Close call. You can't always trust the ILS.

stratobee

stratobee

Cleared for Takeoff
Joined
Dec 18, 2011
Messages
1,112
Display Name
Display name:
stratobee
This is the perfect storm of an accident where almost all holes lined up. Only the crew's last minute gut feeling saved this from being a disaster.

Identifying the navaid was not the problem, nor would a IAF altitude cross check have shown anything out of order either. Only way to catch this would have been to cross check altitudes along the GS all the way down. How many of us do that regularly? Once we're established, we fly the needles. And in fact, that's all the FAA requires us to do.

https://www.youtube.com/watch?v=Wy6BMB3lFEs#t=1146
 
stratobee said:
This is the perfect storm of an accident where almost all holes lines up. Only the crew's last minute gut feeling saved this from being a disaster.

Identifying the navaid was not the problem, nor would a IAF altitude cross check have shown anything out of order either. Only way to catch this would have been to cross check altitudes along the GS all the way down. How many of us do that regularly? Once we're established, we fly the needles.

https://www.youtube.com/watch?v=Wy6BMB3lFEs#t=1146
Click to expand...

Actually for me it's company SOP.
 
But single pilot, normal people, we don't do that. In fact, FAA doesn't even require it because how do you determine established? GS at FAF.

So unless the gut would have told the normal IR pilot otherwise, or there was some visual clue, they would have rode it into the ground.
 
stratobee said:
But single pilot, normal people, we don't do that. In fact, FAA doesn't even require it because how do you determine established? GS at FAF.

So unless the gut would have told the normal IR pilot otherwise, or there was some visual clue, they would have rode it into the ground.
Click to expand...

So you need the FAA to require you on how to enhance your safety?
 
RotorAndWing said:
So you need the FAA to require you on how to enhance your safety?
Click to expand...

All I'm saying maybe the training should allow or emphasize this cross check and verification, don't you think? Today it's not.
 
Last edited:
stratobee said:
But single pilot, normal people, we don't do that. In fact, FAA doesn't even require it because how do you determine established? GS at FAF.



So unless the gut would have told the normal IR pilot otherwise, or there was some visual clue, they would have rode it into the ground.
Click to expand...


Been doing and teaching it for years.
 
stratobee said:
But single pilot, normal people, we don't do that. In fact, FAA doesn't even require it because how do you determine established? GS at FAF
Click to expand...

speak for yourself -

I always verify GS intercept altitude and then what the minimums are . . . unless you want to fly into the ground somewhere . . .
 
Something similar happened to me a few years ago while shooting an ILS to minimums, single pilot without an AP:eek:

http://forums.aopa.org/showthread.php?t=84232

I absolutely credit this video (which I first saw in 2007) with allowing me to quickly recognize my own situation, and take immediate corrective action by automatically going missed, and regrouping.
 
I once had a glideslope receiver fail with no flag. The first clue was that the needle was perfectly centered. As I told the controller, "I'm not that good."
 
Palmpilot said:
I once had a glideslope receiver fail with no flag. The first clue was that the needle was perfectly centered. As I told the controller, "I'm not that good."
Click to expand...

Same happened to me years ago. Recognized that "perfection ain't me" and reverted to Localizer approach (broke out right at Loc mins). Turned out to be a loose connector that caused centered GS needle no flag.
Ever since then, I consider the small deviations and corrections within the inner circle a crucial part of the approach (and always hand fly the final approach).
 
True story: In the 1970s I was the copilot (FO for the status conscious) on a Learjet deadheading into Buffalo one morning in dense fog. I was handling the radios because the captain thought he stood the better chance of not missing the approach due to the weather :)rolleyes:). It probably saved our lives.

I thought it odd that approach was saying we were two miles from the marker and cleared us for the approach because I could hear the OM beeping as he spoke. I think he was covering verbally for a late turn-on. But it caused me, thankfully, to pay closer attention to resolve the ambiguity.

The next odd thing that happened was the middle marker went off while we were still well above minimums. "What's causing these markers to go off early?" I wondered. Even though I didn't like this captain much, I had to commend his ability to keep the needles perfectly centered.

When we reached minimums, the first thing I saw wasn't the threshold lights, but the aiming point 1000' from the end. So, I started counting the touchdown marks straddling the aiming block. There were more stripes on the far side than the near side. We were about to touch down on the last 1000' of the runway at around 120 kts! "Go around, Bud!", I shouted. He pulled the power off instead. "Go around!", I said again, and placed my left hand behind the throttles ready to push them up if he didn't. This time he did.

As we were vectored back for another try, Bud told me how the Learjet can stop in 1200' with thrust reversers and how he could have made it anyway. No real need to go around. The upshot was, after a second approach was successful with no stuck glide slope needle, only one TR deployed. Had we landed, I'm sure he wouldn't have waited for my callout that both sides were deployed before roaring up on the piggybacks. And if he hadn't, we'd have swapped ends, balled up and burned right there off the departure end of runway 23 as surely as another Learjet crashed and burned in Richmond, Indiana.

From that day forward, I always checked the crossing altitude on an ILS and always set both navs to the ILS for cross-checking. The weather was so bad that morning we had the VOR on the copilot's side set up for the missed approach. Never again.

Japan Airlines lost a DC-8 in San Francisco Bay that could have been prevented by checking GS crossing altitudes, too. We live and learn.

dtuuri
 
Last edited:
dtuuri said:
...

Japan Airlines lost a DC-8 in San Francisco Bay that could have been prevented by checking GS crossing altitudes, too. We live and learn.

dtuuri
Click to expand...

Great story, and a lesson well-learned (and taught, too!). My instructor always told me to cross-check altitudes for "logical sense."

---

By the way, the JAL DC-8 that landed in the Bay was actually retrieved, repaired and flew on to a long and productive life. Douglas built 'em tough! :yes:
 
stratobee said:
All I'm saying maybe the training should allow or emphasize this cross check and verification, don't you think? Today it's not.
Click to expand...

:confused: Find better instructors, mine did not fail to convey that importance to me.
 
Radio altimeter helps. Of course SVT is probably the best way to prevent these types of accidents. Also lets not forget, if you're going into a controlled field and the controller is on the ball, they'll give a safety alert if you're significantly below glidepath. Many safety buffers to prevent something like this from happening.

If you have a bare bones panel going into an uncontrolled field, yeah, you better be verifying altitudes on the way down.
 
Velocity173 said:
Also lets not forget, if you're going into a controlled field and the controller is on the ball, they'll give a safety alert if you're significantly below glidepath. Many safety buffers to prevent something like this from happening.
Click to expand...

Works for many/most(?) controlled fields. If it's a non-radar tower then the controller doesn't have any way of helping out. Is there any way of telling that the tower doesn't have radar?
 
Clark1961 said:
Works for many/most(?) controlled fields. If it's a non-radar tower then the controller doesn't have any way of helping out. Is there any way of telling that the tower doesn't have radar?
Click to expand...

Yeah if you're doing an ILS at an uncontrolled airfield and instructed to go to advisory or a controlled airfield where radar coverage doesn't cover within 1/2 mile of the runway end, and they tell you "radar services terminated." In that case, you're own your own. You could still get a safety alert if the controller has time to catch it and relay to tower or over advisory.

Whether the tower itself has a radar feed will be facility dependent. You might have a class D with an uncertified radar display or a class C or B with a certified system that Local can use for limited radar services or other services provided thru an LOA. So yeah, not all controlled airfields will have the capability, but the service does exist.
 
Last edited:
Hopefully with GPS color moving map showing the runway ahead and terrain below, these war stories will become relegated to ancient history.
But as always, you have to keep using all tools at your disposal, trusting none.
 
RotorDude said:
Hopefully with GPS color moving map showing the runway ahead and terrain below, these war stories will become relegated to ancient history.
But as always, you have to keep using all tools at your disposal, trusting none.
Click to expand...

SVT=eVFR, plain and simple. The biggest revolution in situational awareness since Sperry made the gyro pack.
 
SCCutler said:
By the way, the JAL DC-8 that landed in the Bay was actually retrieved, repaired and flew on to a long and productive life. Douglas built 'em tough! :yes:
Click to expand...
I flew that airplane in the late 1990s. It was eventually scrapped and cut up around 2000.
 
RotorAndWing said:
Actually for me it's company SOP.
Click to expand...
Do you guys just have to check the glide slope crossing altitude, or really the whole way down? If so, are you guys just using the RNAV distance to the End of Runway point?
 
Funny, just over a week ago we were discussing whether or not an OM was needed to fly an ILS (http://www.pilotsofamerica.com/forum/showthread.php?t=79875). I incorrectly thought that you needed some sort of range information to make it legal. I was wrong, but after reading this thread and watching the video, I might change my mind back (partway) and say that having range information on an ILS is maybe not required, but probably a good idea. Bottom line is that most of us fly planes now that always have range information in the form of a GPS position. But in the rare case of an ILS with no OM, DME, crossing radial, RADAR or GPS, I'd think twice about flying it.

That's my 2¢... It's worth about what you paid for it... ;-)
 
Sluggo63 said:
Do you guys just have to check the glide slope crossing altitude, or really the whole way down? If so, are you guys just using the RNAV distance to the End of Runway point?
Click to expand...

I use the cross reference points provided on the approach plate.
 
Flew an ILS yesterday for quarterly training. While I have SVT I still did a 3 mile check on the radar altimeter to make sure altitudes matched up. Use everything at your disposal.
 
Sluggo63 said:
Funny, just over a week ago we were discussing whether or not an OM was needed to fly an ILS (http://www.pilotsofamerica.com/forum/showthread.php?t=79875). I incorrectly thought that you needed some sort of range information to make it legal. I was wrong, but after reading this thread and watching the video, I might change my mind back (partway) and say that having range information on an ILS is maybe not required, but probably a good idea. Bottom line is that most of us fly planes now that always have range information in the form of a GPS position. But in the rare case of an ILS with no OM, DME, crossing radial, RADAR or GPS, I'd think twice about flying it.

That's my 2¢... It's worth about what you paid for it... ;-)
Click to expand...


It is definitely a very good idea to have. In the days before GPS, DME was what I looked for in an IFR panel.

Now with glass avionics, you can get the information graphically so it's much easier to maintain up to the second situational awareness, especially with SVT and the predictive indicator. If you are in a bind for info, you can turn on HITS, but I find it kinda annoying and overkill for normal situation use. In an emergency in IMC it's gold.
 
Last edited:
Seems to me that this is why we intercept the loc at an angle and typically well before the GS capture point - so that we intercept GS from below and can verify normal GS capture behavior. This is certainly something everyone is taught to do.

In this incident the GS was immediately captured and centered if I understood correctly. That is not normal behavior. Intercepting from below, the GS indicator should be at full up deflection and come in gradually. And from my understanding, that instantaneous GS capture is what first made the crew do a WTF check.

I think GA pilots are trained to do this much at least. I was further taught to verify the lightning bolt altitude at the FAF, although that isn't very helpful when the GS intercept and FAF are very close to each other.


But stratobee in the OP alluded to checking altitudes "along the GS all the way down". So for those of you who say that you do this as SOP... how do you accomplish it when those altitudes are not published?

The only altitude cross checks printed on the IAP, typically, are the FAF on-glideslope intercept altitude and the DA distance fix. The intermediate distances don't show a corresponding on-glideslope altitude on the charts I'm looking at.

Given the GS angle, it is certainly possible to calculate the altitude cross checks for any distance from the threshold. I wasn't taught to do this and it isn't a tool in my E6B apps, but after watching this it seems worth creating a simple calculation for it.

There is also the table of vertical speeds that should correspond to various GS angles at various ground speeds, but those wouldn't be perfect as vert speeds and ground speeds vary. Altitude xcheck would be exact.

I just want to learn what exact procedure people are following when they say they do "this."
 
Last edited:
petrolero said:
Given the GS angle, it is certainly possible to calculate the altitude cross checks for any distance from the threshold. I wasn't taught to do this and it isn't a tool in my E6B apps, but after watching this it seems worth creating a simple calculation for it.
Click to expand...

Use similar triangles, e.g. If you're half way to the runway then you should be half way to threshold elevation. Easy-Peasy
 
petrolero said:
Seems to me that this is why we intercept the loc at an angle and typically well before the GS capture point - so that we intercept GS from below and can verify normal GS capture behavior. This is certainly something everyone is taught to do.

In this incident the GS was immediately captured and centered if I understood correctly. That is not normal behavior. Intercepting from below, the GS indicator should be at full up deflection and come in gradually. And from my understanding, that instantaneous GS capture is what first made the crew do a WTF check.

I think GA pilots are trained to do this much at least. I was further taught to verify the lightning bolt altitude at the FAF, although that isn't very helpful when the GS intercept and FAF are very close to each other.


But stratobee in the OP alluded to checking altitudes "along the GS all the way down". So for those of you who say that you do this as SOP... how do you accomplish it when those altitudes are not published?

The only altitude cross checks printed on the IAP, typically, are the FAF on-glideslope intercept altitude and the DA distance fix. The intermediate distances don't show a corresponding on-glideslope altitude on the charts I'm looking at.

Given the GS angle, it is certainly possible to calculate the altitude cross checks for any distance from the threshold. I wasn't taught to do this and it isn't a tool in my E6B apps, but after watching this it seems worth creating a simple calculation for it.

There is also the table of vertical speeds that should correspond to various GS angles at various ground speeds, but those wouldn't be perfect as vert speeds and ground speeds vary. Altitude xcheck would be exact.

I just want to learn what exact procedure people are following when they say they do "this."
Click to expand...
I don't know how others do it, but on Jepp plates, they'll put a recommended altitudes box on the plate in case you have to do a LOC-only approach.
Rec%20Altitudes.jpg
 
Sluggo63 said:
I don't know how others do it, but on Jepp plates, they'll put a recommended altitudes box on the plate in case you have to do a LOC-only approach.
Rec%20Altitudes.jpg
Click to expand...

That seems to be the key difference. AeroNav charts don't have that listed.
 
petrolero said:
That seems to be the key difference. AeroNav charts don't have that listed.
Click to expand...

Every plate I can recall always designates two definable locations along the glide slope before you get close in. If I'm on altitude and needle at those 2 points, my confidence is high enough.
 
There is something I'm not understanding here.

For an ILS or LOC approach, don't the LOC minimums account for obstruction clearance past the FAF? That means you shouldn't have a "close call" unless you bust the minimums.

Or is it just the 200 or so feet between the ILS and LOC minimums, when the weather is below the LOC minimum? I'd think the "perfect" GS reading would mean use the higher minimums under that circumstance.
 
MAKG1 said:
There is something I'm not understanding here.

For an ILS or LOC approach, don't the LOC minimums account for obstruction clearance past the FAF? That means you shouldn't have a "close call" unless you bust the minimums.

Or is it just the 200 or so feet between the ILS and LOC minimums, when the weather is below the LOC minimum? I'd think the "perfect" GS reading would mean use the higher minimums under that circumstance.
Click to expand...

I haven't flown an ILS on autopilot, but my impression is that a good autopilot can have the needles very well centered. One needs a way of distinguishing that from the situation the Air New Zealand crew encountered, where the glideslope was malfunctioning in a way that provided no off-center deflection of the needle, without displaying a flag to indicate a failure.

In order to continue the approach as a localizer only approach, you would need to be SURE you knew what components of the system were working and which weren't. There's not much time to make such a determination when on the final segment of the approach doing 140 knots (or whatever speed they use.)
 
Last edited:
I'm sure it depends on the autopilot. Most of the ones I have access to (excepting the CAP G1000) don't even have vertical coupling for the autopilot -- just alt hold and maybe VS. I'm not sure that it's a good idea to couple vertically without an auto throttle, due to the risk of an Asiana accident, especially if there is some wind shear.
 
MAKG1 said:
I'm sure it depends on the autopilot. Most of the ones I have access to (excepting the CAP G1000) don't even have vertical coupling for the autopilot -- just alt hold and maybe VS. I'm not sure that it's a good idea to couple vertically without an auto throttle, due to the risk of an Asiana accident, especially if there is some wind shear.
Click to expand...

:dunno:
 
Palmpilot said:
I haven't flown an ILS on autopilot, but my impression is that a good autopilot can have the needles very well centered. One needs a way of distinguishing that from the situation the Air New Zealand crew encountered, where the glideslope was malfunctioning in a way that provided no off-center deflection of the needle, without displaying a flag to indicate a failure.

In order to continue the approach as a localizer only approach, you would need to be SURE you knew what components of the system were working and which weren't. There's not much time to make such a determination when on the final segment of the approach doing 140 knots (or whatever speed they use.)
Click to expand...

It depends on the winds. In calm conditions they will fly the donut. In stronger gusty conditions they don't do quite as well, but when working, still acceptable in the few I've used. I did an IPC in a Comanche once and it was gusty with about 5°-15° cross wind, and my hand approach was more stable than my coupled one with a 55x.

As for converting an ILS into a LOC approach unexpectedly in mid stream, I would likely prefer to go missed and re brief the approach as a LOC approach to get in the right frame of mind and give me a chance to get a needle up to complete an ILS. Me not flying much IMC has a lot to do with that decision. I am sure that guys who frequently fly both approaches would have no problem.
 
You must log in or register to reply here.
Back
Top