构建科学智能混凝土配制新技术体系的设想和建议（一）

Assumption and Suggestions for Constructing the New Scientific & Smart System of Concrete Proportioning(1)

现在的混凝土配制技术明显落后于时代科技进步的步伐,传统配合比设计的经验方法已与现代混凝土的发展不相匹配;依靠大量试验进行配制与优化,难以应对如今多样性、性质和品种差异化的原材料,以及更高的混凝土性能需求。然而,水泥、混凝土和其他学科的研究至今所积累知识和认知,已为科学配制混凝土指明了技术方向,揭示出混凝土潜在的、未有效利用的性能,以及绿色化或低碳化的潜力;将科学理念与规律贯彻于混凝土配制优化,可以实现“用更少的水泥,获得性能更好的混凝土”。高效率、智能化实施科学配制混凝土,即不依靠大量试验获得最优或较优的混凝土组成与配合比,需要“先做好颗粒堆积体,再用好混凝土化学”,如今可以借助两项关键技术来实现——“颗粒堆积体的智能优化与密实化”和“混凝土新拌、硬化性能的智能预测与优化”。本文综述和分析了相关技术发展现状,建议以固体颗粒粒径分布优化(Dinger-Funk模型)和机器学习为核心,构建混凝土配制的新技术体系,以及具体的技术方案。

Nowadays,the technique&methods of concrete proportioning lag far behind the progressing path of science&technologies in other fields,while empirical,traditional method of proportioning is not suitable for modern concrete and its development.We are facing difficulties of tackling the raw materials with varied,diversified sources,properties or qualities,as well as the higher requirements on concrete performance.However,the present level of scientific unterstanding and accumulated knowledge of cement,concrete and other subjects did indicate the direction of concrete technologies development,the space of improvement for concrete properties,the potential for concrete to be greener or lower-carbon generated.Research and practice have proved the possibility to‘use less cement but obtain better concrete’,given concrete is proportioned and optimized with scientific laws and methods.The scientific,efficient and smart proportioning concrete uses computation to replace most part of tests for obtaining a optimal or good concrete composition,including two basic,important works:‘firstly,achieving dense particle packing,then well understanding&utilizing concrete chemistry’,implemented with the help of‘Computerized optimization of particle packing’and‘Smart prediction and optimization of concrete properties in fresh&harden state’,respectively.This paper summarized and analyzed the current state of related technologies,then suggest to construct a new scientific&smart system for concrete proportioning based on Dinger-Funk model for optimizing the particle size distribution and machine learning for predicting concrete properties.Two frameworks for such system are proposed.Chinese concrete industry is now suffering the shortage of the natural aggregate and minerals or raw materials with good,consistent quality.Other resources,e.g.manufactured sand,industrial solid wastes,tailings,construction wastes,are becoming the alternatives.Constructing the scientific,smart and universal system for concrete proportioning is urgently in need,worth for efforts,important for concrete innovation and improvement of quality and performance.