The purpose of this study is to propose a modified average flow model that can be used for mechanical seals with parallel sealing faces, and clarify their lubrication characteristics. The general average flow model established by Patir et al. could not consider the effects of surface roughness between parallel faces. In the present model, the flow factors with considering side leakage effects between parallel lubricating faces are implemented. In addition, the model allows one to consider the additional expected hydrodynamic pressure generated owing to the relative motion of rough surfaces. The flow factors are evaluated as the average for several different surfaces with the same statistical properties. In this first report, the effects of the root mean square roughness on the lubrication characteristics are investigated, assuming that the probability density function of the roughness follows the normal distribution. The results show that as the root mean square roughness increases, the flow rate increases. In addition, the optimal roughness for the friction coefficient is changed with the rotational speed.

The aim of this research is to establish the modified average flow model for applying to mechanical seals with parallel sealing faces. In the first paper, the effects of the root mean square roughness on the lubrication characteristics were investigated by the model, assuming that the probability density function of the roughness follows the normal distribution. In this paper, the flow factors are evaluated for smoothly finished surfaces with the several different skewness and kurtosis. The lubrication characteristics of mechanical seals are calculated by using these factors. The results show that as skewness decreases, the film thickness and frictional coefficient decrease. In addition, the flow rate has the lowest value at certain skewness. On the other hand, the film thickness, flow rate and frictional coefficient decrease as kurtosis decreases.

We have proposed a fast stair-climbing robot with a simple hopping mechanism. The robot, which consists of two (upper and lower) bodies connected by springs and wire, travels using wheels mounted on its lower body, and hops to climb stairs by releasing energy stored in the springs. The trajectories of the bodies during hopping depend on the design and control parameters. The former are the two body masses and the spring constant, which cannot be changed after fabrication. The latter are the initial spring contraction and horizontal velocity, which can be changed during locomotion. This mechanism allows the robot to climb stairs quickly and to land softly without complicated feedback control. In our previous research, we analyzed a defined soft landing point, i.e., a no-impact landing point (mathematically, a stationary, inflection, and descending point on the lower-body trajectory), using equations of motion, clarified the conditions, and showed the existence characteristics that the point is created by the friction during the spring contraction phase. On the basis of this result, we fabricated the stair-climbing robot and realized fast stair climbing and soft landing. However, the robot cannot climb any stairs with riser heights other than the predetermined riser value. In this paper, to achieve soft landing for multiple riser heights, we analyze how many soft landing points can be creatable by changing only the control parameters using identical design parameters. It is shown that an infinite number of solution sets creating two soft landing points exist, a unique solution set creating three soft landing points exists, and no solution set creating more than three soft landing points exists. Additionally, a numerical simulation showed that these parameters had values in a practical design range.

In order to shorten design period, CAE analysis using 3D-CAD is required to evaluate design results before prototyping. However, designers cannot often use CAE analysis to do a lot of design work. Therefore, knowledge of analysis procedure and analysis know-how from designers is difficult to accumulate. On the other hand, analysis working efficiency has to be improved because CAE analysis working hours for designers are limited. Many studies have been achieved for supporting CAE analysis of designers. Most of them are automation technique of CAE model generation. This paper focused on CAE analysis procedure to improve analysis efficiency. We have developed a CAE analysis procedure standardized method. We applied this technique for thermal analysis. The standardized thermal analysis procedure consists of eight procedures, namely, extraction of analysis object parts from assembly, unit setup, material setup, model simplification, mesh generation, boundary condition setup, calculation execution, and visualization of results. We demonstrated that thermal analysis work time is shortened maximum 50 % by using CAE analysis procedure standardized method. Evaluation result of persons experienced and inexperienced in analysis showed that the procedures of model simplification and boundary condition setup are important.