In this report, the remote non-contact measuring systems which were developed by RTRI were introduced for the health monitoring of railway structures. Three types of ‘U-Doppler’ systems were produced by improving the laser Doppler vibrometer. Furthermore, the unmanned aerial vehicles for structure inspection were developed installing a stereo camera, crawlers, and a hammering test device. By using these inspection systems, we proposed new and efficient inspection techniques for the cable tension of cable-stayed bridge, the harmful alteration of bridge girder surface, the material properties of RC members, and the rockfall risk of rock slope along railway tracks.

　　　In late years optical measurement techniques such as 3D laser scanning or image analysis has been studied to apply for inspection of infrastructure. Authors demonstrated to apply 3D model created by image analysis “Structure from Motion” using images shot with Unmanned Aerial Vehicle (UAV) to inspection of long span bridge without temporary scaffold. However, it is not a method that everyone can easily use the technique such as UAV with risk. Moreover, most of the bridges to be maintained are medium and short span bridge in Japan. Therefore, authors verified the optimal method such as the efficient shooting method and reproducibility of the 3D model and difference in equipment in Structure from Motion for the medium and short span bridge without UAV. In order to make it easy to inspect even those without special knowledge, we aimed at manualization of the recommended inspection method.

　　　Effects of microbubble on immersion cleaning of lubricant oil on a stainless steel plate were investigated to understand basic characteristics for cleaning using microbubble. A stainless steel plate with lubricant oil was immersed in a water tank with microbubble generated using pressurized dissolution method. The immersion cleaning was performed using three kinds of oil with different kinematic viscosity (32, 68 and 220 mm²/s), four surface roughness of the stainless plate (0.047, 3.20, 5.84 and 9.13 μm) and three pH of the cleaning water (4.10, 7.30 and 9.86). In order to understand the removal process of the lubricant oil from the plate, the process was observed using a high-speed camera. The following conclusions were obtained: (1) the cleaning water with microbubble can remove the lubricant oil more than from two to four times of that without microbubble regardless of the oil viscosity, (2) the removal rate increases with decreasing the oil kinematic viscosity, (3) the cleaning water with microbubble can remove the lubricant oil regardless of the surface roughness except for 9.13 μm, (4) the removal rate increases with decreasing the surface roughness of the plate, (5) the cleaning water with microbubble can remove the lubricant oil about two times of that without microbubble regardless of pH of the cleaning water except for 9.86, (6) the oil is removed by the buoyancy force of millimeter order bubble which is formed by coalescence of several small bubbles generated in the oil.

　　　By using the Kinect sensor, we measured desktop reach movements (anterior-posterior back-and-forth movements and interior-exterior ones) that stroke patients and healthy controls performed. Maximum trajectory width was measured in the hand, elbow, shoulder and head. The maximum trajectory width in anteroposterior axis of the shoulder (anterior-posterior, healthy controls) and in horizontal axis of the head (interior-exterior, total subjects) showed significant differences between test score and retest score (p<0.01). This result suggests that we cannot measure close movement precisely by Kinect. But as for maximum trajectory width in anteroposterior axis of the shoulder and in horizontal axis of the head during anterior-posterior back-and-forth movements, the following five results were recognized: (1) a high interclass correlation coefficients of total subjects (0.81 - 0.86),(2) moderately strong correlations with upper limb dysfunction (0.49 - 0.65), (3) significant differences between stroke patients and healthy controls (p<0.01),(4) larger differences between stroke patients and healthy controls than MDC (total subjects), (5)significant differences between participants with severe paralysis and those with mild one (p<0.01). These results show that the Kinect sensor employed in upper-limb training system is useful for assessing compensatory movements of the trunk related to upper limb paralysis.

　　　The generation of compression and expansion waves is one of the biggest issues faced while designing high-speed trains. These waves are generated at the moment the nose and tail of the train enter the tunnel. The vibration of high-speed trains running in a tunnel is known to be greater than that of trains running in an open area. This vibration has been explained by the asymmetrical flow separation and vortex shedding from the surface of the train. In this paper, the vibration might be related to the expansion wave that is generated as the tail of the train enters the tunnel. In this study, a combined simulation of flow and vibration was developed based on computational fluid dynamics and vibration dynamics. The results showed that the pressure difference at the sides of the train caused by the expansion waves passed from the tail to the nose with a time lag. The accuracy of the proposed method was discussed by comparing the results of a real high-speed train. Consequently, we demonstrated that aerodynamic force was strongly associated with lateral vibration when the train entered the tunnel.

　　　Methane diffusion microflame was established on a stainless steel or copper tube burner with the inner (outer) diameter range from 0.7 (1.0) mm to 6 (8) mm and the methane flow rate range from 137 to 275 mm³/s. The temperature profile of burner surface and the heating heat flux to the wall above the burner were measured. Within the measurement conditions, the highest heating heat flux (6.8 W/cm²) was achieved by the SUS tube burner with an inner diameter of 0.7 mm at the distance of 6 mm from burner tip to the wall when methane flow rate was 275 mm³/s. It is shown that a thinner tube has higher burner temperature and higher heating heat flux, except for the copper tube burner with an inner diameter of 1.0 mm. In terms of the heating heat flux to the wall above the burner, the copper tube burner is more strongly affected by the tube size than SUS tube burner.