Compressive strength is the most important durability index for concrete structures and also the key factor to assess the quality or condition of concrete. However, existing estimation methods, which are simple and applicable on site for compressive strength, have not sufficient accuracy due to variety of concrete material characteristics. In this study, a compressive strength estimation method for concrete structures using drilling tests and pattern recognition techniques is proposed. By observing the drilling behavior and comparing the physical quantity such as mean and/or deviation of drilling speed, torque, rotation speed, etc., the parameters highly correlated to compressive strength can be extracted. The feature vectors which have good ability of classifying the compressive strength are then obtained by considering the combination of extracted parameters and drilling sections. Based on it, the Support Vector Machine (SVM) is built. The results show that the classification can achieve higher resolution compared with the results of multiple regression analysis. At last, applicability of the proposed compressive strength estimation method is well demonstrated.

View full abstract

Ground motion records during recent 3 strong earthquakes disseminated from NIED are used to investigate seismic site amplification from bedrock to ground surface to be correlated with ratios of corresponding S-wave-velocity in bedrock, Vs(SUB)b(/SUB), and that in surface layer, Vs. Equivalent S-wave velocity Vs in an equivalent surface layer is introduced and determined from peak frequency of amplification ratio. A clear positive correlation exists between amplification ratio of measured motions (surface/base) and the Vs -ratio. Furthermore, amplification ratio calculated for bedrock modified by the amplification of measured motion is also correlated with the Vs -ratio much better than Vs-ratio using average Vs in top 30 m. The empirical formulation thus developed may be applicable to different earthquakes and site conditions including bedrocks with b Vs_b =400-3000 m/s

View full abstract

This paper studied the utilization method of JMA earthquake early warning system for the university campus in Kusatsu city, Japan. The earthquake waves due to the assumed Tonankai earthquake, which may damage to the campus in the future, were calculated using 1-D nonlinear earthquake response analysis. The predicted waves at the campus continue more than 1 minute and its JMA seismic intensity is 5 lower. Considering the characteristics of the earthquake early warning system, the results showed that we have 30 seconds to prepare for the main shock after receiving the information. We surveyed the time spent in taking refuge from seven-story buildings, the attitude after listening to the earthquake early warning message and the danger spots in the university campus. Using these results, we proposed suitable anti-earthquake measures for the university campus by using earthquake early warning system.

View full abstract