Kyushu J7W1 Shinden/J7W2 Shinden-kai

[AndersJ]

I have heard of a situation I can only partially recall at 'flap reversal' in some aerodynamic conditions. It looks like this could have happened in the 'bank' manouvre.

This is another one I would like to see as a flying replica but then I would say that.

Yes, a J7W1 Shinden replica roaring by the grandstands and flying formation with Mustangs and Thunderbolts at air shows, that would really be something!

 

[Scott Kenny]

About the torque: At first I wondered why they were complaining so much about it, but if you think about it, on a tractor, the rotating slipstream induces angles of attack on the wings and tail section which counteracts the torque the pilot experiences while on a pusher, you’re left to counter the full force of the torque with the ailerons.

In addition, the short moment of arm for the rudders on Shinden don’t help either when trying not to veer of the runway, so I’m not surprised if they had to be even more careful with applying power at take-offs compared to typical tractor designs.

I suspect that the rudders would end up getting enlarged a good bit in the production version.

 

[AndersJ]

I suspect that the rudders would end up getting enlarged a good bit in the production version.

Yes, because not only was the moment of arm short for the rudders, but they were outside the slipstream as well, so basically nothing to work with until you got up to speed.

So when taking off, you would probably have to nurse the throttle quite a bit and avoid opening it up too quickly. I guess they must have used the brakes to keep straight until the speed built enough to give the rudders got control authority.

Connected to this, does anyone know anything about nose-wheel locks or variable gearing on the Shinden as in different modes for taxiing and take-off/landing?

 

[Scott Kenny]

Yes, because not only was the moment of arm short for the rudders, but they were outside the slipstream as well, so basically nothing to work with until you got up to speed.

Oh, ewwwwwwww.....

That would suck in levels I don't even want to think about thinking about...

Connected to this, does anyone know anything about nose-wheel locks or variable gearing on the Shinden as in different modes for taxiing and take-off/landing?

@blackkite ? There are times I wish I lived closer to DC, so I could go diving into the Smithsonian on a regular basis...

 

[Hetzer]

Hi all; what is the difference of angle of attack between the canard and wing (root) typically for a high performance aircraft such as Shinden?

 

[riggerrob]

I suspect that the rudders would end up getting enlarged a good bit in the production version.

Either larger rudders or mount them farther aft to improve the length of the tail moment arm.
The farther rudders are from the center-of-gravity, the more powerful they are.

 

[Justo Miranda]

Hi all; what is the difference of angle of attack between the canard and wing (root) typically for a high performance aircraft such as Shinden?

Hi

 

[AndersJ]

Hi all; what is the difference of angle of attack between the canard and wing (root) typically for a high performance aircraft such as Shinden?

As far as I know, the canard had a 1 deg higher angle of attack than the main wing on the prototypes.

But a number of changes were planned for the series aircraft: "Planned changes from prototype to series aircraft: Main landing gear moved back 10 cm. Propeller changed from 6 to 4 blade with wider blades, possibly 3.4 to 3.5 m diameter. Canard angle of attack versus fuselage reference line increased from 1 to 3 degrees."

Unfortunately, I only have this text in my notes but I'm no longer sure about the provenance. Maybe someone else can confirm and source where the text comes from?

 

[blackkite]

The first prototype of the Shinden, completed in June 1945, failed to take off in July and damaged its propeller.
At the time of the first flight on August 3, a white chrysanthemum tail wheel was added to the lower end of the vertical stabilizer as a temporary measure.
The second test flight was conducted on August 6, and the third test flight on August 8. During the test flight, the engine broke down, and the war ended while repair parts were being ordered from Mitsubishi Heavy Industries for the next test flight.
As a result of the test flight, it was decided to increase the front wing angle of attack from 1 degree to 3 degrees and set the flap operating angle to 35 degrees for the first prototype in order to improve the vertical balance.
In addition, the addition of a large air scoop was considered as a countermeasure against the oil temperature, which tended to rise during flight. It was also considered to incorporate an auxiliary mechanism into the flight control system to counter the tendency of the aircraft to tilt to the right at low speeds due to torque.

 

[blackkite]

[Smithsonian National Air and Space Museum NASM. Steven F. Udvar-Hazy Center] Shinden
14390 Air and Space Museum Parkway, Chantilly, Virginia 20151 U.S.A.

The Smithsonian National Air and Space Museum NASM. Steven F. Udvar-Hazy Center, adjacent to Washington Dulles Airport, has restored and exhibited most of the Japanese Army and Navy aircraft, including those that were transported from occupied Japan for research after the war.

The "Ohka" exhibited next to the Gekko has disappeared somewhere,
and the "Shinden" has come to where the Ohka was!

Steven F. Udvar-Hazy Center · Steven F. Udvar-Hazy Center, 14390 Air and Space Museum Pkwy, Chantilly, VA 20151

âââââ · Museum

www.google.com

 

[blackkite]

Test Flights: (Japanese Wikipedia)
All three test flights were conducted by pilot Miyaishi, who had joined Kyushu Aircraft from the Navy. The engine was not at full power and the landing gear was extended, but the maximum speed during level flight was recorded at 293.5 km/h. During the first test flight (August 3, 1945), the propeller counter torque could not be fully offset, resulting in the plane leaning to the right, which was seen as a major adjustment issue, but other than that, no particular deviations were reported. However, in the second test flight (August 6, 1945), in addition to the right tilt, reports included "during ascent, the nose dropped even when the elevator was fully up," and "when landing, if the elevator was slightly up just before touchdown, the nose suddenly rose, and the plane touched down with the elevator fully down." In the third test flight (August 8, 1945), reports also included "there was no particular change in the nose with an increase or decrease in engine speed, but the nose dropped as the engine slowed down," and "during level flight, if the takeoff flaps were raised, the nose dropped, and level flight was possible with the elevator fully up." Excessive oil temperatures during flight were also recorded, and countermeasures were planned, but the war ended before the plane could be repaired. However, the vibration of the engine extension shaft, which had been a problem with the Raiden, did not occur at low speeds, and directional stability was said to be extremely good.

 

[blackkite]

Complicated aerodynamic characteristics
1.Compared to straight wings, the lift distribution of swept-back wings is larger on the outer wing side and smaller on the inner wing side. This means that the actual aerodynamic mean chord is shifted outward (rearward because it is a swept-back wing) from the aerodynamic mean chord position calculated using normal calculation methods, and the lift of the front wing is insufficient to balance it out. This coincides with the nose-down tendency confirmed in flight tests, and it is possible that the K-16 glider (straight wing), MXY6 (shallow sweep angle, insufficient horsepower to even fly horizontally on its own), and Shinden wind tunnel model (small Reynolds number) used as references for the design were unable to measure the actual situation. As a countermeasure, it was decided to change the canard's angle of attack from 1 degree to 3 degrees. However, the safe stall characteristic of a canard-based aircraft, "the canard stalls before the main wing, so the nose naturally drops and recovers from the stall", came at the expense of the fact that even though the main wing has some leeway before stalling, and potential remains in the maximum lift coefficient, this is not utilized and is cut off as soon as the canard stalls. Therefore, changing the canard's angle of attack by +2 degrees would cause the stall angle to be reached sooner, resulting in more of the main wing's potential being lost

 

[blackkite]

2.The Shinden's flaps are equipped on both the main wing (split) and the canard (slotted) and operate in conjunction with each other, but based on the results of test flights, the flap angle was standardized to 35 degrees (canard/main wing), which is a shallower angle than the Zero Fighter's 60 degrees (split) and the P51's 50 degrees (slotted). This is because there is room for the elevator's "pull" to work (canard) and because the flap position is far from the aircraft's center of gravity and the nose is lowered (wing), which means that it is difficult to adopt flaps with a high lift coefficient (Fowler or double slotted, etc.). This disadvantage is an obstacle to improving the wing horsepower, which is an indicator of not only takeoff and landing runway distance but also maximum speed. In addition, the split flaps on the main wing generate separation turbulence when deployed, which reduces the efficiency of the propeller, so they are not originally suitable for pusher propeller aircraft.

 

[Scott Kenny]

Complicated aerodynamic characteristics
1.Compared to straight wings, the lift distribution of swept-back wings is larger on the outer wing side and smaller on the inner wing side. This means that the actual aerodynamic mean chord is shifted outward (rearward because it is a swept-back wing) from the aerodynamic mean chord position calculated using normal calculation methods, and the lift of the front wing is insufficient to balance it out. This coincides with the nose-down tendency confirmed in flight tests, and it is possible that the K-16 glider (straight wing), MXY6 (shallow sweep angle, insufficient horsepower to even fly horizontally on its own), and Shinden wind tunnel model (small Reynolds number) used as references for the design were unable to measure the actual situation. As a countermeasure, it was decided to change the canard's angle of attack from 1 degree to 3 degrees. However, the safe stall characteristic of a canard-based aircraft, "the canard stalls before the main wing, so the nose naturally drops and recovers from the stall", came at the expense of the fact that even though the main wing has some leeway before stalling, and potential remains in the maximum lift coefficient, this is not utilized and is cut off as soon as the canard stalls. Therefore, changing the canard's angle of attack by +2 degrees would cause the stall angle to be reached sooner, resulting in more of the main wing's potential being lost

A better option would have been enlarging the front wing, even if that would have taken longer to do.

When I was in A&P school, a couple were trying to build a homebuilt crop duster that was a canard pusher. One of my airframe instructors had built his own Long-EZ, so was very familiar with canard pushers. Their plane was a high wing cantilever design powered by a Subaru automotive engine. All of us thought that the canard they had built was much too small, it was about half the size of my Instructor's EZ. My Instructor was not impressed with the attempt at first flight, the nose did not want to leave the ground.

 

[blackkite]

3.The unavoidable fate of a canard aircraft is that the wings are disturbed by the canard and are in a flow with reduced flow velocity, and the wingtip vortex and downwash constantly reduce the efficiency of the wings. At the same time, the wings' angle of attack (= lift) fluctuates due to the operation of the elevators, making vertical stability prone to over-sensitivity. Of course, the effects of turbulence are limited to the inside of the wings, but the angle of attack at the wingtip is relatively high, and the Shinden, which is a swept-back wing, is prone to wingtip stall. To improve this tendency, the wing setting angles of the Shinden are inconsistent at 0 degrees, 3 degrees, and 0 degrees at the root, center, and wingtip, respectively, which puts it at a disadvantage in terms of lift-to-drag ratio compared to a normal canard aircraft, which generates lift at an ideal angle of attack across the entire wingspan.

 

[AndersJ]

I see we have listed about the same issues for canard configurations @blackkite : As in the necessity to have a more heavily loaded canard surface to make sure it stalls before the main wing and that this in turn results in a lower trimmable Clmax on aircraft level. In addition, that the swept wing aggravates the danger of pitch-up at stalls since the tips on swept wings tend to stalls before the root. And this later point makes me wonder why they elected the strange 0 root, 3 mid-wing and 0 tip degrees washout scheme since having a more conventional washout like a 3 deg angle of attack at the root and then 0 at the tip would make more sense not only to lessen the risk of the tips stalling first, but since it would also have provided a better (more elliptical) wing lift distribution. In addition, it would also have helped to provide pitch stability in the same way the Horten brothers used washout and swept wings to attain stability on their flying wings. In fact, I think this, together with the frequent use of the ailerons as elevens, is the main reason so many canards and flying wings have swept wings.

About the wing sweep being the source of the excessive nose-down pitching moment: I’m sure the designers took the main wing’s sweep into account when calculating the aerodynamic neutral point and stability margins, and I would guess that the actual reason the Shinden’s test pilot had to contend with such a nose-heavy aircraft in the early test flights of the prototype was just a prudent precaution since it makes sense to start off with a too forward rather than a too backward center of gravity when you start testing a prototype.

There is however one thing I’m surprised about and that is the test pilot’s statement that directional stability was good. Because looking at the moment of arm of the rudders on the Shinden and comparing that to the destabilizing effect of its forward fuselage and canard, I would from a theoretical standpoint have guessed that the Shinden’s yaw stability would be poor rather than good. In fact, I would guess that the wing’s sweep on the Shinden was to a large extent there simply too increase the rudders moment of arm.

Other than that I’m not surprised to hear about the torque problems on the Shinden since as I pointed out in my previous post, a pusher misses out on the slipstream’s torque cancelling effect on the wings and tail section that is there on a tractor aircraft and thus has to balance the full force of the engine’s torque solely with the ailerons which of course presents problems at lower speeds and at higher power settings.

 

[blackkite]

Thanks for excellent comments!

 

[martinbayer]

The first prototype of the Shinden, completed in June 1945, failed to take off in July and damaged its propeller.

And that's likely why a jet propelled version was considered - I love it when a plan comes together...

 

[AndersJ]

Thanks for excellent comments!

Well thank you for posting all the interesting information on the Shinden in this thread!

 

[Tonton-42]

To illustrate the point about the efficiency of the take-off controls, the accident that Mr. Jacquard’s Spitfire suffered a few years ago is emblematic of what happens when you put too much gas on soft ground with a powerful aircraft. For the record, the pilot was not used to the plane (fighter pilot on jet) and had just flown on a Fury belonging to the same owner as the Spit. The exchange seems to have taken place at a moment’s notice without any prior "public stand", hence the accident. This illustrates in a dramatic way the difficulty of piloting, even by a professional, a powerful propeller machine. In addition to the aerodynamic effects, rotating masses (large propeller turning in one direction !) generate uncontrollable gyroscopic precession effects below the Minimum Control Speed. Considering the size and number of blades of the Shinden’s propeller (which make it a heavy flywheel!) and despite its wide tricycle landing gear being safer than that of the Spit, I suppose this additional effect had to be taken into account as well.

View: https://www.youtube.com/watch?v=pdvZHnklnuI

 

[blackkite]

Prototype modification plan up to mass production
Up to No. 2
Incorporated steering system correction mechanism to counter right tilt
Canard angle of attack 1 degree → 3 degrees
Flap operating angle 35 degrees
Modified lubricant cooler air intake
No. 3
Main landing gear mounting position moved back 10 cm to counter propeller grounding accidents
Temporarily installed side wing wheels abolished
Engine change: Due to Ha 43-43, supercharger air intake only on the right
Machine gun installed
No. 4~
Replaced with 4-blade propeller with 150% area (Plan A: diameter 3.4 m, Plan B: diameter 3.5 m)
No. 8~
Engine change: Abolished Vulcan coupling drive supercharger in Ha 43-44 with single stage three speed mechanical supercharger.　

Aft fuselage extended due to change in center of gravity caused by smaller tank due to reduced lubricant
Mass production
Engine mount raised 10 cm to improve productivity of main wing spar, cowl shape changed

 

[blackkite]

Hi!
Source : MARU ｍagazine in April 2011.

 

[blackkite]

Hi!

 

[blackkite]

Hi!

αcafe「(2024-01-20)十八試局地戦闘機 J7W1 震電」

福岡にある大刀洗平和記念館で展示されている 旧日本海軍の試作戦闘機です ※「局地戦闘機」とは邀撃戦闘機を意味します 映画「ゴジラ-1.0」で使われた実物大の精密レプリカです もうすこし広い場所に展示して欲しかった(;'∀') 17mm超広角でギリギリ撮れましたが、照明が取り付けられているバーが左翼に被ってます(=_=;

acafe.msc.sony.jp

 

[blackkite]

(Photo)　1/1 震電 ＜博物館実機写真＜2024年2月号

We should expect nothing less from the slow Smithsonian.

The biggest fools were the few Japanese who, in their desperation after the defeat, destroyed the Shinden, rendering it unable to fly.
They even burned down the Shinden No. 2 and No. 3, which were under assembly.

However, ABCD are also very to blame for pushing Japan to that point.

 

[blackkite]

View: https://www.youtube.com/watch?v=HmuwEo6R2hU

 

[blackkite]

View: https://www.youtube.com/watch?v=KXPPph5oeDE

 

[AndersJ]

I did not know that Mr Giannetakis built another J7W1 RC-model after the crash with that first incredibly detailed Shinden in 2007. So thanks for posting these new videos @blackkite .

Glad to see that this new version he flew in late 2015 handled so much better. Too bad it too crashed on landing though. But as I understand it the damage was not as bad this time. Granted, I'm not an RC-model flyer myself, but I was surprised to see him attempting to land a model with such small wheels in such high grass.

Another thing that surprised me is that there was such a large variation in elevator angle during the flight. But the model did seem to respond as expected, and that makes me think that maybe a lower gearing on the elevator would have improved the handling of the model and made it easier to control. At least that is what I would have expected from theory. But then I may be misinterpreting the video since my RC-model flying is quite limited to say the least.

However, if he was that far along already in 2015, one wonders if he ever repaired this new J7W1 model and flew it again? Because there seems to be no later videos other than those from about ten years ago?

 

[blackkite]

Hi!

画像ギャラリー | 「ゴジラで使われたものです」震電の展示先ついに暴露！ 監督もさっそく反応 | 乗りものニュース

ようやく胸のつかえがとれたってとこでしょうか。

trafficnews.jp

 

[AndersJ]

While I don’t have a later Thunderbolt modeled in my C++ simulations, it looks like the J7W1 Shinden would have been much faster than the P-47D-30 over the entire altitude band:



The small “kink” in the Shinden’s speed versus altitude from about 8 km up to the critical altitude is due to compressibility effects in the model. However, these effects are difficult to estimate and the result seen in the figure is not specific only to the Shinden’s model, but for the time being lacking other data a more general estimate of at what speed the compressibility drag rise should begin to make itself felt.

In addition, it looks like the Shinden’s ratio of mass to flat plate drag area is significantly higher than on the Thunderbolt meaning that it should also be able to out-dive it, which is really saying something since the Thunderbolt is usually considered to be the “king” of high speed dives.

 

[blackkite]

View: https://www.youtube.com/watch?v=8wVmZhLHyz8

 

[Nicknick]

#64

There is a confusion about Vulcan coupling and Föttinger couplings. The Vulcan coupling is a rubber damped coupling which was presumably used in the drive shaft to decouple vibration and misalignment. It has nothing to do, with the hydraulic Föttinger coupling which was used for realize a variable supercharger speed in the German engines (by a variable filling). The supercharger here, was surly also driven by a Föttinger couling and not a Vulcan coupling.

#134

The wing skin structure with the spot welded inner and outer metal sheets and a corrugated inner sheet is surprisingly similar to the wing of the Junkers J1. The J1 was the first stressed skin all metal plane and it wings even worked completely without an additional wing spar. However, the manufacturing process must have been complicate, especially when several wing segments made of two layer metal sheets has to be connected. Later on, Junkers used only a single layer corrugated metal which was surly much easier to manufacture, but increased drags and lowered the stiffness.

#237

I believe, that big thing in the middle is a dual purpose gear box which reduces the speed for the propeller and the cooling fan. It makes sense to do it in one box.

Because the cooling fan is quite close in its housing, the gear box was very likely stiffly attached to the wing spar and decoupled from the engine. The high speed drive shaft between the engine and the gearbox was decoupled with Vulcan couplings, which is also indicated in the graphic in #253. In this graphic is also the “extension shaft support cylinder” which is a quite unique feature. This could have worked in two ways, not sure, which one is the correct solution:

1. Cars with transaxle drive trains sometimes have a torque tube around the drive shaft which connects the engine with the gear box in the rear. This eliminates the torque transfer between the gear box and the engine trough the chassis. I believe, the “extension shaft support cylinder does something similar, as well as transferring axial forces. But here, unlike in transaxle cars, we will have a fixed gear box with a soft, rubber mounted engine, which means the support cylinder must be attached with joints o the gear box and engine. This could be two gigantic ball joints plus some kind of torque arm (at least, one detail could be interpreted like that with good will) or some kind of kardanic connection (no indication for that can be seen).

2. Another possible explanation, would be a support which keeps the prop axis in a fixed point close to the prop. Here, the Prop, the support cylinder and the engine would be a single stiff block. The engine side would still be supported in rubber bearings but with a limited degree of freedom (2 longitudinal directions and one rotation). This design would be much simpler and also lighter, but would transfer much more vibration from the engine to the plane.

 

[lancer21]

Prototype modification plan up to mass production
Up to No. 2
Incorporated steering system correction mechanism to counter right tilt
Canard angle of attack 1 degree → 3 degrees
Flap operating angle 35 degrees
Modified lubricant cooler air intake
No. 3
Main landing gear mounting position moved back 10 cm to counter propeller grounding accidents
Temporarily installed side wing wheels abolished
Engine change: Due to Ha 43-43, supercharger air intake only on the right
Machine gun installed
No. 4~
Replaced with 4-blade propeller with 150% area (Plan A: diameter 3.4 m, Plan B: diameter 3.5 m)
No. 8~
Engine change: Abolished Vulcan coupling drive supercharger in Ha 43-44 with single stage three speed mechanical supercharger.　

Aft fuselage extended due to change in center of gravity caused by smaller tank due to reduced lubricant
Mass production
Engine mount raised 10 cm to improve productivity of main wing spar, cowl shape changed

That is very interesting info. Looking at the intended production model drawing, the air intake on the right was supposed to be smaller than the one on the left?

 

[blackkite]

That's a very astute observation. The production model needs an ram air intake for the single-stage, three-speed mechanical supercharger, and it may be combined with the left side air intake for engine cooling.