Prediction of dynamic properties of ultra-high performance concrete by an artificial intelligence approach. (January 2019)

• Record Type: Journal Article
• Authors: Khosravani, Mohammad Reza; Nasiri, Sara; Anders, Denis; Weinberg, Kerstin
• Is Part Of: Advances in engineering software. Volume 127(2019)
• Page Start: 51
• Page End: 58
• Publication Date: 2019-01
• Subjects: Ultra-high performance concrete -- Dynamic mechanical properties -- Artificial intelligence -- Case-based reasoning; Computer-aided engineering -- Periodicals; Engineering -- Computer programs -- Periodicals; Engineering -- Software -- Periodicals; Periodicals; 620.0028553
• DOI: 10.1016/j.advengsoft.2018.10.002
• Languages: English
• ISSNs: 0965-9978

Abstract

Highlights: An overview of behavior of UHPC material under dynamic loading. Results of experimental tests on different UHPC under high rate of loading. Implementation of an intelligent system based on CBR approach. Prediction of dynamic properties of UHPC via implemented system. Evaluation of the obtained results and determination of the system accuracy. Abstract: Significant advances in new generation of concrete led to construction of ultra-high performance concrete (UHPC). Since the demands of this engineering cementitious material have been increased, investigation on mechanical behaviour of this material is an interesting topic in today's research programs. Over the years, strain-rate sensitive behavior of UHPC material has been studied by experimental tests. In the current research, an artificial intelligence (AI) system is implemented which can predict dynamic mechanical properties of UHPCs constructed by various mixtures. This system is able to predict elastic modulus, compressive strength and tensile strength, all under high rate of loading. The system is developed within the framework of case-based reasoning (CBR) methodology as a problem-solving AI method. CBR is a learning methodology which utilizes similar previous cases to solve particular new problems. In this paper, case-base of the implemented system is enriched by a literature study of numerous researches which tested various UHPCs. The proposed intelligent system has been applied to reduce human expert dependency and avoid time consuming experimental tests. The implemented system is evaluated by available results from experiments under high rate of loading.