Concrete is a cost-effective building material widely used in various building infrastructure projects.
High-performance concrete has the characteristics of strength and durability, which is crucial for structures that must withstand heavy loads and extreme weather conditions. Accurately predicting the strength of concrete under different mixtures and load conditions is crucial for optimizing performance, reducing costs, and improving safety.
The latest advances in machine learning provide solutions to challenges in structural engineering, including concrete strength prediction. An article evaluates the performance of eight popular machine learning models, including linear, ridge, and LASSO regression methods and tree-based models such as decision trees, random forests, XGBoost, SVM, and ANN.
This evaluation was conducted using a standard dataset containing 1030 concrete samples. In addition, the research team also used SHAP (Shapley Additive explanations) technology to analyze the XGBoost model, providing insights for civil engineers to make wise decisions in concrete mix design and construction practices.
The research results indicate that using artificial neural networks with continuous data is suboptimal due to the extended time required for network training. However, the performance of the XGBoost model is superior to other established benchmark models, and in R? 2? Provide the best results in terms of RMSE. Hence, the research team utilized the SHAP approach, a type of XAI technique, to examine the outcomes of the XGBoost classification and improve the comprehensibility of the model. By employing the SHAP method, the research team gained a thorough understanding and visual representation of individual features, thus allowing them to grasp the impact of each feature on the prediction of concrete strength.
Furthermore, the SHAP method investigated the association and reliance between various features, thereby uncovering the nonlinear connection between features and model outputs. Concrete strength prediction plays a crucial role in practical engineering and construction scenarios. These predictions are used to evaluate whether concrete meets the required strength standards, thereby helping to prevent issues such as structural failure.
Concrete strength prediction is utilized by engineers and architects to choose suitable mixed designs for specific uses and environmental circumstances, ascertain the necessary amount and dimensions of concrete reinforcement for different structural components, guarantee the safety of buildings and infrastructure, minimize excessive design, cut down on material and construction expenses, and contribute to the longevity and resilience of structures.
Using predictive optimization techniques for concrete mix design can effectively decrease the environmental consequences associated with construction activities; this can be achieved by minimizing raw material consumption and the magnitude of waste produced. In future research directions, the research team plans to expand their analysis by considering other characteristics of concrete strength prediction, including the cement fermentation period and the mechanical, physical, and chemical properties of various types of concrete. This extension will help enhance different machine learning models’ analytical and predictive capabilities.
The impact of cement water-reducing agents on the strength of concrete is mainly reflected in the following aspects:
Improve compressive strength
Research has shown that using water-reducing agents can significantly improve the compressive strength of concrete because water-reducing agents can reduce water consumption, improve the compactness of concrete, reduce internal pores and defects, and thus improve its compressive strength.
Improve flexural strength
In addition to improving compressive strength, water-reducing agents can also enhance the flexural strength of concrete; this is important for structures such as roads and bridges that must withstand repeated bending and stress.
Water-reducing agents can reduce the water consumption of concrete and lower the water-cement ratio, thereby reducing the internal pores and cracks of concrete and improving its impermeability and durability; this can make the concrete structure more durable and reduce the frequency of maintenance and replacement.
Enhance frost resistance
In cold regions, concrete structures often need to withstand the test of freeze-thaw cycles. Research has shown that using water-reducing agents can improve the frost resistance of concrete and reduce freeze-thaw damage; this is because water-reducing agents can improve the pore structure and density inside the concrete, reduce the possibility of water migration and aggregation, and thus improve the frost resistance of concrete.
It should be noted that when using cement water-reducing agents, the mix design should be based on specific engineering requirements and concrete raw materials, and appropriate types and dosages of water-reducing agents should be selected to ensure that the performance of concrete meets engineering requirements. At the same time, it is necessary to strengthen quality inspection and control, ensure the quality and uniform dispersion of water-reducing agents, and avoid negative effects on concrete performance.
Concrete water-reducing agent supplier
Luoyang Tongrun Nano Technology Co. Ltd.â€‚ (TRUNNANO) Luoyang City, Henan Province, China, is a reliable and high-quality global chemical material supplier and manufacturer. It has more than 12 years of experience providing ultra-high quality chemicals and nanotechnology materials, including molybdenum Silicide, nitride powder, graphite powder, sulfide powder, and 3D printing powder. If you are looking for a high-quality and cost-effective Concrete water reducing agent, you are welcome to contact us or inquire any time.