Greg's Airplanes: Nakajima Ki-84 Hayate "Frank" History

Ad: This forum contains affiliate links to products on Amazon and eBay. More information in Terms and rules

qaz my understanding is that you can get +/- 50 km/h with a strong tail wing or a strong headwind easily, which is why synthetic bench testing is easier.

So the dimensional analysis that TAIC/TAIU did was after getting their hands on a captured Hayate. If you compare their measurements to the Japanese measurements, they are identical or near identical. In other words, when TAIC did their speed calculation, it was with the correct drag calculation. Furthermore, given that the La-9's 3.1m 3-blade prop could fully absorb an 1,850 HP engine's output, it seems likely that the Pe-32 on the Frank could also absorb the horsepower output. The larger props on the Corsair, Hellcat, and George were likely because of efficiency reasons rather than for absorbing power. In other words, if the Homare 21 could make 1,990 HP and the Pe-32 could absorb that horsepower, than the only thing remaining is the drag calculation.

So first up, and I want to restate this point, TAIC's report on the Homare-21 clearly states that the engine did make its rated output when tested with 96-octane fuel. Of course, this was a top-condition Homare 21 that was measured as having an 8:1 compression ratio engine. So this is not indicative of the 7:1 compression ratio Homare 21 engines. But to the best of what we know, the Homare 21 could make its rated horsepower.

Next up, we have to look at the Pe-32: If we look at some basic calculations, you can see that the Pe-32 prop's length impacts its generated thrust substantially. However, other 3.1m props seem to fully absorb similar horsepower outputs, such as in the La-9. So we can only conclude it's possible that the Pe-32 could fully absorb the Homare 21's output.
Finally, going off drag calculations, we know that at the time TAIC calculated the Frank's top speed, they already had interviewed captured aircrew and knew the Frank's correct dimensions and weights. Therefore, we can only conclude that they had correct drag calculations.

I have to say that the speed differences between what the Japanese say is the top speed and what TAIC claimed is the top speed is likely because of differences in methodology as well as the fact that the US tested a build with superlative build quality. As an example of test methodology, in-air testing is notoriously inaccurate because of the impact of external forces on speed. For example, a head or tailwind can easily throw speed off by 50 KPH, because of the way pitot tubes work. Although there may be some other way of calculating speed in the air during WW2 that I'm not aware of.

Thanks again for digging up that info on the different kinds of Homare engines. Is it correct to say that the big difference between the 7:1 (governed) Homare and the 8:1 standard Homare is their horsepower output? The 7:1 produces around 200 HP less?
 
Or maybe the japanese test was with the bomb/fuel racks installed.
 
Greg is the gold standard. Thanks for posting this.
 
AerialTorpedoDude69 ,
Because TAIC was able to dimensionally measure the airframe by no means determines that they obtained the correct drag.
And even Japanese performance calculations with wind tunnel drag data, correct power numbers, prop efficiency and so on were often at disparity with reality by test flights.

I'm fairly sure that everyone had the means to adequately correct measured airspeed to TAS at the time. Well, I haven't heard that synthetic testing was preferable, except for a situation like TAIC where performance testing all Japanese aircraft adequately was completely impossible.

As for the power numbers of Ha-45-21, I'm hoping to better understand the performance disparity between 'planned output' 'actual output', 'reduced CR output' and so on, but it's hair-pulling as lots of the tables prepared after the end of the war are simply erroneous. Regardless I will post what I gather in the next few days.

I could be misinterpreting you here but, as for "the US tested a build with superlative build quality", the TAIC calculations were not on the basis of a real aircraft's test, and the airframes initially captured in the Philippines were described as 'fair condition' but could not be speed tested later because of constant failures of propeller (and were restricted to operating at 2,900rpm, +250mmHg).
Furthermore the vague guess was made at this time that Ki-84's top speed with Ha-45-21 would be around 400 mph (644km/h).

I was sent this video, I suggest anyone interested in this topic watches as it introduces new materials I had not seen:

View: https://www.youtube.com/watch?v=L12PNyii7hw
The author gives his thoughts on how TAIC arrived at their estimates.
 
That's a great video! His analysis of the prop was new to me as well as the mention of the potential that TAIC may have made an error and then edited out the error by fudging their numbers. Really fascinating analysis. I wish whoever sent you the video would comment here because they probably have a great understanding of the content and their analysis would be significant. Also, I completely don't understand the video's point about the automatic pitch governor breaking on the Frank, leading to an overspeed (or the prop tip exceeding its most efficient speed). AFAIK, Japanese props automatically adjusted pitch to efficiently absorb HP. It's in the equation that we discussed earlier. But my guess is that the La-9 also adjusted its pitch and it must have somehow been able to convert power to thrust by doing so. Couldn't the Japanese prop do the same thing? Or perhaps it was already at the limit for pitch adjustments?

Getting back to your comment about drag coefficient, I'm being borderline incoherent here, so if you've misunderstood anything, that's 100% my fault. I've got to say that I've muddled several facts together:

1. Regarding what I mentioned about the "superlative build quality", I'm referring to the rebuilt engine that was tested outside of an aircraft which showed a power output of 1,990 HP. Greg refers to it multiple times in his video and it's the basis for his analysis. The basic idea is that if the engine was outputting its rated horsepower and the drag coefficient is known, then (let's forget the prop for second) we should be able to calculate speed. It's a mistake to take the 634 KPH number at face value because it's based on military power and an unknown test methodology other than we know it's based on a correct critical altitude and drag coefficient (although as mentioned prior, we don't know if this is a clean configuration or not).

2. The video you linked to though kind of casts a lot of doubt on the 687 KPH TAIC values which seems to be in error and not actually based on an actual understanding of the Frank's drag coefficient. However, the 634 KPH number is not its top speed either.

3. The issue that we all keep running into is how does a 2,000 HP aircraft with low frontal area and very low weight come out with such a low top speed? It doesn't make any sense. The prop could explain it, but other aircraft with small props have made very fast speeds. Could it be the high drag, high lift airfoil? It's possible.

Anyway, I'm inclined to agree that the 687 KPH is not accurate. The 644 KPH with WEP is probably closer to the truth than any other value that I've seen. But I'm still a bit surprised that the George, with the same engine, a supposedly "laminar" flow wing (according to its designer), a similar weight, and a larger prop is substantially slower. Its frontal area seems far larger though, so perhaps that explains it?
 
It is indeed a very great video. I was surprised to see that such a video like this already exists, and the thorough overview of all the relevant US documents.
I don't know if the video's creator has an account here, but I will leave a comment.


All fighters used constantly variable pitch at this time. However the Ki-84 used a different method than the typical Japanese type (which was Hamilton Standard, hydraulic), probably due to the decision to adopt an (unreasonably?) small prop.

My simple understanding of the Pe-32 propeller based on what I've read is like this: In order to minimize overall size and weight, a small Ratier-type electric prop of 3.1m with 4 blades was chosen, with a larger range of pitch and high pitch-change rate. However, due to lack of development experience with electric props, the pitch often changed at the incorrect rate and caused overspeed. If this was corrected at the prototype testing phase, it probably returned in service due to no experience maintaining this type of propeller. Generally this prop is regarded as a practical failure.

As for the points,

1. I do understand that the Ha-45-21 in question was rebuilt to superior standards in 1946. However as I mentioned before, the actual performance result of this engine is not provided in the report, just TAIC's intel numbers from the war are repeated. Of course, when ascertaining the 'true top speed' of Ki-84, I am in agreement to consider the engine with its most ideal state: full CR, full RPM and boost limits. Finding out what these numbers actually were is the problem, as I have a suspicion that TAIC's numbers are planned targets gained from intel. Well, I'm hoping to be able to clarify this soon.

2. I think it's possible that 634km/h, though perhaps not the maximum possible speed, was the highest speed actually ever demonstrated at mil-power with a satisfactory record. The plane that reached this number is also in an unusual starting state according to the report of 3400kg, perhaps a light fuel load or no weapons. Meanwhile the 631km/h number quoted in books and Nakajima's report after the war is specified as full gross weight at takeoff, with prototype No.4 (although critical alt is lower).

However the WEP performance may not be all that superior due to the drop in full pressure altitude to get +500mm. *And also assuming it didn't meet its pitch limit trying to max speed with WEP.

Personally my thought is that the true potential top speed of Ki-84 (WEP, best airframe in terms of aerodynamic improvements) lies in the range of ~635-650km/h. The higher end mainly speculating there was some considerable exhaust thrust gain from the collective -> individual stacks change, or if CR was reduced in our fully-rated Japanese tests.

3. Well, this is a very rudimentary statement, but it should be remembered that the Ki-84 was developed to try and meet some pretty exhaustive performance requirements in all angles. Excellent range and good maneuverability, in addition to the rather unreasonable speed target of 680km/h considering the other factors. It filled a closer role to the Ki-43, rather than the Ki-44 that was essentially a 'heavy fighter' (or interceptor) designed prioritizing speed at the expense of other factors (Army initially was very unimpressed with the Ki-44 overall, until it beat Bf 109E).
The Ki-84 initially was planned with a 19sqm wing, but quickly changed to 21sqm. Even so the Ki-84 was less draggy than Ki-43 as explained in that video, although not as superior as TAIC later expected.
 
Last edited:
qaz

To summarize, the calculation for top speed (if we ignore environmental conditions, test variables, prop, and thrust stacks) is pretty simple. We just need drag and horsepower IIRC.

So, to answer your first point:

TAIC did provide a separate analysis of the Homare (OPNAV-16-V No. T231, February 1945, technical air intelligence center report No. 31, Homare 11 and 21 engines, principal characteristics and performance. Report No. 13-c(26), USSBS Index Section 6) in which their bench testing did result in 2,000HP output.

So if we have the drag, horsepower and critical altitude, we should know the speed. It is unfortunate that the HP rating at the critical altitude was fudged. Because now that greatly complicates matters.

Francillion's book used the speed reported by Nakajima for prototypes. So. as you've said. this does not factor in the jet effect of thrust stacks. But that said, my assumption was that if the allies had the correct dimensions, weight, critical altitude, they could fairly accurately generate the top speed.

But, as the video pointed out, they initially assumed the Ki-84 had a laminar airfoil, which it did not. As Greg's video mentioned, the Ki-84 had a wing that was good for low-speed maneuverability, which means a high-lift, high-drag wing. So it's completely possible that the Frank's airfoil was not suitable for high top speed.

Regarding WEP, you could be right that it's possible that Frank's WEP setting wasn't as efficient as a Corsair because of its weaker mechanical reliability or fuel pressure drop off (which I didn't know about and haven't read anything about this. They did mention in Bueschel's book that the float carb had problems though).

Even so, I'd like to see at least some official data on how much power WEP generated because on the Corsair there were huge increases in horsepower, with a massive drop-off in service life. For example, US aircraft were getting something like 500 HP more out of their engines at WEP with MW50. But of course, they were using 100 octane with low knock characteristics. My guess is that a Homare WEP wouldn't be usable for as long as the R-2800/1800's WEP and that the engine would require overhauls more frequently. But that said, everything I know about engine design tells me that a Homare would be capable of WEP+MW at similar boost ratings as a R-2800 as it's essentially doing the same thing: lowering the rate at which knock occurs and running the engine at its highest boost and RPM. (Certainly, its WEP wasn't as good but it shouldn't have been all that bad as the Japanese had been experimenting with MW as long as or longer than almost any other country)


This may explain what's going on. Planes that can dogfight a low speed aren't particularly good at much else.
 
Last edited:

Some US engines were better in getting full benefit of the ADI, some were lagging behind. Eg. P-47 1st upped the power by ~300 HP, and shortly after that by another ~300 HP. The V-1710 on the P-63 gained about 400 HP. The 2-stage R-2800 B and C series as used on Corsair gained around 200 HP in ww2.
I'm not sure that Corsair's engines were loosing much of their life due to WER with ADI.
 
Where did he say that?
 
Anyways I plotted the TAIC power for climb speed ram and 3000rpm and it seems to reasonably match the NK9B power plant chart, knowing that NK9B power is at 2900RPM and zero ram.
By plotting this we also get that the gear switch is very close (~200m) to the previously mentioned japanese document, so power matches bewteen TAIC and the japanese document.


But if we look closely into the japanese document we see errors like this:

Here it says that engine does 1850hp at 3400m which contrtadicts TAIC and the document itself, because as we saw the gear switch is at 3400m and it would be at about 1570hp and 1850HP crit alt is at 1800m.

Then in the max speed test full throtle crit alt for first gear is said to be 3700m. Which doesnt make any sense if the gear switch for climb is at 3400m and the crit alt for climb speed is at 1800m. The TAIC has a much more realistic full throtle height for the first gear at full speed, 2400m


So we can really trust this japanese document? I dont think so...
Further on the japanese document doesnt say the load/rack conditions of the airplane.

TAIC seems like a much better document to be trusted.
Only thing that I would change in TAIC is MIL max speed to be 677kph at 6700m to make it match with the WEP chart.
 
Metrallaroja said:
Where did he say that?
Click to expand...
It was a really long video, but I believe it was about 14:49 when they were discussing early TAIC calculations before they had a captured aircraft. It was mentioned as an aside and very briefly. Hopefully I didn't misunderstand. But he basically said that TAIC initially assumed (but corrected themselves) that the Frank had a laminar flow wing. Its airfoil is probably based on the Ki-43's and was likely less draggy. At the time, the Japanese were flying out of poorly constructed airbases with short airstrips and so aircraft like the Raiden and Ki-44 could not operate from them.

But the drag coefficient calculation that he uses later on, if I understand it correctly, is very, very low. Lower than any allied aircraft outside of the Mustang. Whatever calculation he uses is embedded in the spreadsheet so we don't see what it is. So going off what I understand, the overestimated performance isn't really related to drag. He's saying that the horsepower at the critical altitude was substantially lower than what TAIC estimated. However, it's possible that his calculation isn't right either as he's comparing that data to what Nakajima claimed the top speed was. This isn't a perfect calculation and it assumes that Nakajima's estimates were accurate.

Also, there is one potential error in the video. When he refers to the electrically operated prop governor overspeeding, that suggests that the engine that was tested was non-standard as the Ha-45-21 had a hydraulically operated prop pitch control (IIRC) and the electric mod was a field mod. Because it was a field mod, it was probably unreliable and likely to lead to incorrect pitch angle, AFAIK.

EDIT: M Metrallaroja Greg's video actually covers the discrepancy between the Nakajima report and TAIC's calculations and the results are very close up until you get to the WEP results. Greg synthetically generated WEP results and says that using a standard calculation for determining speed from horsepower, you get a maximum top speed of around 419 MPH for the Frank at its critical altitude. He said that the results are completely in line with the Japanese results. Because he very clearly shows his work, it's actually more credible than any other estimate out there.
 
Last edited:
Even if Greg is close to being right about the theoretical top speed of the captured KI-84 I still don't agree that he should have used it during his comparisons to Allied aircraft. He incorrectly takes TAIC performance estimates as actual flight test data and with all the exhaustive calculations presented on his video 419 mph should have been considered the "ideal" maximum speed of the aircraft. Add wing racks (which I believe was quite common) and speed probably drops to around 410 mph.

This is a figure I can maybe agree on with aircraft in optimal condition and engine providing maximum possible horsepower.
 
That's a good point. To elaborate on Greg's calculation, he basically said that Nakajima's speed estimates for military power and TAIC's bench test for military power match each other (within a 3% margin of error, which is standard for production aircraft). Greg then takes the horsepower data from TAIC's actual bench of the Homare (which we don't have full access to) and uses a standard calculation for determining top speed to arrive at the 419 MPH speed.

This is a fairly unimpeachable calculation. However, the problem is that it's based on data that is not commonly available as far as I'm aware. Additionally, the same 3% could be applied to US aircraft, so a Hellcat dash 5 (391 MPH as estimated by the manufacturer) could have a top speed of 404 MPH coming off the production line. But we're not talking some rare tuned up model, like with the Blueprint Frank. We're talking about an aircraft that was being pushed out in absurd numbers with unmatched quality and precision. You can imagine that a late-war Frank with a lower compression ratio engine, declining tolerances, and poor alloy materials is likely to have far worse performance than the numbers suggested by Nakajima or TAIC.

But, again, Greg's channel is about the engineering of WW2 aircraft and he's mostly interested in human innovation. So I understand why he undermines his own calculation by accepting TAIC's calculation over his own. More or less, he's trying to frame the performance numbers within the context of engineering and not wartime scarcity.

But that said, Greg's 419 MPH calculation in a clean configuration for the Ha-45-21 Homare is probably the most accurate out there and it matches most of the information from credible sources that we've seen.
 
That's a very good video. Thanks for sharing

I really like his slow analytical style. He really doesn't leave any stone unturned. I need to watch it again because there's so much good stuff packed into it.

He mentioned the propeller as totally inadequate and he made some interesting points to back up this assumption. That could be why calculations that arrive at speeds above the well-established 388 mph is due to an over estimation of propeller efficiency.
 
Last edited:
However, due to lack of development experience with electric props, the pitch often changed at the incorrect rate and caused overspeed. Generally this prop is regarded as a practical failure.
Click to expand...

FWIW I have modelled/modded the Ki-84 in il2-1946 with the correct Pe-32 prop data, and it can have weird overspeed/windmilling prop behaviour if you aren't gentle with throttle and propsettings. il2-1946 is probably at least as advanced as WW2 slide rule estimations.

The Ki-84 I modelled (Ha-45-21 2000hp) ended up with this which is closer to Japanese evaluation than US TAIC. Its nice when the numbers just fall out right
630kph(656WEP)@6000m and 644kph@6500m
(391mph(408WEP)@19685ft and 400mph@21325ft)
and
5:50 to 5000m (MIL)

At the end of the war the US requested materials from Nakajima concerning its aircraft, some were reproduced due to being destroyed or other reasons.
Click to expand...

Its said that Nakajima destroyed all their aircraft data records after the end of WW2, which is why so little Nakajima data is around compared to Kawasaki and Mitsubishi. Some of their engineers claimed they were sad and thought shouldn't we be saving these records for history.
 
Thank you for providing that info!

I'm sorry to ask, but do we know anything about the Pe-32's paddle width or anything about its pitch governor? The only information that seems available online is on its diameter of 3.1-meters. I've read that the TAIC Frank (T-2) model had an electric pitch governor but all japanese aircraft to my knowledge used a hydraulic pitch governor. It just wouldn't make sense in my opinion that the production Frank would use electric because Nakajima tried to reduce all electrical systems, probably in order to keep costs low or perhaps because of the persistent issue with EMF caused by spark plugs. (Although that last point is probably not true).

There are examples of Japanese aircraft that were modified in the field to use electrical pitch control systems but these were unreliable and failed during US testing, just like the Frank's system.

As an aside, Greg's 3% comment puts his numbers in line with your numbers as well. So all of these estimates are within the bounds of reality.
 
I sure wish IL2 would adapt your flight model because as it stands the Frank flies like a rocket ship and can only be handled by allied aircraft of the highest performance. Unfortunately my F6F is routinely left choking on it's prop dust...
 
I forget exactly why the Ratier electrical prop was used on Ki-84, but I think it was as it promised better pitch control for the foreseable problem of the small propellor.

Its definately an electric propellor drawing of circuits with 90W draw. Maru Mechanic has a prop diagram looks like a mechanical spinning governer with electrical contacts if it speeds up or down to close pitch control circuits, and there is a mechanical spring load on the governer weights I guess to set RPM wanted, varible pitch to maintain constant RPM. So its a mechanical system driving electric control. Seems like a good idea, but looks really complex.

early props were maybe 3.0m 32-60 degree pitch and 200kg
later 3.1m were 26-56 degree pitch and 185 kg.

I sure wish IL2 would adapt your flight model because as it stands the Frank flies like a rocket ship
Click to expand...
The first time I flew the il21946 Ki-84 I was amazed at been able to pull away from any Corsair like it was a Tiger Moth. It was made on the US TAIC data
 
Last edited:
Wow, this is a great discussion. Thank you again.

In "theory" an electrically-driven prop governor uses a motor whereas a hydraulic system is going to have not just the fluid but the metal housing capable of containing the fluid under pressure. So there's probably a substantial weight difference between the two.

I like your explanation. An electric governor would probably vary wattage output (or power) depending on the pitch angle, which would explain why the Pe-32 couldn't hit higher pitch angles (because of its low-voltage electrical system). As pitch-angle increases, so too does the required power to turn the blade. The gimped pitch angle is another indicator of the Frank's low-voltage system. A higher voltage system would probably have been able to hit a higher pitch angle.

The Pe-32's electrical system may have had a faster ability to change pitch angle. The lower pitch angle would allow for faster acceleration as well and would have been handy in a dog fight. So this completely makes sense to me. They wanted a better dog fighter.

In other words, Nakajima traded top speed for faster acceleration. If they upped the voltage in their electrical system they might have hit higher pitch angles, which in theory would compensate for the shorter prop blade. But if they never hit that higher voltage electrical system, then the prop never would have reached its maximum potential. That is a serious design flaw.

Were I a mechanic during that time period, I'd probably have tried to implement a hydraulic system as a field mod on the Frank, particularly for bombing runs. Although that may not have been possible on such a compact aircraft.
 
Apologies for double posting, but qaz M Metrallaroja I found the post where krieghund posted the engine test of the Homare. The partial clip only references that the Homare did output its rated power but unfortunately we do not have the full power output tables for it.

It does seem that that the engine did output its rated horsepower, although at what altitudes, we do not know.

I'm curious as to whether later models of the Frank managed to resolve the issue with its propeller because the problem was relatively trivial compared to the other issues that we've seen aviation engineers overcome. To my knowledge, they'd need to buff up the alternator and voltage regulator and then increase the gauge of the wiring and resistors throughout the aircraft in order to compensate for the higher voltage. But I don't know much about this subject. That just with my 101 knowledge of electrical systems.
 
You must log in or register to reply here.

Users who are viewing this thread

Total: 1 (members: 0, guests: 1)

Similar threads

W
Greg's Airplanes & Automobiles: Most Accurate Allied Bomber, or Fighter Bomber of WW2. (European Theater)
Replies
8
Views
736
SaparotRob
M
New Build Copies Of WW2 Luftwaffe Airplanes
Replies
1
Views
755
Gnomey
Greg’s Airplanes Kawasaki Ki-61 history
Replies
2
Views
972
chris ballance
O
Greg of Auto and Airplanes has asked for a Debate
15 16 17
Replies
332
Views
27K
DarrenW
Genius Of The Jet. The Invention Of The turbojet: Frank Whittle. Historical Lost Interviews
Replies
1
Views
568
Gnomey
Menu
Log in
Register