Automated diagnosis systems are necessary for maintenance of superannuated social infrastructures. This paper presents an automated classification method to detect defects of materials using acoustic signals in hammering test. The approach consists of two steps. The first step is extraction of features using Short-Time Fourier Transform (STFT) and the second one is training of classifiers based on AdaBoost which is a kind of ensemble learning algorithm. We use the weak classifiers based on simple template matching method, which can consider both variable scale of amplitude and variable range of frequency. In the experiments, we discriminate between woody and metal materials by different methods of hammering test, which are tapping and rubbing. Furthermore, our method can be applied to actual diagnosis; detection of crack in plaster walls.
This paper deals with design concept of physical measurement standards for optical surface measuring instruments. Using the finite-difference time-domain (FDTD) method, a new simulation tool was developed to analyze optical scattering phenomena on a fundamental laser scanning microscope and to investigate the measured results for inclined plane and sinusoidal, rectangular and triangular cross-section surfaces. Some reasons for outlier inclusion in the measurement results were also made clear. Guidelines for designing the standard surfaces were proposed from a viewpoint of their local geometry within a focused beam spot of the instrument under test.
Machine tool assembly is dynamic and complex because of intrinsic uncertainties in its process. Human dependent production system is often used to deal with uncertainties, but it also increases complexity. Significant research about production scheduling has been done, but effectiveness of its application to human dependent production system is not investigated enough. In addition, not only scheduling but also other countermeasures are applied to improve production performance. This paper presents a framework to analyze how machine tool assembly system responds to supervisors' instruction. The framework includes an event driven simulation, and the simulation model is generated by the facts observed at the real machine tool assembly site. The simulation results reveal conditions which enable to improve production performance effectively at machine tool assembly site.
In this study, we aim to develop innovative stirring technology (electric field non-contact stirring technique (E.N.S.)) applied abrasive control technique with AC electric field. This technique allows for stirring without contact by applying AC electric field to droplets. We examined observation of fundamental behavior and inside behavior for droplet by AC electric field and investigated the feature of electric field non-contact stirring technique as parameter frequency and voltage, quantitatively. As a result, we found that applied voltage affected magnitude of the attractive force acting on droplet, and the resonant frequency of droplet was greatly affected amount of droplet and bottom diameter of droplet, and there is no increase in the droplet temperature, and the motion speed of particles in droplet was accelerated by this stirring technology.
The laser processing with little thermal damage could be achieved by using the femto second laser. One of the unique aspects of femtosecond radiation is the excitation of coherent phonon in the pico-second order duration. We propose a surface processing method with excitation of coherent phonon by using femto pulse train beam. Currently, we are exploring the basic mechanisms of femtosecond laser processing with exicitation of coherent phonon. It is considered that the surface processing during coherent phonon oscillation enables to improve the processing rate of the femto second laser higher than that of the conventional femto pulse laser processing method. In this paper, we report the fundamental experiment to investigate the processing property for Ge(100) with excitation of coherent phonon.
Molybdenum (Mo) is often machined by electrical discharge machining (EDM) because of its high hardness. The cracks did not occur by EDM with a silicon electrode in deionized water. The silicon electrode cannot be used for small-hole EDM because of its heavy wear. Mo was machined by EDM with a Ti electrode in deionized water, while many cracks occurred in the base material and recast layer in the case of the positive electrode polarity. The cracks might generate by crystal grain boundary embrittlement. Because an Mo rolled plate has crystal grain anisotropy, the machining direction has an effect on the crack direction or density. In this study, the effect of the crystal grain anisotropy on the crack generation was investigated. The crack direction and density in the EDMed surface and cross section is parallel to the crystal grain boundary of Mo on the positive electrode polarity. In the case of the machining in the rolling and transvers direction, the crack reached more deeply than that in the normal direction. In contrast, the EDMed surface and cross section indicated crack-less on the negative electrode polarity.