Thoughts on the Nakajima Ki-84 and Kawasaki Ki-100

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

The US Test of Ki 84 engine is actually here on our site. I uploaded it a couple years ago.
The data suggests that the claimed 1990 HP is probably correct. As for the level of maximum speed that would give, I see no reason to believe that the aircraft was not capable of pretty near the claimed performance numbers in some of the Japanese tests. In other words, this was capable of well over 400 MPH if everything was running right which usually was not the case in service conditions.

During US tests, Quality of fuel may have been better, but Octane was not. Testing was done with 92 Octane just like Japanese service fuel. It is described in the test report.
Middletown Test Report

Hello GregP,
The way the Japanese used Water Methanol Injection was way different that just about everyone else. It wasn't just for Emergency power. It was used automatically whenever the Manifold Pressure was pushed beyond something like Maximum Continuous. If you want some details on Homare engine, check out the Middletown Test Report I mentioned earlier. That is why they carried such a large proportion of Water-Methanol for the amount of fuel onboard.
I don't believe the Ki 100 used it though.


Actually someone from the same era who would dispute that claim for speed would be David McCampbell who claimed the Corsair was slightly faster which is why he preferred it. Interesting comment considering what he flew in service.

I believe that although the Corsair and Hellcat were FAIRLY comparable in size, the Hellcat was considerably more draggy.
Here is my THEORY as to what was really happening:
The Engines and Propellers were identical. That does not mean that their efficiency was the same though. What is behind the propeller has some influence on efficiency.... but I have no call on which was better. (!)
The larger Wing Area and greater CD0 would already indicate Hellcat is more draggy, so why isn't it considerably slower???
I believe the reason is Exhaust Thrust. The Exhausts of Hellcat are positioned to generate more Thrust. The Exhausts of the F4U-1 Corsair are not positioned in a way to generate significant Thrust. The extra thrust made up for the extra drag.
Now when the engine was upgraded in the F4U-4 and XF6F-6, the Corsair gained significantly more speed while the Hellcat did not.
I believe the reason was that the -4 Corsair gained significant Exhaust Thrust as compared to -1 Corsair while the Hellcat did not gain anything compared to earlier birds.

I haven't seen anyone else propose this idea before but the idea seems pretty reasonable to me.

- Ivan.
 
Regarding Flight Simulators:
I believe the Microsoft Simulators dating as far back as Combat Flight Simulator 1 (Circa 1999) were reasonably capable of representing the performance of WW2 piston engine aircraft though with some serious limitations. The question isn't really whether the simulator is capable but whether the author is using good performance reports and is any good at knowing how to tweak a flight model. Most people aren't very good at it.
Some of the other simulators are not so good at representing WW2 era piston engine aircraft. I don't believe IL2 as it first came out was really entirely capable. The reason why I say this is because I have followed some discussions in which folks have described having to reprogram things to handle Compressibility. This isn't something that comes up much but it does come up.

I am not describing the Eye Candy aspect. IL2 wins there. I am describing features available for flight models.

As with typical PC simulators, things are good if you are "Flying" nose forward at a "reasonable" angle of attack. If you happen to get into a situation at say negative 90 degrees AoA, then all bets are off. The author of the flight model is not likely to have enough data to KNOW how to properly model that situation. Most likely you are experiencing some fellow's view of the world.
As with most computers, Garbage In ==> Garbage Out. If folks can't find test data because there isn't any available, then what data is that flight model based on?

There is also the question of how different authors interpret the same flight report. When the report describes instability, how much are we talking about? My GUESS might be quite different from yours.

- Ivan.
 
In the WWII sims I have flown, including some of the best, the performance was never all that realistic.

The planes take off easy, land easy, and have very good manners, even when you apply full throttle to go around. Try that in a real WWII fighter and they'll bury you quickly.

There are numerous other "let's make this easy and fun" characteristics that make them games rather than simulators. In a real WWII fighter, try extending full flaps in a terminal dive and find out what part of the wings break off. Try using full rudder to get around a slow turn. It works. In the real plane, you'd be in a spin so fast your head would swim.

Don't get me wrong, they WWII sims are fun. But realistic? No way. If you service the main landing gear of a Corsair to WWII Navy specs, it flat doesn't land straight down ANY runway. To make them friendlier to non-military users, they typically run only half the oleo air pressure specified in the wartime manual. Many more faults to rectify before you could call it a "simulator," at least in my opinion.

I've never seen a software sim that simulates starting of a radial engine very well, either. But it DOES make it a easier to start without damaging the engine.

I know, opinions vary like gas mileage.

Cheers.
 
Ivan1GFP said:
I haven't seen anyone else propose this idea before but the idea seems pretty reasonable to me.

- Ivan.
Click to expand...
You could very well be on to something.

Early Corsairs pretty much just dumped the exhaust.

Perhaps they were better than an F4F-4 but the exhaust thrust vector is at an angle to the planes line of flight and not nearly parallel. It was also found that 3 cylinders per pipe was the MAX for getting decent exhaust thrust and 2 cylinders per pipe was much better.
 
Greg,

How much difference is there in weight between a combat equipped F4U and a as currently flown version? I would think quite a bit (no armor, guns, new lighter IFF / Radios). Would that also be part of the reason the struts are serviced to a lessor degree? Also did the WW2 users change the strut servicing depending on the load out (bombs and drop tanks) vice none? On the Eagle the mains were serviced differently when we carried more external tanks (also the quality of the tires were looked at).

Cheers,
Biff
 
Last edited:
Hello GregP,
There are certain features in various simulators are that just outright wrong. Being able to deploy flaps at maximum speed is pretty silly. Blow-Up flaps that cannot deploy because of aerodynamic pressure often cannot be deployed. The engine starting cycle is way over simplified..... etc. etc. Being able to land on water and take off from water isn't right.... Or the alternative of exploding when touching the water is also not right. The sounds are not right, It is difficult if not impossible in some sims to implement a Coffman starter. Those are certainly limitations just like not having to reach under your desk chair to operate a control lever or not having the same level of control over your fuel tanks, or not having drop tanks.
One of the things I hate the worst is that in CFS1, the world is basically a Mercator projection map. (Oops!!!!) Flying near the Poles can be really amusing!

As long as you understand and are willing to accept the limitations, there is something to be gained here.
It is very much like a conversation I had with a friend a couple days ago:
We calculate Means and Standard Deviations and use them as a measure of the quality of ammunition. The problem is that a Standard Deviation is really a technique that is applicable to a Gaussian Distribution.
I asked him: How many rounds of 5.56 mm ammunition do you think the US Military has used since the round was adopted?
Billions, a Trillion? What is a reasonable Mean velocity for all those rounds? Perhaps 3100 FPS? OK..... Good so far....
With that many rounds, if they really fell into a Gaussian Distribution, we probably should have seen a couple that had velocities that were about 20 FPS or maybe a few that were at 8000 FPS, Right?
Do you really believe that to be the case? If not and this is not a real Gaussian Distribution, then why do we use techniques as if it were????

Sometimes a tool even with limitations can give you an insight that you can't get any other way. I don't think any of us will ever see a combat equipped FW 190D fight a Corsair. We can speculate and argue until whenever but we will never see it.... Ever!
In a simulator, we can get an IMPRESSION of how that MIGHT go. If you don't think things are right, feel free to tweak either aircraft until is suits you. You might not get the G-forces or any of the nasty effects of real flight, but you can also try the scenario 1000 times and refine it as much as it suits you. I would not use the results to argue that one is superior to the other.
I have flown simulators for years and tweak the heck out of the flight models, and certainly have formed a few opinions, but I don't think I have ever tried to claim those opinions were fact either here or anywhere else. At the very best, they are just my view of the world and most of the time they don't even reach that standard because of the "Garbage-In ==> Garbage-Out" aspect of simulators.

A lot of what you described as faults are really not faults in the simulator but faults in the flight modeling by authors who simply do not know what they are doing. It isn't hard to create enough engine torque in a flight model to make a take off very difficult. I intentionally did this when working on a Messerschmitt 109. With full engine power at the start, you simply don't have enough control authority to hold it on the runway. Directional instability at landing is also not difficult to do. I actually made it so bad in testing that I was only able to land successfully about 1/3 of the time even when I was anticipating what would happen. The old Microsoft simulator provided the features, the author just has to know how to use them and most authors do not know or do not take the time to refine the flight model.
Did I get it right???? Maybe. Maybe not. It was my best GUESS from what information I had and I don't expect anyone is about to let me kill myself and destroy their 109 in the process of finding out for sure.
This is why you get a stock Spitfire Mk.IX that flies like a Mk.XIV and a 480 MPH Messerschmitt 109G and a P-51D that weighs about a ton less than it should. There are a lot of other much more sneaky ways to give a flight model way superior performance that is much harder to measure if the author knows what s/he is doing, and along the same lines one can tweak a virtual aircraft to handle much better than the real thing or much worse than the real thing.

It is much like saying Wood is a terrible material for building furniture because all the furniture I have from Ikea and Walmart is garbage.... It isn't necessarily a good conclusion.

- Ivan.
 
Ivan1GFP said:
The way the Japanese used Water Methanol Injection was way different that just about everyone else. It wasn't just for Emergency power. It was used automatically whenever the Manifold Pressure was pushed beyond something like Maximum Continuous.
Click to expand...
I wonder how this system compared to the supercharger on the Nakajima Ki-87.
 
Some years ago I read an article about the testing of the Ki-84 and when US fuel was put into it the top speed was about 427 mph. This high speed surprised every one. It's a good thing they had lower octane fuel as they would have been way more competitive and in some cases superior!
 
Hello Dcazz7606,
The fuel used to test captured Ki 84 was no higher octane than what the Japanese were using: 92 Octane.
Putting higher octane fuel into an engine that has not been re-tuned for it really won't make any difference.
The link to the test report is in a prior post.

As I understand it, the Homare was designed to operate with 100 Octane fuel. That quality of fuel was anticipated but never became available except in very very small quantities and only for test aircraft, so in service, it had to operate on 91-92 Octane and used Water-Methanol (a LOT of it) to allow the higher boost settings the engine was designed to use. Just about all the late-war Japanese fighters did the same thing.
The kinds of things that were done to Ki 84 to achieve the performance they got was to basically bring the engine and other aspects of the airframe up to the proper specifications. Things like Ignition systems, Fuel pumps and Oil pumps and the like were overhauled or replaced by equivalent American pieces where necessary. Basically they were correcting the problems with late war Japanese manufacturing. Fixes are detailed in the report.

- Ivan.
 
Admiral Beez said:
I wonder how this system compared to the supercharger on the Nakajima Ki-87.
Click to expand...
They would be entirely different things. The supercharger determines how much boost can be supplied to the intake manifold.
The Water-Methanol delays detonation to allow that boost to actually be used without destroying the engine.
 
I imagine if given a total overhaul with new US-made, cost is no object pistons, cranks, rods, rings, valves, etc. set to a higher tolerance and compression that >430 mph should be feasible.
 
Not too sure the muzzle velocity of a military 30-caliber round is a normal distribution, but it well might be. But, as you acquire sampling distributions, the distribution of sampling distributions tends to be normal, assuming math hasn't changed in the last few years.

I'm pretty sure they have several million samples of 3 or 5 rounds, but you can get a pretty good normal approximation using 2 or 3 boxes of ammunition, with sample sizes of 3 or 5 rounds each. Of course, it will not exactly be a random sampling distribution, but you COULD approach that by getting 3 or 4 boxes from people who have had them for several years, taking a new box, and maybe one from 10 years back and one from 3 and 5 years back. They would be guaranteed to not be from the same lot. Might prove interesting.

Your argument above about flight sim games, while technically correct, really does not really hold up since the sims everyone is saying they get their impressions of WWII airplanes from are the PC sims available to the public. Ergo, they ... the public ... are using games to make interpretations of real airplanes that do NOT fly like the games, almost in any way. My conclusion is simple, their opinions are flawed in the extreme and do not even approach real airplanes. Real WWII fighters are nowhere NEAR as forgiving as game sims.

A Cessna 172 might be well simulated, sure. And maybe you could learn something from it by flying a C-172 game sim.

The WWII fighter sims are generally NOT well simulated and you cannot survive trying to fly a real North American P-51D because you can fly a CFS1 P-51D. Maybe once it is airborne and flying, you could keep it in the air, but good luck landing it without crashing or even properly operating the engine and prop!

That's all I am saying. I'm not putting down the game sims, they are fun!

But it's like going to a western movie, seeing a Hollywood gunfight, and concluding that the way it happened in the movie was the way it happened in the old West. It is entertaining, to be sure, but real gunfights didn't happen that way. At least, not very often. I'm sure one or two did, but an "outlaw" very probably had no sense of "fair play." Similarly, a real WWII fighter will not be "forgiving" of incorrect handling, particularly during takeoff and landing, and playing a game sim will not prepare you for flying one, other than maybe showing you where the instruments are. But, since most warbirds do not have "stock" instrument panels, that might not even be the case.

I've seen one two-place Hawker Hunter with a full glass panel! Bet there are NO sims showing THAT. Have sat in a Super Tucano with two full glass panels along with front and rear Garmin GTN 650/750 combos! I KNOW there are no sims for that one!

Doesn't mean the PC sims aren't fun. Many are, at least to me. Just not very realistic. Flew an F-15 sim last week at a museum in a "simulator." Had no trouble pulling 18 - 25 g's! Man, could that baby TURN!

And the fuel gauge never gave me the slightest trouble, even after 5 minutes in full afterburner! In a real F-15, after 5 minutes on full afterburner, you'd be dead or really low on fuel and looking for a runway or a tanker! Just saying.

Cheers.
 
If Wiki is correct the Homare was a 2187 cu in engine, It already used a rather high compression ratio in the cylinders, which may have helped cruise settings but hurt max power settings.
The main problem with air cooled aircraft engines is getting rid of the heat.
The water alcohol helps with this in several ways, one is as an internal coolant that absorbs some of the heat and carries it out the exhaust. another way is by lowering the peak temperatures in the cylinder.

In the American and British radials each major increase in power was accompanied by changes in the fin area on the Cylinders heads and barrels.
Without it you have things like the P-47 using both 150 octane and water injection at the same time going around 20mph faster than the same plane suing 100/130 and no water.
P-47 with 150 octane and water still could not keep the cylinders cool when climbing and thus limited the amount of time the climb could be performed at the higher power settings.

Using higher boost settings does two things, It raises the temperature of the cylinder because the more highly compressed air will be hotter when it goes in the cylinder and the the more highly compressed air and fuel mixture is burning more fuel per cylinder stroke/ unit of time and creating more heat that way. A double wammy.
 
I followed everything you wrote except why higher compression may help in cruise but hurt max power. Is that do to cruise power being close to max power in a higher compression example or is it do to heat rise?

Cheers,
Biff
 
BiffF15 said:
I followed everything you wrote except why higher compression may help in cruise but hurt max power. Is that do to cruise power being close to max power in a higher compression example or is it do to heat rise?

Cheers,
Biff
Click to expand...

Possibly because high compression limits the amount of boost that can be thrown at the engine.
 
Hello GregP,
The point I was trying to make was that the muzzle velocity of a 5.56mm (.223) round is NOT really a Normal distribution and yet small samples are treated as if they were because it then makes the tools for Gaussian distributions available for comparison. The same applies for quite a few rules for Biology: They are not linear though we simplify things to assume they are in a small sample range to more easily grasp a concept.

Regarding Cessna 172 versus P-51D, there is no real difference. The same physical laws apply and it is only a matter of degree.
Just about every complaint you have made is probably due to the game authors not having actual experience flying the aircraft in the game and guessing wrong. FWIW, There are enough really cool effects to simulate the nasty handling of rotary engine from the Great War.
The simulators for the most part CAN handle them IFF the flight models are done correctly. It is a pity that you and I do not live closer together where we could sit down at the same PC and you can show me what is wrong with a particular simulator aeroplane and I can correct a few of the problems for you. You have the knowledge of how the actual warbird handles. I have at least some knowledge to tweak a flight model to do what I want but just don't know what is "correct".
I can tweak a flight model to let you do a 15G turn on a prop plane or never let you exceed about 4G as I might do on a heavy bomber. It is just knowing where to mess around.

I don't believe a two place cockpit is that hard for the newer simulators to handle. It is a bit beyond the scope of my interest though. I am mostly interested in flight modelling and 3D modelling. As for game control panels, for the most part, I don't believe they are really accurate for the simple reason that scattering instruments and controls all over the place doesn't work unless you have a purpose built cockpit. People have different opinions on that.

- Ivan.
 
Hello Shortround6,
The Homare 21 had a compression ratio of 8:1.
Maximum manifold pressure on Take-Off was +500 mm or 49.6 inches Hg so in comparison to American engines, this was not so high. Regular maximum (Military) power was +350 mm or 43.7 inches Hg.
Water injection automatically started at +180 mm boost or 37.0 inches Hg.

For a full internal fuel load of 697 Liters, 130 Liters of Water-Methanol were carried.
That is a pretty high proportion of Water to Fuel.

- Ivan.
 
The American radials were around 6.7 compression ratio, give or take .1-.2 depending on model.
Switching to liquid cooled for just a moment the Merlin was 6.0 and the Allison was 6.65, the Merlin would (officially) tolerate higher boost and make more power but had slightly higher fuel consumption per horsepower hour.
Compression ratio is a nominal number in that in measures the volume of the cylinder at both bottom dead center and top center but makes no allowances for valve (or port) openings which decrease the effective compression ratio somewhat.
That being said an engine with 8 to 1 compression ratio (unless the valves stay open for a very long time) will not be able to use the same boost using the same fuel as an engine with under 7 to 1 compression. The water injection helps boost the "effective" octane rating of the fuel and it allows (perhaps depending on the engine) the engine to run leaner at high power settings (many engines were using extra fuel as a coolant at high power settings).
The R-2800 used about 52-54in Hg at take off depending on model for 2000hp. so it wasn't that much higher than the Homare. Anything above 52-54in was WEP and yes they could get to 65-70in using water injection, 150 fuel two stage superchargers and intercoolers.
A Wright R-2600 of the 1700hp variety used 43-42in at low altitude and 44 1/2 to 45 1/2 in high gear for 1450hp with no water injection.
 
SR6,

If I understand you correctly, the higher the compression a motor has, the less boost it can take (it's by default of higher compression already near it's max capability)? That sounds correct as I have read if you want to turbo or supercharge a motor the compression ratio is usually lowered.

Cheers,
Biff
 
You must log in or register to reply here.

Users who are viewing this thread

Total: 1 (members: 0, guests: 1)

Similar threads

J
Ki-84 vs P-51 engagemants 1945
2
Replies
25
Views
3K
Jerrybandit67
J
The real combat history of the Ki-43
19 20 21
Replies
409
Views
28K
Wild_Bill_Kelso
KI-84 vs soviet aircraft 1945
Replies
11
Views
2K
Vincenzo
S
Comparison of the Gloster F.5/34 and the Mitsubishi A6M2.
Replies
6
Views
569
tomo pauk
N
He 100,Space age construction?
Replies
4
Views
885
OldGeezer
O
Menu
Log in
Register