|
Title |
Combined Non-Destructive Test Strength Estimation Model for Aged Concrete via Age-Correction and Sensitivity Enhancement
|
|
Authors |
조현석(Hyeon-Seok Jo) ; 노영숙(Young-Sook Roh) |
|
DOI |
https://doi.org/10.11112/jksmi.2026.30.4.38 |
|
Keywords |
노후 콘크리트; 복합 비파괴 시험법; 강도 추정 모델; 재령 보정; 변동계수(COV); 평균제곱근오차(RMSE) Aged Concrete; Combined NDT method; Strength estimation model; Age-correction; Coefficient of variation (COV); Root mean square error (RMSE) |
|
Abstract |
Accurate estimation of the compressive strength of aged reinforced concrete structures is a critical element for evaluating structural safety and predicting the life-cycle management of facilities. This study proposes a “three-stage integrated combined non-destructive strength estimation model” that couples field data from safety diagnostics of large-scale buildings with age-specific physical characteristics. To this end, the statistical characteristics of different age groups were analyzed using a total of 395 sets of field non-destructive testing (NDT) data - specifically rebound hardness and ultrasonic pulse velocity (V) - sourced from domestic detailed safety diagnostic reports, covering structures aged 7 to 50 years with design compressive strengths of 21 MPa and 24 MPa.The proposed model incorporates three distinctive stages: deriving a baseline multiple linear regression equation based on ordinary least squares (OLS) (Stage 1), introducing age-specific correction coefficients through measured strength ratio analysis (Stage 2), and implementing design strength-centered sensitivity control (Stage 3). To validate the practical applicability and effectiveness of the developed model, a comparative error analysis was conducted using a total of 43 independent data points from destructive testing of actual core compressive strengths, where design information was strictly excluded.The final performance validation demonstrated that the proposed model achieves high accuracy and reliability, yielding an average strength ratio of 0.979, a coefficient of variation (COV) of 3.10%, and a root mean square error (RMSE) of 1.29 MPa. Compared to widely used domestic and international conventional formulas - such as the Architectural Institute of Japan (AIJ) equation (COV 24.41%, RMSE 8.98), the Korea Authority of Land & Infrastructure Safety (KISTEC) equation (COV 18.14%, RMSE 18.36), and the European RILEM equation (COV 20.27%, RMSE 5.34) - these results represent a significant improvement, reducing the error variance range to approximately 10% of the conventional levels. The outcomes of this study are expected to maximize the reliability of quantitative judgments made by field engineers during detailed safety diagnostics of long-term operational aged facilities, significantly contributing to the advancement of structural maintenance systems.
|