186 ALQUIST GEARING FOR SHIP PROPULSION. 
that under which a rigid gear of the same dimensions usually is operated. I can assure 
Mr. Emmet that the “fearsome” picture he has conjured up, in the paragraph preceding 
that from which I have just quoted, of the disasters likely to follow the adoption of such 
gears from “the vibrations of supporting structures” have never been even suggested in a 
now extended experience. 
My criticism of the Alquist gear will be directed to the form shown diagrammatically 
in the lower part of Fig. 2, Plate 101, but will apply largely to the other two forms. The 
high-speed pinion I will call Pinion A, and each gear meshing with it, Gear 4; those of 
the second reduction I will call Pinions and Gear B. Those who have not studied high- 
speed reduction gears carefully may think that in what follows I split hairs and refer only 
to negligible values. But in this subject 1/1000 inch is a large and important quantity. 
1. Pinion 4 is the well-known De Laval pinion. Mr. Emmet, near the bottom of 
page 183, claims that he can use Pinions A smaller in diameter and longer than is possible 
in other types of gear (though this is not borne out by his Fig. 5, Plate 104, which shows a 
Pinion A short in proportion to its diameter). Let us examine the result of adopting such 
a long, slender pinion unsupported in the center. Unless nearly equal powers are taken off 
by the two gear wheels meshing with it, this pinion bends excessively. This difficult con- 
dition has restricted its success, but the success of the Alquist gear depends on the condition 
being fulfilled almost perfectly. Mr. Emmet does not draw attention to this great difficulty, 
but it has been fully recognized by the designers of the Alquist gear and the means by 
which it is attempted to overcome it, as stated in the patents, is interesting. 
In considering this it is first to be noted that the structure of the complete gear is re- 
dundant. This is evident since the gear would still run after we had removed one Gear A 
of the first reduction together with the Pinion B attached to it. But the power transmis- 
sible by the gear would then be greatly reduced, as the slender high-speed pinion would have 
no support against cross-bending. The number of teeth in Gears A and Pinions B are 
made prime to one another, so that if one of these gears is running as an idler it can be 
made to take stress by advancing it one or more teeth in its engagement with Pinion A. 
But, in a redundant structure, it would require the measurement of microscopic strains to 
tell if each part were taking its proper load and slight calculation, for a particular case will 
show that the adjustment provided is far too coarse, since the advance per tooth would be 
much greater than the strain under full load; and also much greater than could be compen- 
sated by any elasticity due to the laminations of the gears, so that this device is of little 
service here. In some earlier designs, Gear A drove Pinion B through a flexible shaft which 
helped matters a little, but that complicated the design and seems to have been dropped. 
Again, even if once adjusted for equal load on Gears A, slight straining of the gear case or 
wear or heating of a bearing would completely upset the equality. I regret that Mr. Emmet 
has not told us how this difficulty is overcome. 
2. It would have been interesting had Mr. Emmet told us how the very flexible plates 
in Fig. 1, Plate 100, were machined so that the touching surfaces, at the center and outside, 
were all true and in the same plane. Ordinary methods of turning would hardly suffice 
even if the flexible plate were bolted down on a stiff, flat surface, for the side first bolted 
down is not finished true and the plate will usually be sprung. Should this not be accom- 
plished with great accuracy the plates will be sprung when bolted together for tooth cut- 
ting; then when the parting tool separates them the strains will be relaxed and the teeth 
on contiguous disks will not be truly spaced, as when cut. 
Oe dy 
