Using Artificial Intelligence Techniques to Predict Punching Shear Capacity of Lightweight Concrete Slabs

Area of Science:

• Civil Engineering
• Materials Science
• Structural Engineering

Background:

• Lightweight concrete is crucial for mega-projects, yet research often overlooks its punching shear behavior.
• Punching shear failure in concrete slabs poses significant risks, necessitating reliable prediction models.

Purpose of the Study:

• To investigate and predict the punching shear strength of lightweight concrete slabs.
• To develop accurate artificial intelligence models for predicting punching shear capacity.
• To identify key parameters influencing punching shear strength and guide future design codes.

Main Methods:

• Compiled an extensive experimental database of lightweight concrete slabs under punching shear loading.
• Utilized statistical methods to determine effective parameters: concrete density, column dimensions, slab depth, concrete strength, reinforcement ratio, and steel yield stress.
• Developed and compared three artificial intelligence models: Genetic Programming (GP), Artificial Neural Network (ANN), and Evolutionary Polynomial Regression (EPR).

Main Results:

• Identified key parameters influencing punching shear strength.
• Developed three AI models (GP, ANN, EPR) capable of accurately predicting punching shear strength.
• Provided an expression detailing the inter-relation between influential variables.

Conclusions:

• The developed AI models accurately predict the punching shear strength of lightweight concrete slabs.
• This research offers valuable insights for future lightweight concrete design codes.
• The study highlights the importance of considering specific parameters for safe and economical mega-project construction.