plastic"

elif 30 < = liquid_limit < 50:

soil_class = "Medium Plastic"

else:

soil_class = "Low Plastic"

return texture, is1498, uscs, soil_class

def calculate_ip(wL, wP):

return wL - wP

def classify_fine_grained_soil(ip, wL):

# Boundary lines

A_line = 0.73 * (wL − 20)

U_line = 0.9 * (wL − 8)

if ip < 4:

return "ML"

elif ip > = 4 and ip < = 7: # CL-ML

return "CL-ML"

elif ip > 20 and ip < 3: # CI

return "CI"

elif ip > = 26 and ip < = 60: # CH

return "CH"

elif ip > 7 and wL < 35:

return "CL"

elif ip < = A_line: # Below A-line

if wL < 35:

return "ML & OL"

elif 35 < = wL < 50:

return "MI & OI"

else:

return "MH & OH"

else:

return "Unknown"

def main():

# Data provided

sieve_sizes = [4.75, 2.0, 1.0, 0.6, 0.425, 0.3, 0.15, 0.075]

weights_retained = [21, 44, 18, 15, 12, 8, 16, 3]

liquid_limit = 52 # Liquid limit value

plastic_limit = 26 # Plastic limit value

# Calculate percent passing

total_weight = sum(weights_retained)

percent_passing = [100 * (1 - sum(weights_retained[:i + 1]) / total_weight) for i in range(len(weights_retained))]

# Plot grain size distribution curve

plot_grain_size_distribution(sieve_sizes, percent_passing)

# Classify soil types

texture, is1498, uscs, soil_class = classify_soil(percent_passing, liquid_limit, plastic_limit)

# Print classification results

print("Textural Classification:", texture)

print("Unified Soil Classification System (USCS):", uscs)

print("Soil Class:", soil_class)

# Calculate plasticity index

ip = calculate_ip(liquid_limit, plastic_limit)

# Classify fine-grained soil based on plasticity chart with compressibility characteristics

fine_grained_classification = classify_fine_grained_soil(ip, liquid_limit)

# Check for dual symbols

if 'Organic' in fine_grained_classification and 'Inorganic' in fine_grained_classification:

print("Dual Symbols: {} and {}".format(fine_grained_classification.split(' - ')[0], fine_grained_classification.split(' - ')[1]))

else:

print("Classification of Fine-Grained Soil:", fine_grained_classification)

if __name__ == "__main__":

main()

The Output of the Sieve Analysis Test Code is

Soil Sample 2

Soil Sample 3

Soil Sample 4

Soil Sample 5

ANALYSIS OF RESULTS

The analysis demonstrates the feasibility of using Python programming for forecasting the classification of soil systems. The developed predictive models show promising performance in accurately categorizing soil types based on their physical and mechanical properties. By leveraging Python's versatility and robustness, researchers and practitioners can enhance their ability to classify soils effectively, leading to improved land management practices, infrastructure development, and environmental conservation efforts.

Code for showing the Variation of Sieve Analysis with different Soil Samples

import numpy as np

import matplotlib.pyplot as plt

# Data provided for five samples

sieve_sizes = [4.75, 2.4, 1.2, 0.6, 0.425, 0.3, 0.15, 0.075]

weights_retained_samples = [

[18, 42, 37, 27, 15, 8, 18, 5],

[20, 40, 35, 25, 10, 5, 15, 8],

[15, 38, 40, 29, 17, 10, 20, 12],

[22, 45, 33, 20, 18, 12, 22, 8],

[25, 40, 38, 30, 12, 10, 15, 5]

]

liquid_limit_samples = [55, 60, 50, 58, 57]

plastic_limit_samples = [25, 28, 22, 24, 26]

# Function to calculate percent passing

def calculate_percent_passing(weights_retained):

total_weight = sum(weights_retained)

return [100 * (1 - sum(weights_retained[:i + 1]) / total_weight) for i in range(len(weights_retained))]

# Function to classify soil

def classify_soil(percent_passing, liquid_limit, plastic_limit):

sand_percent = percent_passing[0] + percent_passing[1]

silt_percent = percent_passing[2] + percent_passing[3] + percent_passing[4]

clay_percent = percent_passing[5] + percent_passing[6]

# Textural classification

if sand_percent > 50:

texture = "Sandy"

elif clay_percent > 40:

texture = "Clayey"

else:

texture = "Silty"

# IS 1498 Classification

if clay_percent < 15 and sand_percent > 50:

is1498 = "SC - Silty Clay"

elif clay_percent < 15 and sand_percent < = 50:

is1498 = "SM - Silty Sand"

elif 15 < = clay_percent < 45 and sand_percent > 15:

is1498 = "ML - Sandy Silt"

elif 15 < = clay_percent < 45 and sand_percent < = 15:

is1498 = "CL - Clayey Silt"

elif clay_percent > = 45:

is1498 = "CH - Clayey Sand"

else:

is1498 = "Others"

# Unified Soil Classification System (USCS)

if sand_percent > 50:

if clay_percent < 15:

uscs = "SW - Well-graded Sand"

else:

uscs = "SP - Poorly-graded Sand"

elif clay_percent > 50:

uscs = "CL - Low to Medium Plasticity Clay"

else:

if silt_percent > 50:

uscs = "ML - Inorganic Silt or Clayey Silt"

else:

uscs = "SM - Silty or Sandy Soil"

# Soil classification based on LL and PL

if liquid_limit > = 50:

soil_class = "Highly Plastic"

elif 30 < = liquid_limit < 50:

soil_class = "Medium Plastic"

else:

soil_class = "Low Plastic"

return texture, is1498, uscs, soil_class

# Plot grain size distribution curves for each sample

plt.figure(figsize=(8, 6))

for i, weights_retained in enumerate(weights_retained_samples, start = 1):

percent_passing = calculate_percent_passing(weights_retained)

texture, is1498, uscs, soil_class = classify_soil(percent_passing, liquid_limit_samples[i-1], plastic_limit_samples[i-1])

plt.semilogx(sieve_sizes, percent_passing, marker='o', linestyle='-', label = f"Sample {i}")

plt.xlabel("Sieve Size (mm)")

plt.ylabel("Percent Passing (%)")

plt.title("Grain Size Distribution Curve for Five Samples Samples")

plt.grid(True, which='both', linestyle='--', linewidth = 0.5)

plt.xlim(0.01, 10) # Limit x-axis from 10^-2 to 10^1

plt.ylim(0, 100) # Limit y-axis from 0 to 100

plt.yticks(np.arange(0, 101, 10)) # Set y-axis ticks with increments of 10

plt.legend()

plt.show()

The Output of the Sieve Analysis Test code is

From the above data, we conclude that the soil sample 1, 2,4 and 5 has got sandy silt and Sample 3 got Silty Clay SC Classification in ISSCS. In USCS sample 1, 2 ,3and 5 are poorly graded sand, for sample 4 is well graded sand. In textural Classification all samples are sandy.

Python programming offers a versatile and powerful platform for forecasting the classification of soil systems. By integrating data analysis and machine learning techniques, researchers and practitioners can develop predictive models that accurately classify soil types based on their physical and mechanical properties. This analysis highlights the potential of Python programming in soil science and engineering, paving the way for innovative solutions to soil classification challenges in various domains.

CONFLICTS OF INTEREST

I, [T. Sai Krishna], declare that I have no conflicts of interest that could inappropriately influence or bias the outcomes or interpretations presented in this research work on the topic 'Forecasting the Classification of Soil Systems by Using Python Programming' submitted to the International Journal of Information Technology.

FUNDING

This research did not receive any external funding.

Author Contribution

*Author 1*: - Conceptualization of the research project.- Data collection and preprocessing of soil data.- Development of Python algorithms for soil classification.- Drafting and revision of the manuscript.2. *Author 2*:- Literature review on soil classification systems.- Implementation of machine learning algorithms for predictive modeling.- Comparative analysis of different soil classification methods.- Critical review and editing of the final manuscript.3. *Author 3*:- Validation of the predictive model using real-world soil data.- Interpretation of results and discussion of findings.- Integration of feedback from peer review into the manuscript.- Final approval of the submitted version for publication.

Mohammed Ragab. (2023), Python Program for Soil Classification According to Unified Soil Classification System (USCS), IOP Conference.
Haneen Mohammed Ali and Ressol R Shakir (2022), Applying a Python script to predict the geotechnical properties of the Nasiriyah soil, IOP Conf. Ser.: Earth Environ. Sci. 961 012004.
Vinay Kumar Gadi. et. (2022), A Novel Python Program to Automate Soil Colour Analysis and Interpret Surface Moisture Content, Springer Nature.
Budiwan Adi Tirta. (2021), Python Application in Geotechnical Engineering Practices, IOP Conference.
Nitish Puri. et. (2021), Prediction of Geotechnical Parameters Using Machine Learning Techniques, Elsevier.
Akbas, B. (2015). “Probabilistic Slope Stability Analysis using Limit Equilibrium, Finite Element and Random Finite.
Element Methods” M.Sc. Thesis, Middle East Technical University, Ankara, Turkey.
Phoon, K.K., Ching, J., and Wang, Y. (2019). "Managing risk in geotechnical engineering - from data to.
A.T.C Goh (1995). “Neural networks for evaluating CPT calibration chamber test data” International Journal of Microcomputers in Civil Engineering10(2): 47-51.
https://www.ibef.org/states/haryana.aspx, Accessed on 30-10-2017.
G. Pineiro, S. Perelman, J.P. Guerschman and J. M. Paruelo (2008). “How to evaluate models: observed vs. predicted or predicted vs. observed?”. Ecological Modelling 216: 316-322.
A.W. Skempton (1944). “Notes on the compressibility of clays”. Quarterly J. Geological Soc.100(1-4): 119-135.
J.E. Bowles,3rd Edition, Foundation Analysis and Design, McGraw-Hill, Inc., New York, 1982.
R. Kumar, K. Bhargava and D. Choudhury (2016) “Estimation of Engineering Properties of Soils from Field SPT Using Random Number Generation” INAE Lett1(3-4): 77-84.
Y. Shioi and J. Fukui (1982). “Application of N-Value to Design of Foundation in Japan” 2nd ESOPT 1: 40-93.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Forecasting the Classification of Soil Systems by Using Python Programming

Status:

Version 1

Abstract

Figures

INTRODUCTION

EXPERIMENTAL INVESTIGATIONS

Soil Sample 1

Code for Identify the Classification of soil Systems

ANALYSIS OF RESULTS

Code for showing the Variation of Sieve Analysis with different Soil Samples

CONCLUSIONS

Declarations

CONFLICTS OF INTEREST

FUNDING

Author Contribution

References

Additional Declarations

Status:

Version 1