Why were push-pull designs so rare? It seems like a really good

Sounds like a maintenance nightmare tbqh.

>>33304361
Jet engines murdered them in the crib.

>>33304406
But why didn't people play around with it in the 1930s when Germany was developing the BF-110, the US the P-38 etc

>>33304437
I'm gonna assume the technology simply wasn't mature enough

It took the greatest conflict in Human history and the cold war that followed to initiate most of the technological revolutions we saw. 30s was a different time

Rear engine overheating.
Problems with stability during landing and take off due to heavy tail.
Rear propeller prone to damage during take off
Overcomplicated fuel and hydrolic systems

Bottom line: concept that was looking good on paper, but was riddled with problems when turned into real plane

File: 3FDB3037-7F65-4992-9D7F-2B4F1B570DAD-20824-00001893EE1D07D2_tmp.jpg (93KB, 600x400px) Image search: [Google]

93KB, 600x400px

Does the O-2 suffer from all these problems?

Saw Bat*21 the ither day, they did some awesome low flying stunts with it.

>>33304406
This desu

File: Fokker D23 Soesterberg.jpg (90KB, 1200x758px) Image search: [Google]

90KB, 1200x758px

Push-pull designs tend to add a lot of complexity that you don't get if you just go with a conventional side-by-side arrangement.

They tried them a lot in the interwar period on multi-engined aircraft, but they ended up losing efficiency in the rear propeller from the mounts. That's not to say that it's an unsurmountable problem, but it was enough at the time for them to drop it.

Later, the concern became things like engine cooling and managing the center of gravity of the aircraft. An engine placed in the aft of the aircraft is going to have trouble finding enough airflow to keep it cooled, complicating the design process. Plus, because your engine is now so far back, stability becomes an issue - you want to have the center of gravity forward of the center of pressure on a conventional plane, so a rear-mounted engine pushes the wing back, which in turn forces you to make the tail bigger, which again shifts center of gravity back. Dornier solved this with a long driveshaft, but Dornier was also known for remotely driven propellers and had been working on such systems since the '20s.

Then you've got practical concerns in operation, with things like tailstrikes to worry about. Usually the rear-facing propeller is placed in such a way that conventional taildragger gear isn't feasible anymore, and the tail has to be designed in such a way that it protects the rear-facing prop - pretty much, you make it so the tail hits the ground before the prop. The rear-facing prop also presents a danger to the pilot in the event of an emergency, as he's now going to have to bail out into the path of the prop. That issue has been dealt with plenty of times, but again it's needless complexity compared to a conventional design.

In the end, it just really isn't worth all the extra complexity. If you absolutely need tandem propellers, the better option is contra-rotating props.

>>33306660
Thanks

Extremely difficult to build a lightweight thrust bearing that can take 1500 hp and the propeller column has to be rigid or vibration destroys it.

File: De Havilland Hornet.jpg (86KB, 1181x887px) Image search: [Google]

86KB, 1181x887px

>>33304361
>It seems like a really good way to make a two-engine interceptor?
Why would you say that?

>>33304486
Don't forget propeller clearance issues on rotation, forcing higher takeoff and landing speeds.

Another thing, two coaxial props won't produce as much static thrust as two adjacent props of the same size. The rear prop is only taking the propwash of the front prop and accelerating it more, which is less efficient than drawing additional stagnant air and accelerating it through an independent actuator disc. The latter is effectively half the disc loading.

>>33304361
that plane is so shit in war thunder

>>33306660
/raises hand

You seem like you know about this exotica so... what are the merits and demerits of a pushrotor design, given the technology of the period. I recall one Air Corps fighter, prewar design, that looked like a P-38 with one big huge engine behind the cockpit. It worked great until they fucked it all up by increasing the requirements and making it too heavy, IIRC.

File: XP-55 Ascender.jpg (65KB, 750x351px) Image search: [Google]

65KB, 750x351px

>>33308315
Short version is that there's just nothing that it really offers over a conventional fighter that makes it worth the effort. Push prop fighters end up with a center of gravity very far back, really messing with stability. Cooling's a huge issue, and you're forced to come up with some convoluted method to allow pilots to escape - either an emergency propeller stop or a way to blow off the propeller so the pilot can bail out.

Finding space for the tail is also a problem - most pusher prop fighters ended up using canards, which have their own stability issues, and the only other realistic option is a twin-boom layout (like the Saab 21).

And for all of that, the only real advantage you get is that you've now got free space in the nose for guns. The Army ran a competition that called for push-prop fighters in WW2, producing three designs - the XP-54, XP-55, and XP-56. Generally what they found was a bunch of stability issues, underwhelming performance, and difficulties on takeoff. The issues were solvable, but with jet fighters coming up within a couple years, it ended up not really being worth it.

>>33308315
>merits
No bulky engine and cowling up front, making for excellent helicopter-esque forward visibility. No forward propeller obstructing centerline-mounted guns. Potentially less parasite drag due to laminar flow over the forward fuselage (propellers tend to turbulate everything downstream of them).
>demerits
Propeller is working in the accelerated and turbulated flowfield around the fuselage, costing a significant amount of efficiency. CG is unfavorable like Anon already mentioned, and inertia moments can be similarly unfavorable. Low-speed control authority does not benefit from propwash over tail surfaces. Propeller is more susceptible to FOD. Aircraft egress/bailout is considerably more risky. Propeller diameter is often severely constrained by ground clearance at takeoff/landing AOA (and/or takeoff attitude compromised by propeller clearance). Cooling can be challenging.

File: p-38-lightning-2.jpg (1MB, 2805x1959px) Image search: [Google]

1MB, 2805x1959px

>>33308814
>>33308963
Thanks. As long as we're on the topic: merits and demerits of one engine versus two? The krauts seemed to favor twin-engined birds a lot, and most like the me-110 couldn't fly on one engine.

>>33308993
>me-110
REEEEEEEE
The Bf 110 was capable of limping home on a single engine. Practically all multi engine aircraft are designed to be flyable with one engine out.

>>33309044
Hm. Okay. Apparently I was misinformed.

>>33308993
>merits
Same visibility and obstruction benefits as a pusher. Left-turning tendencies (torque, P-factor, etc.) can potentially be cancelled out if props counterrotate (assuming both engines are RUNNING, of course). Possibility of OEI operation in an emergency (though it's still not a sure thing). Ground clearance is rarely an issue. Two engines is a straightforward way of doubling your horsepower and thrust. Two propellers tend to offer more disc area than one for a given ground clearance or whatever, making for better acceleration and climb efficiency.
>demerits
Twice as many engines to operate and maintain, twice the failures. Asymmetric thrust causes major control difficulties in the event of a single-engine failure. Higher roll inertia impairs handling. Control authority may or may not be degraded due to absence of prop blast, depending on control surface positioning.

File: Example Constraint Diagram.jpg (52KB, 1093x669px) Image search: [Google]

52KB, 1093x669px

>>33308993
From a preliminary design standpoint, you decide on the layout of an aircraft based on requirements for things like
>Climb rate
>Takeoff/landing requirements
>Top and Cruise Speed
>Maneuvering
>Engine out for multi-engined aircraft
Which defines a design space in terms of wing loading (weight/wing area) and thrust to weight ratio/power to weight ratio.

Then you plug that into initial sizing analysis, which takes into account
>Efficiency of engines available
>Payload
>Range
>Mission profile
To give you a projected weight. With wing loading and thrust/power to weight ratios, you get the engine requirements. If the plane's big enough and thrust/weight ratio requirement is high enough, you select a multiple engine layout.

Ideally, though, you want as few engines as possible, even if you could theoretically get the same thrust from multiple powerplants.

>>33309743
What might an ideal WW2 fighter look like with the benefit of hindsight, assuming a design process that began in the mid-late 1930s?

What about a hindsight bomber?

File: hawker-tempest-fighter-02.png (456KB, 1000x700px) Image search: [Google]

456KB, 1000x700px

>>33309949
Depends entirely on our requirements. An escort fighter is going to be very different from a short-ranged interceptor, and a fast medium bomber is going to be very different from a nighttime strategic bomber.

More generally, the late war fighters are more like what you'd end up designing with the benefit of hindsight - the only real change being that you'd be working with the ~1,000hp class engines available at the end of the '30s rather than the 2,000+hp engines they had by the end of the war.

I've got a corsair H40 and I put a second 120mm fan on the radiator
Knocked my idle temp down by 15 degrees

>>33304361
iv seen the last one of these in existence in person...one of the fastest prop planes ever made.

Why didn't they just put air scoops on the rear engine?

File: kyushu-j7w1-shinden-navy-type-18-shi-01.png (205KB, 1002x502px) Image search: [Google]

205KB, 1002x502px

>>33314149
They do, but air scoops cause drag. The issue tends to be that you're reliant more on a specialized cooling system as opposed to regular airflow over the engine (for liquid cooled engines) or that you need larger (and thus higher drag) scoops to provide enough cooling flow (for air cooled engines).

It's doable, but it's generally needless complication of the design process for negligible benefit.

>>33308814
>Push prop fighters end up with a center of gravity very far back, really messing with stability.

You design the aircraft so the cg is in the proper location.

>most pusher prop fighters ended up using canards, which have their own stability issues

no more than any other design

>And for all of that, the only real advantage you get is that you've now got free space in the nose for guns.

as well as a reduction in drag since you are re-energizing the boundary layer and no issues of propwash over the wing/fuselage

for example this
>>33306166

flew better on the rear engine than on the front in the case of single engine operation

>>33314195
>The issue tends to be that you're reliant more on a specialized cooling system as opposed to regular airflow over the engine (for liquid cooled engines)

liquid cooled engines use radiators in ducts, its really no different then pusher engines

>that you need larger (and thus higher drag) scoops to provide enough cooling flow (for air cooled engines).

not necessarily true

>>33304361

Okay so you know how planes get lift? with their wings?

So you put a push-pull plane together, same amount of lift as a regular single engined monoplane, but with the weight of two engines.

Put a more regular twin engine design together, you put a monohull in the center, wing spar, engine, then more wing than on a single engined monoplane.

and then you have to remember that a plane becomes more maneuverable as the center of lift is nearer to the center of mass, but also slight behind the center of lift.

So put a large weight at the front and back of a plane, you need to put the wing exactly center, which in turn makes the placement of wheels trickier than if the wings were further back (like in jets) or further forward (like in most prop planes).

And remember that because of the added weight of the second engine you need more rugged wheels too.

And the engines use more fuel despite, again, not increasing the space for fuel so you'd end up with very short range interceptors.

>>33314392
>So you put a push-pull plane together, same amount of lift as a regular single engined monoplane, but with the weight of two engines.
Oh yes... I forgot how all planes go the same speed once in the air...

>>33314392
>same amount of lift as a regular single engined monoplane, but with the weight of two engines.

this is dependent on wing area and airfoil section, not number of engines

>and then you have to remember that a plane becomes more maneuverable as the center of lift is nearer to the center of mass, but also slight behind the center of lift.

this is what causes negative or neutral stability, the opposite of what you want in an aircraft with no fly-by-wire and designed for high speeds

>So put a large weight at the front and back of a plane, you need to put the wing exactly center, which in turn makes the placement of wheels trickier than if the wings were further back (like in jets) or further forward (like in most prop planes).

not really.

>not increasing the space for fuel so you'd end up with very short range interceptors.

fuel fraction isnt as simple as available volume


Structural issues related to vibration and weight of engine mountings are the real killers

>>33314289
>You design the aircraft so the cg is in the proper location.
That's really oversimplifying it. Engine location is probably the biggest single contributor to Cg after structural weight on most aircraft. The further aft your engines, the further back your Cg is, shortening the moment arm for your tail and forcing you to use a larger tail for the same amount of control.

Even aircraft with mid-mounted engines like the P-39 were noted for their stability issues and peculiar handling.

>no more than any other design
Canards tend to produce stability issues with aircraft. They may not make a system completely unstable, but the general trend is to reduce the stability of a system (whereas a conventional tail increases longitudinal stability). The most common solution for this - having the canard stall before the wing - tends to reduce efficiency of the system (higher induced drag in the canard), and for peak efficiency in a canard system, the ideal layout would be unstable.

Good point on the boundary layer though - that's something I hadn't considered.

>>33314554
>The further aft your engines, the further back your Cg is, shortening the moment arm for your tail and forcing you to use a larger tail for the same amount of control.

which translates to an issue of higher wetted area

>Even aircraft with mid-mounted engines like the P-39 were noted for their stability issues and peculiar handling.

The p-39 had an issue with the longitudinal and vertical cg being within an inch or two of each other. This was the primary cause of its poor spin recovery.

>The most common solution for this - having the canard stall before the wing - tends to reduce efficiency of the system (higher induced drag in the canard), and for peak efficiency in a canard system, the ideal layout would be unstable.

Yeah, this was one of the problems with the assender. The canarder was really more like a flying (all moving) trim surface. One of the big issues with canards though is that it may seem better on paper, but the wake from the canard is spoiling airflow over the main wing.

Another problem with the ascender is they were counting on the side area from the prop doing a lot of work to correct yaw. This worked fine when the engine was running but in engine out scenarios they had yaw instability.

>>33314418
>I forgot how all planes go the same speed once in the air...

All planes go through the same velocity envelopes during landings and take offs.

Do you see the problem that neccesitates longer wing spans for heavier planes now?

>>33314554
>(higher induced drag in the canard)
Still not as bad as conventional layout that horizontal stabilizer produces literally negative lift.

>>33315104

just reflex the trailing edge of the wing

;^)

>>33315104
Not necessarily. A "stable" canard system (where the canard stalls before the wing) requires the canard to be producing a higher lift coefficient than the wing. Higher lift coefficients lead to higher induced drag. Conventional tails have no such restriction, and thus can work at much lower lift coefficients to keep drag low. Depending on how you design your aircraft, the extra lift the wing has to make up will likely be significantly less than what you're losing from the tail.

>>33304437
Engines weren't all that powerful in the 30s, and since Germany wasn't in desperate-make-lots-of-prototypes mode, and the US wasn't in kill-all-the-fascists mode mode, the need wasn't there.

File: P-38.jpg (39KB, 667x478px) Image search: [Google]

39KB, 667x478px

>>33304437
>the US the P-38 etc

most countries did have manufacturers looking at push pull configurations during that time

but they tended to lose out to more conventional designs

>>33314554
>Canards tend to produce stability issues with aircraft. They may not make a system completely unstable, but the general trend is to reduce the stability of a system (whereas a conventional tail increases longitudinal stability).
Might be true if designers were just slapping canards on existing airframes all willy-nilly with no regard for stability, but that's not how it works. Usually with a clean-sheet canard design like the XP-55, the main wing will be positioned further aft and possibly swept, so that the static margin is in the desired place once main wing, canard and CG are all accounted for. Even with canard retrofits like certain Flanker or Mirage variants, it's usually done to bring the static margin back down following some other modification that increases nose weight (i.e. a larger radar).
>>33315248
>Higher lift coefficients lead to higher induced drag.
Assuming the same aspect ratio - which is a silly assumption to make.

>>33318384
>Assuming the same aspect ratio - which is a silly assumption to make.
Not necessarily. Generally when you're comparing lifting surfaces you want to assume similar planforms. Of course the solution to the aforementioned drag problem on canard aircraft is to make the canards much higher aspect ratio than a similar tail, but then it complicates the analysis because now we have to throw in structures and airfoil selection into the equation.