Validation of an Impulse Response Filter for Impact Force Reconstruction on a Hammer Drill

Abstract

AbstractLike many mechanical systems hammer drills are subject to high-frequency external loading. For a proper design it is key to get a good grasp on the commonly encountered loads during operation. However, due to the strongly varying boundary conditions in which they are operated, e.g., hand-held, and the high-frequency range of interest model-based input estimation methods such as Kalman filtering are difficult to exploit for these applications. In this study, an impulse response matrix deconvolution approach has been adopted to estimate the impact forces applied to an existing hammer driller system. This approach relies purely on affordable sensor data for its setup, avoiding the need of calibration of complex numerical models. The impulse response procedure was validated against an experiment with complex boundary conditions with the objective of demonstrating the effectiveness of the method to retrieve accurate estimates in difficult-to-model configurations. The experiments result in high frequency excitations of the hammer and system responses with a high modal density, thus requiring a filter implementation with a high resolution to capture all the system dynamics and prevent numerical instabilities. The experimental responses used in the validation were time history data of strain and hammer impact force. The validation showed that the impulse response filter is superior to model-based estimation techniques and is quite robust to the system intrinsic non-linearities arising from complex damping mechanisms, wave propagation phenomena and boundary conditions.KeywordsBoundary conditionsHigh modal densityImpulse response filterStrain sensorsBroadband excitation