Aerodynamic Drag Properties of the A6M

[tomo pauk]

It may have been possible.
Trying to figure exhaust thrust with any accuracy is just about impossible without a lot more data than we have for most planes. The only really good figures I have seen are for a Merlin XX in a Hurricane under test by RR.
...

There is a some data about exhaust thrust and/or equivalent of power gainedvia use of exhaust thrust for German engines. The DB 601A was making extra ~100 HP worth up to 7km altitude, provided aircraft is flying at 600 km (this is NOT due to ram, but, as far as I can get it, higher speed decreases back pressure), decreasing : graph (lowest part of the graph is for power-equivalent of exhaust thrust: 'Strahlleistungen in PS bei 600 km/h' ). Similar is for DB 601Aa (graph), the uppermost, thin line that ends at ~4.4 km shows total power at 600 km/h (shaft power + 'thrust power' equivalent) at max RPM and boost available.
For BMW 801D, exhaust thrust is expressed in kg (also noted in lbs here): graph
For Jumo 213A: graph. Exhaust thrust is 'Abgasstrahlschub', again expessed in kg.
I can't lay my hands at Jumo 213E thrust graph right now.

 

[Shortround6]

The RR data included the mass air flow, the exhaust gas velocity, the atmospheric pressure at altitude (and temperatures) some this was calculated. They had also rigged an engine in a test house/cell where they could but instruments on one cylinder bank to measure some this stuff while controlling the air pressure/temp. They then checked flight performace vs calculations.

Just saying engine X picked up 100hp due to exhaust thrust actually doesn't tell us much
photos for the earlier mentioned DB 601 exhausts

Now how much thrust do you get from a DB 601 engine??????
Or rather how much useful thrust if the axis of the thrust is 45 degrees from the aircraft's line of flight?

Kind of depends on the exhaust pipes doesn't it?

DB 605 on a Bf 110 night fighter.

Not only more drag but what is the velocity of the exiting exhaust gas at the rear after it is mixed with the cool/cold ambient air? Lowering the temperature of the exhaust gases lowes the volume and thus the pressure in the pipe. Lower pressure means lower velocity.

 

[tomo pauk]

The RR data included the mass air flow, the exhaust gas velocity, the atmospheric pressure at altitude (and temperatures) some this was calculated. They had also rigged an engine in a test house/cell where they could but instruments on one cylinder bank to measure some this stuff while controlling the air pressure/temp. They then checked flight performace vs calculations.

Just saying engine X picked up 100hp due to exhaust thrust actually doesn't tell us much
photos for the earlier mentioned DB 601 exhausts
Now how much thrust do you get from a DB 601 engine??????
Or rather how much useful thrust if the axis of the thrust is 45 degrees from the aircraft's line of flight?

Kind of depends on the exhaust pipes doesn't it?

DB 605 on a Bf 110 night fighter.

Not only more drag but what is the velocity of the exiting exhaust gas at the rear after it is mixed with the cool/cold ambient air? Lowering the temperature of the exhaust gases lowes the volume and thus the pressure in the pipe. Lower pressure means lower velocity.

Obvously, one is going to get less than specified if exhausts are convoluted, or merge in a common manifold. My take is that specified kg or PS gain is for the best possible layout of exhaust stacks - probably the 1-per-cylinder type we can see on many ww2 aircraft not intended to fly during the night, with exhausts pointed as backwards as possible. Perhaps the RR data was also for best-case scenario (exhausts as used on Spitfire IX, or P-51, not for the early type of exhausts as used on Hurricane I/II, Spitfire I/II/V, nor for the type used on Lancasters or Beaufighters)?

 

[Shortround6]

Perhaps the RR data was also for best-case scenario (exhausts as used on Spitfire IX, or P-51, not for the early type of exhausts as used on Hurricane I/II, Spitfire I/II/V, nor for the type used on Lancasters or Beaufighters)?

Test aircraft was a Hurricane II with standard (?) exhaust. test was done in high gear only from 15,000ft to 35,000ft in 5,000 ft increments.

There are a few other tests/charts were experiments were done with different sized exhaust openings. again on engine with standard or open exhaust on one side and test exhaust on the other while in the test cell/house.

ANd we are talking about exhaust thrust and not exhaust Horsepower which is inversely proportional to the speed of the aircraft.
The closer the plane's speed matches the speed of escaping gases the more "power" you get from the same thrust.
Exhaust (or jet thrust) with the plane parked on the runway with brakes set is zero thrust horsepower.

 

[DarrenW]

My sympathies for your loss. I saw your announcement but don't know you well enough to have anything more meaningful to say.

Thank you Ivan.

You accused me of being a "proponent of the Zero's virtues"

Sorry, it wasn't meant as something demeaning. You seemed to be providing me with a point/counterpoint situation. I find it's more fun to have a little healthy competiveness while exchanging information (must be the military guy in me). I was taking the pessimistic approach concerning maximum level speeds attained, while you seemed to be more optimistic IMHO. This obviously didn't turn out for the best though...

The "Normal Rated" Power for Hellcat isn't really the same thing as what was called "Normal Maximum" for the A6M series.

If you look at the closest engine settings to the "Normal Maximum" for Sakae engines and meet the criteria of Boost Pressure barely above Sea Level ambient and RPM about 250 below maximum, what you get for an equivalent for the Hellcat is "Maximum Cruise"

So let me see if I have this straight. The speed listed in the A6M5 manual is for "Normal Maximum", which is basically like a high cruise setting for the airplane, correct? But doesn't 338 mph at altitude seem rather excessive to you in this setting, especially with boost running barely above normal S/L pressure? There has to be more to this that I'm honestly not getting at that moment. Does the manual happen to give the altitude for that particular speed?

 

[fliger747]

The unmarked boxes of fruit contains apples, oranges and bananas add to the confusion. For US reciprocating engines of the era the limits were; Takeoff or Military power; Max continuous and War Emergency or (wet) power. The speeds noted for the aircraft are often not clearly mated with the appropriate power regime. The excellent book "Americas Hundred Thousand" does have graphs showing the speeds vrs altitude for both Mil and WEP.

As all of the engines considered have some form of Super/turbo charging, limiting MP will be available to some value above SL. With a multi stage supercharger the HP and speed curves will have a definite jag in them at the point of blower shifts. The upshot of this is that plane will higher speeds available up to a critical altitude, dependent on the power level selected. An aircraft can maintain MP for max continuous higher than Mil or WEP.

And yes, as was the bane of early jets, the thrust conversion to some equivalent HP value will increase with he aircrafts velocity.

 

[taly01]

Many interesting questions here I worry about also, I mainly have a Ki-43 interest but A6M been so similar I follow it also.

Speeds Zero 21/22/52

Interesting someone calculated Zero.21 as less drag than Zero.52. I suspect the smaller diameter cowling on 21 flowed better along fuselage cigar profile, Zero.21 is the efficiency peak of the design. The apparently similar speeds for A6M2 vs A6M3/5 is only valid for the A6M2's single speed supercharger rated altitude 14,000ft, above and below that the two speed A6M3/5 is faster peaking at 9,000/20,000ft.

Quotes for the Zero.22 been faster than the Zero.32 are an example of a test anomaly as the 32 was originally made because the long 2x series wings were felt to slow the A6M3 too much! Given the data sources we have on other Japanese planes I suspect the often quoted 351mph speed for A6M5 was a pre-production test machine.

Thrust exhausts , jet thrust effect?

I have struggled to understand the Ki-43 jet effect exhausts and I summarise here to the similar A6M development.

Here is a typical early war radial engine exhaust collector,

+ easy to fit
+ cheap to make
+ compact space efficiency

- poor exhaust flow path reduces cylinder gas scavenging that
- increases exhaust valve heat load
- reduces power due to sharp gas bends


Here is the A6M5 with a very nice jet effect exhaust

+ increases HP due to better cylinder gas scavenging
+ reduces heat load on exhaust valves
+ has more rearward thrust from direct gas path
+ reduces skin friction along fuselage
+ jet engine effect at higher plane speeds
+ engine more responsive
+ engine can speed up faster and higher rpm limit

- more difficult to make
- requires airframe changes to fit
- can burn airplanes skin

So my theory is on a RADIAL engine the jet effect exhaust noticeably increases HP and rev capability, and this is worth as much gain as the "jet-effect".

 

[Ivan1GFP]

ANd we are talking about exhaust thrust and not exhaust Horsepower which is inversely proportional to the speed of the aircraft.
The closer the plane's speed matches the speed of escaping gases the more "power" you get from the same thrust.
Exhaust (or jet thrust) with the plane parked on the runway with brakes set is zero thrust horsepower.

Hello Shortround6,
I do not believe you are correct in your explanation.
Power = Force * Velocity
so the Power is proportional (not inversely proportional) to aircraft speed.
The second issue is that if the aeroplane's speed matches the speed of the exhaust gas, then the force being applied to accelerate the gas would be zero thus force would be zero and there would be no power at all.
Regarding the Power at Zero forward speed: With Propeller thrust, there is something called "Low Speed Theory" which doesn't use the typical HP * Efficiency to calculate Thrust.

Sorry, it wasn't meant as something demeaning. You seemed to be providing me with a point/counterpoint situation. I find it's more fun to have a little healthy competiveness while exchanging information (must be the military guy in me). I was taking the pessimistic approach concerning maximum level speeds attained, while you seemed to be more optimistic IMHO. This obviously didn't turn out for the best though...

Hello DarrenW,
I have spent most of my career working with the military and I don't see that kind of attitude with people I have encountered regardless of whether they were American or allied personnel. There have been a couple, but they were the odd ones and we just avoided them when possible.
Regarding optimism versus pessimism, just observe what kind of reaction my comment about the condition of the typical Hellcat aboard a carrier got as a response. If we are arguing about condition, then I know I can provide no real evidence to justify any guesses and guesses are all that they would be.
If you believe that I was arguing that the A6M was superior to the Hellcat as a fighter, I definitely was not. I figure that the A6M was the dominant fighter in the Pacific for about two years (1941-1942) and the Hellcat was also pretty dominant for about two year (1943-1944). After that, each was essentially left behind in performance.

So let me see if I have this straight. The speed listed in the A6M5 manual is for "Normal Maximum", which is basically like a high cruise setting for the airplane, correct? But doesn't 338 mph at altitude seem rather excessive to you in this setting, especially with boost running barely above normal S/L pressure? There has to be more to this that I'm honestly not getting at that moment. Does the manual happen to give the altitude for that particular speed?

I posted these engine settings from the manual back at Post #30:
Engine Settings

From the numbers listed one could argue that they represent equivalents of Max Continuous or Max Cruise.
I don't know which is closer or if either is an equivalent. You need to find someone who knows more about the Sakae engine.
I am just pulling data from my notes and from a manual which really doesn't explain.
These settings really ARE barely above WOT at sea level with an unsupercharged engine.
Figure that
+50 mm = 31.9 inches Hg or not quite +1 pounds boost.
+75 mm = 32.9 inches Hg or +1.45 pounds boost.

Does this kind of speed seem reasonable at those power settings?
I believe so.
I doubt that they would be included in a manual on how to operate the aircraft if they were not reasonably accurate.
Remember though that these Sakae engines and Japanese engine as a rule were much less supercharged than Allied engines and their maximum manifold pressures were much lower.
Another thing to keep in mind is that both models of the Sakae described here had single stage superchargers.
The Sakae 21 has a two SPEED supercharger, but it was still single stage.
These engines would be losing less power to drive more supercharger as on the R-2800.

Model Speed Altitude
A6M2-21 275 Kts 4400 m
A6M3-32 290 Kts 6150 m
A6M3-22 292 Kts 5900 m
A6M5-52 294 Kts 5900 m

- Ivan.

 

[DarrenW]

we just avoided them when possible

From here on out consider yourself ignored.

 

[DarrenW]

Interesting someone calculated Zero.21 as less drag than Zero.52. I suspect the smaller diameter cowling on 21 flowed better along fuselage cigar profile, Zero.21 is the efficiency peak of the design. The apparently similar speeds for A6M2 vs A6M3/5 is only valid for the A6M2's single speed supercharger rated altitude 14,000ft, above and below that the two speed A6M3/5 is faster peaking at 9,000/20,000ft.

Quotes for the Zero.22 been faster than the Zero.32 are an example of a test anomaly as the 32 was originally made because the long 2x series wings were felt to slow the A6M3 too much! Given the data sources we have on other Japanese planes I suspect the often quoted 351mph speed for A6M5 was a pre-production test machine.

Nice summation of the presented facts. This is what I will take away from this discussion as they seem to be based more on logic, rather than what makes someone "feel good".

 

[Ivan1GFP]

Interesting someone calculated Zero.21 as less drag than Zero.52. I suspect the smaller diameter cowling on 21 flowed better along fuselage cigar profile, Zero.21 is the efficiency peak of the design. The apparently similar speeds for A6M2 vs A6M3/5 is only valid for the A6M2's single speed supercharger rated altitude 14,000ft, above and below that the two speed A6M3/5 is faster peaking at 9,000/20,000ft.

Quotes for the Zero.22 been faster than the Zero.32 are an example of a test anomaly as the 32 was originally made because the long 2x series wings were felt to slow the A6M3 too much! Given the data sources we have on other Japanese planes I suspect the often quoted 351mph speed for A6M5 was a pre-production test machine.

Hello Taly01,
Glad to see you're involved here. The Ki 43 and A6M do seem to share a lot of similarities.
I still wonder about the drag comparison between A6M2 and A6M3/A6M5. I am not sure exactly what you are saying here.
As I see it, the A6M3/5 had a cowl of the same width but slightly higher because of the change in carburetor intake.
The lower side was cleaner because there was no longer an external carburetor intake ahead of the oil cooler scoop.
The cowl also had a longer chord.

The quote of the A6M3-22 being faster than the A6M3-32 is actually from the aircraft manual which is about as official as it gets.
As for the history of the A6M3, note that the Model 32 came first, so there was no prior experience with the older 12 meter wing configuration on the A6M3 airframe. The Model 22 came about to address issues encountered with Model 32.
I also would have expected Model 32 to be slightly faster, but that is not how it is listed in the manual.
The difference is very small and the altitudes are not quite the same even for the same engine, so perhaps that affected the results.

Regarding the differences with an ejector exhaust configuration:
I can see the possible improvements in scavenging and engine response, but that would depend on the length and volume of the runners along with the engine's firing order.
What I don't see is how it would reduce the heat load on the exhaust valves and how the engine's RPM limit would be raised.
As I see it, if scavenging is improved, the temperature of combustion gas would remain about the same but the volume may increase slightly and thus increase the heat load on the exhaust valves. As for RPM limits, that would be a limitation of the internal stresses on the reciprocating parts of the engine and that would not change at all.

Regarding pre-production A6M5, there appear to have been quite a few service test examples with different configurations. Are you suggesting that the definitive production variant was faster or slower?

- Ivan.

 

[taly01]

I also would have expected Model 32 to be slightly faster, but that is not how it is listed in the manual.
The difference is very small and the altitudes are not quite the same even for the same engine, so perhaps that affected the results.

Yes we are looking at a <2% difference in speeds, not statistically significant unless they tested 10 of each plane, which i doubt! Logically the short wing Model.32 should be faster but perhaps at around the 20,000ft top speed height the longer winged Model.22 has some advantage.

What I don't see is how it would reduce the heat load on the exhaust valves and how the engine's RPM limit would be raised.

I found some papers about metallurgy of aircraft exhaust steels by Roll Royce where they say collector rings run at 800'C and straight exhausts run at 400'C. My piston engine experience is in street racing car engines, and a free flowing exhaust does improve throttle response, and alone they can improve power by ~10% over simple "log" manifolds even on a common engine. Its not really a higher RPM limit, but as more power now comes at higher RPM that makes that high RPM usable. If the exhaust is very restrictive all the 1000'C exhaust gases will not be expelled and some remain in the cylinder, and gases in the exhaust pipe may even flow back to the closed exhaust valve after it closes due to back pressures!

I was shocked when I first saw a typical radial engine exhaust collector, its about as bad a design for efficiency as you could make. There is no doubt a free flow exhaust on a radial would increase HP, I am surprised there is no test data on this?

Regarding pre-production A6M5, there appear to have been quite a few service test examples with different configurations. Are you suggesting that the definitive production variant was faster or slower?

I think 351mph is the speed for a good A6M5, The 1944 US tests A6M5 was part of a batch of ~12 they brought from Saipan, so I give them credit that they built a decent one out of that and they got 335mph which is only <5% slower.

 

[Ivan1GFP]

Yes we are looking at a <2% difference in speeds, not statistically significant unless they tested 10 of each plane, which i doubt! Logically the short wing Model.32 should be faster but perhaps at around the 20,000ft top speed height the longer winged Model.22 has some advantage.

Hello Taly01,
I would hope that if a military service sees the point in listing a speed difference in two nearly identical aircraft types in their operations manual, there actually is some basis for it and it is more than a test of a single example of each aircraft. Keep in mind that these were not full power speed runs and not likely to be abusive to the aircraft, so to get a bunch of test runs is easy. Then again, these are just assumptions and I don't know that this was actually done. Be pretty silly if it were not though.

I found some papers about metallurgy of aircraft exhaust steels by Roll Royce where they say collector rings run at 800'C and straight exhausts run at 400'C. My piston engine experience is in street racing car engines, and a free flowing exhaust does improve throttle response, and alone they can improve power by ~10% over simple "log" manifolds even on a common engine. Its not really a higher RPM limit, but as more power now comes at higher RPM that makes that high RPM usable. If the exhaust is very restrictive all the 1000'C exhaust gases will not be expelled and some remain in the cylinder, and gases in the exhaust pipe may even flow back to the closed exhaust valve after it closes due to back pressures!

I am in agreement with your points here. My own experience is that quite a lot can be done with an automotive exhaust system even if one restricts modifications to behind the exhaust manifolds. Running without mufflers and only little stubs behind the cats on each side made for some serious differences in launching. The car squatted so much it seemed to me that it might have been lifting the left front wheel. Replacing the catalytic H pipe with a X type crossover pipe also seemed to seriously improve things though there wasn't much ground clearance (about 3 inches). Although the engines rev faster and mid range torque improves, there isn't any change in RPM limits though.
The point that I was really getting at is that scavenging, reducing back pressure and such are not going to reduce the heat load on the exhaust valves even if they reduce the heat of the exhaust manifold. If some combustion gas remains in the cylinder, all it is going to do is reduce the temperature of the next combustion cycle. That is the principle of the EGR system in modern automobiles.
If you take apart a few engines, you will find that typically there are carbon deposits on the intake valve stems (especially with carburetor and throttle body injection) and a couple places in the combustion chamber but never on the exhaust valve. They tend to run so hot that everything is burned off and if one is not careful, even the edges of the valve can get burned off. Typically aircraft engines and high performance auto engines hollow out the exhaust valve stem and fill the cavity with sodium to improve heat transfer.

I was shocked when I first saw a typical radial engine exhaust collector, its about as bad a design for efficiency as you could make. There is no doubt a free flow exhaust on a radial would increase HP, I am surprised there is no test data on this?

I guess what we would have here is a nice balancing act. There is no question the exhaust collector ring is non-optimal.
The question is whether very short open exhaust stubs or tuned exhaust runners are better. I believe exhaust runners would be better for scavenging, but would reduce exhaust thrust because the temperature would be reduced.

I think 351mph is the speed for a good A6M5, The 1944 US tests A6M5 was part of a batch of ~12 they brought from Saipan, so I give them credit that they built a decent one out of that and they got 335mph which is only <5% slower.

Lets say that 351 MPH is the "maximum speed" for a A6M5 in good condition.
Presumably that would be at their Military Power setting of +200 mm and 2700 RPM at around 6000 meters altitude.
The next question is whether there was a higher manifold pressure setting available at that altitude.
Perhaps the supercharger has remaining capacity or perhaps ram effect has raised the critical altitude significantly.
Certainly the engine can be operated at 2750 RPM without destruction. Would this make a difference?

Regarding the A6M5 restored for testing as compared to the earlier A6M2 restored for testing:
The A6M2 was able to achieve very near its original maximum speed even though it started off as a crashed aircraft,
The A6M5's from Saipan were captured in a group and transported along with spare engines back to the states.
Were they worn out by the time of their capture on Saipan or perhaps the journey by ship damaged them?

- Ivan.

 

[taly01]

Quoted Aircraft Speeds

I agree there would be some test basis, but I bring up the Me109G-2 tests the Germans did in 1942, from memory they tested some 20 planes coming of the production line and the speeds varied 390-410mph!

exhausts & A6M2 vs A6M3/5 performance

I may overemphasise the exhaust valve issue but straight through thrust-pipes would remove some heat from that region. I would love to see some HP test data on the change from collector ring to thrust-pipes for the Sakae engine, stating 1130hp for all definately is an example of book figures rather than tested figures.

In Jiro Horikoshi's book he says the A6M3 was the first plane he designed that was slower than design calculations, as planes in flight had more HP due to ram effect than by dyno figures, his theory was that the A6M3 air inlet in upper cowl was not as effective as the A6M2 one. With the later Ki-43's Nakajima had at least 3 variations of their air inlet with the Ha-115 "Sakae.21", they enlarged and pushed it just outside the cowl ring and each variation did boost the planes speed. Its surprising that Mitsubishi did not try this "simple" trick also?

A6M3/5 air intake

Ki-43-III

 

[Ivan1GFP]

I agree there would be some test basis, but I bring up the Me109G-2 tests the Germans did in 1942, from memory they tested some 20 planes coming of the production line and the speeds varied 390-410mph!

Hello Taly01,
I don't actually know much about the production of the Me 109G-2, but I do know that later versions were made by several different factories and the performance of some factories' aircraft was better than others. The Erla 109s were notably faster that the others but I don't know if they were building them as far back as the G-2.
I wonder what the manuals would have listed for performances of the G-2?

I may overemphasise the exhaust valve issue but straight through thrust-pipes would remove some heat from that region. I would love to see some HP test data on the change from collector ring to thrust-pipes for the Sakae engine, stating 1130hp for all definately is an example of book figures rather than tested figures.

I was really commenting specifically about the heat load on the Exhaust Valve. I am not disagreeing that the exhaust manifold would be cooler or the area under the cowl / hood would be cooler. Basically the amount of heat the valve is subjected to is determined by the combustion temperatures and the amount of gas flow. The only way the valve can reject heat is to the cylinder head via contact with the valve seat and through contact between valve stem and valve guide.

That 1130 HP figure gets tossed around a lot for the Sakae 21 engine but I don't believe it is really representative of the type.
The only time the Sakae 21 is supposed to make 1130 HP is at its Take-Off setting at Sea Level.
See this post for some comparisons:
Nakajima Sakae Engine Settings

In Jiro Horikoshi's book he says the A6M3 was the first plane he designed that was slower than design calculations, as planes in flight had more HP due to ram effect than by dyno figures, his theory was that the A6M3 air inlet in upper cowl was not as effective as the A6M2 one. With the later Ki-43's Nakajima had at least 3 variations of their air inlet with the Ha-115 "Sakae.21", they enlarged and pushed it just outside the cowl ring and each variation did boost the planes speed. Its surprising that Mitsubishi did not try this "simple" trick also?
[/QUOTE]

Is this the book "Eagles of Mitsubishi"? If so, what page is the description on? (I just want to read about it for myself.)
My own belief is that the actual power increase of the Sakae 21 over Sakae 12 turned out to be much less than expected.
(Please refer to link above)
The Sakae 12 / Ha-25 is often listed as a 940 HP or 950 HP engine.
It makes 940 HP @ 2550 RPM at +250 mm for Take-Off
It is capable of achieving 950 HP @ 2500 RPM at +150 mm military power at its critical altitude of 4200 meters.

The Sakae 21 / Ha-115 is typically described as a 1130 HP engine.
It makes 1130 HP @ 2750 RPM at +300 mm for Take-Off (So far, so good.....)
At its critical altitude of 6000 meters, it makes 980 HP @ 2700 RPM at + 200 mm for military power.
With the supercharger in low speed, it does come close.
At 2850 meters, it would make 1100 HP @ 2700 RPM at + 200 mm.

Regarding differences in the carburetor intake:
Note that the Ki 43-I had the same intake configuration as A6M2 and was a LOT slower with basically the same engine.
Typically listed speed is 308 MPH versus 335 MPH.
The Ki 43-II had several intakes and cowl shapes as you mentioned but although it had the same engine as the A6M3, it was still slower.
Typically listed speed is only 329 MPH versus 340 - 345 MPH.
The Ki 43-III added ejector exhaust and is the first model of the Hayabusa that is actually faster than the A6M (A6M5) with an "equivalent" engine.
Typically listed speeds are 358 MPH versus 351 MPH.
IS this the full story though?
The typically listed power output for the Ki 43-III is over 1200 HP and apparently the Army managed to get a Water Methanol injection system working on it while the Navy didn't with the A6M5.

I believe the numbers are pretty representative but please correct them if you have better numbers.
I am not convinced that this "simple" intake change was a significant performance enhancement.

- Ivan.

 

[taly01]

I don't actually know much about the production of the Me 109G-2,

I was just giving an example of production variation, a complex machine like an aircraft assembled by several different shifts from different production batches will not all be +/-1% to book values.

Is this the book "Eagles of Mitsubishi"? If so, what page is the description on? (I just want to read about it for myself.)
My own belief is that the actual power increase of the Sakae 21 over Sakae 12 turned out to be much less than expected.

Ask Hiromachi
yes the Sakae.21 only got some 200rpm more and +2psi extra boost, the 2 speed supercharger helped low alt and high alt. The 980 vs 1130hp is misleading as its a low altitude peak. But it was ~50kg heavier and needed a larger cowling. I find the A6M3/5 cowl ugly

Note that the Ki 43-I had the same intake configuration as A6M2 and was a LOT slower with basically the same engine.

The Ki-43-I had a huge wing and apparently not as smooth riveted skin as Zero, also the fuselage design was not a high speed cigar, and Army used worse fuel than Navy before 1942! Its interesting that the Ki-43 followed much the same development path, but improved much more over its development than Zero to catch up and even slightly overtake it. The Ki-43 with 2 speed ha-115 "Sakae.21" did 1 major cowl redesign to taper it better and 3 redesigns of air intake to enlarge it and move into air stream more.

Although they say air ram effect only really matters over rated altitude of supercharger speeds, so maybe thats why zero did not get a better one.

The typically listed power output for the Ki 43-III is over 1200 HP and apparently the Army managed to get a Water Methanol injection system working on it while the Navy didn't with the A6M5.

The water-injected Sakae.31 is a bit of a mystery, it apparently needed a larger cowling, but the Ki-43-III's water-injected ha-115-II did not? Perhaps the Navy used direct manifold injection ?!?

 

[Ivan1GFP]

yes the Sakae.21 only got some 200rpm more and +2psi extra boost, the 2 speed supercharger helped low alt and high alt. The 980 vs 1130hp is misleading as its a low altitude peak. But it was ~50kg heavier and needed a larger cowling. I find the A6M3/5 cowl ugly

Hello Taly01,
I actually like the shape of the A6M3 cowl better than the others. The A6M5 is actually shaped just a little differently according to Aero Detail and I can see that there are differences, but I can't really tell where they are (besides the ejector exhausts of course).
The later cowl is undoubtedly higher because of the relocated intake, but is it also wider?

The Ki-43-I had a huge wing and apparently not as smooth riveted skin as Zero, also the fuselage design was not a high speed cigar, and Army used worse fuel than Navy before 1942! Its interesting that the Ki-43 followed much the same development path, but improved much more over its development than Zero to catch up and even slightly overtake it. The Ki-43 with 2 speed ha-115 "Sakae.21" did 1 major cowl redesign to taper it better and 3 redesigns of air intake to enlarge it and move into air stream more.

FWIW, the A6M2 actually had a larger wing than the Ki-43-I
A6M2
Wing Span 12.000 meters
Wing Area 22.44 meter^2 (241.5 feet^2)

Ki 43-I
Wing Span 11.437 meters (11.500 meters orthogonal)
Wing Area 22.00 meter^2 (236.8 feet^2)

How was the fuel different other than the 91 octane versus 92 octane standard?
I had always figured that the difference in performance was due to the two blade adjustable pitch propeller which wasn't really enough to absorb the power of the engine. The interesting thing is that the A6M also started life with a two blade propeller.
Although the Ki 43 always had an agility advantage, I don't see that it actually caught up in maximum speed until the Ki-43-III and there was enough of an engine power increase that its speed advantage may be attributed to that rather than aerodynamic improvements.
Did the changes in intake configuration actually improve performance of the Ki 43-II and if so, by how much?

Although they say air ram effect only really matters over rated altitude of supercharger speeds, so maybe thats why zero did not get a better one.

Ram effect is basically additional supercharger capacity, so if Mitsubishi had been able to raise their critical altitude it would have been silly not to attempt it. My guess (emphasis on GUESS) is that altering the intake to a more prominent version might improve ram effect but would also increase drag and cost performance.

The water-injected Sakae.31 is a bit of a mystery, it apparently needed a larger cowling, but the Ki-43-III's water-injected ha-115-II did not? Perhaps the Navy used direct manifold injection ?!?

I believe the photographs I have seen of Sakae 31 aircraft have had more of a gap between cowl flaps and fuselage but otherwise look the same. As for why the Navy didn't get a Water Methanol system working on the Sakae 31, it does seem strange because they didn't seem to have trouble with other aircraft such as the J2M and N1K series.

- Ivan.

 

[taly01]

Thanks Ivan i am starting to relook at some things I had ignored for a while.

The later cowl is undoubtedly higher because of the relocated intake, but is it also wider?

The A6M3/5 cowl is higher as the cowl MG openings get covered by it, where in the A6M2 they are open channels. Whether the cowl is a perfect circle though?
In comparison the Ki-43-I -> II cowl did not get wider, just longer and MG tubes appear same.

Ki-43-I.... and there was enough of an engine power increase that its speed advantage may be attributed to that rather than aerodynamic improvements.
Did the changes in intake configuration actually improve performance of the Ki 43-II and if so, by how much?

Apparently early IJ army fuel was only 87 octane up to 1942. The two blade propeller may be a case of cheapness rather than performance like using old tubular gunsights on Ki-43-I But the Ki-43-I was a "bad" plane, Nakajima got free contract from Army without competition, (and this at around time they said Zero was impossible to build but Horikoshi did it!) Zero should have replaced Ki-43 by Pearl Harbour if they made honest appraisal and no Army vs Navy rivalry!

I rechecked my references and some do state the Army Ha-115 (Sakae.21) take off HP increased from 1120 -> 1130 -> 1150 -> 1190 during the different Ki-43 series. The 1130 -> 1190hp increase actually matches my guess-estimated 5% increase from straight pipes vs collector ring!

Cowl Ki-43 -IIb -II Kai

cowl for Ki-43-III

With a series of constant refinements during 1943 the Ki-43-II went from around 320 to 335mph on the same power, the Ki-43-III may have tested 358mph in clean configuration but all came with bomb racks (see above) that cost ~15mph, so it also was around 340mph true speeds.

I believe the photographs I have seen of Sakae 31 aircraft have had more of a gap between cowl flaps and fuselage but otherwise look the same

Well spotted, its just an optical illusion from slightly shorter cowl flaps? Maybe this was a way to get some more cowling without drag?

I think also Sakae.31 was a case of why bother....it would not have been enough help and their were other more promising projects to do in 1945.

 

[Ivan1GFP]

The A6M3/5 cowl is higher as the cowl MG openings get covered by it, where in the A6M2 they are open channels. Whether the cowl is a perfect circle though?
In comparison the Ki-43-I -> II cowl did not get wider, just longer and MG tubes appear same.

Hello Taly01,
I looked through the drawings in AeroDetail Number 7 and found nothing that indicated actual diameter of the cowl
The height is obviously greater but from the plan views, the width appears to be the same. If there are differences, I don't think they would be visible without dimensional labels.
If anyone would know, Shinpachi should. He has built some very detailed 3D models. My own 3D models are much less precise and back when I did them around 15 years ago, they were based on drawings that were not so good.

Apparently early IJ army fuel was only 87 octane up to 1942. The two blade propeller may be a case of cheapness rather than performance like using old tubular gunsights on Ki-43-I But the Ki-43-I was a "bad" plane, Nakajima got free contract from Army without competition, (and this at around time they said Zero was impossible to build but Horikoshi did it!) Zero should have replaced Ki-43 by Pearl Harbour if they made honest appraisal and no Army vs Navy rivalry!

The easy way to check whether the lower octane had any real effect is to check the manifold pressure settings on the early aircraft. If they are capable of running the same manifold pressure and RPM without detonation and without Water-Methanol, then obviously it wasn't necessary. I collected a bunch of drawings and a lot of dimensional data on the Hayabusa in preparation for building a 3D model, but didn't collect as much performance data.
Absolutely agree that Ki 43-I was a "bad plane" that had a tendency to break up in the air even without battle damage. I am not so much in agreement that the Ki 43-II was inferior to A6M though. Its firepower wasn't so good but its high speed handling was much better.

I rechecked my references and some do state the Army Ha-115 (Sakae.21) take off HP increased from 1120 -> 1130 -> 1150 -> 1190 during the different Ki-43 series. The 1130 -> 1190hp increase actually matches my guess-estimated 5% increase from straight pipes vs collector ring!

With a series of constant refinements during 1943 the Ki-43-II went from around 320 to 335mph on the same power, the Ki-43-III may have tested 358mph in clean configuration but all came with bomb racks (see above) that cost ~15mph, so it also was around 340mph true speeds.

5% is an awfully big change for an alteration of that type in my opinion, but I haven't looked at your details either.
The Refinements for Ki 43-II were not just cowl shape, but also a reduction in wing span and wing area and canopy improvements which along with some minimal power increases would make that level of speed increase pretty reasonable.
I am actually a bit surprised that there was such a great loss in speed with the wing racks on Ki 43-III.
That would actually make the factory Ki 43-III slower than a factory new A6M5.

Well spotted, its just an optical illusion from slightly shorter cowl flaps? Maybe this was a way to get some more cowling without drag?

I think also Sakae.31 was a case of why bother....it would not have been enough help and their were other more promising projects to do in 1945.

AeroDetail states that the cowl was larger for Sakae 31 and seems to indicate that Model 62 would have had Water Methanol injection but Model 63 had Sakae 31 without Water Methanol. I take it that for more promising, you are referring to A6M8 with Kinsei engine?

- Ivan.

 

[taly01]

AeroDetail states that the cowl was larger for Sakae 31 and seems to indicate that Model 62 would have had Water Methanol injection but Model 63 had Sakae 31 without Water Methanol. I take it that for more promising, you are referring to A6M8 with Kinsei engine?

I checked in Squadron/Signal Zero and the A6M3 -> A6M5 did change air intake slightly higher and wider on same cowl shape. The A6M7 cowl appears have an even higher upper cowl line and a bulge in lower cowl, but I had to look so hard my eyes hurt! Model 62/63 is a very confusing model, and I think engineers rather work on A6M8 or A7M!

I am actually a bit surprised that there was such a great loss in speed with the wing racks on Ki 43-III. That would actually make the factory Ki 43-III slower than a factory new A6M5.

Osprey Aces Ki-43 has pilots claims that the last Ki-43-II which came with wing racks was slower than earlier Ki-43-II (no racks) and state "some 15mph slower" (although japanese army would have used kph). It must have been with the bomb claws attached.

Found some new comprehensive Jp.wiki data

栄 (エンジン) - Wikipedia

Ha-115/Sakae.21 all used +200mm as standard boost and +300mm as T.O. boost, so the Anti-Detonant-Injection water only increased power by "evaporative cooling density change of compresed intake air". All 7.2:1 compression all peak T.O. at +300mm boost and 2750rpm

A6M6 Sakae 31 T.O. 1300hp ***failed engine***
A6M7 m.62/3 Sakae 31a/b T.O. 1210hp

Ki-43-IIa (early) Ha-115 T.O. 1130hp
1st stage peak 1100hp @ 2850m +200mm
Ki-43-III Ha-115-II T.O. 1300hp
1st stage peak 1230hp @ 2800m +200mm

So at 200mm boost Ki-43-IIa vs Ki-43-III is 1100 vs 1230 = 12%
So at 300mm boost Ki-43-IIa vs Ki-43-III is 1130 vs 1300 = 15%

Nice improvement!