Below the GS at SFO again

[Cows getting bigger]

Hey, I'm not disagreeing with there being huge advantages from a stable approach. All I'm saying is that your assertion that "the object of a Non-precision approach is to arrive at the MAP and MDA at the same time" doesn't work in a lot of circumstances.

I would prefer to say something like "The object of a non-precision approach should be to arrive at a decision point whereby the approach can be continued having achieved the required visual references without any significant change in configuration/power setting." All that said, 'dive and drive' is acceptable and works quite well with smaller GA aircraft types. I have no idea about wide-bodies but I suspect it isn't a good idea. :eek:

[framer]

It's pretty simple. If you're still doing ' dive n drive' then you are increasing the risk of an accident if you become distracted or disoriented.

[bubbers44]

My approach into TGU, Tegucigalpa, Honduras came through a valley at a descent clearance to 2700 AGL, 6,000 MSL. If you didn't get down in the valley at 6,000 you would probably still be in the clouds because that is where they started to lower. You could do the constant descent approach and go around or get down a bit early and land, your choice. Once you had the airport in sight you could descend to 5,000 ft and make a normal visual landing. If you did the constant profile of 3 degrees with your profile descent you may have to go around. In over 600 approaches I only had to go around once because of the ceiling and no other airliner got in that day.

Dive and drive isn't exactly what we were doing but reaching MDA early let us land many times where the constant descent wouldn't have worked. Both are safe but both are legal. Once you are cleared to the next altitude use your own judgement on how rapidly you want to do it because the airspace is clear. Captains should use good judgement on how to conduct their approaches and not be forced to use SOP's when other options will work much better. Captains are in charge and shouldn't be handicapped by being forced to not use their skills because others can't do it safely.

[Capn Bloggs]

A Squared -"Frighteningly low" was the term used by bug smasher. Now if dragging a large aircraft in low and slow (MDA and Vref) "a very long way from the airport" is frightening for the pilot then, surely, this is not a good thing.
Bugg smasher had a large lump on the side of his face: a certain amount of tongue-in-cheek, as it were.

There is no question that, with cloud right on the minima, you stand a better chance of getting in if you are lower than the 3° profile by a bit (say 1-200ft). That allows you to actually get right to the MDA, have a look and then GA if you can't stay on slope. This is impossible on a true 3° CDA. I doubt Bugg Smasher is suggesting miles and miles at the MDA, except when he wants to annoy his mate's BBQ.

[armchairpilot94116]

So I gather "dive and drive " is fine if :

a. you know what you are doing
b. you are not 600 feet below the glide slope
c. the tower does not freak out
d. you are not an Asian airline (because then please see "a" above)

[RetiredF4]

One more input:
When flying a descent profile which brings you down to the MDA at the MAP or a point last suitable for landing, then you either have to make the go around decision prior to the MDA or you will drop below the MDa during go around.
If you fly a descent profile which brings you down to MDA early, you can level off on the MDA and make the go around later prior or at the MAP, but some feet lower then in the previous case.

[BuzzBox]

Flight Safety Foundation article on the dangers of 'dive and drive':

http://www.flightsafety.org/asw/nov07/asw_nov07_p13-17.pdf?dl=1

[RetiredF4]

Nice read, although in its context it deals with the task and asociated risks of NPA approaches in general. And due to the fact (as the article states) that they are not often flown they should be set up as close to a PA approach as possible.

One main cause of the dangers of a NPA is named as crew overload. I wonder, what they are saying about a NPA or even an PA followed by circling to another runway.

Bottom line i think it's again a training issue.

[BuzzBox]

In a study performed during development of the FSF ALAR Tool Kit, several pilots were asked when they last conducted a nonprecision approach. Most pilots, especially those from the United States, replied, “When I had my last simulator ride.” Some pilots said that the dive-and-drive technique is markedly different from the way they normally fly approaches and that they get very little practice in this procedure to maintain proficiency.

Training and recency methinks. No doubt they CAN be flown safely, but there's a much greater risk of a stuff up and most of us don't get anywhere near enough practice (if any) to maintain proficiency. Further:
Dive-and-drive is the antithesis of the stabilized approach recommended by the Flight Safety Foundation (FSF) Approach-and-Landing Accident Reduction (ALAR) Task Force.

I wonder, what they are saying about a NPA or even an PA followed by circling to another runway.

I don't know what the FSF has to say, but the airline I work for (major international, widebody) banned circling approaches not so long ago. A study revealed that most crew rarely (if ever) flew them, except in the simulator. Consequently, they decided it was one more risk they could do without.

[Deep and fast]

CDFA on a non precision approach, add 50 feet to MDA and treat as a DA. You won't bust minima and if you are visual you should be stable to continue.

This has been a policy at both of my jet employers.

[Capn Bloggs]

CDFA on a non precision approach, add 50 feet to MDA and treat as a DA. You won't bust minima and if you are visual you should be stable to continue.
As pointed out before, best by RetiredF4, using this method you cannot get to the MDA and still land if the cloudbase is just above the MDA. In fact, it puts more pressure on crews because they have to make a snap decision with fewer protections than afforded by a precision approach where you are allowed to go below the DA and won't hit anything. NPAs by definition may well have poorer tracking positioning and offset alignment with the runway.

It is better to get to the MDA a little early, below the "glideslope", have a good look, and then GA if necessary. Now if you don't know how to organise/do that, then a database 3° CDFA NPA is the next best option.

[Lonewolf_50]

Captain Bloggs, your summary sounds a lot like what people are taught on their way to an instrument rating. :ok::cool: The key to success in that method is staying ahead of the aircraft (mentally) as a crew, and a good "who is looking out, who is on gauges" CRM standard. And of course, if at MAP all is not well, GA and try again, or try another approach, or head to alternate ...

[golfyankeesierra]

It is better to get to the MDA a little early, below the "glideslope", have a good look, and then GA if necessary
Sure, for the little planes that might work, but we are talking about 777's here...
Personally I abandoned your approach when I left the pistons behind.

[West Coast]

Capt Bloggs

Increases in safety often comes with a price. If the once in a blue moon consequence is I don't get in because the wx is at the difference between dive and drive and a constant angle altitudes, so be it.

I like the constant angle NPA procedures, my workload is lessened and the bigger picture is easier to maintain.

[bugg smasher]

Regarding the original subject of this thread, after having spoken to a few guys who were around that day, and I stress, the official investigation is still pending, but this being a rumor board; the EVA aircraft crossed the San Mateo Bridge slightly above 600', a full 1300' below the recommended altitude. It is not clear if they were descending, or level. ATC alerted the aircraft twice, with no response, and then issued the Go Around instruction, to which they responded.

I would say, and this is my personal opinion only, they were frighteningly low.

[sb_sfo]

I live abeam the bridge, and if you're right, they were only about 400' above it, google lists 135' to the water. Frightening to the traffic on the bridge, too, cause that approach is usually about 1615 lcl, early rush hour!

[Capn Bloggs]

Increases in safety often comes with a price. If the once in a blue moon consequence is I don't get in because the wx is at the difference between dive and drive and a constant angle altitudes, so be it.

I like the constant angle NPA procedures, my workload is lessened and the bigger picture is easier to maintain.
Fair enough. :ok:

[bugg smasher]

The traffic pattern at San Carlos is 800', so at least they were underneath that..., although it might be disconcerting for a student pilot in a 152 to see a triple seven slide by below him...

[BOAC]

What happens to San C traffic when SFO 28 traffic flies the RNAV or ILS?

[bubbers44]

RNAV 28L for KSFO restricts arrivals to no lower than 1800 ft until well past San Carlos so with their 800 ft pattern at least 1,000 ft separation exists.

[BOAC]

So why 1900 for a visual?

[olasek]

It is logical that in VMC they want to increase vertical separation between SFO arrivals and San Carlos pattern traffic but in IMC everybody is under positive radar control plus no one at San Carlos is flying traffic pattern hence no longer need for extra buffer in separation.

[BOAC]

but in IMC everybody is under positive radar control plus no one at San Carlos is flying traffic pattern hence no longer need for extra buffer in separation. - why do you assume it is 'IMC' before aircraft fly the RNAV or ILS - not so. I thought also we had seen a restriction now on 'no visuals' for 'foreign' pilots regardless of weather?

[underfire]

Helmet cam images from firefighters responding

Video from firefighter's helmet cam sheds new light on death of teen at Asiana Airlines crash site (http://news.yahoo.com/video-fire-helmet-asiana-crash-teen-144658815.html)

[bugg smasher]

BOAC, today we were offered the choice of LOC/DME Y or RNAV (GPS) Z to 28R. The FAF in both cases is AXMUL, which has a crossing restriction of 1800', and also happens to be conveniently, and not so surprisingly, co-located with the San Mateo Bridge.

A brief examination of both these charted procedures should answer any remaining questions you may have. Happy to assist with any further queries, as and where I am able.

[Auberon]

So why 1900 for a visual?

It's a valid question. The RNAV (GPS) RWY 28L has an 1800 minimum altitude between HEMAN and DUYET. DUYET is actually slightly before the bridge, so aircraft on the RNAV are crossing the bridge below 1800.

[bubbers44]

BOAC, you can see that there is at least 1,000 ft separation so why did you bring up the question?

[bubbers44]

How do I know, now you have 1100 ft clearance so it is safer. That is what the RNAV approach says versus the visual. Ask Bay approach.

[bubbers44]

BOAC, 1,000 ft separation is normal. All ILS and RNAV approaches to 28L have 1,000 ft or greater separation. The 800 ft traffic pattern will have 1,000 ft separation.

[Auberon]

How do I know, now you have 1100 ft clearance so it is safer. That is what the RNAV approach says versus the visual. Ask Bay approach.

Is it really safer? 1900' at the bridge puts you above the PAPI. 1800' is closer to being on the PAPI. There seem to have been a lot more issues with unstable visual approaches to KSFO than conflicts with KSQL traffic.

[West Coast]

A 100 ft difference (if accurate) at the bridge isn't the difference between a stable and unstable approach. There's larger issues if you can't lose an extra 100 ft that far out.

[bubbers44]

I heard they were at 400 ft over the bridge so guess we are not talking about 100 ft.

[bubbers44]

They obviously need some flying lessons because any US pilot would look out his window and say what am I doing at 400 ft 6 miles out.

[Capn Bloggs]

what am I doing at 400 ft 6 miles out
Not quite that low, Bubba. They were 600 feet at 3.8 miles.

Is it really safer? 1900' at the bridge puts you above the PAPI. 1800' is closer to being on the PAPI.
The bridge is about 5.2nm from the threshold. Given the PAPI on 28L is 2.85°, on-slope would be about 1540ft. So we are now at 1900ft (350ft-odd high) with the PAPI showing 4 whites, with less than 2 minutes to get down to be stable by 1000ft AGL (assuming). What's going to happen? They are going to stuff the nose down. I suspect that in all the hullaballoo, they forgot to pull back when they got on-PAPI.

They should have coped but it appears the system set them up.

[olasek]

1540ft. So we are now at 1900ft (350ft-odd high)
Your calculations are seriously OFF, it was calcuted before, it is not 1540 but 1740 instead, this is how high you would be crossing the bridge shooting the ordinary ILS to 28L, so at 1900 you are only 160 ft higher, nothing that even a rookie commercial pilot could not handle. From the bridge it is 5.4 nm to the touchdown zone and the GS is at 3 deg. and the trigonometry is very easy. Also the Asiana 777 crew was nowhere near 1900' when crossing the bridge, they were at about 2400' :ugh: :} so the system did not 'set them up', if there was any 'set up' here it was of their own making. Also your assumption about stability at 1000 ft is wrong, in VMC you can get stable as far down as 500 feet. Lets dispell any myth that this was a difficult approach, comparing with what's out there in the world of aviation it doesn't even qualify as a moderately difficult approach.

[BOAC]

bubbers/bug smasher - you have both failed to explain why a greater separation is required on a visual approach against that required on the other two. All three may be flown (and no doubt are) in VMC. (The visual glideslope is 2.85 degrees, by the way, (VGSI) so 1540'ish at the bridge would be correct. and would obviously conflict with San C, but why make it 1900?)

The whole thing sounds like a 1960's Haight-Ashbury put together.

[bubbers44]

I agree. I have done lots of approaches to 28L with no ILS and it isn't even worth talking about. It is so easy. Just look out the window and line up. Any private pilot could do it easily.

Automation has made pilots into monkeys. Not all pilots, just some who need automation because they can't do a simple visual approach without it.

[bubbers44]

I think any competent pilot could be 3,000 ft at the bridge and be stabilized at 500 ft so being a bit high should not be a problem unless you have a problem flying. 1900 ft at the bridge is a non event.

[Lord Spandex Masher]

They obviously need some flying lessons because any US pilot would look out his window and say what am I doing at 400 ft 6 miles out.

I love your unwavering belief that any US pilot is infallible. I wonder why any US pilot would allow themselves to get to 400' at 6 miles?

[bubbers44]

LSM I don't think any pilot on earth would be at 600 ft 6 miles out and look out his window and think that looks about right. I will keep descending. Do you?

[Cows getting bigger]

Or indeed why any US pilot would slam a nosewheel into a runway? There are people on here who talk as if we should still be sky gods, capable of wrestling victory from defeat, rather than recognising the harsh reality that we are merely part of a system. Sure, if the Korean guys need their parts feeling for aspects of their airmanship then so be it, but let us not lunge at this as the only causal/contributory factor.

[BOAC]

LSM - what these gallant superb aviators cannot understand is that - shall we assume that it would be 'nice' to be stabilised on the visual g/slope at 1000' keeping all FOQA readers happy? - to achieve this but to be at 1900 at 5.2 from touchdown requires a rate of descent of almost 1300fpm or almost twice the correct rate.

The whole thing could be made far better if the VGSIs were at 3 degrees and 1800' at the bridge was 'standard', when I reckon 900fpm would crack it - and is far more 'acceptable' in FOQA terms and for 'seaparation' on g/a traffic.

I think we should quietly leave Bubbers in his own happy world at 3000' at the bridge....

[bubbers44]

We did th 3,000 ft outer marker into SJC for decades with no problem with our 737's. It was a noise abatement arrival that all of us did with no problem. It would be even easier into 28L at SFO with the long runway. The physics was about the same. Idle power until 500 ft which went away about 10 years ago.

[rudderrudderrat]

Hi BOAC,
shall we assume that it would be 'nice' to be stabilised on the visual g/slope at 1000' keeping all FOQA readers happy? to achieve this but to be at 1900 at 5.2 from touchdown requires a rate of descent of almost 1300fpm or almost twice the correct rate.
Even that is well within the guidance of our FCTM when doing an ILS in IMC.
So it certainly is not a problem when visual.
"GLIDE SLOPE INTERCEPTION FROM ABOVE
The following procedure should only be applied when established on the localizer. There are a number of factors which might lead to a glide slope interception from above. In such a case, the crew must react without delay to ensure the aircraft is configured for landing before 1 000 ft AAL.
In order to get the best rate of descent when cleared by ATC and below the limiting speeds, the crew should lower the landing gear and select CONF 2. Speedbrakes may also be used, noting the considerations detailed in the sub-section "Deceleration and configuration change" earlier in this chapter. When cleared to intercept the glide slope, the crew should:
• Press the APPR pb on FCU and confirm G/S is armed.
• Select the FCU altitude above aircraft altitude to avoid unwanted ALT*.
• Select V/S 1500 ft/min initially. V/S in excess of 2 000 ft/min will result in the speed increasing towards VFE"

[BOAC]

Bubbers - things have changed a bit. The emaphasis now is on steady, measured flying as an aim.

RRat - it was not in the 737 FCTM in my time, but a 'special brief' was required if more than 1000fpm was required.

I still cannot understand why SFO has done it this way when 1800 and 3 degrees would make life a lot easier..

[Auberon]

As they said in school, show your work. :)

The 28L extended centerline crosses the bridge at approximately 37°34'21.00"N,122°15'49.20"W.
http://goo.gl/maps/sC6la

The 28L displaced threshold is at approximately 37º36'43.54"N,122º21'33.35"W.
http://goo.gl/maps/xJP48

The distance between those two points is approximately 31,227 feet.
Great Circle Mapper (http://goo.gl/7FdChk)

The 28L PAPI has a 2.85º glide path and a TCH of 64 feet, so add 1,286 feet to get the touchdown point.
64/tan 2.85º - Wolfram|Alpha (http://goo.gl/GVm29a)

So we have 32,513 feet between the bridge and the touchdown point.
31227+1286 - Wolfram|Alpha (http://goo.gl/kDCLrC)

That means to be on the PAPI at the bridge you would cross at 1,619 feet.
32513*tan 2.85º - Wolfram|Alpha (http://goo.gl/bHN07g)

[DOVES]

To BOAC
to achieve this but to be at 1900 at 5.2 from touchdown requires a rate of descent of almost 1300fpm
Probably I did not understand correctly
But I would like to point out humbly that:
With 120 knots to lose 1900 ft in 5.2 NM a variometer of a little more than 730 ft/min is needed, with 150 knots 760 ft/min.

[BOAC]

shall we assume that it would be 'nice' to be stabilised on the visual g/slope at 1000' keeping all FOQA readers happy? - ie not threshold.

[bubbers44]

BOAC, Please don't worry so much because when you retire it doesn't matter if it is 2.85 or a 3 degree glide slope. You can intercept a glide slope from either above or below. Who cares what the new hire qualifications are. You won't have to fly with them. Just kick back and enjoy life.

[BARKINGMAD]

Is it not up to us airframe managers to have the cohones to say to ATC;

"Regret, unable due energy management" when invited/instructed to place our craft in a potentially UNSTABLE approach, a proper response to which is usually enough to permit a more measured and controlled manoeuvre?

So why are we not doing so?

There are a large number of "cowboy" airframe managers out there who regard that as "cissy" and the "we can hack it" attitude seems to be reflected in the spate of overruns at worst or FDM events at best.

Maybe the HR wallahs could be invited to throw away their useless psychometric tests and search for a more reliable testing method to exclude these dangerous characters, possibly before the insurers force them to??

Similarly, training departments and staff should be up to the mark when they are asked for guidance on the subject of E M, a subject which is so sadly neglected or is liable to a variety of methods which only serve to confuse.

My recent conversations with "newbies", otherwise very well-prepared for the line, shows this essential aspect of flying is either not taught well, badly imparted or a confusing variety of methods of (mentally) computing and managing the problem just makes life more difficult and liable to error.

Furthermore, the majority of these "rules of thumb" only refer to the combination of potential (altitude) energy, (kinetic) speed energy, of aircraft and (occasionally) the instantaneous wind velocity. None I have heard of include the weight (sorry, mass!) of the 'frame, which can have a significant effect on the trackmiles required to dissipate this energy, hopefully in an efficient, quiet, smooth and fuel-cheap manner.

Proof of the "m" effect on the 1/2mV squared part of the total is hidden away in the OMs various contained in the EFB or in the drawer beneath the jumpseat in the "NG", sadly it has recently been removed from the QRH Performance Inflight (in my current company), where it should be readily available on a sector by sector basis, especially for the "newbies" as they get to grips with the job.

I suspect that some of the "oldies" are not aware of this significant factor either, but I place the blame fairly and squarely in the hands of the relevant TRTOs.

Rant over, I really must get out more...............................................:(

[roulishollandais]

@Doves
Speed 150 kts results not 750 ft/mn but 913 ft/m

@BOAC
5.2 nm 3° 900 ft/mn results in speed 170 kts, Z=1656 ft not 1800

1900 ft, 5.2 nm, 1300 ft/mn results in speed 215 kts at the bridge decreasing with speed

[fireflybob]

When I read these posts about how "difficult" Energy Management is I wonder how we managed all those approaches in basic aircraft like the B 737-200 when due to lack of ATC coordination and/or radar you were often held high!

Anybody would think it was some sort of "black art".

[737-NG]

Good to see there are still a few real pilots left in this world who can actually make real and simple calculations.
Ca se perd, mon dieu. Merci mon ami!

[rudderrudderrat]

Hi BOAC,

On 16 July you said, Returning to this SFO "180 to 5" - I consider that if flown EXACTLY that way it is un-doable (I await the 'aces' cries). Most manufacturers' FCOM do not encourage using flaps as 'speedbrakes' which is what is necessary here.
Today you said it was not in the 737 FCTM in my time, but a 'special brief' was required if more than 1000fpm was required.
When capturing the glide slope profile from above, RODs of more than 1,000ft / min will be required and we've always had SOPs to achieve this.
May I suggest you obtain up to date copies of FCOM etc.

[BOAC]

I wonder how we managed all those approaches in basic aircraft like the B 737-200 - indeed, but in our days there was a little less 'electronic' monitoring so it was probably 'easier' to hack it. Come to that, I can go back a few more years to land at Vref-5 on the 'numbers' (threshold), throttles closed - every time:) Try that in your modern company jet? As I said, things have changed.

[fireflybob]

The basics of safely operating an aircraft never change - that's the point!

[BOAC]

No, but the ability to do so appears to have.

[rudderrudderrat]

Hi BOAC,
I can go back a few more years to land at Vref-5 on the 'numbers' (threshold), throttles closed - every time
Why on earth would you try to do that with passengers on board?

[BOAC]

Why on earth would you try to do that with passengers on board? - who said I did? (Actually with a 'pax' in the two-seater it would still be the same.) That was mil swept wing. Just trying to show that it was indeed "safely operating an aircraft...." in that environment. The environment has changed. Just as the way we 'got in' in the Jurassic might not please the modern FOQA readers every now and then.

[millerscourt]

I thought we all landed with the throttles closed except on an aircraft carrier:confused:

[BOAC]

Yes - I think you are right.

[BARKINGMAD]

As BOAC said;


- indeed, but in our days there was a little less 'electronic' monitoring so it was probably 'easier' to hack it. Come to that, I can go back a few more years to land at Vref-5 on the 'numbers' (threshold), throttles closed - every timehttp://images.ibsrv.net/ibsrv/res/src:www.pprune.org/get/images/smilies/smile.gif Try that in your modern company jet? As I said, things have changed.


Omigawd, I'm now flagellating myself with guilt at the memory of placing a Bae146, with SLF aboard, on the "keys" at NCL, having shut the thrust levers at 20,000ft and not reopened them. What naughty boys and girls we were then?
And that in a 'frame with NO IRSs, wind readouts, not even the early GPS boxes were available.

Seriously though folks, have the skills deteriorated since then, or has the FDM environment shoe-horned us all into a mindset and skillset where such practices are not only frowned upon but may be impossible to repeat when the chips are down and superior skills are required following some catastrophic failure such as Sioux City??

Before the arrows start flying I'm just asking the question, NOT proposing deadstick arrivals in future, though they did save fuel and noise....................;)

[BizJetJock]

having shut the thrust levers at 20,000ft and not reopened them.
We used to do that when we suspected the natives were less than friendly...

[aterpster]

Letter to editor in the Aug 5-12 issue of Aviation Week. Sounds like he might belong to Prune. :)

http://i201.photobucket.com/albums/aa214/aterpster/AvLeakAUG5-12_zps01e0ca5b.jpg

[Capn Bloggs]

Your calculations are seriously OFF, it was calcuted before, it is not 1540 but 1740 instead, this is how high you would be crossing the bridge shooting the ordinary ILS to 28L, so at 1900 you are only 160 ft higher, nothing that even a rookie commercial pilot could not handle. From the bridge it is 5.4 nm to the touchdown zone and the GS is at 3 deg. and the trigonometry is very easy. Also the Asiana 777 crew was nowhere near 1900' when crossing the bridge, they were at about 2400' so the system did not 'set them up', if there was any 'set up' here it was of their own making. Also your assumption about stability at 1000 ft is wrong, in VMC you can get stable as far down as 500 feet. Lets dispell any myth that this was a difficult approach, comparing with what's out there in the world of aviation it doesn't even qualify as a moderately difficult approach.

Rubbish, rubbish, rubbish. The ILS was off. PAPI is 2.85°. I said 1540, Auberon says 1619 (thanks). Not your 1740. This thread is not about Asiana, it's about EVA. And finally, what is the real stabilised policy for EVA?

We used to do that when we suspected the natives were less than friendly...
The hair dryers on the swinebat didn't put out enough heat to get shot at, be them at full power or idle... :}

Letter to editor in the Aug 5-12 issue of Aviation Week. Sounds like he might belong to Prune.
One would have to suspect the original article? After all, if such an esteemed publication could be so lead astray as to come up with this about Asiana and FLCH on the 22nd:
One group of pilots has concluded this based on intimate knowledge of the 777-200ER's automation systems; the other by flying scenarios in a 777 simulator
one should question whether they actually know what is going on.
That "intimate knowledge" was on Prune within a day or two of the prang! :ouch:

[bubbers44]

I started flying with the airlines in 737 100s and 200s, then the 300. It seemed so simple then in the 80's. SJC being 3,000 ft 6 miles out was a lot of fun but not a real challenge. Now somehow it is. Why? Different pilots with different experience?

We didn't have GPS then so this FMC constant descent approach wasn't possible then. Now some piilots add 50 feet to MDA so they don't have to level off but go around 50 ft above MDA? I usually did a circle to land approach where I flew so adding 50 ft made no sense because I was going to level off anyway. If you want to be really safe only land when it is VFR. Now the dumbest pilot in the system probably can handle it. We have minimums on all approaches that are safe so unless you are a new captain getting your first 100 hrs at higher minimums, fly the approach to minimums like everybody else.

[A Squared]

If you want to be really safe only land when it is VFR. Now the dumbest pilot in the system probably can handle it.

Did you hear about the Asiana crash at SFO? It was in all the papers.

[EGPFlyer]

We didn't have GPS then so this FMC constant descent approach wasn't possible then. Now some piilots add 50 feet to MDA so they don't have to level off but go around 50 ft above MDA? I usually did a circle to land approach where I flew so adding 50 ft made no sense because I was going to level off anyway. If you want to be really safe only land when it is VFR. Now the dumbest pilot in the system probably can handle it. We have minimums on all approaches that are safe so unless you are a new captain getting your first 100 hrs at higher minimums, fly the approach to minimums like everybody else.

No one adds 50ft to an MDA for circling do they? Isnt it only for straight in CDFAs with no DH listed?

[bubbers44]

I didn't say it, somebody else did. Yes I thought of the Korean airline having a lot of problems landing VFR when I made that post. I meant pilots looking out the window approaches.

[bugg smasher]

The whole thing sounds like a 1960's Haight-Ashbury put together

Dope-smoking hippies BOAC? How very condescending of you. One would think your airline had never landed a triple seven frighteningly short of the runway. 'Gallant superb aviators' indeed.

[bubbers44]

BS, are you talking about the 777 with no engines? I think they did a great job of not killing anybody. I put them up in the Sully catagory.

[Lord Spandex Masher]

LSM I don't think any pilot on earth would be at 600 ft 6 miles out and look out his window and think that looks about right. I will keep descending. Do you?

Well, I'm aware of at least one...or two.

[flarepilot]

many years ago I read quite a bit about the visual miscues of flying...one of the wisest was one of the most basic...and that ''depth perception'' really doesn't work, or else no one would ever bump wingtips on the ground.having read these interesting books and words like...above 800 feet or so, nothing makes sense to the tired eye, I vowed to always check myself on visual approaches, esp in large planes


I am not alone in this as many pilots have landed at SFO without much of a problem. So, are we all super pilots and asiana pilots are average? No...somewhere these men entrusted with a very large plane and 300 people missed some basics.

whenever a baseball team sucks, its back to the basics

so to a pilot...are you screwing up? go back to the basics.


Stable approaches hopefully by 1000' and certainly no lower than 500'. USE all available aids, known distances, 3 to 1 rule, DME, glideslopes of ILS or PAPI or kinds of computerized.

Looking out the window does work esp with papis.

A student of mine from many years ago wrote me...he is now flying transatlantic runs...he reminded me of the visual scan for any landing...runway airspeed, over and over. Add to that descent rate, pre calculated based on groundspeed and you should not be crashing.

one hand on control yoke, one on the throttes/thrust levers and never let the plane fly you.

Crossing the San Mateo Bridge at 1900 feet is not out of the question for any plane...configure early if you have to, I"ve seen L1011s with their gear down 16 miles from the airport.

[jandakotcruiser]

No one adds 50ft to an MDA for circling do they? Isnt it only for straight in CDFAs with no DH listed?

True, most operators do not add 50 ft for a level segment!

When I flew the B737 classics and old classics, I flew thousands of visual approaches and circling approaches too after the DME step downs! Heck, I dove and drove too!

However once on the heavies and widebodies where the characteristics dictate a saner approach would be constant angle approach to MDA+50', I have no problems with that too. No mucho mumbo jumbo for me because on the heavies, I would be lucky to have 2 approaches and landings a month to airports where I hardly remember.

If I am not mistaken in the U S of A, the heavies are not to circle if the CIG is less than 1000' and vis below 2 sm.

[OK465]

Crossing the San Mateo Bridge at 1900 feet is not out of the question for any plane

:)

I don't think many are questioning this in terms of either what any aircraft is capable of, or what the pilot should be capable of....the question is....WHAT exactly IS 1900' actually providing relative to aircraft in the Class B airspace that goes up to 1500' in some proximity.

It was pointed out that traffic pattern altitude for both fixed-wing and rotary is 805' MSL at KSQL. Rounded upward to 900' + 1000' is indeed 1900', but....

all traffic patterns at KSQL are performed southwest of RWY 12/30 and even the IF legs to the RNAV finals for both 12 & 30 are southwest of the both FACs. The distance from DUYET to KSQL is 3.5 NM and from HEMAN is 3.6 NM. Is this pattern traffic really a factor for the 28L or 28R centerlines?

It would appear that any significant separation requirement for SFO traffic would arise due to GA aircraft that are transiting the Class B between SFO 7 DME & 10 DME inbound from, or outbound to the north or northeast of KSQL, i.e. traffic which actually crosses the 28L/R centerlines.

So....are they issued a restriction to "maintain 900' or below" by NORCAL at certain times in this area of Class B airspace that goes up to 1500' for this reason? Or what?

The 'why' is still a reasonable question.

(BTW b44 I was going to call 'Bay' approach but that number's been disconnected)

[MarcK]

So....are they issued a restriction to "maintain 900' or below" by NORCAL at certain times in this area of Class B airspace that goes up to 1500' for this reason? Or what?
I've never had a restriction, other than staying clear of class B. That's 1500' outside of the 10NM ring. Class B is at the surface at the San Mateo bridge. Typical GA crossing the bay are at 9 DME when crossing the extended centerlines if going to Hayward, and more like 12 DME if going to the central valley via SUNOL.

[bugg smasher]

Whilst this does not address the 1900' vs 1800' issue we are here addressing, the complexity of the airspace is worth noting. From the FAA-issued Southwest A/FD Supplemental;


"SAN FRANCISCO INTERNATIONAL AIRPORT EXPANDED CHARTED
VISUAL FLIGHT PROCEDURES
(Until Further Notice)

***GENERAL***

San Francisco International Airport (SFO) is subject to stratus moving slowly from West to East, creating a reportable weather ceiling over the airport, while the final approach area for Runways 28R and 28L have no significant ceiling or visibility conditions. And expanded charted visual flight procedure (E/CVFP) has been developed to maximize the level of airport efficiency during the unusual weather conditions described above.

***MINIMUMS***

The E/CVFP incorporates the following weather minimums:

SFO ceiling 2100 feet and visibility 5 miles; or,
SFO ceiling 1000 feet and visibility 3 miles, and, visibility 5 miles in the Eastern quadrant (030–120), and, ceiling 2400 and visibility 5 miles at the automated weather observing system (AWOS) located at N37°34.332 W122°15.592 LOM. In the event the AWOS is inoperative, weather at San Carlos (SQL) is required to be at least ceiling 2400 feet and visibility 5 miles.

Although the listed weather minima are in effect aircraft should not expect simultaneous E/CVFP approaches unless N37°34.332 W122°15.592 AWOS ceiling is at least 3500 feet and visibility is at least 5 miles.

***SPACING AND SEQUENCING***

Controllers will clear aircraft for the E/CVFP in accordance with the provisions of Order 7110.65, Air Traffic Control. They will not utilize phrases requesting or requiring aircraft to ‘‘fly right alongside’’, ‘‘wingtip to wingtip’’, or ‘‘directly abeam’’ other aircraft. Additionally, controllers will not assign instructions or require aircraft to pass and/or overtake other aircraft on the adjacent final approach course. Preferably, aircraft will be vectored to achieve a slightly staggered position of approximately 1/ to 1⁄4 mile behind the aircraft on the adjacent final approach course. Heavy aircraft and B757’s will not be
authorized to overtake another aircraft on the adjacent final approach course. Wake turbulence cautionary advisories will be issued, as appropriate.

***GO-AROUND PROCEDURE***

The Tipp Toe and Quiet Bridge approaches are visual approaches, and as such have no missed approach segment. If a go-around is necessary, aircraft will be issued an appropriate advisory/clearance/instruction by the tower or tracon. To
ensure standard separation from other traffic, these instructions will include the assignment of a specific heading and altitude, Normally, the following procedures will apply:

Tipp Toe Visual Runway 28L
In the event of a go-around turn left heading 265, climb and maintain 3000; or as directed by Air Traffic Control.

Quiet Bridge Visual Runway 28R
In the event of a go-around turn right heading 310, climb and maintain 3000; or as directed by Air Traffic Control."

In any and all events, ladies and gentlemen, let us not allow ths very interesting discussion to descend, predictably, into a trans-Atlantic slagging match. There are very real issues here at SFO, what better way to sort through them, than with the combined and focused intelligence of PPRUNES' own illuminati.

[BOAC]

It was pointed out that traffic pattern altitude for both fixed-wing and rotary is 805' MSL at KSQL. Rounded upward to 900' + 1000' is indeed 1900', but.... - I'm relieved to see you are questioning this too. In the case of RNAV or ILS this separation is NOT achieved, so why is it there? Now you are telling us the patterns do not conflict anyway? Why 2.85 PAPIs (an angle that no-one needs)?

[FullWings]

If you came in on a 3deg approach to 2.85deg PAPIs, you'd still see two whites and two reds. AFAIK, the transition to three whites/reds is normally set 1/6 (0.166) to 1/5 (0.2) of a degree above/below the datum slope. Four whites/reds is 0.5deg off the datum.

As to WHY they're set at 2.85 @SFO, I have no idea... :confused:

[BOAC]

Thanks, FW - put in the fading memory bank:)

[aterpster]

BOAC:

- I'm relieved to see you are questioning this too. In the case of RNAV or ILS this separation is NOT achieved, so why is it there? Now you are telling us the patterns do not conflict anyway? Why 2.85 PAPIs (an angle that no-one needs)?

These are the folks who take input and can provide such explanations:

[email protected]

[underfire]

Fullwings,

It looks like a mathematical and visual 'equivalent'.

VGSI is set at 2.85° GPA with TCH=64,
ILS is 3° GPA with TCH=57

These two GPA's cross at about 2700 feet from threshold at about 200 feet above threshold, so at that point the 3° ILS GS is on the 2.85° VGSI GS.

At 1800 feet the ILS will be about 80 feet high above the VGSI GS.

The RNAV procedure which is being recommended, has a 2.85° GPA and TCH=53.

I am wondering if the 28L GS is different from the 28R GS was an early attempt to provide some aircraft offset between the parallel runways.

Edit: FW, looking at your .2° number. When you are at 1800 feet on the ILS 3° with TCH=57 GS:
the VGSI GS 2.85° TCH=64 altitude is 1719, (81 feet high on the VGSI GS)
and the +0.2° altitude is 1836. (36 feet low on the VGSI GS +0.2°)
(+.16° altitude is 1812, so you are 12 feet below the VGSI GS +0.16°)

Would this altitude difference be enough to show with the lights you are above the GS?

[fleigle]

For your edification;
http://www.youtube.com/watch?v=q4j3fzJ-8eo&feature=share&list=UU32r0TkhAGOxRiE52Uu2Kqw

Enjoy,
f

[Basil]

For your edification;
Izzat LHR on a quiet day? :p

[Basil]

New SFO approach light option for Asiana pilots......................

http://lists.kjsl.com/pipermail/beech-owners/attachments/20130719/7a1b0737/attachment-0001.gif

[bubbers44]

LSM, I said pilots looking out the window would be at 600 ft 6 miles out saying that looks about right. Obviously nobody was looking out the window on those two Asian approaches. Normal pilots on a visual approach look outside because they want to actually pilot the aircraft to landing not program it to landing. It has been done quite easily for over 100 years.

[Jaair]

Daylight as well. Kudos to the tower.

[flarepilot]

how did lindbergh manage a safe landing with no navaids, at night, at lebourget, after over 33 hours at the controls, no flight attendants, no relief pilot, no movies, rnav, authothrottles and NO flight attendants...

oh, and he was ahead of schedule with reserves to make rome.

[AirRabbit]

Hey Coagie: Sorry for the delay on this ... life goes on ... but ... I wanted to respond to your comments ... I couldn’t agree more with most of your statements. My experience has been focused almost exclusively on training and qualification issues, methods, and practices and I, too, would not want to cast dispersions on those members of airline community who do not have previous military background/training … but, those who do not and are still competent pilots must have paid dearly for the experience they do have and/or they’ve spent some significant amount of time in another context.

However, instead of looking in the direction of Pipers and Cessnas for that “back to basics” kind of training (while there are likely some who may be better off in going to that extreme) I think the better choice would be the proper and competent use of airplane flight simulation where specific training program requirements (in the form of specifically named and specifically described flight training tasks) are taught by competently trained instructors (and by that I mean trained on how to teach – not teach like they were taught) who know what tasks are to be taught, how to understand what may be or may not be “getting through” to the student, how to structure task assignments using directions that rely on the student to do something one his/her own. By that I mean … instead of telling a student learning to land the airplane, “flare to the landing attitude,” or more precisely “flare to a level flight attitude” - as the student will very likely have to be told what specific attitude that is and rely on “hints” provided by more experienced pilots as to what to do to find that attitude – all of which will provide opportunities to get it wrong or do it by rote memory (and usually will – as no two pilots fly the same and use exactly the same airplane input stimuli in exactly the same way in exactly the same sequence) which will almost always lead to wrong applications. Instead, a really good instructor (the kind of instructor I’ve attempted to describe above) will tell the student to do something that the student should already know how to do. You and I know that airplanes land from a “level-flight attitude – that is … all airplanes, all the time, no exceptions … unless there are unusual circumstances (and those should be covered AFTER the basics are learned). So, when landing training is needed (and it is almost always needed) good instructors should tell the student to fly down to the runway from the final approach … and level off at a specific height above the runway such that the wheels will be (pick a number) feet above the surface – preferably a rather small number, depending on the airplane – and if you’re unsure of the student’s ability to do this – pick a larger number). Then fly at that altitude AND that airspeed all the way down the runway. Also, tell the student “do not climb, do not descend, do not accelerate, and do not decelerate – also do not raise the nose and do not let the nose drop.” An experienced pilot will recognize that these “do not” instructions cannot be maintained without moving the throttles. A good instructor will point out to the student that a small amount of additional power will be required, but should be reminded to not allow the airplane nose to rise or fall and not to let the airspeed climb above or fall below what it was at the end of the flare … until reaching the end of the runway – then a go-around will be conducted.

Recall that the original reason for going through all this was to teach the student how to find what the landing attitude was and how THAT student should be able to find it. So, once the instructor and the student have reached that point, instructing the student to complete the landing, by gradually reducing the throttle(s) to idle at the end of (or throughout) the flare (ensuring “idle” is achieved not later than at touchdown), maintaining the airplane nose at THAT attitude/position), noting, of course, that the airspeed will decay and will allow the airplane to descend to a proper touchdown (which should be properly firm – not hard and not a “grease job”) should be able to be accomplished rather easily.

If the simulator used is properly defined, constructed, programmed, and tested then the above kind of training can (and should) be able to be used to achieve a student who has been properly and completely trained to land the airplane – and that technique is equally applicable to zero wind, up wind, down wind, cross wind, ice, snow, rain, or dry. Of course there would be more specifics when winds are involved – the above will always remain basic. And - with respect to the simulator – like I said, the simulator would have to replicate the airplane – to the largest extent possible … in every area where simulator performance or handling differs from that of the simulator, additional training and/or experience will be required.

The only logical way for every one of the kinds of training tasks, expected outcomes, methods of achieving such training outcomes would logically require some sort of standardized training instructions and rules. From this it is a rather short distance to understand that the more standardized the basic rules - the more standardized the training - and the more standardized the day-to-day operations. If we go back to what we were discussing initially – I think everyone will understand that we are interested in getting well trained, well qualified pilots into the cockpits of today’s airline operations. To me … this means well planned and well executed training and evaluation standards. And, while I may be stepping out a bit here ... I think that is logically in the hands of the regulator.

[Ye Olde Pilot]

Sops are great until it all goes wrong when you need someone who has basic flying skills and some experience:ok:

[Coagie]

AirRabbit, I think you have something there with the use of the simulator. When I thought of the Cherokees or 172's, I was thinking, in many cases, simulator time is at a premium, and simulators and their use, may be actually, more expensive to the bean counters, than small piston aircraft. I could be wrong. The simulator, would be a more accurate representation of the dynamics of the aircraft the big jet pilot flies, than a Cherokee, etc.. Maybe designing simulator testing, using your common sense approach, as a triage, to separate out, only the pilots that have lost their edge (and maybe some newbies, that never had it), could actually cut down on total simulator time, as only the pilots, who need it, would have remedial simulator training. AirRabbit, your approach may be the best way. That is, working within the framework, that's already there, and making meaningful changes. The KISS method rides again! Bravo!:ok:

[DozyWannabe]

@Coagie

There's a practicality consideration though - namely that there are only so many simulators in the world, but there are a hell of a lot more single-engined trainers.

Also, I think focusing on hand-flying finesse is only one facet of the issue. Another biggie for me is the situation whereby recurrent training at ATPL level can assume familiarity with the basics, when those basics can atrophy without revision. As an example, I'm thinking in terms of the airline industry focusing so hard on stall prevention training that those techniques supplant stall recovery training, which many pilots hadn't practiced since their PPL days.

[Coagie]

There's a practicality consideration though - namely that there are only so many simulators in the world, but there are a hell of a lot more single-engined trainers.
I figured that. Just wasn't sure of myself.

[Auberon]

Well, it looks like no one could really answer why the charted visual procedures crossed the bridge at or above 1900, so they've been changed to cross at or above 1800.

But there's still a discrepancy between the charted visual and instrument procedures. The charted visuals cross MENLO at or above 5000, while the instrument procedures cross MENLO at or above 4000.

http://155.178.201.160/d-tpp/1309/00375QUIETBRIDGE_VIS28LR.PDF
http://155.178.201.160/d-tpp/1309/00375TIPPTOE_VIS28LR.PDF