Ridiculously hard landing on LAND 3 (-771fpm)

[Andrew Hafey 1]

Coming in runway 15 at Anchorage (yes I know sim calls it 14)...777F with 228k, so running heavy.  Intercepted localizer, was on approach mode, flaps 30, manual speed holding at vref+5.  Yes, I had LAND 3, G/S and Flare engaged..but for whatever freaking reason, I hit hard (-771 fpm)!!!  I was watching for the runway (saw it right at minimums) and didn't notice a flare, but I know for a fact it was armed.   I actually came on to the runway heading at a YOHNN +4nm waypoint at 3500 feet...so I know this isn't an issue relating to too short of an approach.  I don't even use it all that much except when needed (very low vis)...according to my logbook, only had 6 autolands on it, all between -65fpm and -400.  I know many people here love to bash on landing rates and I haven't complained about it except that my VA registers anything over -500 as a hard landing with a very expensive repair bill.

Question is, could this be a scenery issue?  I run Orbx with openLC NA, Vector and Southern Alaska sceneries.  Vector has had autoconfigured ran...and there are no visual anomalies to indicate a scenery issue, but I'm at a complete loss as to why it just caused a $2.4M repair bill because of an autoland?  If not scenery, what else could be the issue?  This is the first 777F flight I've ran since upgrading from P3D v4.1 to v4.2...but I have done several 737 NGX flights since then with 1 or 2 autolands and never and issue.

Any suggestions???

Edited April 28, 2018 by Andrew Hafey

[BlueStar 323]

Pilot error. 

[harrry 322]

Can you use manual speed, or do you mean you have auto throttle with the speed window open, with flare?  Did you apply power or remove it?

Edited April 28, 2018 by harrry

[Dave_YVR 973]

Why all of the Autolands? Just do it yourself like they more often do in the real world.

[G-RFRY 1,711]

In the real world I think the tower has to be informed if a full Autoland is intended, they then have to increase the aircraft spacing for safety and keep ground traffic away from ILS transmitters, and airlines will only allow pilots to use it when condition dictate, most will be vectored to the alternate on the flight plan if possible.

I don`t think Flare will have very little affect on a hard landing, airspeed and GW and Trim will.  

Edited April 28, 2018 by rjfry

[iwebber 207]

10 hours ago, Andrew Hafey said:

Any suggestions???

Adjust your expectations.

You're using a product (PMDG 777) that has been painstakingly recreated to perform just like the real thing. You're then measuring it with a product designed for a $50 flight sim (this is not a criticism of either).

Autolands are not usually smooth, that's a simism, autolands are designed to put the aircraft on the end of the runway in a wide variety of weather conditions in such a way that you can use it again. In my experience, they're pretty agricultural but they nearly always meet that brief.

Landing 'smoothness' is measured in G, not fpm. While I appreciate the two are related, they use G in real life and there is a reason for that.

2 hours ago, rjfry said:

In the real world I think the tower has to be informed if a full Autoland is intended

No, they don't, but if you want to use minima less than CAT 1 then you need LVP (Low Visibility Procedures (or LVO sometimes depending on the week - LV Operations)) in force. Some places require the tower to be notified if you're doing an approach to a minima lower than CAT 1.

If LVP(O)'s aren't in operation you should be ready to go around (not visual) or disconnect the AP and take over (visual) if you're not happy with the approach. Actually, you should always be ready for that, it's just much more likely if you're doing an autoland without protections. I've done a few CAT 1 (without protections) autolands without incident (back when we used to have to do a certain number every month).

I'm afraid your $2.4 million repair bill is an unfortunate side effect of the unrealistic way your VA is measuring your landing performance. My recommendation, don't autoland unless it's absolutely necessary... then hope, which, to be fair, is exactly what we do in real life.

HTH

Ian Webber

[skelsey 765]

Steady on guys... Andrew has already said that he doesn't use autoland all the time, and implied that the visibility was such that one was warranted.

Regardless, it sounds as though for whatever reason the aircraft failed to flare properly, an issue which has been reported, sporadically, by other users in the past. I don't know the technical details of how the PMDG 777 autopilot is coded but from what I know about P3D in general I would suggest that it is possible this could be scenery related in some way.

Technical issues aside - it does highlight the need to be poised and ready to take action if the automation doesn't function as expected! Malfunctions can and do occur and it is vital to be in the mindset to take prompt action if e.g. no flare occurs (even if everything annunciates correctly). Not easy and there is very little time to identify the problem and intervene!

[Andrew Hafey 1]

7 hours ago, iwebber said:

Adjust your expectations.

You're using a product (PMDG 777) that has been painstakingly recreated to perform just like the real thing. You're then measuring it with a product designed for a $50 flight sim (this is not a criticism of either).

Autolands are not usually smooth, that's a simism, autolands are designed to put the aircraft on the end of the runway in a wide variety of weather conditions in such a way that you can use it again. In my experience, they're pretty agricultural but they nearly always meet that brief.

Landing 'smoothness' is measured in G, not fpm. While I appreciate the two are related, they use G in real life and there is a reason for that.

No, they don't, but if you want to use minima less than CAT 1 then you need LVP (Low Visibility Procedures (or LVO sometimes depending on the week - LV Operations)) in force. Some places require the tower to be notified if you're doing an approach to a minima lower than CAT 1.

If LVP(O)'s aren't in operation you should be ready to go around (not visual) or disconnect the AP and take over (visual) if you're not happy with the approach. Actually, you should always be ready for that, it's just much more likely if you're doing an autoland without protections. I've done a few CAT 1 (without protections) autolands without incident (back when we used to have to do a certain number every month).

I'm afraid your $2.4 million repair bill is an unfortunate side effect of the unrealistic way your VA is measuring your landing performance. My recommendation, don't autoland unless it's absolutely necessary... then hope, which, to be fair, is exactly what we do in real life.

HTH

Ian Webber

in real life, you would be using autoland under those conditions (no vis, lots of rain, pretty good crosswind).  I had no visibility on the runway until about 60ft AGL...I was actually half a heartbeat from aborting the landing.  If that's not conditions for a CAT III autoland...by all means, please tell me what would be???

As far as the repair bill, 2.4M is unrealistic most likely, but a hard landing at almost 800fpm will almost certainly cause damage to the landing gear IRL.  

Edited April 28, 2018 by Andrew Hafey

[Ephedrin 418]

A few months ago I landed the 737 manually in Sitka, Alaska.. Or better, I attempted to do so. The visibility was fine, very low crosswind.... I flared... and slammed into the ground with 1200fpm (VA measurement...). I know it would have been a good landing but this must have been a scenery thing.. and I think it's the same on your side. The 777 commonly lands with what people call a greaser. All above with low to no winds which I suppose you have had when the visibility was that low.

[downscc 1,327]

PANC ILS15 is a CAT I approach, C&D mins are 250 ft AGL, I wouldn't be using autoland there.  Were you landing at the ORBX FTX SCA PANC? I'm not aware of an elevation problem with the scenery, I'm using the new Aerosoft PANC without issues.

An 800 fpm landing is hard but within the rating of those amazing landing gear assemblies.

[BrianW 124]

Is there an approach for runway 15 that lets you go below 401’ in a 777?  If not you should be more concerned about busting minimums than the autoland.  While it certainly is possible to autoland using a CAT I ILS there are no guarantees it will be successful. Here’s what one training guide has to say about it:

 

“Certain features of the Category II ILS installations ensure optimum

performance for the autoland system. They include localizer and glideslope

quality, signal interference protection from ground vehicles and aircraft,

glideslope angle, glide path intercept point (GPI) with the runway, and final

approach area terrain.”

 

“Autoland performance on non-Category II ILS installations may tend to be

somewhat erratic. The pilot is the monitor. He/She should closely monitor the

approach and flare progress and be prepared to immediately revert to manual

control of the aircraft in any questionable situation.”

 

While I don’t think FSX or P3D do anything with the ILS signals, terrain and glideslope antenna position may play a role in the sim.  Radar altitude rate of change plays a big role in the 777 flare operation, so if there are rapid changes this could influence the flare.

 

1 hour ago, Andrew Hafey said:

in real life, you would be using autoland under those conditions (no vis, lots of rain, pretty good crosswind).  I had no visibility on the runway until about 60ft AGL...I was actually half a heartbeat from aborting the landing.  If that's not conditions for a CAT III autoland...by all means, please tell me what would be???

 

Actually autoland has restrictions that manual landings don’t, for example if you have more than 25 knots of headwind you can’t autoland, but you can do a manual landing.  So if it was very windy this could also be a factor in your landing rate.

[Ephedrin 418]

If it had had more than 25 knots of wind I doubt the visibility would have been that low in Anchorage   (edit: except from a blizzard -.- to make everyone happy in advance...) That has so obviously been a technical issue, either with the scenery or with the airplane.. while I find the scenery more likely.

Edited April 28, 2018 by Ephedrin

[30K 72]

If I recall correctly then there is a "cliff" shortly before the runway. So we have a fast change in RA and that could be the thing.

At SFO 28R  I usually get harder landings than at KDEN or somwhere w/o a fast drop in RA.

No guaranties as my Last autoland is 2 month ago and the one before pff idk.

Edited April 28, 2018 by 30K

[scandinavian13 5,192]

Also of note, that I don't think anyone has mentioned:

Weather add-ons. I believe Active Sky still models wake turbulence, and I'm not sure if they've ever fixed the issue where planes sitting on the side of the runway will still generate a wake. If you have AI traffic on, and are using AS, it's possible that the aircraft generated a wake that caused the fall.

As Dan noted, though, if it's a CAT I only, then there's usually a reason for that.

[BlueStar 323]

On ‎4‎/‎28‎/‎2018 at 1:20 PM, Andrew Hafey said:

in real life, you would be using autoland under those conditions

In real life you would have most likely lost your job and your license for busting minimums and bending an aircraft. 🙂  At  MGLW and close to 800 fpm rate of descent there will probably be some wrinkled sheet metal.  

That's the great thing about Flight Sim, just reset, have another beer and go on down the road. 🙂

 

 

Edited April 30, 2018 by Bluestar