A controversial Whirlwind Theory..

I find these peregrinations fascinating, but two esteemed members of the forum butting heads like elk in the forest, distracting. Gentlemen, keep playing the ball, it is fascinating. Still, it would be fun to be a second to either of you in a duel, with historically accurate ball and powder, some wretched maiden rushing on a foaming horse in the distance to stop her lovers in their madness, gratefully succeeding and allowing us all to retire to a few pints in the pub.

Yes, I wanted to establish what changes caused the Whirlwind to pass testing at Martlesham in March 1940 (I think, but I am in a van in Dorset with no references) and be given a service ceiling of 31,000 ft and then cause the first 263 WW Squadron Leader to call it useless above 26,000 in a communique to Dowding. There was no second set of flight tests, the aircraft passed the first and only and then entered service.

The whole point is that it passed with a Clark Y Rotol prop of 8% t/c ratio at 70% and entered service with a RAF-6 DH prop of 9.6% and yet Martlesham said to the RAF that the aircraft were identical so that no-one seemed to pick up on the real difference – at the time or since.

The aircraft in service was reported anecdotally as useless at height and this went to Dowding. Dowding, who was concerned for his pilots more than anything else then told Beaverbrook that he didn’t care for the Whirlwind because of the danger from 109’s at height and Beaverbrook halted production. thus anecdotes are very important to the story – more so than absolute figures. Ever since the engines had been blamed – and by association Rolls Royce – even though no one could be specific as to how and why. I have set out to prove this wrong and unfair on Rolls, by virtue of understanding what really changed between acceptance into service and what the service then said about the aeroplane.

As there was no second trial, instead we have to look to tests with mods for any DH-equipped WW’s performance curves. The one that shows performance with height was in – I think – Jan 1942, but this isn’t too relevant. Of course it shows slower speeds than a Spitfire I (the aircraft was carrying two 250lb bombs externally), but that’s not the point, it’s what we can glean (when we know what other equivalent graphs look like and know what to look for) from the shape of the graph.There’s no need to get too bogged down in this.

Far from adding two and two and making five as would seem to be the accusation, I have made four. It’s actually more of a stretch to say that an engine that was essentially a Kestrel, a tried and tested design with a tried and tested supercharger should suddenly develop ‘non-specific’ yet altitude-related ailments when applied to the Whirlwind (but not the tested prototype). That really is taking two and two and making five, though without an appreciation of propeller Mach issues it’s an understandable leap. Once made, this guess at the causes quickly became ‘fact’ – my aim was to question this, and then tell what I found when I did.

You called the theory controversial and wrote that you prepared for the arguments. Think of this as nothing more than testing those statements, not because I have any personal point to make, it is because I cannot understand what you are are wanting to prove or the relevance of the information you introduced to prove your claim.

As I now understand it, you wish to establish what changes to the WW design caused such a drastic change in top speed at altitude between two flight test dates? Is this correct?

If so, what were the dates of the tests?

The vertical axis is altitude and the horizontal one True Air Speed. I am sure you knew that really. It is a poor reproduction but as I am working from a camper van in Dorset while on a family holiday I cannot provide what you are rather obstinately demanding. The numbers, as I have already explained, while available do not matter as much as the shape of the curves on the graphs. If it still means nothing to you then there is no more I can do. For what it’s worth the vertical runs from 10,000 to 22,000ft and the horizontal I have marked up.

No, I am not trying to establish why the WW was slower at altitude than the Spitfire. I am trying to establish why the Whirlwind was much slower and apparently less powerful at height than a) the prototype was, b) the specification dictated and c) it should have been given it’s performance lower down. I use the Spitfire for a comparison between drop off in performance with height without altitude-related problems (other than progressively less dense air into the supercharger) and drop off in performance with height with some additional factor exacerbating, as per WW.

This has been understandably attributed to the superchargers alone and Rolls blamed since 1940 but it has never made sense to me as it was all proven and predictable engineering, developed from the supercharged Kestrel, all other Rolls products behaved similarly to each other and not like this and what’s more it only manifested itself on the production aircraft with DH props. That was my first clue.

Yes, it was down to Mcrit. This of course reduces with altitude. What actually happens in the real world is more and more of the blade, moving inboard, passes through Mcrit as one climbs – and fatter blades do this sooner than thinner ones. Also RAF section blades do this sooner than Clark Y ones.

Constant speed units do not care what the source of the drag is. They can’t filter out wave drag. They will react by reducing the AoA of the blades. This reduces forward speed. The more wave drag on the blades, the slower the aircraft. Changing perspective for a moment, this actually applies to several aircraft I am studying in this context, and was major driver for jet development.

I have calculated the areas beyond Mcrit for WW blades based on the US Hamilton equivalent on which DH based their blades, and also the drag rise on those areas. It’s a bit ‘back of fag packet’ but the principle is demonstrated.

This problem with US 9.6% thick blades was the reason DH (working with Vickers Supermarine and the RAE) thinned the Hamilton 6503 to make the 7.6% DP55409B of the Spitfire I. They didn’t seem to feel the need for the WW, but it would have helped if they had.

Just briefly, but unfortunately exactly when the WW flew, we were applying 1920’s blade design to aircraft encountering very different speeds and altitudes.

Apologies for not doing all of the maths again here and now for you. The data I used is out there and you are welcome to check it all adds up if you remain incredulous – though I can assure you it does.

I have of course considered the hp and drag of the aircraft – it’s a starting point in all this. There are some very nice calculations done by the Farnborough wind tunnel team who investigated a full sized one – and even found something odd about propeller efficiencies against flight tests. These are also available from the National Archives.

Wing planform hasn’t been a focus for me – can you expand on that, in terms of how it affects performance with altitude? Thanks.

The graph in your post last evening means nothing without labels. As I now understand it, your research is trying to establish;

Why was the maximum speed of the WW at altitude less than that of the Spitfire I – is that correct?

And that you put this lack of comparative performance down to your idea that this was because of the propellers fitted to the WW reduced the available thrust by their having higher drag than the one installed on the Spitfire, and the reason for the greater drag (you say) was due to the blades becoming Mcrit? Am I correct?

And you have not considered;

Did you consider the comments made by Harold Penrose regarding modifications made to the WW after it had undergone testing at Farnborough? [QUOTE]In June the first prototype was ready for testing with a new nacelle-ducted exhaust system, but decreased top speed was inevitable because of the greater drag and loss of injector effect.[/QUOTE]

To add – propellers have always had what Douglas Adams called an ‘SEP field’ around them. Everyone from the A&AEE when they claimed their tested WW was no different to those in service to Mr. Brat today seem to believe it is the engine which provides the ‘thrust’ component of the famous four-arrow lift/drag/weight/thrust diagram for a propeller aircraft. It is not, it is the propeller. Change the propeller and you might as well have changed the engine.

In order to prove that the idea does not work, rather than staring at graphs one would have to demonstrate one of the following:

Until one can show this, I would suggest that this has more bearing than wing loading – which did not change between prototype as tested at A&AEE and production.

The problem we have in quantifying the problem with absolute figures is that there are no plots of performance apart from those for the Rotol prototype or variations on production – wing thickening and bombs. However, these latter DO show an increased gradient in the max speed curve compared to prototype and other comparable aircraft, regardless of absolute figures (which of course should be how we look at his anyway).

It is the difference in the rate of change, not the absolute figures, which is the ‘smoking gun’ when we come to consider the physics. I hope I am explaining this point well enough – what someone who does not understand the principle would think was a weakness (no absolute figure comparison) is in fact just the correct approach as one cannot usefully compare absolute figures. One can look at “if and how” those figures change as external parameters change and then look to mechanisms that might cause this.

This is a basic principle of engineering and applied maths – calculus over algebra, If you want to apply it to the graphs that do exist, be my guest, but it won’t prove anything that a visual check on the contemporary graphs and a grasp of what is going on with the aircraft won’t.

You think I’m making this up? Why would anyone do that? OK. Let me show you these, at the risk of boring anybody else not so determined to demonstrate my inability to read a graph.

[ATTACH=JSON]{“alt”:”Click image for larger version Name:tww.png Views:t0 Size:t296.6 KB ID:t3870853″,”data-align”:”none”,”data-attachmentid”:”3870853″,”data-size”:”full”}[/ATTACH]

The background image is from an RAE test of a standard Spitfire I R6770, added is Whirlwind from trials in 1942 with bombs. You can find the data through the National Archives, apologies for no reference here. Both have equivalent performance with the same curve up to FTH. Above this there is variation – the absolute maximum speed drops off faster with height with the WW than it does with the Spitfire. The difference is in the degree to which altitude impacts performance. but only above FTH. This last tells us it’s related to thrust, not aircraft drag. So, what changes with altitude? The speed of sound. What was critical? Mach around the prop blades. That’s why it’s called critical Mach. What affects critical Mach on prop blades? Thickness. What was different between these two aircraft (and also the difference between this production WW and the ‘performing as expected’ prototype)? Blade thickness.

Blade – or wing – thickness doesn’t make a difference to drag until it does. It does when transonic speeds are reached. You put excess drag on a CP prop blade and what does it do? It turns. It reduces pitch. That reduces thrust. It does take a lot of mental picturing to get the whole thing, but it really does work and the texts are out there.

Sorry to present it like this, but you wanted to follow my reasoning. Here you have the data and a simplified version of my thought process. As I say, I have written about this in greater detail elsewhere, and all of the references are available. If you struggle to understand after absorbing all of this that is not my fault, please make more effort. If you need more data – and I do have it – I suggest you do the rest yourself. Yes there is something odd about the difference beginning at around FTH, but if you think about it enough you will understand that with a CP prop this is how it is going to manifest itself – this and a slight variance in apparent FTH from that specified by the engine manufacturer (which also is in the data). No, I will not work this out here for you, it’s tedious – as an expert able to dismiss others so readily I am sure you can work it out yourself.

Of course we could look at all aspects of the aircraft, but this is a theory about one of them. It would only be worthwhile to explore other things in relation to it if you were desperate to find some other reason why the theory isn’t feasible. Maybe you will, but wing loading won’t cut it. Personally I can’t see why you have such a fly up your intake about this.

I do appreciate that you are a very experienced and knowledgeable person. Sometimes, however, this leads to a form of arrogance that means that anyone who comes up with something you hadn’t thought of must a) be wrong and b) be slapped down. Please make sure you understand the principle already put forward here and in much greater detail elsewhere (rather than demanding I show you my working like some teacher who doesn’t really get it himself), before dismissing further. I say this with all due respect.

The performance drop off of the Whirlwind with height when compared to other types in production is there in the figures provided by the RAE and respective manufacturers. It is extant as a series of figures and graphs which are public record and which do not require regurgitation in order to then explore reasons for an unexplained difference.

Yes you do need to show your working and thought process.

How can you compare one aircraft type with others then in production without making a through investigation and comparison of ALL aspects of each aircraft type? You are not quantifying the problem, simply supplying the reader with supposition.

*******
You say that the Whirlwind performed less well at altitude than other aircraft types of the time. One very important factor that will limit an aircraft’s ceiling is wing loading. It’s not the only one, but for the purpose of this argument I think it is useful. Did you consider that?

Compare the Whirlwind, Spitfire Vb and Hurricane 2c using wing loading data from about the same time period:

Whirlwind 45.78 lbs/sq ft.
Spitfire Vb 27.68 lbs/sq ft
Hurricane 2c 33.89 lbs/sq ft.

The three types that different wing sections; Clark YH, NACA 2213/2209.4, and NACA 23015. All three aerofoils have very similar max Cl at comparable Re. No real surprise then that the Whirlwind had a ceiling some 6,000 ft lower than the other two is it?

You will note that I haven’t shown my working, but since “figures and graphs which are public record, and which do not require regurgitation…”.

TT, that’s interesting, it’s always good to get first hand info. In the wrong order – the slats were locked shut fairly early on in the aircraft’s career at which point this stopped being a problem – whether this was ‘known’ by those encountering the Whirlwind would depend upon when they did so.

Single-engine performance wasn’t so much lack of power as a lack of feathering capability on the bracket-type props – there were no 4,000 series Hydromatic equivalents available as the design just hadn’t come from the US.

Engine fires were, it would seem, largely rumour and bar anecdote – I cannot find one example actually documented.

Ex Brat, I have just seen this old thread promoted by a recent post, which I will come to next.

The performance drop off of the Whirlwind with height when compared to other types in production is there in the figures provided by the RAE and respective manufacturers. It is extant as a series of figures and graphs which are public record and which do not require regurgitation in order to then explore reasons for an unexplained difference.

Dowding did say in writing that he was unwilling to commit the aircraft in its primary role because of the altitude problems – which is pretty terminal for a new type. He then entered into correspondence with Beaverbrook, who, from the correspondence, was the one to order a halt in production somewhat arbitrarily as was his style – though for sound reasons that emerge from the correspondence, all of which is on record at the National Archives.

The largest operational problem with the Whirlwind was clear and is a matter of record.

If you do not understand the argument then please read it again, or perhaps take a look at my TAH article on the subject which spells it out very carefully, at length and with full references. It also explains that the aircraft did meet its agreed specifications at test, and this was because the aircraft tested had Rotol propeller blades and not the de Havilland ones that were a problem. It is alarming but absolutely the truth as documented in black and white that the two aircraft (tested prototype and production) were deemed ‘identical’ and the problems with performance at altitude that dropped below those at test were was called ‘inexplicable’.

All I have attempted to do is explain them. Please don’t be so derogatory, as you have been in this post and others. I have been carefully researching this for years and much as it might pain you to accept it it does actually stack up as a theory. It is not arm-waving or some kind of paranoid conspiracy – just something that was overlooked because trans-sonic theory wasn’t really where most minds were ‘at’ at the time. Some were, and the Spitfre’s 55409 was thinned as a result. It was generally known that the blades as they came from US Hamilton blanks were too fat and thus inefficient, though it was less well understood at the time why.

I haven’t done any research on the Whirlwind, never had the time, but my impression has always been that the aeroplane was made of some pretty exotic materials for the period and that these had to be imported. Given Beaverbrook’s passion for limiting production of aircraft to a few important ones (the right decision at the time), the Whirlwind just had to go.
Having said that, the few aeroplanes produced lasted in squadron service until (I think off the top my head) 1944 so the extraordinary performance that these materials bestowed on the fighter so just what could have be achieved with a single seat twin engine fighter that could escort bombers to Berlin and back. A pity that Petter thought only in terms of metal and not wood.
Wasn’t there a proposal to develop the Whirlwind with Merlins?

My old mate F/L D.W. ‘Doug’ Sturgeon was an unusual chap in many ways not least because he flew Whirlwinds then Mossies, then Hornets so could compare the lot. He noted that whilst the single engine handling caught a lot of people out and is given as the principal reason for their failure (apart from engine fires, underpowered etc) what people never commented on were the HP slots – he told me that these were cable operated and could be locked but the cables had a nasty habit of snapping. It was a while ago now so don’t quote me but I think he said the cable routed around the engine and the heat could cause one to snap – – one would then deploy with the other retracted – leading to a flick roll when one wasn’t expecting it.

This is from memory now so I may be misremembering… he was ex- 263 sqn and a top bloke

TT

I fail to understand the argument the OP is trying to make. He has launched into trying to establish a theory using peripheral RAE and NACA documents without actually stating what the problem with the Whirlwind was. A dropoff in performance as altitude increases is usual for all aircraft types. The reasons can be multiple.

Can the OP quote actual documentation and numbers to this perceived performance shortfall of the Whirlwind, or is the evidence simple anecdotal?

Did the type achieve its agreed specifications before service entry, and if not, what official reasons were given for failing to do so?

I have convinced myself. Critical Mach at ‘tip’ is in fact at 0.91 radius, after which the profile misbehaves as the thickness tapers towards the rounded tip.

You can see the maximum measured speed converging with the theoretical point at which the props will stop helping you go forward, starting at Full throttle height and agreeing -showing the limiting factor – above 20,000 feet.

This is why squadron pilots were on record as being confused as to how Martlesham got theirs to 31,000ft, when their aircraft stopped being of much use above 26,000.

All of the performance testing was done on prototype L6845, which had Rotols. The squadrons got De Havillands:

[ATTACH=CONFIG]241161[/ATTACH]

Ah, I see.

The problem, if it’s real, did affect its future. It made it sluggish at height, and this counted very greatly against continuing with production (I have that in writing), and development – and thus being thought a ‘success’. The Spitfire was different – The Spitfire was a proven aircraft with a development ‘roadmap’ attached by the time the disappointing altitude performance of the MkII (a fast aeroplane, otherwise) came to light. By contrast the Whirlwind hadn’t even been delivered to squadron strength at the time, and the programme was eminently ‘haltable’ – and it was.

Yes, fascinating material but I meant that it didn’t affect the production of the Spitfire Mk.II. Similarly, had the Whirlwind been more successful, the tip problem may not have affected its future.

..doesn’t seem to have affected the Spitfire Mk.II too adversely.

It did a bit – it would have been 5mph faster 500ft higher with the DH blade. This table also handily includes the compressibility factor for each blade, you can see the issue with the Rotol. The one I highlighted was (I think) the production standard. It was the most efficient of the Rotols tested, and as is often the case it was also the one most affected by compressibility. This might be why the MkII was faster lower down, but slower higher up, than the MkI:

[ATTACH=CONFIG]241147[/ATTACH]

It is actually surprisingly difficult to find a precise date on which the Whirlwind was ‘written off’. Thanks to one of our researchers I am now looking at probably the entire record set held by the National Archives on the Whirlwind. The short version of the timeline is that on 8th May 1940, in response to the concern expressed by George Bullman (the Air Ministry’s Engine ‘Head’) about the effect of continued Peregrine production on Rolls’ output, WR Farrar, the AD to the DTD simply said in a memo ‘No New Whirlwinds’. This is as close to an official decision as I can get.

[ATTACH=CONFIG]241143[/ATTACH]

However, before, during and after this decision was being made there was a discussion carrying on about the Mark II (AVIA15/317). This would have had Morris radiators, 100-octane-ready Peregrines, greater tankage, revised cannon feed arrangements and Hydromatic props, no mention of blade changes, just the greater pitch range imparted by the Hydromatic mechanism.

No decision was made on this until Dowding wrote to Beaverbrook in October 1940 saying (in response to a direct question) that although it was still to early to say for definite he’d not be unhappy if no further Whirlwinds were produced. I doubt very much that Whirlwinds were anywhere near the front of Dowding’s mind at the time.

Beaverbrook replied with characteristic unilateralism (and in line with Alertken’s post above) that he would therefore not produce any more Whirlwinds. It looks like the axe was ultimately wielded by the Minister for Aircraft Production acting independently and without official sanction (beyond a three-word memo five months earlier), but all official parties were happy to tacitly let it go, meaning they could stop thinking about the Whirlwind and focus on other things.

There was a last comment on the matter made at a meeting in February 1941, more a post-rationalisation of the decision made by Beaverbook for the Ministry three months earlier:

[ATTACH=CONFIG]241145[/ATTACH]

I would still argue that the propeller tip story is not irrelevant, however. It was a limiting factor on the aircraft before any decisions were made, it didn’t just kick in after the event. Decisions were made upon the usefulness of the aircraft as against its impact on production capabilities. Performance at height, or rather the lack of it, figured very largely on the ‘balance sheet’.

Decisions on munitions production – what to build, where, in what quantity – were not made by industry. RR had no say in dumping Peregrine (or Boreas, Exe, Vulture) in favour of Merlin: nor was it up to them to churn out Merlin, or anything. The Customer – us – paid the piper; organised supply, shipping, labour; built and equipped factories.

The twin-engined, 4 cannon fighter awarded 2/37 to little Petter’s/Westland as more promising than others’ schemes, was a challenge – new airframe, systems, engine, soon (1/41, when high altitude was added as Welkin) the novelty of pressurisation. There was no Prime Contractor in those days – Project Management was by the Customer, who, 7/38, caused the business insertion of John Brown and AEI (Westland was near-bankrupt), and planned to produce Whirlwind in Nuffield’s Castle Bromwich (turf turned 8/38). AMDP Freeman added Spitfire II to CBAF, 4/39; then, as has been infamously misrepresented, its output was slower than hoped…but not because RR was overloaded trying to prove and build multiple products. See these contradictory reports:
– “this decision (i.e: to rely on N) cost (at) least 1,000 (Spits) which should have been in reserve when the (BoB) started” (Air Minister Swinton, 60 Years of Power, Hutchinson, 1966, P119, blaming N for output shortfall). cf:
– Spitfire was “in danger of being eliminated from the re-armament programme (due to delay, not at CBAF, but at V-S. Its) future was assured (due to the) stubbornness of 1 man (N, insisting) on producing (all of all) 1,000 (at CBAF, so Whirlwind was) squeezed out”’ A.M. memo, 11/7/39, in E.B.Morgan/ E.Shacklady, Spitfire, Key, 1987, P51.

The fall of France caused Minister of Aircraft Production Beaverbrook to “dump” awhile anything undeliverable NOW! Whirlwind thereafter did not recover any priority, and Westland was consumed in building and repairing Spitfires and acting as Design Authority on Seafire…not because Whirlwind, Peregrine or propellors were bad (or victims of conspiracies), but because lots of (cheapish) proven Spitfires were of higher priority than future better…maybe.

I don’t know the date at which the Whirlwind was “written off”. If you think that had already happened by April 1940 then the propeller tip story becomes irrelevant. If it hadn’t then it becomes one more black mark, although it doesn’t seem to have affected the Spitfire Mk.II too adversely. The timing and sequence of events is pretty important. We are approaching the time when priorities were given to a handful of types, at the expense of development of others. The Whirlwind certainly wasn’t on that short list, so new props were not going to be considered during the crisis, but by the time priorities were relaxed it surely had been discounted.