Our posts crossed – sorry.

The DH was selected over the Rotol based entirely upon Petter’s personal preferences. It does indeed seem an odd selection – maybe DH had a better salesman? There is a ‘rule of thumb’ that one finds recurring in the technical reports – the more efficient the prop section is at subsonic speeds, the earlier it hits compressibility issues. It may be that the nice fat DH blade even performed slightly better during initial testing? But I don’t know the answer.

Both props were exactly 10ft in diameter.

DH were not asked to do anything to the 54409 blade as fitted to the Whirlwind, as far as the available record shows, anyway. They WERE asked to alter the 55409 for the Spitfire (from 7.6% down to 6%) to increase efficiency at altitude, as such things as Mcrit were increasingly understood.

I’d argue the relevance of the Spitfire data – These were the only operational types really working in this ‘envelope’ (though I hate that word) up to 1940. While WW wasn’t really explored, the Spitfire was tested within an inch of its life – so its the best analogy we can now get when it comes to prop compressibility effects, for example.

I have a fair amount of data on the DH blades – we measured one – which I will dig out this lunchtime. Essentially the WW DH blade is RAF6 section, 9.6% at 0.7r. the rest of the blade curve (t/c, chord and twist) I will get to you. It is very similar to the Spitfire DH blade in many respects, but has a planform identical to an earlier US Navy design, and is thicker. In fact, really very unusually thick.

I have nothing on the Rotol blades used at all, apart from their diameter, a probable Clarke-Y section (Higher Mcrit for thick sections – this was standard for Rotol) and circumstantial evidence that they might have had a 9% t/c ratio (the RAE 24ft wind tunnel report and Vickers test data, again).

No aerodynamic changes.

The Spitfire figures are all TAS, and are from here – http://www.spitfireperformance.com/spit2prop-b.jpg. They were compiled by Vickers in their standard tests, the methodology of which did not vary allowing for direct comparisons. In fact most of the propeller trials plotted on the graph were carried out on the same aircraft – K9791. That’s why the data is so significant in showing the effect of propeller type – and crucially thickness – on performance when operating at these speeds. This trial data shows a variation in max TAS, and a variation in the height at which this attained of over 2,000 feet – on an otherwise identical aircraft.

The logical point I was making there was that full throttle height and altitude of max TAS are not the same thing, and on an otherwise identical aircraft the former remains constant and the latter varies with blade design.

..and the implication I was then going on to make was that performance drop-off with altitude might have as much to do with the prop hitting compressibility issues – becoming less and less efficient, as local Mcrit is approached by the blade. This is an altitude effect as much as boost pressure drop, but depending upon the blade design can be more or less marked, and the effects become significant at lesser or greater altitude.

I don’t know whether you waded through the whole thing, but the point was there were no tests of the DH-bladed production Whirlwind. All of the altitude issues were reported in service. The only altitude tested prototype had thinner Rotol prop blades, and that was given a service ceiling of 31,000 feet. In service it was said that the aircraft would simply not go over 27,000, and it’s effective ceiling was 25,000. It was NOT significantly heavier – and the problems at altitude were with ‘boost and propeller revolutions’

So there were no ‘like for like’ tests, but you will note that the Rotol bladed prototype was called effectively the ‘first production’ aircraft as it had, in the view of Martlesham Heath, no differences. It really seems that blades were not considered significant, which when you think that they are the only means of transferring an engine’s potential power into actual thrust and that their efficiency in doing so was already known to vary greatly (and in fact NACA data shows that 9.6% thick blades were doomed to hit compressibility problems at WW climbing rpm around 25,000), was an odd omission. But it happened.

I know that I have not proved anything, but the Spitfire test data appears sound.