There are two types of stress happening to a turbine blade: an alternating, or vibration, stress and a centrifugal static stress. As blades rotate, they experience natural resonant frequencies at different rpms. While vibrating, the blade is also under a centrifugal load as if it is being pulled from root to tip. For a complete stress test, all stresses should be tested for simultaneously.
There have been attempts to pull on the end of a vibrating blade to test the interaction of both static and dynamic stresses, which has proven unreliable due to failure of the blades at the tip before reaching adequate loading or cycles. Pulling on the blade also creates a constant stress from root to tip. Actual blades experience a stress gradient (high at the root and zero at the tip) since one end of the blade is free. This fact also adds to experimental error.
“Other than using dynamic spin testing, the only alternative to test for both accurate static and dynamic stress is to run the blade in an actual jet engine. This is not only extremely expensive, but to test the part to the point of failure will ruin the engine,” said Robert Murner, President, Test Devices Inc. (TDI). “What we developed was a way to induce customer-specified vibrational modes and associated amplitudes in a spinning environment that require very tight speed control and adequate excitation force to produce the right test conditions for those blades.”
The process basically imparts dynamic (vibrational) and static (centrifugal) stresses in jet turbine engine blades to validate the predicted blade life. The interaction between static and dynamic stresses is often depicted in a Goodman diagram, where the vertical axis represents dynamic stress and the horizontal axis represents static stress. The Goodman line typically intersects the vertical axis at 107 reverse bending cycles (static stress=0) and the horizontal axis at the yield strength of the given material under test (dynamic stress=0). The area under the Goodman line represents safe life for the blade, and the area above the Goodman line represents failure.
Static stress is produced by spinning the component at a prescribed high speed, which varies by component. Dynamic stress is produced via a patented oil jet excitation method. The impact of the oil on the rotating blades is what forces the amplitude of the resonant vibrational stress to a level sufficient to reach the Goodman line. Natural vibrational frequencies are achieved by replicating the appropriate engine order in the spin rig.
Instrumentation typically includes a non-intrusive stress-measurement system (NSMS) and strain gauges. If the customer desires a heated test, thermocouples may also be involved. An NSMS plot of the excitation mode, stress amplitude (1500 microstrain), frequency, and rpm is generated. This particular test requires exceptionally tight speed control (±0.25 rpm) to hold on resonance for 3-4 h at a very high Q factor (poorly damped) mode. Conditions inside the spin rig change during the test, which produces slight variations in the natural blade frequency(s). These variations require altering speed slightly during the test to hold resonance.
Investigation of crack initiation and propagation is currently under development. Murner noted that TDI’s dynamic test method is well suited for blade crack growth studies.
“Understanding how a crack progresses, specifically the time it takes to reach a critical length, in a specific type of blade drives the frequency of inspection intervals, which drives the cost to maintain a particular turbine engine,” he said.
Testing various blade damping methodologies and anti-wear coatings, such as anti-fretting coatings, are good opportunities for this test method as well. Also, testing to determine foreign-object damage (FOD) effect on blade life and helping to design blades that are more FOD tolerant are continuing activities.