Vibration Behaviour of Servohydraulic Dynamic Tensile Testing Machines

Abstract

Abstract The oscillation behaviour of the Schenck and Roell-Amsler servohydraulic dynamic tensile testing machines used at the RWTH IEHK has been investigated [1]. The dynamic longitudinal and bending oscillations, propagating through the test set-up, have been characterised with a Fast Fourier Transformation (FFT) and a simple analytical model. When increasing the strain rate, the Eigen frequency of the test set-up becomes a limiting factor. With higher testing speed the oscillations amplitude and inertia effects also increase, which makes it difficult to determine accurately the mechanical properties as usually done with quasistatic tensile tests. A set-up optimisation can be reached by decreasing the mass between the load cell and the sample, either by moving the load cell closer to the sample, or with a lighter machine design and grip shape. A combination of both possibilities has been followed at IEHK. A global piezoquartz load cell, a quasilocal calibrated strain gage load cell on the grip, and a local calibrated strain gage load cell on the sample can be used as load measurement systems. The corresponding load signals have been compared with regard to initial oscillation amplitude, frequency, and damping factor for different strain rates. The positive effects of a lightweight sample grip design are also shown and quantified. The initial amplitude can be reduced easily by using damping rubbers. Advantages and draw backs of this method are shown. Short sample geometries are also beneficial to increase the vibration frequency. Knowing the vibration behaviour of the machine, it is then easier to classify the materials to be tested dynamically, based on the quasistatic yield ratio and fracture elongation values. Since vibrations cannot be avoided with conventional servohydraulic machines, an adequate smoothing method must still be chosen to determine the mechanical properties and flow curves. Low pass filtering results are compared with analytical polynoms and the traditional spline smoothing method used at IEHK.