Please Improve WW2 Ground Physics and Prop + Prop-to-Tail effects

[FusRoPotato]

Requesting some basic functions to make WW2 aircraft behave more like the real thing, and in some cases, to improve controllability aspects that render parts of the sim unattractive.

Tail Wheels
Although a caster wheel doesn't have to have it, a caster wheel on an aircraft typically has what's called a caster angle. This caster angle ensures that when weight is applied to the wheel, the wheel attempts to reach its most recessed upward position. For most, this caster angle leans forward. For some, like the Thunderbolt, it leans back (and thus why it has shimmy and stability problems that a lock helps prevent).

 

Bf-109

Fw-190A

P-47

The Spitfire is a bit misleading because it looks like it might be mounted to a 45 degree caster axis, but it actually is not. It's much less, but still present and leans forward.

The basic idea of this norm in geometric design is that when the tail wheel pivots away from its forward aligned position, the caster angle lifts the tail. Thus in reverse, gravity straightens it out. An approximation of the moment returning it might look something like
M = F * sin(c) * sin(r) * d

Where M is the moment around the y axis , F is the sum of force (weight + aerodynamic down/up force), c = caster angle relative to the ground,  r is the rotation angle from the aligned position, and d is the offset of the tire from the caster axis.

Some aircraft like the Fw190's take it even further and incorporate a return spring, where not only is there a caster angle assisting in ground stability, but also a spring that tries to push it towards center.

I'd like to see caster angles modeled more cleanly in DCS. Without them, ground stability during taxi requires locking the tail. This may be more expected for a P-47, but it seems unreasonable for the rest, especially because of how much more difficult it makes ground handling. 

Prop + Prop-to-Tail effects

I'd like to see P-factor and Gyroscopic procession implemented. It's not yet implemented on any aircraft but is a significant effect. Currently we only have torque roll and prop wash, but even those two effects have a few problems.

P-factor is a moment that is applied when there is AoA or sideslip velocity and power applied. High AoA with power translates to left yaw. Right sideslip (left yaw angle) translates to negative pitch. Negative AoA translates to right yaw. 

Gyroscopic procession is almost the same, except instead of being based on the difference in thrust across a prop blades rotation, it is based on the RPM, rotational inertia, and pitch/yaw angular rates of the aircraft.

When I pitch any aircraft up and down rapidly, or hold high angles of attack/sideslip, I experience neither of these effects.

Torque Roll

Some planes like the Fw190, when stuck in a vertical climb, will begin torque rolling as if the plane doesn't have wings. They will spin up like a helicopter. It's too fast. The wings should be applying some drag and dampening that prevent 4 rolls per second. Other planes, like the TF-51, don't have this (or prop wash) modeled at all. It just gets locked into position. This seems like an unfortunate mistake because it should be nearly identical to the P-51 model and it's also the introductory aircraft that people will judge the whole sim on.

Prop Wash

Prop wash otherwise looks really well done and I'm impressed with how it can be used to control a tail stand. There is just one minor problem with it, and it's that it doesn't deprecate in any way during a tail slide. Some might argue this isn't typically how the aircraft are flown but, that's not entirely true. It's just weird to be falling backwards at 50 kts and still have full prop wash control like a vectored thrust jet.

Edited October 19, 2023 by FusRoPotato

[Art-J]

On 10/13/2023 at 2:54 AM, FusRoPotato said:

I'd like to see P-factor and Gyroscopic procession implemented. It's not yet implemented on any aircraft but is a significant effect. Currently we only have torque roll and prop wash, but even those two effects have a few problems.

P-factor is a moment that is applied when there is AoA or sideslip velocity and power applied. High AoA with power translates to left yaw. Right sideslip (left yaw angle) translates to negative pitch. Negative AoA translates to right yaw. 

Gyroscopic procession is almost the same, except instead of being based on the difference in thrust across a prop blades rotation, it is based on the RPM, rotational inertia, and pitch/yaw angular rates of the aircraft.

When I pitch any aircraft up and down rapidly, or hold high angles of attack/sideslip, I experience neither of these effects.

Torque Roll

Some planes like the Fw190, when stuck in a vertical climb, will begin torque rolling as if the plane doesn't have wings. They will spin up like a helicopter. It's too fast. The wings should be applying some drag and dampening that prevent 4 rolls per second. Other planes, like the TF-51, don't have this (or prop wash) modeled at all. It just gets locked into position. This seems like an unfortunate mistake because it should be nearly identical to the P-51 model and it's also the introductory aircraft that people will judge the whole sim on.

P-factor and gyro have been implemented since forever, though. At least in ED warbirds (I don't own I-16 and CE-II so can't comment on these). They differ a lot from plane to plane and one might argue about how strong they "should" be (though posting just "feelings" alone won't get anyone anywhere on these forums), but they're there allright.

Just did a refreshing rides in both P- and TF- Mustangs. As far as P-factor is concerned, apart from obvious pull to the left on takeoffs, high AoA slow flight results in left yaw, throttling up makes it even deeper. Gyro precession has always been noticeable in all warbirds especially while raising the tail during takeoff roll quickly (mostly in 109, which requires full right rudder kick to counteract the effect). Haven't noticed any changes today either. Pitch changes in flight - up yawed left, down yawed right. 

As for the vertical climb with constant power, well, both Mustangs do start rolling left when the speed drops really low, even more so when throttling up. Not much difference compared to 190s apart from slower roll rate at which it happens.

All and all, I just can't agree with all of your observations. Are you sure you don't have game flight mode and rudder & takeoff auto-helpers reset to ON after one of the last game updates? 

 

[FusRoPotato]

Don't perform your assumptions during takeoff. At slow speeds and slow rotation rates, your prop wash will produce more dominating moments and will mask the other effects. P-factor and Gyroscopic procession effects are better isolated in flight later on when speeds are much higher. 

Gyroscopic procession isolation requires high speed, high rpm, low power, and high angular rate movements in pitch or yaw. Pulling back a bit on throttle and going into a dive is a good method.

P-factor isolation requires moderate speed, high power application (large difference in air speed vs prop true speed), and large alpha/beta. Reaching a moderately high speed and then pulling into a slight climb at full throttle is an ideal start to test. Manually choosing a lower RPM or higher prop pitch should usually produce a more pronounced effect. P-factor is about the difference in prop lift from one side to the other, which takes a parameter of free stream velocity projected in a direction perpendicular to the prop axis, meaning it's not just about the AoA. P-factor is very commonly misidentified as the yaw induced when punching throttle to start takeoff, but that is prop wash hitting the tail. 

Edited October 19, 2023 by FusRoPotato

[RedBjorn]

One of the issues that the Mustang always seemed to have throughout development was torque rolling in a landing configuration (flaps down, gear down, roughly 25 inches MP, high 3,000 RPM). It was infact so bad, along with yaw issues, that the designers added a dorsal fin as early as the B/C models as the aircraft (designed to combat this somewhat, and reduce adverse yaw), during a quick application of power, could roll over onto the cockpit and kill the pilot. There’s a crash not long ago that did exactly that, killing said pilot and damaging another Mustang, even with the aforementioned installation of the fin. Try it sometime. Go up to altitude, setup yourself in a landing configuration and speed, and throttle up as if you were going around. It waffles, and the torque movement should be more significant, and require far more pilot awareness, and rudder and aileron input. This is a very dangerous condition, and many pilots have been trained to apply the power far more smoothly and anticipate this movement in the case of a go-around. This is exacerbated with fuel in the aft tank or with lots of fuel in the wings, and a partial reason why most Mustangs today have the aft tank removed.

Now, I’ve worked with a LOT of companies such as A2A Simulations for years in the past, have hours of archived cockpit footage in these birds, and have extensive experience with the Mustang, writing technical manuals, and even working with multiple pilots such as Dudley Henriques and Vlado Lenoch (RIP). Frankly, I’ve come two two conclusions. First, things such as p-factor are severely underwhelming, not particularly dynamic, and it’s most apparent at low speeds with the onset of high power at low speeds, and second, ground handling is not particularly good - so bad in fact that I nearly didn’t play DCS initially due to poor ground physics. 

I would like to see these topics revisited. Further, having conversed about the 190, and studying the rear wheel schematics, I fully support what Potato is saying here. That wheel, as far as I can tell, will right itself with little to no input from the pilot after a small rollout.

Edit: The control locking of the 190 during a torque roll also seems incorrect to my intuition. During this maneuver, I’d expect “some” control with aileron, but as is, it’s like you’re completely locked out and the aircraft spins like a top. If true, this is the kind of thing you’d see in pilot reports of the aircraft. I’m not aware of any such reports.

Edit 2: To clarify, in my initial observation I’m not strictly discussing p-factor. I am just using the Mustang as an example that the effects of propellor driven aircraft with this kind of power don’t seem particularly dynamic and should be revisited.

Edited October 19, 2023 by RedBjorn

[draconus]

Perhaps it'd be better to report bugs in specific modules' subforums, esp. if you're SME, like @RedBjorn, please make the detailed Mustang reports (with tracks and related docs) here:

https://forum.dcs.world/forum/203-bugs-and-problems/