Development of Expert System to Predict Mechanical Properties of Concrete Made with Recycled Concrete Aggregate
Description: The availability of user friendly guidelines for use of waste streams in construction
industry will promote further recycling of construction and demolition waste. This can lead to
increased acceptance and implementation of recycled concrete aggregate (RCA) in concrete
construction. The performance of the concrete is highly dependent on the properties and content of
RCA. The heterogeneous nature of RCA stemming from the adhered residual mortar, origin of the waste
concrete source, recycling process, level of chemical contamination, etc. can lead to wide range of
performance when RCA is incorporated. The degree of heterogeneity of RCA can be reflected in key
physical properties, including variability in specific gravity (2.0-2.5), water absorption (2%-8%),
and Los Angeles abrasion mass loss (20%-50%) (FHWA 2008). Such variability of RCA characteristics
can impact engineering properties, e.g. compressive strength, modulus of elasticity, splitting
tensile strength, etc. that can decrease by 0-40% (NCHRP 2013).
Proper strategies enabling the prediction of the performance of concrete based on key
characteristics of RCA are required to enhance the reliability of using RCA in transportation
infrastructure applications. The present research proposal seeks to use artificial intelligence for
development of guidelines to predict the performance of concrete made with RCA.
The main objectives of the research are to:
1. Develop a smart system using artificial intelligence techniques (e.g. artificial neural
networks) to predict the key mechanical properties of concrete incorporating RCA.
2. Validate the developed system using additional datasets obtained from laboratory investigation.
3. Develop guidelines for selection of RCA to produce reliable concrete with RCA in transportation
The project will develop guidelines to predict concrete performance based on RCA
characteristics. The research program targets concrete used in pavement applications as well as
structural concrete that is of great interest to owner agencies and engineers considering the
and use of sustainable concrete for infrastructure applications.
The proposed research is expected to include the following tasks:
Develop a comprehensive database of key properties of concrete incorporating fine and
coarse RCA. This will include an extensive review of published materials and the preparation of
several concrete mixtures using different RCA materials procured from a variety of sources across
the country. The properties of the RCA materials that are of
special interest are:
physical properties of RCA, including water absorption, specific gravity, Los
Angeles abrasion resistance, Micro-Deval, and deleterious materials content;
b. raw materials and mixture design of concrete where RCA is employed (binder type and content,
w/cm, RCA replacement level, properties of virgin aggregate); and
c. concrete performance, including mechanical properties and durability.
Employ artificial neural networks and statistical data analysis techniques to analyze the
Develop models to predict compressive strength, splitting tensile strength, flexural strength,
and modulus of elasticity of concrete made with RCA.
Validate models by testing concrete with different RCA materials used at replacement levels of
0, 20%, 35%, 50%, 70%, and 100% of coarse RCA.
Repeat Task 4 for concrete made with 0, 10%, 20%, and 30% fine RCA and concrete containing both
fine and coarse RCA.
Proposed recommendations and guidelines for the evaluation of key RCA materials and
prediction of mechanical properties of concrete made with such recycled materials.
Index Terms: Portland cement concrete, recycled concrete aggregate, sustainability,
construction and demolition waste, artificial intelligent, neural network modeling, statistical
analysis, concrete mechanical properties.
|Sponsoring Committee:||AFN40, Concrete Materials and Placement Techniques
|Research Period:||24 - 36 months|
|RNS Developer:||Kamal H. Khayat|
|Index Terms:||Artificial intelligence, Expert systems, Mechanical properties, Concrete, Recycled concrete aggregate, Concrete aggregates, Recycled materials, |