AI-based performance prediction for 3D-printed concrete considering anisotropy and steam curing condition

Xiaofei Yao, Xin Lyu, Junbo Sun, Bolin Wang, Yufei Wang, Min Yang, Yao Wei, Mohamed Elchalakani, Danqi Li, Xiangyu Wang

Research output: Contribution to journalArticlepeer-review

55 Citations (Scopus)

Abstract

The 3D concrete printing (3DCP) technique piques the curiosity of several researchers and enterprises. However, there are few systematic investigations into how curing conditions influence the mechanical performance of 3DCP. This study aims to investigate the effect of various steam curing conditions (temperature rise rate, retention capacity, and sustained temperature) on the performance properties of 3D printing concrete materials at various ages of curing. A thorough test comprises macroscopic and microscopic analysis was conducted. In addition, the best conditions for steam curing are established for compressive characteristics in different directions. Then the anisotropy of mechanical properties of printed materials are studied under various curing settings. This study has contributed to the theoretical research on the influence of steam curing conditions on printed components. In addition, the experimental results were used to create two machine learning (ML) models and the beetle antennae search (BAS) technique was utilised. According to test data, the model is carried out to achieve the mechanical performance prediction of steam curing concrete. To automatically find optimal hyperparameters of ML models, the BAS algorithm was proposed, providing a solid guarantee for the rapid construction of the model.

Original languageEnglish
Article number130898
JournalConstruction and Building Materials
Volume375
DOIs
Publication statusPublished - 24 Apr 2023

Fingerprint

Dive into the research topics of 'AI-based performance prediction for 3D-printed concrete considering anisotropy and steam curing condition'. Together they form a unique fingerprint.

Cite this