Following specification of design requirements by the Sir Lawrence Wackett Centre for Aerospace Design Technology, work was completed on the evaluation of a new means of exciting aircraft structures in flight, for Federal Aviation Regulations Part 23 (FAR23) flutter certification. A secondary goal was the redesign of the existing system.
Design, experimental and theoretical work was carried out, with an aim to developing the knowledge base for a new form of aerodynamic exciter unit.
While the experimental work proved to be unsuccessful, the theoretical work produced a number of important results.
The most significant of these results was a finding that the flat plate present in the model had only a minor effect. This result was, however, not entirely unexpected. It was anticipated that the bulk of the force would be generated through a basic momentum transfer from the rotating vane.
The effect of cylinder gap spacing was also investigated, with the finding that a close gap provides a greater force output. However, it can be expected that small gaps would provide a force that is greatly limited by rotation angle.
Sufficient theoretical data has been produced during this investigation, for design work to commence upon the aerodynamic excitation unit.
The author wishes to thank the following for their efforts in making this thesis possible:
Mr. Sean O’Meara, for his frequent help in overcoming difficulties. It was greatly appreciated.
Mr. Ken Lee, for his generous assistance with the wind tunnel.
Mr. Adam Randall, for his advice and assistance with the wind tunnel balance calibration.
Mr. Steve Kilpatrick, for the manufacture of the test unit.
The Sir Lawrence Wackett Centre for Aerospace Design Technology regularly conducts flutter certification of FAR23 aircraft. To this end, they require reliable means of exciting aircraft structures in flight.
This report is an investigation into in-flight vibration excitation systems for certification of general aviation type aircraft. Various methods of excitation have been evaluated, with an aim to improving the methods currently in use by the RMIT Aerospace Department.
The main aims of this investigation were:
Flutter is a dangerous phenomenon encountered in flexible structures subjected to aerodynamic forces. This includes aircraft, buildings, telegraph wires, and bridges. In the case of an aircraft, flutter is of particular concern, due to the innate flexibility of the structure and extreme aerodynamic loads experienced.
Flutter occurs as a result of interactions between aerodynamic, stiffness, and inertial forces on a structure. For an aircraft, flutter may occur when the aircraft is accelerated to a speed where, when disturbed, the wings flex, and the resultant vibrations do not have sufficient damping. The damping of an aircraft’s vibrations is a function of the speed at which it is flying. [13]
Figure 1: Torsion/bending flutter
Illustrated above is the simplest, two-dimensional case of bending/torsion flutter. When a wing hits a gust it experiences an increase in lift. This causes the wing to flex upwards, as shown. However, for many wings, due to the location of the centre of pressure and elastic axis of the wing, such bending is combined with torsion. The resulting torsion causes a change in lift, causing the wing to swing back downwards. At the lowest point, the wing beings to twist back up again. This happens naturally to an aircraft in flight, but for the case of flutter, such vibrations are divergent, and hence unsafe. The resulting wing flexure may cause structural failure.
In-fight flutter testing is required for certification of an aircraft. The aircraft is required to be safe up to its design dive speed.
To verify the safety of an aircraft requires its designers to test fly the aircraft up to this speed, taking all reasonable measures to ensure sufficient excitation is given to elicit any potentially dangerous vibrations. This involves flying the aircraft near the design dive speed and providing sufficient excitation that, if flutter were possible at this speed, it would be induced.
Flying the aircraft at this speed is insufficient, since flutter requires some form of gust or turbulence to occur, even if the system is unstable. A disturbance, such as a wind gust or turbulence, must be applied to trigger this phenomenon.
In practice, flutter testing involves progressively increasing the speed of the aircraft until either the design dive speed is reached, or the test engineer believes that further flights would be unsafe.
During the test flights, accelerometers or strain gauges mounted on the wing, tailplane, fuselage, and control surfaces, provide data for the test engineer. Accelerometers are generally applied in places where deflections are expected to be high (the wingtips for example), and care is taken to avoid placing them near node points. Feedback from these sensors is then spectrally analysed to determine whether system damping has become unstable. [13]
Current methods of exciting airframes
The current methods for excitation in flight-testing are:
Pyrotechnics involve attaching a charge to the airframe, and setting it off to provide a thrust. The rockets themselves produce an initial disturbance rather than a direct oscillation.
This method is rarely used, due to safety reasons, and because a very limited number of charges may be carried in each test run. Several charges may be required to be triggered simultaneously in each test. [1]
Aerodynamic excitation involves attaching some form of aerodynamic exciter to an airframe. This could be in the form of a rotating vane or oscillating airfoil, or more exotic mechanisms. This method of excitation is dependent on both the airspeed and angle of attack of the aircraft, and hence can have difficulty in producing sufficient excitation at low speeds. An advantage in this method is that it may be attached to the wingtip, hence providing maximum excitation, with minimum force. [1]
Mechanical excitation involves attaching a rotating, unbalanced mass to the interior of the airframe. Such a mass provides an oscillatory force. Drawbacks to this method are that the force produced increases as the square of the excitation frequency. Given that most flutter tests require excitation through a wide range of frequencies (typically between 5 and 50Hz), it is difficult to both produce a significant excitation at low frequency, and a structurally safe load at high frequency. Another disadvantage of this method is that it requires large excitation forces, since forces are generally applied within the fuselage, which makes adequate excitation of the wings and empennage difficult. [1]
Control surface excitation involves a slight input of a control surface. This could involve either a pilot tapping the stick or rudder pedals, or modification to the flight control system. This method has been found to be unreliable, as it only excites low frequencies. [1] Crashes due to over-reliance on this excitation method have been reported. [2]
Spectral analysis and the power spectral density function
Spectral analysis involves use of a transform to convert time domain data into the frequency domain. For this investigation, a computational method known as the Fast Fourier Transform (FFT) was used. [10]
A more useful means of viewing frequency data is the power spectral density function. This is essentially the absolute value of the FFT of a sample. [10]
Source-sink panel methods are a means of computationally modeling the aerodynamics of a body. The basis of this model is:
