| Project Id
|
BITS025F001507
|
| Project Detail
|
| Project Title
|
AI and ML models for predicting the volume change behaviour of problematic soils
|
| Senior Supervision Team (BITS)
|
| Supervisor name and Title
|
Dr. Sayantan Chakraborty
|
School or Department (or company, if applicable)
|
BITS PILANI, PILANI CAMPUS
|
|
|
|
Email ID
|
sayantan.chakraborty@pilani.bits-pilani.ac.in
|
| URL for more info
|
https://www.bits-pilani.ac.in/pilani/sayantan-chakraborty/
|
| a) Are you currently supervising a BITS or RMIT HDR student?
|
YES
|
| Please comment how many you are supervising
|
2
|
| b) Have you supervised an offshore candidate before?
|
NO
|
| If no, what support structures do you have in place?
|
|
| If yes, please elaborate
|
|
| Senior Supervision Team (RMIT)
|
| Supervisor name and Title
|
Dr Jaspreet Pooni
|
School or Department (or company, if applicable)
|
STEM
|
|
|
|
Email ID
|
jaspreetsingh.pooni@rmit.edu.au
|
| URL for more info
|
https://academics.rmit.edu.au/jaspreetsingh-pooni
|
| a) Are you currently supervising a BITS or RMIT HDR student?
|
YES
|
| Please comment how many you are supervising
|
2
|
| b) Have you supervised an offshore candidate before?
|
NO
|
| If no, what support structures do you have in place?
|
|
| If yes, please elaborate
|
|
| Other Supervisors (BITS)
|
| Supervisor name and Title
|
|
School or Department (or company, if applicable)
|
|
| Phone Number (Optional)
|
|
Email ID
|
|
| URL for more info
|
|
| Other Supervisors (BITS)
|
| Supervisor name and Title
|
Dr. Dilan Robert
|
School or Department (or company, if applicable)
|
STEM
|
| Phone Number (Optional)
|
399251908
|
Email ID
|
dilan.robert@rmit.edu.au
|
| URL for more info
|
https://www.rmit.edu.au/contact/staff-contacts/academic-staff/r/robert-associate-professor-dilan
|
| Field of Research (For Codes)
|
| 4005 | Civil engineering | 100.00 |
|
| Project Description
|
|
Problematic soils are frequently encountered worldwide. These soils usually have low bearing capacities and undergo significant volume changes when subjected to moisture variations. It is reported that damage caused by expansive soils exceeds $13.5 billion dollars annually to infrastructure. Chemical stabilization of these problematic soils is often employed to improve their engineering properties. Even though traditional calcium (Ca)-based stabilizers are used extensively, these treatment techniques are often found to be ineffective for sulfate-rich soils due to the formation of ettringite. Overall, expansive characteristics may be imparted due to clay mineral-induced swelling before chemical treatment and ettringite-induced swelling after the treatment of sulfate-rich soils with Ca-based stabilizers. Soil behaviour is complex; the physical, mechanical, and chemical behaviours significantly vary from one location to another, which is further altered by stabilization. The swelling characteristics of problematic soils depend on several parameters, including clay mineralogy, stabilizer type and dosage, sulfate content, and curing time before moisture intrusion. Currently, 1D free swell strain tests or swell pressure tests are performed to estimate the extent of swelling expected after moisture exposure. However, these tests are time-consuming and require a few weeks to obtain the swell strain or swell pressure data. Developments in AI and ML present an opportunity to analyze and model these swelling soils and predict the volume change characteristics as a function of the abovementioned causal factors. This will improve the reliability and accuracy of volume change prediction without the need to perform time-consuming laboratory tests.
This research project will identify and collate relevant datasets (untreated and stabilized soils) from published literature and comprehensive laboratory tests that will be conducted as a part of this study. Different problematic soils with various predominant clay minerals will be prepared and modified with gypsum to represent a wide range of sulfate levels. These soil groups will be treated with stabilizers. The swelling behaviour of the untreated and treated soils will be studied after various curing periods. ML and AI algorithms will be employed, and their suitability for predicting the volume change characteristics will be evaluated. Correlations will be captured to develop, train, and validate the prediction model.
|
| Project Deliverable/Outcomes
|
|
Project Deliverables/Outcomes
The research outcomes are expected to include:
• Feature importance and sensitivity analysis of factors influencing volume change behaviour
• Rapid and reliable prediction of volume change
• Identify AI and ML methods that are best suited for prediction of volume change
• Improved pavement designs tackling problematic soils
• Evaluate stabilizer suitability and expected performance
• Model the SWCC of clay soils (untreated and stabilized)
This study will enhance geotechnical engineering by optimizing soil treatment strategies, reducing laboratory testing, and improving prediction accuracy for better decision-making in infrastructure development.
|
| Research Impact Themes
|
| AI/ML and Data Analytics / Data Science with a focus on applications/translation in | Construction |
|
| Which RMIT Sustainable Development Goal (SDG) does your project align to
|
|
INDUSTRY, INNOVATION, AND INFRASTRUCTURE
|
| Which RMIT Enabling Impact Platform (EIP) does your project align to
|
|
URBAN FUTURES
|
| Which RMIT Program code will this project sit under?
|
|
DR218 (CIVIL)
|
| Student Capabilities and Qualifications
|
|
Excellent written and verbal communication skills, Capacity to work independently and as a part of a team
|
|
Strong computational, programming, algorithms, and data analysis skills
|
|
MTech
|
| Preferred discipline of Student
|
| Civil Engineering, Structural Engineering |
|