The growing demand for sustainable building materials has motivated significant research into alternative

cementitious binders. Corn Cob Ash (CCA), an agricultural by-product rich in silica, has potential as a partial

cement replacement in concrete. This study investigates the mechanical performance of M20 concrete

incorporating CCA at replacement levels of 0%, 10%, 20%, and 30% by weight of cement. Compressive strength

was evaluated at 7, 28, 56, and 90 days. Results indicate a gradual reduction in strength with increasing CCA

content, although mixes containing up to 20% CCA demonstrated comparable strength development to

resource-efficient construction.

Concrete remains the most widely used construction material globally due to its versatility, durability, and

structural performance. However, conventional cement production is energy-intensive and contributes

approximately 7–8% of global CO₂ emissions, creating an urgent need for environmentally responsible

alternatives to conventional binders. Sustainable cementitious materials derived from agricultural and industrial

by-products have gained considerable attention as they offer both environmental and economic benefits while

promoting circular economy principles in construction.

Previous studies indicate the feasibility of using agricultural ashes such as rice husk ash, sugarcane bagasse ash,

and palm oil fuel ash as supplementary cementitious materials. However, comprehensive research on CCA

remains relatively limited, particularly concerning its long-term mechanical behavior and predictive modeling

using artificial intelligence. While empirical laboratory testing remains the principal method for assessing

concrete performance, the adoption of machine learning techniques—especially Artificial Neural Networks

(ANN)—has emerged as a promising approach to predict material behavior based on mix design variables.

This study investigates the mechanical performance of M20 grade concrete incorporating CCA at 0%, 10%,

20%, and 30% cement replacement levels. Compressive strength was evaluated at 7, 28, 56, and 90 days to

assess short-term and long-term strength characteristics. Furthermore, a multi-output ANN model was developed

to predict compressive strength at multiple curing ages simultaneously, providing a data-driven tool to support

4. Evaluate the suitability of ANN as a prediction tool for sustainable concrete mixtures.

The findings of this study contribute toward advancing low-carbon construction materials and demonstrate the

potential of integrating artificial intelligence for predictive assessment in sustainable concrete technology.

Growing global emphasis on sustainable construction has accelerated the search for low-carbon supplementary

cementitious materials (SCMs). Agricultural ashes such as rice husk ash, sugarcane bagasse ash, palm oil fuel

ash, and corn cob ash have shown significant pozzolanic potential when appropriately processed and

incorporated into cement-based composites.

Pozzolanic agro-wastes can refine pore structure and contribute to secondary C-S-H gel formation, enhancing

long-term strength and durability (Singh & Siddique, 2021; Samadi et al., 2022). Corn cob ash (CCA) contains

performance when utilizing CCA as SCM in structural concrete. Conversely, higher CCA contents beyond 25–

30% have been associated with retarded early-age strength due to reduced clinker content and slower pozzolanic

reaction kinetics (Ibrahim et al., 2020).

Alongside sustainable materials development, machine learning-based predictive models have gained

prominence for forecasting mechanical properties of concrete. Artificial Neural Networks (ANN) have exhibited

high accuracy in predicting compressive strength based on mix variables and curing period (Zhao et al., 2021;

Chou et al., 2022). Multi-output ANN architectures have been successfully used to predict compressive, tensile,

and flexural strengths simultaneously (Yoon et al., 2023). Recent studies emphasize ANN superiority over

multiple regression and classical empirical models, particularly when dealing with nonlinear material

interactions (Bui & Nguyen, 2023; Li et al., 2024).

Ordinary Portland Cement (OPC) conforming to IS 8112 (Grade 43) was used. The physical and mechanical

properties complied with standard requirements for construction applications.

Natural river sand passing 4.75 mm sieve and conforming to Zone II of IS 383:2016 was used. The specific

gravity of sand was 2.65 and fineness modulus was 2.70.

Crushed angular aggregates with a maximum nominal size of 20 mm were employed. The specific gravity and

As no chemical testing was performed in this study, typical composition values from literature are adopted,

summarised in Table 1.

Table 1 Typical Chemical Composition of Corn Cob Ash (CCA) (as per literature)

Ordinary potable water free from impurities was used for mixing and curing, conforming to IS 456:2000.

Concrete was designed for M20 grade following IS 10262:2019 guidelines. CCA replaced cement at 0%, 10%,

20%, and 30% by weight.

Table 2 Concrete Mix Proportions

and cured in water at 27 ± 2 °C until testing.

A multi-output Artificial Neural Network (ANN) was developed to predict compressive strength at 7, 28, 56,

and 90 days.

 Inputs: Cement, Water, Sand, Coarse Aggregate, W/C ratio, CCA%

 Outputs: 7-day, 28-day, 56-day, 90-day compressive strength

 Architecture: 16–8–4 neuron layers

 Activation: ReLU

 Optimizer: Adam

 Normalization: StandardScaler

 Validation: Leave-One-Out Cross-Validation (LOOCV)

This ANN structure is consistent with approaches adopted in recent concrete AI research (Yoon et al., 2023; Li

Additionally, a multi-output Artificial Neural Network (ANN) model was successfully implemented to predict

compressive strength at 7, 28, 56, and 90 days using mix parameters as input features. The ANN framework

performance of geopolymer concrete. Journal of Building Engineering, 52, 104497.

networks. Sensors, 23(4), 2259.

10. Samadi, M., Ghasemi, M., & Ramezanianpour, A. (2022). Pozzolanic agricultural ashes in sustainable

concrete: Mechanical and durability performance. Journal of Cleaner Production, 379, 134806.

11. Singh, B., & Siddique, R. (2021). Utilization of agro-industrial waste as supplementary cementitious

material: A review. Resources, Conservation and Recycling, 168, 105287.

12. Udoeyo, F. F., & Dash, N. (2020). Engineering properties of corn cob ash-modified concrete.

Construction and Building Materials, 234, 117246.

13. Yoon, Y., Choi, J., & Lee, S. (2023). Multi-output neural networks for predicting early-age and long-

term concrete strength. Automation in Construction, 155, 104173.

14. Zhao, X., Yang, H., & Liu, Y. (2021). Artificial intelligence techniques for predicting compressive

New Delhi, India.