Search

Gas atomized powder is commonly used for Laser powder bed fusion (L-PBF) additive manufacturing because of its good flowability. On the other hand, water atomized powder is inexpensive and has a great potential for application to middle range manufactured products. In this study, SUS316L water atomized powder was applied to L-PBF additive manufacturing. To solve its poor flowability and excessive spatter, we investigated these problems and modified the additive manufacturing machine. Our test results revealed that a laser energy density of 57.8 J/mm3 is one of the best laser conditions for water atomized powder additive manufacturing. To apply the design for additive manufacturing (DfAM), the mechanical properties and overhang process window were revised. Finally, an extruder die plate with better pellet productivity was designed and produced by using SUS316L water atomized powder. The results showed that, it produced better quality pellets than the original one. From this research, we found that water atomized powder is useful for L-PBF additive manufactured products.

This paper proposes an identification method for the excitation force of a rigid body vibration source. This is done by constructing a characteristic matrix without a force signal during a preliminary excitation test. In mechanical products, if an abnormal vibration such as excessive vibration occurs after machine assembly, reworking is necessary to rectify the problem. This extends of the development period and increases cost. Therefore, quantitative vibration prediction at the design state is important. In order to predict quantitative vibration, it is necessary to understand the excitation force of the vibration source. The proposed method of excitation force does not use a force signal during the preliminary excitation test. Therefore, there is no restriction on the vibration method in the preliminary vibration test, and it solves the problem in the case where the excitation point cannot be secured. In this method, the modal mass is calculated using a rigid body mode vector of 6 degrees of freedom (acquired during the preliminary vibration test) and the mass of the vibration source. Then, after determining the mass normalized mode vector from the calculated modal mass, the characteristic matrix of the rigid body vibration source is identified. Finally, the excitation force when the machine is running is identified from the equation of motion in the frequency domain. As a result of testing the verify the validity of the method, the proposed method can identify the excitation with the same precision as when using the force signal and has demonstrated its usefulness.

This paper proposes identification method of excitation force of rigid body vibration source by physical model identification. If the reworking occurs after the assembly of the prototype in mechanical product, the cost increases due to the retrofit countermeasures and development period extension. Pre-prediction of vibration at the design stage is important to avoid these problems. Prior prediction of vibration needs to grasp the excitation force of the vibration source. As conventional methods, the mount stiffness method and the matrix inversion method have been proposed. However, mount stiffness method calculates the mount transmission force. Therefore, if the development machine changes the structure, preliminary evaluation does not apply. The matrix inversion method can solve this problem. However, when the frequency response function contains a measurement error, the error spreading propagates in inverse matrix calculation. Therefore, in order to avoid the inverse matrix calculation, we propose identification method of excitation force of rigid body vibration source by physical model identification. In this paper, it was investigated features of the inverse matrix method and the identification method of excitation force by physical model identification using the basic experiment. As a result of study, the method by physical model identification showed that the influence of measurement error is smaller than the matrix inversion method. In addition, it showed that it is possible to identify the excitation force with fewer excitation points than the matrix inversion method.