Robust Updating of Operational Boundary Conditions for a Grinding Machine

摘要

Operational boundary conditions of actual structures are often a crucial and difficult aspect in model updating of CAD/CAE models based on dynamic tests and experimental modal identification. In this work two grinding machines are experimentally tested through roving hammer impact tests. Due to the importance of comparing two different solutions of grinding structures involving traditional cast-iron or a new polymeric concrete, actual operational boundary conditions, consisting in industrial supports acting on a not-well-known pavement, are taken into account. Their effects affect the dynamic behaviour, in particular at low frequencies and they are the main updating parameters of the CAD/CAE models. Gradient-based optimization and genetic algorithms are implemented to reduce modal parameters errors in the identification and updating phase of parametric models to demonstrate feasibility and improvements of robust numerical models, in order to evaluate effective improvements of the material choice for the grinding machine.