The dynamic behavior of three different designs of a cantilever boring bar are compared by means of forced vibration tests. These are a bar with diametrically opposed flats machined along its length, a bar fabricated from laminates of steel and a damping compound, and a bar fitted with an impact damper. The impact damper boring bar is found to be the most effective, and an improved design, giving increases in stable metal removal rates of more than 100 percent, is outlined and tested. A theoretical analysis is presented for predicting the effectiveness of the impact damper with a spring supported impacting mass. This analysis enables the optimum mass ratio and gap setting of the damper to be selected for specified characteristics of the vibrating systems to which the damper is fitted.

This paper examines the effect of a spring-supported impact mass on the performance of an impact damper. This technique was developed after it was found that gravity detracts from the performance of a normal impact damper. The springs are used to support the impact mass at the mid point of its stroke so that the effect of gravity is negated. The analysis presented in this paper is based on the assumption that the motion of the impact mass is stable and that there are two impacts in each cycle. The assumption that the impacts are equi-spaced is not made, and it is shown that such behaviour occurs very rarely in the resonant region. Theoretical predictions are verified experimentally, and the analysis is then used to develop design charts for the impact damper at various damping ratios, mass ratios and gap ratios. It is shown that the spring is detrimental to the performance of the damper if gravity is not present, but that the performance under gravity is improved to that of the device without gravity. It is shown that very light supporting springs make the accuracy required in centring the impact mass too high, so that the performance in practice increases with spring stiffness. The performance of the device at natural frequency ratios greater than 0.5 is not studied.

No abstract