欢迎访问《硅酸盐通报》官方网站,今天是

硅酸盐通报 ›› 2026, Vol. 45 ›› Issue (7): 2561-2572.DOI: 10.16552/j.cnki.issn1001-1625.2025.1285

• 道路材料 • 上一篇    下一篇

再生骨料水泥稳定碎石力学性能分析与最优配合比预测

申彦利1,2(), 魏冠超1, 王鹏1(), 王嘉炜1, 徐磊3, 王辉3, 王志岭3, 卫爱魁3   

  1. 1.河北工程大学土木工程学院,邯郸 056038
    2.河北省装配式结构技术创新中心,邯郸 056038
    3.河北光太路桥工程集团有限公司,邯郸 056000
  • 收稿日期:2025-12-23 修订日期:2026-02-11 出版日期:2026-07-15 发布日期:2026-08-13
  • 通信作者: 王 鹏,博士,讲师。E-mail:wangpeng269@hebeu.edu.cn
  • 作者简介:申彦利(1977—),男,博士,教授。主要从事固废路基材料的研究。E-mail:shenyanli@hebeu.edu.cn
  • 基金资助:
    国家自然科学基金(52308272);河北省自然科学基金(E2024210089);冶金减排与资源综合利用教育部重点实验室开放课题(JKF20-07);邯郸市科学技术研究与发展计划(24422023200ZC)

Mechanical Properties Analysis and Optimal Mix Proportion Prediction of Cement-Stabilized Crushed Stone with Recycled Aggregate

SHEN Yanli1,2(), WEI Guanchao1, WANG Peng1(), WANG Jiawei1, XU Lei3, WANG Hui3, WANG Zhiling3, WEI Aikui3   

  1. 1.School of Civil Engineering,Hebei University of Engineering,Handan 056038,China
    2.Hebei Province Prefabricated Structure Technology Innovation Center,Handan 056038,China
    3.Hebei Guangtai Road and Bridge Engineering Group Co.,Ltd.,Handan 056000,China
  • Received:2025-12-23 Revised:2026-02-11 Published:2026-07-15 Online:2026-08-13

摘要:

为实现废旧道路混凝土再生骨料在道路基层中的资源循环利用,采用试验分析和人工智能相结合的方法,研究不同水泥掺量(3.50%、4.50%、5.50%,质量分数)和再生骨料掺量(4.15%、56.20%、78.10%、100.00%,质量分数)对废旧道路混凝土再生骨料水泥稳定碎石的最大干密度、最佳含水率、无侧限抗压强度、抗压回弹模量和劈裂强度的影响,基于机器学习与遗传算法(GA)构建再生骨料水泥稳定碎石7、90 d无侧限抗压强度预测和级配优化模型。结果表明:水泥掺量越高,再生骨料水泥稳定碎石的力学性能越好;再生骨料掺量提高会降低水泥稳定碎石的最大干密度并提高最佳含水率;适量再生骨料可提升水泥稳定碎石的7、90 d无侧限抗压强度和90 d劈裂强度,而抗压回弹模量随再生骨料掺量的增加逐渐降低,当再生骨料掺量为78.10%、水泥掺量为5.50%时,7、90 d无侧限抗压强度代表值分别为4.24、7.07 MPa,90 d劈裂强度与抗压回弹模量分别为0.64、1 925 MPa,均满足相关规范要求。根据再生骨料水泥稳定碎石力学性能的分析结果,选择随机森林模型结合遗传算法的方式进行级配优化,得到最优配合比下7 d无侧限抗压强度预测值为5.82 MPa。

关键词: 再生骨料, 水泥稳定碎石, 力学性能, 机器学习, 随机森林, 遗传算法

Abstract:

To achieve the resource recycling of waste road concrete recycled aggregate in road base courses, this study combined experimental analysis and artificial intelligence methods were employed to investigate the effects of different cement content (3.50%, 4.50%, 5.50%, mass fraction) and recycled aggregate incorporation rates (4.15%, 56.20%, 78.10%, 100.00%, mass fraction) on the maximum dry density, optimum moisture content, unconfined compressive strength, compressive resilient modulus, and splitting strength of cement-stabilized crushed stone containing recycled aggregates from waste road concrete. Based on machine learning and the genetic algorithm (GA), prediction models for the 7 and 90 d unconfined compressive strength and a gradation optimization model for recycled aggregate cement-stabilized crushed stone were established. The results indicate that higher cement content improves the mechanical properties of recycled aggregate cement-stabilized crushed stone. Increasing the recycled aggregate content reduces the maximum dry density and raises the optimum moisture content. An appropriate amount of recycled aggregate can enhance the 7 and 90 d unconfined compressive strength as well as the 90 d splitting strength, whereas the compressive resilient modulus gradually decreases with increasing recycled aggregate content. When the recycled aggregate content is 78.10% and the cement content is 5.50%, the representative values of 7 and 90 d unconfined compressive strength are 4.24 and 7.07 MPa, respectively, and the average values of 90 d splitting strength and compressive resilient modulus are 0.64 and 1 925 MPa, all of which meet the relevant specification requirements. Based on the analysis of the mechanical properties of recycled aggregate cement-stabilized crushed stone, a random forest model combined with the genetic algorithm (GA) is selected for gradation optimization, yielding a predicted 7 d unconfined compressive strength of 5.82 MPa for the optimal mixture proportion.

Key words: recycled aggregate, cement-stabilized crushed stone, mechanical property, machine learning, random forest, genetic algorithm

中图分类号: