Nakajima Ki-84 advanced versions and derivatives

[HoHun]

Hi Tomo,

'Lower' means 'lower flow rate' there, for how much is needed at 20000 ft vs. how much is needed at 16000 ft, for max boost at 3000 rpm.

Well, at MIL, the injection rate seems to be independent of altitude, and I assume the same is true for WEP (as the data points we have for WEP are identical). However, as the reported flow rate seems to be consistent with the flow rate your statement suggests, I am not quite sure why you brought it up.

WEP does not provide anti-detonant, it requires it.

My implicit assumption was that the engine would set the required flow rate if told so by the pilot, providing the required anti-detonant automatically I haven't checked this against the test flight report, so that assumption might not be accurate.

Regards,

Henning (HoHun)

 

[tomo pauk]

Well, at MIL, the injection rate seems to be independent of altitude, and I assume the same is true for WEP (as the data points we have for WEP are identical). However, as the reported flow rate seems to be consistent with the flow rate your statement suggests, I am not quite sure why you brought it up.

As you can se, injection rate is very dependent of altitude, going from the max (79.2 gals/hr at 3000 ft) down as altitude increases. The next data-point we see is at 5900 ft, where the injection of ADI is at 39.6 gals/hr.
We can expect the same for the high gear.
Take-off setting uses the max flow so the max boost can be used, the BHP being lower than at 3000 HP due to throttling losses.

ADI flow rate is dictated by the boost the S/C can provide. S/C will be providing more boost as altitude is lower.

BTW - Americans using American term for non-American engine is a bit misleading and arbitrary - Japanese were not using terms like 'WER', WEP' or 'military power'. Using the term 'military power' when the power setting requires ADI - we probably will never know why the Americans did that.

 

[HoHun]

Hi Tomo,

As you can se, injection rate is very dependent of altitude, going from the max (79.2 gals/hr at 3000 ft) down as altitude increases. The next data-point we see is at 5900 ft, where the injection of ADI is at 39.6 gals/hr.

Hm, I'd group the data points as follows:

MIL:

00000 ft - 3000 rpm - 43.7" Hg - 39.6 US gallons/h
05900 ft - 3000 rpm - 43.7" Hg - 39.6 US gallons/h
00000 ft - 3000 rpm - 43.7" Hg - 39.6 US gallons/h
19680 ft - 3000 rpm - 43.7" Hg - 39.6 US gallons/h

Take-off/WEP

00000 ft - 3000 rpm - 49.6" Hg - 79.2 US gallons/h <- Take-off Power
06000 ft - 3000 rpm - 49.6" Hg - 79.2 US gallons/h <- WEP
16000 ft - 3000 rpm - 49.6" Hg - ??? US gallons/h <- WEP

This looks to me as if the only varying variable here that influences water-methanol injection rate is boost pressure, so I am quite confident that my guess for the ??? data point of 79.2 US gallons/h is sensible.

ADI flow rate is dictated by the boost the S/C can provide. S/C will be providing more boost as altitude is lower.

Hm, I am not sure I follow. If you've seen the Me 109 climb graphs, the supercharger pressure and the boost pressure are different variables ... "Gebläsedruck" is supercharger pressure, "Ladedruck" is manifold pressure: http://www.wwiiaircraftperformance.org/me109/14026pg5.jpg

As the DB 601E is a direct-injection engine, I'd imagine it can serve as an analog for the Ha-45, but with the latter engine, I am actually not entirely sure how the boost control works, so maybe you're right when it comes to this engine.

BTW - Americans using American term for non-American engine is a bit misleading and arbitrary - Japanese were not using terms like 'WER', WEP' or 'military power'. Using the term 'military power' when the power setting requires ADI - we probably will never know why the Americans did that.

Well, to their credit, the TAIC manual actually starts off with an introduction that defines the terms they use. Still, terminology is a bit of a minefield, especially as the US changed their rating system in mid-war, and I am not even sure that Army and Navy always had the same idea about how to rate engines either.

Regards,

Henning (HoHun)

 

[tomo pauk]

This looks to me as if the only varying variable here that influences water-methanol injection rate is boost pressure, so I am quite confident that my guess for the ??? data point of 79.2 US gallons/h is sensible.

Bingo.

Hm, I am not sure I follow. If you've seen the Me 109 climb graphs, the supercharger pressure and the boost pressure are different variables ... "Gebläsedruck" is supercharger pressure, "Ladedruck" is manifold pressure: http://www.wwiiaircraftperformance.org/me109/14026pg5.jpg

This is what happens if the throttle plates (or more of them) is after the supercharger - the S/C provides much more pressure than the engine can withstand, thus the throttle plate(s) must reduce the pressure to a value suitable for engine.

What we call 'boost' is indeed manifold pressure - Ladedruck.
The Gebläsedruck and Laderdruck should became of about same value at the rated height, at least by looking at this picture.

That layout (S/C -> throttle -> cylinders) was probably not judged as very efficient ( the impeller's work is wasted at lower altitudes despite the variable drive) since the DB 605L and 603L were supposed to gain the very efficient swirl throttle instead.

As the DB 601E is a direct-injection engine, I'd imagine it can serve as an analog for the Ha-45, but with the latter engine, I am actually not entirely sure how the boost control works, so maybe you're right when it comes to this engine

The boost control was automatic via the boost control device, with three throttle plates between the S/C and 1st pair of cylinders. See here.

 

[HoHun]

Hi Tomo,

Bingo.

So we don't actually disagree on anything? I wasn't sure of that

That layout (S/C -> throttle -> cylinders) was probably not judged as very efficient ( the impeller's work is wasted at lower altitudes despite the variable drive) since the DB 605L and 603L were supposed to gain the very efficient swirl throttle instead.

Well, the swirl throttle works by imparting a spin on the intake air stream to relieve the centrifugal supercharger, which has to do the same, from a bit of work, so it needs to go before the supercharger. I don't think it's possible to make any conclusions about the efficiency of the old throttle setup from that. (If you've decided on a draw-through setup on a carburetted engine, to increase supercharger efficiency as pointed out by Hooker, you can't really decide freely where to put the throttle either.)

The boost control was automatic via the boost control device, with three throttle plates between the S/C and 1st pair of cylinders. See here.

This seems to be the same link on the DB 605 again ... perhaps a copy/paste problem? I was actually looking for details on the Ha-45 boost control system.

Regards,

Henning (HoHun)

 

[tomo pauk]

So we don't actually disagree on anything? I wasn't sure of that

I'd say that we're above the average, when it is about how much people agree on such matters on the internet

Well, the swirl throttle works by imparting a spin on the intake air stream to relieve the centrifugal supercharger, which has to do the same, from a bit of work, so it needs to go before the supercharger. I don't think it's possible to make any conclusions about the efficiency of the old throttle setup from that. (If you've decided on a draw-through setup on a carburetted engine, to increase supercharger efficiency as pointed out by Hooker, you can't really decide freely where to put the throttle either.)

We can see that both Jumo and DB (and Klimov, with VK-107 and -108) were going with swirl throttles, so the proof is in the pudding?

The carburetor-before-S/C layout predates Hooker.
He is, unfortunately, very silent on how much a float-type carburetor is detrimental to the engine's performance in the 1st place, and he is not the only one.

This seems to be the same link on the DB 605 again ... perhaps a copy/paste problem? I was actually looking for details on the Ha-45 boost control system.

Doh
Here it is, for the DB 601E/605A (scroll down a bit).

 

[HoHun]

Hi Tomo,

I'd say that we're above the average, when it is about how much people agree on such matters on the internet

LOL, probably true! :-D But I thought about the specific issue of the Ha-45 Model 21 WEP full throttle height being underestimated at 16000 ft, if the 19680 ft are accepted as realistic for the MIL full throttle height.

We can see that both Jumo and DB (and Klimov, with VK-107 and -108) were going with swirl throttles, so the proof is in the pudding?

Well, the old arrangement still might have been the most efficient solution before the swirl throttle was invented ... that it was replaced by a new and better solutions doesn't tell us how good it was compared to contemporary solutions.

The carburetor-before-S/C layout predates Hooker.
He is, unfortunately, very silent on how much a float-type carburetor is detrimental to the engine's performance in the 1st place, and he is not the only one.

I believe there's something like engineering culture that leads to an over-appreciation of the technological solutions one is used to. In wartime (as evident from articles in "Flight" magazine), I suspect some amount of posturing is added to that, in order to keep up morale In contemporary German publications, I've found statements like, "Frequently, the engineer discovers new and superior materials when in wartime, the customary ones aren't available" ... obviously, that was mostly wishful thinking, too.

Here it is, for the DB 601E/605A (scroll down a bit).

Ah, thanks - cool stuff! The little extra throttle for the low power end is neat ... I think it was Jeffrey Quill who pointed out how difficult it was to get precise power control on the Griffon. I don't think he would have minded Rolls-Royce taking a leaf from Daimler-Benz there! (The historical solution was to increase the movement range of the throttle quadrant ... simpler, but maybe less effective.)

Regards,

Henning (HoHun)

 

[tomo pauk]

LOL, probably true! :-D But I thought about the specific issue of the Ha-45 Model 21 WEP full throttle height being underestimated at 16000 ft, if the 19680 ft are accepted as realistic for the MIL full throttle height.

FWIW, the 16000 ft rated height for WER is noted as 'estimated' on that Homare report. It is also noted as estimated at 16500 ft on the big TAIC manual.
OTOH, the 'Middleton test' says WER was 1850 HP at 17650 ft(!); mil power was supposed to be 1695 HP at 21000 ft!!

I don't have any confirmation that any of these figures, or even the mil power, are the result of a bench test.

Book about Nakajima engines gives 1650 CV at 20000 ft (6100m) for the Mod 21 (7:1 CR is noted for it).

My questions: whom to believe, and what actual horsepower was attained with in-service engines; how much might be the tolerance in power level per a chosen power setting?

 

[HoHun]

Hi Tomo,

I don't have any confirmation that any of these figures, or even the mil power, are the result of a bench test.

Book about Nakajima engines gives 1650 CV at 20000 ft (6100m) for the Mod 21 (7:1 CR is noted for it).

My questions: whom to believe, and what actual horsepower was attained with in-service engines; how much might be the tolerance in power level per a chosen power setting?

It's always difficult, even when it comes to period bench tests. If you read Hooker's booklet, they weren't really equipped to measure a full performance-over-altitude curve, and at least at the beginning of the war made do with approximations that weren't all that accurate. His mathematical approach gave fair matches to Merlin tests run on more advanced US test stands, but test stands had their limitations as well, so apparently the US considered it best to develop torque sensor to have the ability to measure power in flight. I believe that technology was available for quite a few engine models during the war, but only used in a couple of experimental situations, for example by NACA.

So I suspect that quite often, manufacturers weren't quite so sure how much power their engine was actually providing in a certain situation in flight, not only for the Japanese engines.

The full report you linked has a pretty good looking engine power curve fo the Ha-45 Model 11, but it looks to as if it's from Japanese sources, and the ATIG Report 45 http://www.wwiiaircraftperformance.org/japan/ATIG-Report-45.pdf states:

"For engine tests the only equipment available was the simple electric dynamometer with no provision for altitude simluation of carburetor air temperature or exhaust pressure. Sea level power calibrations were run and alttiude ratings determined from standard correction formulaes furnished by the Army and Navy. The Nakajima company apparently attempted to give most of their new engines a flight test that was more extensive than that of the other engine companies. This was particularly true for carburetor settings were dynamometer results were checked by flight test and the final setting determined froma combination of dynamometer and flight results. Relatively simple instrumentation was used in these flight tests, including a form of Benix flowmeter. Carburetor adjustements which flight tests showed to be necessary were made direcctly on the installed carbureters."

Accordingly, I keep a mental note for generally every engine that it might not perform exactly as expected by the manufacturer even when running according to specifications, and that there might be systematic differences beetween manufacturers, countries and, as evident from the Japanese manual, even individual services, as well over time, because the methodology for establishing power curves might have been changed at one or more points before and in the war.

The Nakajima Ha-45 Model 21 is a difficult engine to assess because it was produced under very difficult conditions, and we don't have much information on how the engine was expected to perform even when running according to specifications. However, if you look at other engines, there's good reason to understand most the published information as approximations, usually to the best of the ability of the parties involved, though they sometimes might be conventional simplifications owed to the rating system used.

Regards,

Henning (HoHun)

 

[HoHun]

Hi again,

The Nakajima Ha-45 Model 21 is a difficult engine to assess because it was produced under very difficult conditions, and we don't have much information on how the engine was expected to perform even when running according to specifications.

I looked at the various documents I found online, and noticed that a set of figures marked as originating from ADVATIS translation no. 92 might be helpful as it shows a clean shape around the high gear full throttle height, unlike the figures from the "Garber facility" manual.



The problem with the figures from the ADVATIS translation is that in the fragments of the original document I found online, there is no information on the exact engine model (and none on the configuration of the Ki-84 it's flown in, either).

Of course, any information on that would be welcome!

Regards,

Henning (HoHun)

 

[tomo pauk]

It's always difficult, even when it comes to period bench tests. If you read Hooker's booklet, they weren't really equipped to measure a full performance-over-altitude curve, and at least at the beginning of the war made do with approximations that weren't all that accurate. His mathematical approach gave fair matches to Merlin tests run on more advanced US test stands, but test stands had their limitations as well, so apparently the US considered it best to develop torque sensor to have the ability to measure power in flight. I believe that technology was available for quite a few engine models during the war, but only used in a couple of experimental situations, for example by NACA.

So I suspect that quite often, manufacturers weren't quite so sure how much power their engine was actually providing in a certain situation in flight, not only for the Japanese engines.

RR expected that Merlin XX will be doing 1175 HP at 3000 rpm and +9 psi at 21000 ft, no ram effect, and such data points were stated eg. in the Beaufighter II data sheet. However, the tests of the Hurricane II note that it took full ram, as present at almost 330 mph, to make that kind of boost at 20000 ft - meaning that earlier data was overly optimistic.
The sibling of the Mk.XX, the V-1650-1, was credited to only 1120 HP at 18500 ft - a much more realistic value, that aligns well with the much later graph posted here.

So yes - a small dose of skepticism is always a good thing to have

 

[HoHun]

Hi Tomo,

RR expected that Merlin XX will be doing 1175 HP at 3000 rpm and +9 psi at 21000 ft, no ram effect, and such data points were stated eg. in the Beaufighter II data sheet.

That's a can of worms by itself, as the Merlin XX was projected and tested with various different gear ratios before one was settled on for production. The tests on the Merlin XX also fell in the period in which Hooker improved the British methodology for calculating altitude performance ... I haven't seen the Beaufighter data sheet you mention, but if there's a date on it, that would be interesting to know.

Regards,

Henning (HoHun)

 

[tomo pauk]

I haven't seen the Beaufighter data sheet you mention, but if there's a date on it, that would be interesting to know.

I've posted a lot of the data sheets here, including for the Beaufighter.
Merlin XX was supposed to be good for 1175 HP at 20500 ft for the Beau Mk.II per the data sheet.

 

[Noveos]

Did the IJAAF have any plans to replace the Ki-84 with another design?

 

[blackkite]

You mean Ki-117?

https://www.secretprojects.co.uk/threads/nakajima-ki-84-advanced-versions-and-derivatives.4155/post-87010

 

[Noveos]

You mean Ki-117?

https://www.secretprojects.co.uk/threads/nakajima-ki-84-advanced-versions-and-derivatives.4155/post-87010

I mean as in how the A7M was the replacement of the A6M, I was wondering if they were designing something to replace this Ki-84 with a new design