Any aerodynamics in real life is time-dependent. No technical fluid flow is really constant in time. This starts at smallest scales with random molecular movements, goes over turbulence and flow phenomenon like vortex shedding and ends in flow conditions like compressor surge that affect a whole turbomachine. Thus, steady-state flow conditions are only a special and simplified approximation of general aerodynamics. However, admittedly often a good one.
Fluid flow around bodies creates aerodynamic forces. Thus, time-dependent flow over bodies (like airfoils for example) causes time-dependent (unsteady) aerodynamic forces. And, as nothing is really rigid the unsteady forces cause time dependent deflections of the airfoils. In big commercial airplanes the wingtips move up and down up to 2 meters due to gust response of the wings (look out of the window next time you fly). Turbomachine-blades experience similar unsteady forces. As parts of a turbomachine rotate the blades pass periodically through the wakes of upstream lying blades or obstacles (struts, combustion chambers etc.). Thus the airflow in a turbomachine is by default time-dependent.
The periodic aerodynamic forces start vibrations of the blades. Usually the vibrations are small and should stay this way. But, under special conditions and when the excitation frequency is close to a natural frequency of the blades the vibration amplitudes become quite large. When they become to large the material of the blades finally will rupture. This can cause a failure and even the destruction of the turbomachine. If the machine is a steam-turbine for power generation this can spoil your evening because you will have momentarily no electricity. If the machine is a jet-engine attached to the plane bound for your holiday destination this can spoil more than just the rest of your vacation (but don't worry, flying is still the most secure means of transportation).
Even without a catastrophic failure unsteady aerodynamic forces may lead to vibration levels of airfoils that considerably reduce the lifetime or the necessary maintenance intervalls of a turbomachine. In recent years the turbomachine blades have become more slender and the loading has increased in order to improve the efficiency. But, that resulted also in a higher sensitivity towards mechanical and aerodynamic excitation. Any small increase in efficiency demands more and more financial efforts for a relatively small fuel saving. However, the fuel savings can easily be annihilated if a lower life time due to blade vibration demands more down times of a machine which can cost a lot of money. Thus, it is obvious that one is interrested to predict and to prevent operation conditions where excessive blade vibration occurs.
The difficulty to predict time-dependent aerodynamic forces consists in the fact that unsteadiness in turbomachines is a global problem in contrast to a local problem like heat transfer for example. To predict the heat transfer for a certain spot in a machine one has to know the flow properties at this specific point. If the flow velocity, the viscosity of the flow, the pressure, the turbulence level, the temperature etc. are known for this spot one can give the heat transfer as a function of these parameters. To establish this function is certainly not an easy task. But, once this function is known all other points with the same flow parameters will have the same heat transfer rate. Thus, heat transfer is a local problem and thus can be predicted when the steady-state flow is known.
To predict the unsteady-pressures and forces at a certain spot on the other hand it is not enough to know all the steady-state flow parameters. One has to know also the geometry and the state of the flow at each point of time in the whole flow field. The unsteady pressures generated on an oscillating surface depend on the relative geometry changes during the oscillation cycle, pressure waves propagate in the flow field from one spot to another and vibrating bodies in a flow field or phenomenon like vortics and flow separations modify the time-dependent flow field itsself. Thus, unsteady flow is a global problem because to predict it one needs not only to know the local flow properties but one needs also a lot of information about the world around. Simplifications are possible. But, strictly spoken each flow problem is unique as regards to its unsteady behaviour.
Even with the most advanced numerical codes it is not possible to completely solve general time-dependent flow cases. And even with advances in hard- and software it won't probably be possible for a long time to come. However, with some simplifications it is possible to predict a lot of flow cases that are important for turbomachinery. A major task of the research in unsteady aerodynamics is to find out what are reasonable simplifications in numerical codes and to provide experimental data to check these codes.
The main objectives of the research work at LTT are:
|
Annular cascade experiments |