IMPLICATIONS OF COMPUTER-BASED SIMULATION MODELS, EXPERT SYSTEMS, DATABASES, AND NETWORKS FOR CEMENT RESEARCH

Abstract

The use of new materials, construction practices, and designs have prompted a re-examination of traditional approaches to concrete technology. In addition, concrete structures are often used beyond their planned lifetimes, in functions for which they were not originally intended, or in harsher conditions than originally expected. There is, therefore, a need to develop Improved methods and procedures for predicting the performance of cement and concrete. At the same time, rapid growth in the power and availability of computers raises questions about how they might be used to make concrete a more predictable and better performing material. As will be seen, the answers to these questions can suggest alternative ways to plan and organize cement research. The significance of research on cement, as on other materials, depends on the questions asked, and both the questions and answers are changing because of computers. Paradoxically, the questions i,n cement research are becoming more oriented towards the end use of cements in concrete and, at the same time, becoming more fundamental. This is because computers can be used as links between technology and science. Thus, it is no longer unrealistic to expect scientific researchers to address such complex technological questions as "How can the performance of concrete be predicted?" which, even ten years ago, were remote from their interests and experience. The objective of the present paper is to suggest how research questions might be addressed using computers, specifically using a computer-based knowledge system (Fig. 1) consisting of 1) models to simulate chemical and physical relationships, 2) knowledge-based expert systems, and 3) databases. In the figure, the complementary parts of the knowledge system are connected through a global computer network linking collaborating researchers. Although there is other research in this area, we are using specific examples from investigations currently being carried out at the National Bureau of Standards (NBS) to illustrate the technical feasibility of the general approach