Some results of GARTEUR Action Group HC-AG 19 on methods for improvement of structural dynamic finite element models

Thumbnail Image
Coppotelli, G.
Conti, E.
Tongeren, H. van
Journal Title
Journal ISSN
Volume Title
The issue of vibration in helicopters is of major concern to operators. This requires close attention to the vehicle dynamics. The ability to faithfully simulate and optimise vehicle response, structural modifications, vehicle updates, the addition of stores and equipment is the key to producing a low vibration helicopter. GARTEUR Action Group, HC-AG14, concluded that helicopter dynamic models are still deficient in their capability to predict airframe vibration. The AG looked at the methods for improving the model correlation with modal test data along with the suitability of existing shake test methods. The helicopter structure tested in AG14 was suspended in the laboratory. However, this is not the operational environment where there are very significant mass, inertia and gyroscopic effects from the rotor systems. Nowadays, modal analysis consists of two principal approaches: experimental modal analysis (EMA) and operational modal analysis (OMA). The EMA evaluates the modal parameters by considering that the excitation and the response of the system are both measurable. The OMA evaluates the modal parameters using only the measured response. The lack of knowledge of the input is replaced by the assumption that the input is a distributed stochastic load, constant in a broad frequency band, e.g. white noise, and uncorrelated in space. This hypothesis, nevertheless, is restrictive in rotorcraft applications, because in these cases the load is characterized by harmonic components, i.e. deterministic signals, originating from the rotating parts. A new action group HCAG19 was formed to study the benefit of using in-flight dynamic data for improving finite element models. Methodologies were assessed to evaluate vibration measurements from flight tests. The objective is to extract modal parameters and demonstrate that the dynamic model can be updated using this data. This paper presents one of the approaches developed by the University of Rome �La Sapienza".