The longer I think about this generator, the more I get the feeling that there is something fundamentally wrong with figure 2 of KD 718. So discussion about this subject on this forum helps to develop the idea. I now think that the magnetic flow goes directly from a magnetic north pole to a magnetic south pole. So only four instead of eight magnetic loops are coming out of the armature and the iron poles in between the magnet poles simply don't become magnetic at all. So the armature is much less stronger than expected. It might be possible to solve this problem but therefore I have to go back to the original idea about this 16-pole generator.
For the original idea, all sixteen poles had a row of magnets. So the poles N1, N2, N3 and N4 were a row magnets with the north pole to the outside and the poles S1, S2, S3 and S4 were a row of magnets with the south pole to the outside. For this construction, you get the magnetic flow lines as given in figure 2 but now every magnetic loop goes through two magnets. The calculation of the magnetic flux density in the air gap as given in chapter 3, shows that the magnetic flux in the air gap is very high, even if there is only one 5 mm thick magnet in a loop. So this is one of the reasons why I cancelled half of the magnets. But instead of cancelling half of the north poles and half of the south poles, I should have cancelled all four magnetic south poles. So if all four poles N1, N2, N3 and N4 contain a row of magnets, the magnetc flow is forced to go through the four iron south poles. This will give the same magnetic flow pattern as shown in figure 2.
Using only magnets for the north poles has some disadvantages. One is that the distance in between adjacent magnets becomes very small at the bottom of a magnet groove. So it might be necessary to use a larger frame size for the same magnet dimensions. Another is that now the magnets can be scratched if the armature is mounted in the stator. Having an iron pole in between each magnetic pole makes that the magnet can't touch the stator during mounting. I still have to find a methode to prevent scratching of the magnets during mounting if only the north poles have magnets. A 6-pole armature is less sensible to this problem as it will have three groups of south poles under 120°.
If there are 16 mechanical armature poles and 36 stator poles, there is a preference position every 2.5° and so there are 144 preference positions per revolution. This might be enough to flatten the fluctuation of the sticking torque such that the peak on the sticking torque is low enough. But it seems possible to double the number of preference positions if the group of four mechanical south poles is rotated 1.25° left or right hand with respect to the group of four magnetic north poles. The peak on the sticking torque caused by the north poles now doesn't coincide to the peak on the sticking torque caused by the south poles. The four south poles must have grooves in between the poles to make that a south pole also has a width of 15 mm.
As figure 2 for the 16-pole generator with frame size 100 is wrong, it means that the description of all five other frame sizes is wrong too and report KD 718 has to be reviewed completely. Frame size 112 wil be chosen instead of frame size 100. I will do this in the coming days and I will mention it on this forum when the old report is replaced by the new one. I still think that it is worth while to develop this idea because the standard motor winding can be used and because the total magnet costs are rather low for this big generator.
