I sometimes need air core transformers for contactless energy transfer of small powers (usually less then 1 watt). However, on rotating machines, sometimes a bearing is needed in proximity of the energy transformer. The bearing than gets some coupling to the transformers field and as it's steel just forms a shorted winding, consumes some energy.
Now I have the case of a larger bearing, spanning about half of the transformer coils plane. Obviously, it will short out much of the energy transmitted.
I wonder if it would be possible to remove the inner part of the field, by adding some reversed windings of the bearing diameter to the primary coil. The primary field would then be constricted to the outer part of the primaries core area I guess, leaving the bearing uncoupled. But how many windings would that take? The same amount as there are forward windings in the outer coil?
The situation is most likely this:
red: coils, grey: interfering metal, non-metal parts omitted.
Best Answer
I have to tackle similar problems on rotating machines - sending power and receiving data. Do you use power resonant coils? That's the first question. What sort of power transmission frequency do you use? Steel doesn't have to act like a shorted turn and if it does, it won't completely kill the mag field - with resonance you can re-tune and use the metal as part of the tuned circuit.
Trying to cancel parts of the magnetic field is a fruitless task - Sometimes I use ferrites when the metal (such as stainless with poor conductivity) is heating up too much. Normal steel is better but it still has losses and ferrite does the job of masking areas that would otherwise drain all the energy.
Some diagrams/pictures would help but the above are my thoughts to start off with.