Integrated Approach for the Structural Random Vibration and the Associated Structural Fatigue Life Prediction Due to the Launch Acoustic Load

Abstract

During the Launch phase, Spacecraft structures are submitted to heavy acoustic excitation mainly generated by the engines and the aerodynamic forces over the whole the vehicle. The acoustic excitation creates a vibration environment that must be characterized to correctly design the secondary structure for a multimission operational profile. Various design approaches for predicting the random fatigue life of structural items have been proposed and developed. These methods generally include predicting the random loads, estimating the fluctuating stress response of the structures and predicting the life from stress versus cycles to failure curves for the relevant material. As the random loads increase, the stress response and the structures fatigue behaviour become more non linear and very difficult to predict with the classical computation methods, because the theory of non linear random vibration has not reached a good state of maturity. Although many methods of solution exist, there can be no general rule about the suitability and reliability of any methods. Moreover, the fatigue process which reduce strength in composite materials are generally very complex, involving the accumulation of multiple damage modes which may combine in a variety of way to produce numerous failure modes. For the fatigue life assessment, in the literature there are many methodologies based on simple analysis for evaluating response prediction, using fundamental mode approximation for estimation of induced stress and strain caused by random loads. The derived stress levels in rms form for particular load cases, are used together with rms stress endurance data to estimate fatigue damage for each case. The overall structural damage can then be used to obtain the fatigue life of the relevant component. However, life estimates are not so accurate and often factors correlating predictions with tests are enough large. Possibly reasons for these discrepancies might be first the use of rms levels in the damage analysis; they cannot be expected to reflect the more complex behaviour of the range of stresses they are representing. Second cause might be the use of available endurance design data to cover a too wide range of structural geometry. A probabilistic model for damage accumulation in composite is therefore the most suitable one. Consequently, probabilistic concepts and methods are needed for an assessment of the safety and reliability of metallic and composite under fatigue. In the CEE BRITE EURAM RESEACH “AMADEUS”, an analytical formulation based on a statistical method has been proposed to evaluate the “expected” fatigue damage on structures subjected to random loads during their service life. The target of the present paper is to show the reliability of the simulation techniques based on Finite Element Method (FEM) and the statistical Energy Analysis (SEA) modelling, adopted for the derivation of spacecraft Random Vibration Environment (RVE) integrated with a newly developed algorithm to predict the multiaxial Random fatigue for structural panels models. Based on this Design verification approach ,the potential Design criticality can be identified in the early design stage of a project and where required local structural modifications can be implemented during the development phase ,resulting in an optimized design at reduced cost.