##plugins.themes.bootstrap3.article.main##

L. Allegrucci F. De Paolis A. Coletta M. Bernabei

Abstract

A Light Utility Helicopter (LUH), in the course of a training flight, leaving the ground during the taxi to take off, went into an uncontrolled rolling to the right; consequently the helicopter gradually laid down on the right side. The impact with the runway destroyed the rotating blades up to the hubs rotor. The accident investigation focused on main rotor oscillatory plate servo actuators . These components, directly linked to the cloche movements, regulate main rotor blades plane tilt and pitch. Following the preliminary examination, only front servo actuator attachment was found to be broken in two parts. In detail, the present paper deals with the fracture analysis results. The servo actuator attachment material is a 2014 Aluminum alloy extrudate,  undergone to T651 heat treatment. Fracture surfaces were examined by optical and electronic microscopy in order to determine the main morphological features and consequently to trace the origin of failure mechanism and causes. The accordance with the specification requirements about alloy composition was verified by quantitative elementary analysis through inductive coupled plasma spectroscopy (ICP); furthermore, semi-quantitative elementary analysis was locally verified by Energy dispersion spectroscopy X ray (EDS_RX). Finally, the hydrogen content of the material was evaluated by the total hydrogen analysis. Microstructural and
technological alloy characteristics were verified as well by using metallographic microscopy and hardness testing of the material.
Macroscopic fracture surfaces evidences were characterized by the lack of any significant plastic deformations
and by the presence of symmetry compared to the servo actuator axis. Microscopic fracture features of both the investigated surfaces were not coherent to the hypothesis of an impact of the main rotor to the soil. Further
achieved evidences, such as grain boundary fracture propagation, the presence of corrosion products, were all in accordance with a Stress Corrosion Cracking (SCC) progressive fracture mechanism.
Finite Element Analysis (FEA) located the highest tensile stress value, when the servo actuator is in its nominal
working condition, at the same points where the corrosion products were more concentrated (i.e. in the part of
the fracture exposed to oxidative air effect for the longest time). The good agreement between FEA and
morphological evidences allowed to determine the progressive fracture origin area, though it was not possible to individuate the crack initiation point. In fact, in correspondence to the initiation area of both the fracture
surfaces, shining and flat morphology was found; then there were evidence of plastic deformations, due to the
detachment of a servo actuator part.


The ICP analysis and hardness testing results were in accordance with the material specification requirements.
However, the hydrogen content was one order of magnitude greater than the required value and many and
unexpected globular formations were observed on the fracture surface. Part of these were dendritic formations, while the others looked smooth and shining. Further, FESEM boundary grain observation gave evidences of a high presence of precipitates on the investigated surfaces. Hence, observed microstructural characteristics, boundary grain precipitates and globular formations allowed to hypothesize possible overheating/eutectic melting phenomena, occurred during manufacturing processes.
As widely reported in literature, the AA 2014 alloy is one of the aluminum-copper-magnesium-silicon type,
employing copper aluminide (CuAl2 ) as the primary precipitation-hardening agent. The need for a maximum
Cu phase dispersion in solid solution requires a heat treatment range with an upper limit (507°C) that is near to
the melting of the eutectics (510°C). Moreover, since the 1960s, AA2014 has been defined as sensitive to SCC.
This condition is mainly related to the presence of coarse-grained and aligned CuAl2 precipitates. This
arrangement is due to an overheating (more than 507°C) or to a cooling process carried out too slowly.
Microstructural analysis was carried out on three items: 1) a large portion of the broken actuator attachment; 2)
on a servoactuator coming from the same production batch; 3) on a servo actuator coming from a different
production batch. The microstructure from the broken actuator attachment showed a great amount of precipitates (second phases) lengthwise aligned to the boundary grain, pores, and also cavities and dendritic globular formations. Analysis results, morphology evidences and reference images available on scientific literature were found to be in excellent agreement and validated the embrittlement and subsequent SCC mechanism hypotesis (intergranular failure propagation).
In conclusion, flight accident causes are attributable to main rotor actuator attachment failure. Failure mechanism is classifiable as SCC supported by microstructural anomalies of the material. The investigation of the manufacturing process highlighted how one of the servo actuator batches was not properly produced due to poor control and accuracy of heat treatment temperature and/or cooling time. This led to hydrogen embrittlement and to a microstructural problem (globular formations and boundary grain precipitates). The combination of those phenomena caused an increase of the SCC sensitivity and were the basic progressive failure driving forces. Nevertheless, as above mentioned, alloy composition was found compliant with the material specification requirements and this just because none of the scheduled quality control tests is able to determine the peculiar microstructural anomalies reported.

Comments

  1. Latest Oldest Top Comments

    ##plugins.themes.bootstrap3.article.details##

    Section
    Articles

    How to Cite

    Allegrucci, L., De Paolis, F., Coletta, A. and Bernabei, M. (2013) “Crack of a helicopter main rotor actuator attachment: failure analysis and lessons learned”, Frattura ed Integrità Strutturale, 7(26), pp. Pages 104-122. doi: 10.3221/IGF-ESIS.26.11.