Is the Weis-Fogh principle exploitable in turbomachines?

Change log
Furber, Stephen Byram 

In this dissertation we ask whether the Weis-Fogh effect, by which a hovering insect flies with an aerodynamic performance superior to anything previously known, can be exploited in turbomachinery. We think the answer is yes.

Normal turbomachinery design is based on the analysis of isolated cascades of blades with steady entry and exit flows. The interactions between adjacent cascades, and non-uniformities of the flow, are regarded as serious problems which have to be minimised. Unsteadiness gives rise to noise and causes damaging vibration. Despite these known disadvantages we take here the view that some benefits might accrue from controlled blade interactions, and that these advantages should be quantified. They might possibly outweigh the disadvantages that normally drive the designer to minimise blade interference. We therefore take the positive view of proposing a novel type of turbomachinery stage which depends on the interaction between the rotor and the stator for its normal operation. The stage exploits the Weis-Fogh principle, and has the unusual property that when started from rest it generates a pressure rise without shedding any vorticity into the fluid. We conclude from both our analysis and our experiments that there are definite performance advantages for stages of this new type, and that the noise and vibration problems they are likely to cause are not significantly greater than those found in existing high performance aero-engines.

Our experiments were performed to check the validity of a theoretical model, and these are described in detail. The results seem to show that strong blade interactions can cause aerofoils to work usefully at higher mean loadings than those in conventional designs, and that the increased loading is also associated with an improved aerodynamic efficiency. Our analysis seems to predict the experimental performance very satisfactorily, and we see no fundamental reason why the beneficial effects should not carry over to the higher speed larger machinery of engineering importance, but we realise that the difficulties of engineering our device will be far from trivial! We would look to our principle for a measurable increase in the efficiency of such machines together with an improved surge margin. In our experiment these amounted to 3 percent and 10 percent respectively, and we are encouraged to expect these figures to be relevant to real machines from our interpretation of some test data on large scale machines with low blade clearances.

Doctor of Philosophy (PhD)
Awarding Institution
University of Cambridge