Optimization and Modelling of the Physical and Mechanical Properties of Grass Fiber Reinforced with Slag-Based...

The construction industry's high energy consumption and carbon emissions negatively impact the ecological environment; large-scale construction projects consume much energy and emit a significant amount of CO₂ into the atmosphere. Statistics show that 30% of energy loss and 40% of solid waste in the construction industry are generated during construction. Therefore, reducing emissions during construction has significant research potential and value. Many scholars have recently studied eco-friendly building materials to facilitate the use of high-carbon emission materials like cement. Adding fibers to composite materials has become a research hotspot among these studies. Although adding fibers to composite materials has many advantages, it mainly reduces the compressive strength of the composite material. This research used the response surface methodology (RSM) to optimize the raw material ratios and thus improve the performance of plant fiber composite materials. Single-factor experiments were conducted to analyze the effects of grass size, grass content, and quicklime content on the composite materials' compressive strength, flexural strength, and water absorption. The influencing factors and levels for the response surface experiment were determined based on the results of the single-factor analysis. Using the response surface methodology (RSM), a second-order polynomial regression model was established to analyze the interaction effects of the three factors on the composite materials' compressive strength, flexural strength, and water absorption rate. The optimal ratio was determined: the optimized options for grass size, grass content, and quicklime content are 2.0 mm, 8.2 g, and 38 g, respectively. The actual values of compressive strength, flexural strength, and water absorption rate of the composite materials made according to the predicted ratio are 11.425 MPa, 2.145 MPa, and 21.89%, respectively, with a relative error of 8% between the actual and predicted values. X-ray diffraction and scanning electron microscopy were also used to reveal the factors contributing to the relatively high strength of the optimized samples.