This study proposes a quantitative and systematic evaluation framework for rationally determining investment priorities in maintenance projects for heterogeneous road infrastructures such as bridges and tunnels. In Korea, conventional maintenance decision-making relies significantly on empirical judgments and policy-driven preferences, thus resulting in inefficiencies, inconsistencies, difficulties across facility types, as well as limited transparency in budget allocation. Hence, a multicriteria decision-making model integrating three key indicators–defect (performance), economic value (asset-based benefit), and risk (safety)–is developed. In particular, the economic evaluation introduces the concept of asset-value recovery and employs artificial intelligence-based machine-learning models (i.e., random forest, light gradient boosting machine, and extreme gradient boosting) to estimate reasonable replacement costs and quantify benefits in monetary terms. The proposed model enables the correlation between these quantitative indicators with maintenance project types to prioritize investments by combining benefit scores and risk indices. The case study demonstrates that the proposed framework enhances the objectivity and efficiency of budget allocation and enables data-driven investment prioritization instead of policydependent decisions. Moreover, this approach provides a foundation for transitioning from experience-based decisions to data-driven infrastructure management. This methodology can be further expanded to include probabilistic risk assessment and life-cycle cost-based management frameworks, thus ultimately contributing to sustainable evidence-based decision support systems for national infrastructure asset management.

In this study, we investigated the unit process parameters in spherical kernel preparation. Nearly perfect spherical microspheres were obtained from the 0.6M of U-concentration in the broth solution, and the microstructure of the kernel appeared the good results in the calcining, reducing, and sintering processes. For good sphericity, high density, suitable microstructure, and no-crack final microspheres, the temperature control range in calcination process was , and the microstructure, the pore structure, and the density of kernel could be controlled in this temperature range. Also, the concentration changes of the ageing solution in aging step were not effective factor in the gelation of the liquid droplets, but the temperature change of the ageing solution was very sensitive for the final ADU gel particles

The effects of thermal treatment conditions on ADU (ammonium diuranate) prepared by SOL-GEL method, so-called GSP (Gel supported precipitation) process, were investigated for kernel preparation. In this study, ADU compound particles were calcined to particles in air and Ar atmospheres, and these particles were reduced and sintered in 4%-/Ar. During the thermal calcining treatment in air, ADU compound was slightly decomposed, and then converted to phases at . At , the phase appeared together with . After sintering of theses particles, the uranium oxide phases were reduced to a stoichiometric . As a result of the calcining treatment in Ar, more reduced-form of uranium oxide was observed than that treated in air atmosphere by XRD analysis. The final phases of these particles were estimated as a mixture of and .

Incoloy 800H, KSA (Kaeri Superalloy)-6, Inconel 600 및 Hastelloy C-276 합금의 용융염에서의 부식거동을 650~850˚C 온도범위에서 조사하였다. LiCl-Li2O혼합용융염에서의 부식은 Li2O에 의한 염기성 용해 기구에 의해 진행되며, 부식속도가 LiCl에서보다 훨씬 빠르게 나타났다. 혼합용융염 LiCl-Li2O에서는 Ni기 합금의 부식속도가 Fe기 합금보다 빠르고, Mo와 W의 함량이 높은 Hastelloy C-276이 가장 빠른 부식속도를 나타내었다. 용융염 LiCl에서는 LiCrO2의 단일 부식층이 형성되고, LiCl-Li2O 혼합용융염에서는 산화물과 Ni의 2상구조의 다공성 부식층이 형성되었다.