New behaviors at the edge of the speed envelope in 2.8?

[DJBiscuit1818]

Running recon work in the Mirage F1 on the ECW server, I've been working on finding the ideal profile to quickly recon as much of the frontline as possible.

Since the max alt is 4km AGL (13k ft), I generally aim for 12k ft over the lowest point on my planned run to maximize coverage while also avoiding dead spots. This is usually ~13-15k ft on the altimeter.

At this altitude in level flight, the max speed is usually M1.3. However, if I take it up to 30k+ on my cruise in, then make a shallow dive to start my run, I can go much faster. If I get survitesse to kick in (which is oddly enough better served by a shallow dive than a steep one), the limiting factor becomes going too fast, as the aircraft will continuously accelerate to and past its limits. Because I'm trying to stretch my precious seconds of film/recon time (and not get shot down), the faster I can go, the better. Thus, I'll generally push the aircraft as far as I can in that range. Additionally, since both the airbrake and afterburner are either on or off, I have to repeatedly switch either one or the other to modulate speed. (as an aside, is this correct? The A/B section of the throttle is something like 40% of axis travel when stock, and also has a visibly large travel in the cockpit. This doesn't really seem to make sense, but I don't know enough to know if this is correct or not). The result of this is that I often push the aircraft until it complains, cut AB (and thus speed) until just over the survitesse cut-off, then go back into full AB. Rinse and repeat. The aircraft is relatively docile at M1.4 (~774 KCAS / ~725 KEAS), but encounters buffeting from airspeed somewhere around 750-800KIAS, then two things happens as you further increase speed. The aircraft starts to roll and/or pitch of its own accord (mach tuck?), and the "LIM" light turns on and an alarm begins to sound.

Before 2.8, the former issue (mach tuck) used to be handled quite nicely by the autopilot. In both Attitude and Altitude modes, the aircraft would stay solidly on course (albeit with plenty of buffeting over the aforementioned ~750KIAS). In Heading Hold Mode, the aircraft would often oscillate slightly in roll, but it would still usually maintain roughly on course. However, as of 2.8, this is no longer the case. In all modes, as soon as Mach tuck(?) begins, the autopilot fails to compensate, and will disengage after ~60° of bank. This seems to be due to both more severe tucking, and the changes to autopilot, which now inevitably lead to overcompensation and/or heavy oscillation if you demand a moderate to large change in heading or attitude. (As an aside, I do appreciate that trim inputs will now vary commanded attitude/heading in attitude hold mode, it's a massive improvement—when it works)

As such, formerly the LIM light was the real limit I'd have to respect. I'd line up on a sector, dive from 30 to my recon alt, level out, and command autopilot to altitude hold (and often heading hold as well). The tuck wasn't severe enough that it caused issues, so I'd simply accelerate until the plane started chirping at me (usually ~1.6M), deploy airbrake until speed decreased to ~1.45, rinse and repeat. While I repeatedly triggered the LIM light/alarm, I respected the warning, and never really tried to push past it. That being said, the airbrake takes a couple seconds to deploy, so I'd spend several seconds in the alarm zone whenever I triggered it, and sometimes trigger that alarm a dozen times in a flight or more. That's ~24-36 seconds spent in the alarm area in a flight, with no adverse effects.

However, as of recent patches, I encountered a concerning new issue. If the LIM light/alarm kick on, you're seemingly at serious risk of damaging the engine the instant that warning triggers. Due to the worse tuck/broken autopilot, I've been controlling my recon flight manually. This usually means that I'm not pushing into the alarm zone nearly as often (since the limiting factor is usually fighting mach tuck and trying to maintain a good flight path). Additionally, the way I'm slowing down is more often cutting throttle, since using the airbrake causes a strong pitch-down moment, which is something I don't like when I'm already having to fight to maintain course. When I do hit the LIM alarm, this also causes a quicker decrease in speed, as the AB cuts out instantly when throttle is reduced.

However, I've now run into the following issue several times: Upon hitting the LIM and slowing down, going back to full AB won't allow you to maintain survitesse.

Indeed, you can't even maintain M1.3.

Or Mach 1.

In full afterburner, with an almost completely clean jet (just one empty centerline pylon) in level flight, at 10-15k feet, you max out at mach 0.9. *Climbing to ~35k (optimal alt for acceleration), you still won't build any more speed. You barely break mach in a dive. Your non-AB performance also takes a hit, with a level flight max of somewhere in the .75-.8 range at mil power.

It's not like I push until something goes "bang" or I'm ignoring the alarm. The first time I crippled an engine, on the the first time I hit the limit, I was slow in I got the throttle down. So I decided to be more careful. So as soon as I hear that chirp, I'm pulling back on the throttle/hitting the airbrake. The second time it happened, I hit the limit once, and was fine. "Ah, so I just have to be fast, I guess they made it more sensitive" I hit the limit again, and it didn't even chirp twice before I cut the throttle. Then I reapplied throttle, and my speed kept dropping. Down, down, down. I turned for home and ran, because at M 0.9, I was a sitting duck. I barely limped home and only narrowly avoided getting jumped by a Fishbed because a friendly happened to be able to vector in and save my ass.

What is the point of a warning alarm that doesn't warn you? You need time to respond before stuff explodes/breaks. The landing gear doesn't instantly snap or jam at 2 KIAS above the limit of 240. You start to risk failure when you exceed the limit, with probability of failure scaling as you depart further and further from the limit. Additionally, once the engine is damaged, there's no indication that it's happened besides being completely unable to break mach. No master caution, no lights on the panic panel, only neutered performance. Again, if you snap your gear off, the plane isn't going to pretend your gear are actually up. To me, this behavior seems wrong (and isn't mentioned in the patchnotes), so I'm wondering if this was an intentional change or not. I'll admit, my sample size is small, so maybe I just got incredibly unlucky.

Has anyone else encountered this behavior?

Edited November 26, 2022 by DJBiscuit1818

[Rudel_chw]

On 11/25/2022 at 9:59 PM, DJBiscuit1818 said:

Has anyone else encountered this behavior?

 

I'm learning / studying the Autopilot system and your thread came up during my research. I have not encountered the issues that you mention, but I'm amazed that such a detailed and clearly writting issue report, has received absolutely no comment from Aerges nor been refuted by them.

I must admit that it is discouraging to see so little involvement from Aerges on this part of the Forum.  

[fausete]

Hi, 

Most/none of this most likely no longer applies as the FM has been extensively tweaked since then, so reaching those speeds in a continuous manner is more difficult.

The answer to this is basically that the aircraft was being oversped. The LIM warning light is triggered by reaching an impact temperature of more than 135 °C. However, at low altitudes, the IAS limits are exceeded by a fair amount before those temperatures are reached. The roll he's describing is the inversion of control surfaces due to deformation of the wing and the loss of thrust is the damage to the engine.

[Rudel_chw]

1 hour ago, fausete said:

The answer to this is basically that the aircraft was being oversped. The LIM warning light is triggered by reaching an impact temperature of more than 135 °C. However, at low altitudes, the IAS limits are exceeded by a fair amount before those temperatures are reached. The roll he's describing is the inversion of control surfaces due to deformation of the wing and the loss of thrust is the damage to the engine.

 

Hello,

Thanks a lot for the answer, it clears up why I hadn't seen the issue myself.

Best regards,

 

Eduardo

[fausete]

De nada, regards to you too!