The 8th International Conference on Engineering Innovation and Seismic Mitigation of Bridges （ICEISMB 2025） (ICEISMB 2025)

To study the influence of grouting defects on the anchoring performance and failure modes of large-diameter grouting sleeves, defect size (40 mm, 60 mm, and 80 mm) and defect location (middle, end) were selected as influencing factors for combination. Twenty-four grouting sleeve connectors with diameter 40 mm HRB400E steel bars as anchoring bars were designed and fabricated. Uniaxial tensile tests were conducted. A finite element model of large-diameter grouting sleeves was established based on the tests. The simulation results were verified with the test results. A Support Vector Machine (SVM) model database with steel bar diameter, defect size, and defect location as variables was established. The SVM algorithm was used to predict the failure modes. The results show that the specimens mainly exhibit two failure modes: steel bar fracture and steel bar pull-out. When the defect length is 40 mm, the steel bar pull-out failure occurs in large-diameter grouting sleeves with diameter 40 mm steel bars. The increase in defect length leads to a reduction in the ultimate bearing capacity and ultimate displacement of the specimens. When the defect length is within 60 mm, the reduction in both is small. When the defect length reaches 80 mm, both decrease significantly. Under the same defect size, end defects have a greater impact on the anchoring performance of large-diameter grouting sleeves than middle defects. The finite element model can effectively simulate the test results. The model exhibits the same failure mode as the specimens. The maximum error in ultimate load is 5.9%. The load-displacement curves show good agreement. The prediction accuracy of the SVM algorithm for the failure modes of grouting sleeves with middle defects and end defects both reaches about 90%. Meanwhile, the algorithm performs better in predicting the failure modes of grouting sleeves with middle defects than those with end defects. This algorithm provides an intelligent diagnostic tool for predicting the failure modes of grouting sleeves.

bridge engineering; large diameter grouting sleeve; uniaxial tensile tests; grouting defects; numerical simulation; support vector machine