This study proposes a machine learning algorithm-based model for flexural capacity of corroded prestressed concrete (PC) structures, which considers the feature parameters including geometric parameters, material parameters and corrosion parameters. First, three black box models are utilized to predict the flexural capacity of corroded PC structures. Then, the sensitivity of feature parameters to the flexural capacity is quantified using the Shapley Additive Explanations (SHAP) additive model, and the model’s key features parameters are extracted. Following this, a practical model is established using Genetic Programming (GP) based on these key feature parameters. The model is verified and compared with the existing models. The results show that the interpretable machine learning model can quantify the influence of the feature parameters on the flexural capacity and obtain the feature importance ranking. The model can fit the calculation formula for the flexural capacity of corroded PC beams by symbolic regression method, which effectively reveal the potential nonlinear relationship between basic feature and flexural capacity. The accuracy and interpretability of the calculation formula are between the existing empirical model and black box models.
Amir HaeriMEbadzadehMMFolinoG (2017) Statistical genetic programming for symbolic regression. Applied Soft Computing60: 447–469. DOI: 10.1016/j.asoc.2017.06.050.
2.
AngelinePJ (1994) Genetic programming: on the programming of computers by means of natural selection. Biosystems33: 69–73. DOI: 10.1016/0303-2647(94)90062-0.
3.
BelgiuMDrăguţL (2016) Random Forest in remote sensing: a review of applications and Future Directions. ISPRS Journal of Photogrammetry and Remote Sensing114: 24–31.
4.
CevikAArslanMHKöroğluMA (2010) Genetic-programming-based modeling of RC beam torsional strength. KSCE Journal of Civil Engineering14: 371–384. DOI: 10.1007/s12205-010-0371-6.
5.
ChenR (2021) Research on Bearing Capacity of Acid Rain Corrosion Prestressed Concrete Beam and Existing Prestress Test. Thesis (M.D). Nanchang: East China Jiaotong University. (in Chinese).
6.
ChenTGuestrinC (2016) Xgboost: a scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining785: 794.
7.
Code for concrete structure design (2011) GB50010-2010. Beijing: China Architecture & Building Press. (in Chinese).
8.
De DomenicoDMessinaDRecuperoA (2022) Quality Control and safety assessment of prestressed concrete bridge decks through combined field tests and numerical simulation. Structures39: 1135–1157. DOI: 10.1016/j.istruc.2022.03.086.
9.
FengQVisintinPOehlersDJ (2018) Effect of steel bar corrosion on flexural behaviour through bond. Proceedings of the Institution of Civil Engineers–Structures and Buildings171(5): 380–394.
10.
FuB (2017) Study on Bending Performance of Prestressed Beams in Simulated Acid Rain 571 Environment. Thesis (M.D). Nanchang, China: East China Jiaotong University. (in Chinese).
11.
HearstMADumaisSTOsunaE, et al. (1998) Support vector machines, IEEE Intelligent Systems and Their Applications13: 18–28. DOI: 10.1109/5254.708428.
12.
HuangHBurtonHV (2019) Classification of in-plane failure modes for reinforced concrete frames with infills using machine learning. Journal of Building Engineering25: 100767. DOI: 10.1016/j.jobe.2019.100767.
13.
HuiYLLinZSLiR (1997) Experimental study and analysis on the property of corroded rebar. Industrial Construction27(6): 10–13. (in Chinese).
14.
JeonCHShimCS (2020) Flexural behavior of post-tensioned concrete beams with multiple internal corroded strands. Applied Sciences10: 7994. DOI: 10.3390/app10227994.
15.
KioumarsiMMHendriksMANKohlerJ, et al. (2016) The effect of interference of corrosion pits on the failure probability of a reinforced concrete beam. Engineering Structures114: 113–121. DOI: 10.1016/j.engstruct.2016.01.058.
16.
KnightDVisintinPOehlersDJ, et al. (2013) Incorporating residual strains in the flexural rigidity of RC members with varying degrees of prestress and cracking. Advances in Structural Engineering16(10): 1701–1718.
17.
KnightDVisintinPOehlersDJ, et al. (2014) Short-term partial-interaction behavior of RC beams with prestressed FRP and steel. Journal of Composites for Construction18(1): 04013029.
18.
LiFM (2008) Effect of Steel Strands Corrosion on Prestressed Concrete Structures under Chloride Environment. Xuzhou: China University of Mining Technology. (in Chinese).
19.
LiQ (2017) Study on Behaviors of Partially Prestressed Concrete Beams with 600 MPa Non-prestressed Steel Strands. Tian jin: Hebei University of Technology. (in Chinese).
20.
LiFMYuanYS (2010) Experimental study on bending property of prestressed concrete beams with corroded steel strands. Journal of Building Structures32: 78–84. (in Chinese).
21.
LinEC (2006) Experimental Research on the Properties of Prestressed Concrete Beams after the SCC of Prestressed Bars. Thesis (M.D). Yangzhou: Yangzhou University. (in Chinese).
22.
LiuYYFanYFYuJ, et al. (2020) Flexural behavior test of corroded prestressed concrete beams under chloride environment. Acta Materiae Compositae Sinica37: 707–715. (in Chinese).
23.
LuZHYangZNLiCQ (2023) A new empirical model for flexural stiffness of corroded post-tensioned concrete beams. Engineering Structures295: 116860. DOI: 10.1016/j.engstruct.2023.116860.
24.
MaYFLiuYGuoZZ, et al. (2020) Calculation method for flexural capacity of corroded PC beams considering bond degradation. Disaster Prev Mitig Eng40: 1–8, (in Chinese).
25.
MaYFLiuYWangL (2020) Consider bond ability degradation method for calculation the corrosion of the PC beam bending strength. Journal of Disaster Prevention and Mitigation Engineering40: 639–646. DOI: 10.13409/j.cnki.jdpme.2020,04.019.
26.
MaCWangSZhaoJ, et al. (2023) Prediction of shear strength of RC Deep Beams based on interpretable machine learning, Construction and Building Materials387: 131640. DOI: 10.1016/j.conbuildmat.2023.131640.
27.
MangalathuSJeonJS (2018) Classification of failure mode and prediction of shear strength for reinforced concrete beam-column joints using machine learning techniques. Engineering Structures160: 85–94. DOI: 10.1016/j.engstruct.2018.01.008.
28.
MousaviSMAminianPGandomiAH, et al. (2012) A new predictive model for compressive strength of HPC using gene expression programming. Advances in Engineering Software45: 105–114. DOI: 10.1016/j.advengsoft.2011.09.014.
29.
QiuYG (2014) Experimental and Numerical Investigation of Flexural Performance of Corrosion-Affected High-Speed Rail Box Beam. Thesis (MD). Changsha, Hunan: Central South University. (in Chinese).
30.
RinaldiZImperatoreSValenteC (2010) Experimental evaluation of the flexural behavior of corroded P/C beams. Construction and Building Materials24: 2267–2278. DOI: 10.1016/j.conbuildmat.2010.04.029.
31.
SuXMaYWangL, et al. (2022) Fatigue life prediction for prestressed concrete beams under corrosion deterioration process. Structures43: 1704–1715. DOI: 10.1016/j.istruc.2022.07.043.
32.
TangXP (2015) Study of the Corrosion Damage Performance of Unbonded Prestressed Concrete Beam. Thesis (M.D). Chongqing: Chongqing Jiaotong University. (in Chinese).
33.
VecchiFFranceschiniLTondoloF, et al. (2021) Corrosion morphology of prestressing steel strands in naturally corroded PC beams. Construction and Building Materials296: 123720. DOI: 10.1016/j.conbuildmat.2021.123720.
34.
WangXHZhongTY (2005) Relation between the loss coefficient of the corroded rebar's cross-section in concrete and that of its weight. Research & Application of Building Materials1: 4–6. (in Chinese).
35.
YangXD (2016) Study on Flexural Performance and Reliability Analysis of Corrosion Affected Simply-Supported Box Girder for High Speed Railway. Thesis (MD). Changsha, Hunan: Central South University. (in Chinese).
36.
YeMLiLYooDY, et al. (2023) Prediction of shear strength in UHPC beams using machine learning-based models and Shap Interpretation. Construction and Building Materials408: 133752. DOI: 10.1016/j.conbuildmat.2023.133752.
37.
YuF (2013) Study on Mechanical Performance of Partial Prestressed Concrete Beams with Corroded Strand. Ph. D. Thesis. Dalian, Liaoning: Dalian University of Technology. (in Chinese).
38.
YuQQGuXLZengYH, et al. (2022) Flexural behavior of corrosion-damaged prestressed concrete beams. Engineering Structures272: 114985. DOI: 10.1016/j.engstruct.2022.114985.
39.
ZengYHHuangQHGuXL, et al. (2010) Experimental study on bending behavior of corroded post-tensioned concrete beams. Earth in Space2010: 3521–3528. DOI: 10.1061/41096(366)336.
40.
ZhangXWangLZhangJ, et al. (2017) Corrosion-induced flexural behavior degradation of locally ungrouted post-tensioned concrete beams. Construction and Building Materials134: 7–17. DOI: 10.1016/j.conbuildmat.2016.12.140.
41.
ZhangXWangLZhangJ, et al. (2017) Flexural behavior of bonded post-tensioned concrete beams under strand corrosion. Nuclear Engineering and Design313: 414–424. DOI: 10.1016/j.nucengdes.2017.01.004.
