Development of Recurrent Neural Network Model for Bridge Effect Detection by Various Fibers Incorporated in Fiber-Reinforced Concrete

Abstract:

The purpose of this study is to detect the bridging effect formed at cracks after external forces act on various types of fiber-reinforced concrete reinforced with microfibers ranging in size from 10 μm up to 40 μm using a recurrent neural network (RNN). The bridge effect, which is an advantage of fiber-reinforced concrete(FRC) and a criterion for judging usability, can prevent cracks and brittle failure of concrete by forming a net with fibers from cracks formed in concrete to transmit stress. In this study, concrete surface image data taken from crack sections of various fiber-reinforced concrete were collected to create a crack and bridge effect exploration model based on RNN. Afterwards, the features of the part where the bridge effect appeared were directly labeled to enable high accuracy detection in the data added after model production. As a result of detecting surface cracks and bridge effects after modeling, the more sophisticated the labeling, the more accurate image data analysis was possible.

• Periodical: Key Engineering Materials (Volume 972)
• Pages: 99-104
• DOI: https://doi.org/10.4028/p-GEQ5Cd
• Online since: December 2023
• Authors: Dongwook Kim*, Sung Gul Hong
• Keywords: Bridge Effect, Crack Detection, Durability, Fiber-Reinforced Concrete (FRC), Image Data Amplification, Recurrent Neural Network (RNN)

References

[1] Afroughsabet, Vahid, Luigi Biolzi, and Togay Ozbakkaloglu. "High-performance fiber-reinforced concrete: a review." Journal of materials science 51 (2016): 6517-6551.

DOI: 10.1007/s10853-016-9917-4

[2] Kwan, A. K. H., & Chu, S. H. (2018). Direct tension behaviour of steel fibre reinforced concrete measured by a new test method. Engineering Structures, 176, 324-336.

DOI: 10.1016/j.engstruct.2018.09.010

[3] Surface of FRC with CFS 100 – 2 ASTM Type Ⅱ Concrete Reinforcing Steel Fibers, (http://www.concretefibersolutions.com/)

[4] Tai, H. W., Chen, J. H., Cheng, J. Y., Hsu, S. C., & Wei, H. H. (2021). Learn Curve for Precast Component Productivity in Construction. International Journal of Civil Engineering, 19, 1179-1194.

DOI: 10.1007/s40999-021-00621-z

[5] Chen, Y., & Qiao, P. (2011). Crack growth resistance of hybrid fiber-reinforced cement matrix composites. Journal of aerospace engineering, 24(2), 154-161.

DOI: 10.1061/(asce)as.1943-5525.0000031

[6] Lin, S., He, Z., & Sun, L. (2019). Defect enhancement generative adversarial network for enlarging data set of microcrack defect. IEEE Access, 7, 148413-148423.

DOI: 10.1109/access.2019.2946062

[7] Wang, B., Li, Y., Zhao, W., Zhang, Z., Zhang, Y., & Wang, Z. (2019). Effective crack damage detection using multilayer sparse feature representation and incremental extreme learning machine. Applied Sciences, 9(3), 614.

DOI: 10.3390/app9030614

[8] Yang, D., Gu, C., Zhu, Y., Dai, B., Zhang, K., Zhang, Z., & Li, B. (2020). A concrete dam deformation prediction method based on LSTM with attention mechanism. IEEE Access, 8, 185177-185186.

DOI: 10.1109/access.2020.3029562

[9] ACI 544.4R-18, Guide to Design with Fiber-Reinforced Concrete.