how high can you get in an approach on the 737

[mamad]

Hi guys

Sorry if this have been discussed before.

Let say you are shooting an approach on a 737 Cl ou NG the controller brings you quite high and clears you for the approach.

As per your experience how do you know if you can still make it or not ( request a 360 ) in other words what are the "high " tolerance of the 737

thank you

[TypeIV]

Depends on company sop, by what altitude must you be stable and what is the maximum sink rate.

With a normal load, slight headwind on a 3deg gp at 6,5nm out blazing in at 190-200kts you're usually stable by 1000' if you are with flaps 5 and start getting the a/c dirty but it varies alot from time to time. If its a hot summer day with alot of thermal activity it gets more difficult. The 737 is not forgiving.

[mamad]

thank you

Company SOP requires to be stable by 1000 ft IMC and 500 ft VMC

[BOAC]

Gear and full flap will allow 3000fpm (normal operating limit) at around 140kts groundspeed. You can work it out.

[HazelNuts39]

3000fpm @ 140kts - does that qualify as a steep approach?

(THY at AMS didn't quite achieve that level of performance:
It went from 140 kts @ 700 ft to 90 kts @ 400 ft in 27 seconds in LDG config @ idle, equivalent to 1800 fpm at constant speed).

[Centaurus]

I read of one Singaporean 737 captain that was very high and instead of going around (real men don't go around, syndrome) he simply whacked on full aileron left and right alternately in order to get flight spoiler drag. Passengers were upset understandably.

He still was 900 ft high over the threshold continuing with the wing waggle technique until his F/O had enough of this crap and shoved open the thrust levers thus forcing a go-around. The F/O saved the day on that occasion. The captain was later killed in another accident.

The moral is if the aircraft is clearly uncomfortably high on glide slope then don't risk crashing just to big note yourself in front of the co-pilot. Go around early - not later. The later you leave the decision to give it away and go-around, the more is the temptation to believing you can just sneak it in and continue the unstable approach. It is commonly known as piss poor airmanship. There is no shortage of cowboys who are tempted to give it a go.

[BOAC]

HN39 - it is about 4-5 times the normal ILS slope. I call that steep. The difference is this is planned, THY was not.

[HazelNuts39]

BOAC - Having been involved with qualifying aircraft for steep approach, I too thought it was steep. What do you do to descend more steeply than THY at idle?

[BOAC]

Remember they started 'on the glidepath' so the rate would be building - and I'm afraid I don't think your calculation is meaningful.

[HazelNuts39]

We'll have to differ then. The performance of the 737 demonstrated in the accident is quite similar to that of the aircraft I referred to, i.e. an L/D of around 6.5 in the landing configuration at Vref.

[BOAC]

I don't understand the point you are making. Over 27 seconds it went from a 3 degree Vref approach to a stall. There is no way you can get anything from an 'average' RoD. Show a plot at 27x1 second intervals and we can talk. I don't think the RoD was particularly 'steep' there anyway until the very end. Just in the wrong place. Also they had full power at a high pitch angle to cushion.

[HazelNuts39]

EDIT::

Over 27 seconds it went from a 3 degree Vref approach to a stall. There is no way you can get anything from an 'average' RoD.
In those 27 seconds the RoD was approximately constant at 11.11 fps. At the average airspeed of 115 kts the flight path gradient was 5.72 % (3.28 degrees).

The airspeed dropped at an approximately constant rate of 1.85 kt per second (0.0972 g).

The average drag-to-weight ratio is then 0.0572 + 0.0972 = 0.154 and the lift-to-drag ratio is 1/0.154 = 6.5

The argument is about the rate of energy dissipation.

Total energy is the sum of potential energy per unit mass (height) and kinetic energy (speed). Within certain limits, potential and kinetic energy are interchangeable.

The rate of dissipation of total energy is a function of thrust minus drag per unit weight. With the engines at idle, thrust is negligible, so the rate of energy dissipation is determined by drag-per-unit-weight, the reciprocal of L/D.

An L/D of 6.5 at zero thrust produces either a flight path angle of 8.8 degrees (15.4%) at constant speed or a deceleration of 3 kt/second (0.154*g) at constant height, or anything in between.

What is the basis of your 3000 fpm?

[BOAC]

What is the basis of your 3000 fpm? - see posts 2 and 4. The normal max in most airlines is 3000fpm within 3000' ft of MSA. The only way of achieving that without accumulating too much KE is as described.

[HazelNuts39]

The point that I'm making is that full flaps, gear down, idle thrust at 140 kts will give you a steady rate of descent of 2200 fpm.

At 3000 fpm your KE will increase at the rate of 68 kt per minute. Perhaps you should use the airbrakes.

P.S.
I'm pursuing this discussion only because I don't feel it would be good advice to the OP to accept an ATC request that requires him to descend in the landing configuration at 3000 fpm.

[BOAC]

I agree with your caution but better to know how to do it when it happens, and it does.

[john_tullamarine]

Appears that we have two discussions in parallel ?

(a) steep descent prior to conducting a normal ILS or similar.

This is something we often had to contend with on the Classic in Oz. One watched what ATC was doing to the profile like a hawk as the aircraft got closer in ... and the calculation below was run constantly in the mind so that the need for an orbit was known throughout ..

Easiest way to plan for it was to

(i) start from the desired configuration and speed spun up in the slot at one's desired distance to run to touch,

(ii) work the calculation backwards

(iii) include an appropriate distance/height (level always made it much easier all round) for landing onfiguration change and slow down to final approach speed

(iv) figure on about 1nm/1000ft gear down, approach flap, idle thrust and a little below limit flap speed - our SOP rules required us to be spun up with land flap regardless of height .. which defeated the aim of the exercise .. hence approach flap to get down. The airline had been a heavy user of B727 for which land flap with idle thrust was frowned upon by all. Habits die hard ...

Generally worked out real fine and just needed the pilot to massage speed a knot or two up or down during descent to stay on the desired profile. Mind you, the pilot had to be able to do simple arithmetic.


(b) steep approach once on the final approach.

This depends on the SOP configuration sequence.

In general we rolled over into the final descent (eg ILS) at 3000ft agl 210kt clean with a requirement to be at final approach speed in the landing configuration and spun up by 1500ft.

Not much room for excess height but the sequence generally worked like a dream unless there was more than a modest tailwind with which to contend. That just meant starting with some flap and the appropriate speed for that flap.


If the approach were commenced in a part flap and reduced speed configuration, eg as per standard OEM, one could trade some initial excess height without compromising the final gate requirement.

As a general rule, not a good technique as it just loaded the crew up and increased the chances of an unsatisfactory outcome with the need for a miss .. which wasted a lot more time than requesting an orbit first time around.

[HazelNuts39]

Hi John,

Appears that we have two discussions in parallel ?

Not really, we're discussing:
(a) steep descent prior to conducting a normal ILS or similar.

We just have three gradients in reply to the original post:

BOAC: 0.78 NM/1000 ft
JT: 1 NM/1000 ft
HN39: 1.07 NM/1000 ft

Maybe someone can be interested to try it in a simulator?

[BOAC]

NB JT's 1/1000 is based on approach flap not full, so may well be closer to my figure.

[HazelNuts39]

It 'worked out really fine' for the B727. Would it work out equally well for a B737?

[john_tullamarine]

My comments were relating only to the 737. For the 727, the boards did us proud. For the 737, however, they were pretty useless.

[HazelNuts39]

BOAC,

Is your figure based on any configuration in particular?

[BOAC]

See post 4?

[HazelNuts39]

So what would be your figure for the clean aircraft?

[BOAC]

No idea - what speed?

[HazelNuts39]

Green dot?

[BOAC]

A guess - 1000fpm'ish. How will this help mamad?

[HazelNuts39]

From your post #13:

The normal max in most airlines is 3000fpm within 3000' ft of MSA.
The point is that this limit does not define the RoD achieved at constant airspeed by a particular airplane, in a particular configuration, at a particular speed, at idle thrust.

[Lord Spandex Masher]

Why would it need to?

Not only that but for every Unit of weight change the RoD will vary for any given config and speed. The amount of data in there is vast! And useless.

[HazelNuts39]

It doesn't need to because it has a different purpose. It is a manoeuvre limit that is not based on performance capability.

My understanding of the original question is that it is a performance question. Therefore the 'normal max in most airlines' is not a good basis for a reply.

If you want a simple rule-of-thumb, then JT's 1/1000 is hard to beat!

Show a plot at 27x1 second intervals and we can talk.

Here it is. The data points in the graphic are at 1 second interval, starting at 10h25m25.5s.

http://i.imgur.com/sdcBAnl.gif?1

[5420N]

THY operates NG's. BOAC may well be talking Classics...........

[BOAC]

If you want a simple rule-of-thumb, then JT's 1/1000 is hard to beat! - I have no reason to question JT's figures for that flap setting in a 727, but:
1) Add at least 20% (probably more) drag to go to full flap (40) from 'approach flap' ((? 5? ?15? )
2) Reduce speed from that speed (probably either 170 or 150) to Vref+5 (?130?)

Go back to your whizz-wheel and tell me what gradient you get - purely out of interest.

I cannot see rate of descent on your graph anywhere which is what I thought we were discussing.

[HazelNuts39]

The beauty of JT's 1/1000 is that it doesn't need a whizz-wheel.

'nuff said ...

[BOAC]

Come on - don't be shy - we want to know the numbers.

[john_tullamarine]

BOAC .. gradient suggested was for the -300.

Interestingly, I can't recall ever having had to dirty up at height in the 727-100/-200 to get down in normal ops.

Reasonably frequent need in the 737 if ATC had to hold you high back at 40-50 miles or so and then forgot about you for a while ...

Slow down to max flap speed in anticipation, dirty up on further descent clearance, go down, final slow down and into the slot .. worked very fine and was quite predictable.

One needed to factor in the wind, of course, and the mental arithmetic on a very long extremely tight straight in ILS saw steam coming out of the ears on the way down ... DME was king of the calculation inputs .. only very occasionally did one get caught out close in and still have to do an orbit or miss.

I recall one approach SY 16 ILS where we were still up near FL200 after coming around the corner at MQD onto the runway extended centreline .. and ATC just kept us there ... with a fairly high traffic density... lots of mental arithmetic until in the slot on the ILS but it worked out just fine.

Flap 25 was the most we were permitted to use without spinning up. Can't get that certification engineering hat off .. approach referring to last before land.

I was never really interested in ROD while poling (other than as a normal I/F scan input) unless the clearance required something specific .. it was all about gradient capability against requirement to end up wherever I needed to end up ...

Of course, I routinely ignored the FMS so still had enough grey matter in reserve to run the mental calcs .. had I tried to factor in the lies the box told about descent back in those days it would have been a foregone waste of time trying to make a tight descent requirement ....

[HazelNuts39]

I cannot see rate of descent on your graph anywhere which is what I thought we were discussing. I'm so sorry. I thought that if you had understood the explanation I gave in previous posts you would be able to see the relation between L / D and RoD.

The following graph shows for the 737-800 (NG) in the conditions of the accident:

A = actual rate of descent
B = loss of airspeed expressed as an equivalent rate of descent
A + B = the rate of descent achievable at constant airspeed

http://i.imgur.com/jwNb9X3.gif?1

[BOAC]

Therefore the 'normal max in most airlines' is not a good basis for a reply. - not from an engineer's point of view, I appreciate, but for a pilot (which I assume the OP is) to know how you can achieve the max permitted is really quite useful.

Now, if the mods will accept a diversion from the topic, I assume on your two plots that the timeline starts at the right hand end with CAS and RoD 'on spec' for an approach. Thus it presumably finishes at impact at the left hand end, with an actual RoD at impact of less than 400fpm, fully stalled, and a rate never exceeding 1000fpm (normal operating maximum) at any point?

[HazelNuts39]

BOAC,

Not sure why you're asking and how much detail you are interested in, but:

I was looking for the longest interval where the engines were at idle (or almost) and the RoD and deceleration were approximately constant. The first 1 second interval starts at 10:25:25 and the last one ends at 10:25:48 (yes, from right to left in the two graphs). Impact was at about 10:26:01.5 at a RoD of about 3000 fpm. A few seconds prior to impact TOGA was selected and pitch control inputs were made.

Hope that answers your question.

[BOAC]

It does.....

[FlightDetent]

2: HzN / BOAC

I know this is not a 737, yet still:
(please ignore the unrealistic speed profile speed restrictions, I needed to set them up in order to avoid the SW's "out of envelope" warnings.)

http://imageshack.us/a/img35/9616/8egt.png

Result (certified AFM data):

http://imageshack.us/a/img38/457/9op.png

The first screen shows the init conditions for A320 with flaps 3 / SPD BRK etc.. and the second provides results for idle descent from 3000 to 1000 ft.

Cheers,
FD.

[john_tullamarine]

The early 733 FMS descent profile was an exercise in fairytales for the little ones at night. I can't recall anyone who did other than ignore it other than for the TOD marker as a prompt for those too engrossed in their coffees.

Does the Airbus do it better in reality ?

[FlightDetent]

mamad:

It is quite hard to say. If you fly the A/C within modern day SOP brackets, as a simple line pilot you will never know. The SOPs are built with a margin, and you never get to learn where the hard limit is because you never get to experience it.

Sort of Catch 22, or the traditional superior knowledge vs. superior skills situation.

But para 1 above is partly bs, pilots happen to travel outside SOP profile now and then. Hence the answer to your question could go like this: Each pilot's limits are different, since they are based on how badly we mismanaged before, and what the outcome had been.
a) couldn't make it to stable approach gate at 1000 and went around -> will never start an approach with that much energy again;
b) made it to to stable approach gate at 1000 sweatting all over -> could do it again, but nothing more.

My personal limit on 9 NM final would be 1500 on top (some excess speed included), if 2000 on top at 9 NM with slow speed and good "drag conf" I may give it a try and see how it ends up. Anything more - "unable for approach".

Love your original question.

[HazelNuts39]

The early 733 FMS descent profile was an exercise in fairytales for the little ones at night.I can't see any reason why an FMS would not be able to accurately predict the descent profile in a perfect atmosphere, The problem is that the FMS does not know the changes in wind that lie ahead.

The only way to get around that uncertainty is to build in conservatisms which result in the aircraft using more fuel.

[Denti]

Simply download the winds into the descend forecast page? Just a click of a button, nothing more.

[john_tullamarine]

FMS and winds

Of course .. as we only had a rough idea from the forecast which, itself was a rough idea .. ie fantasyland. Easiest real time option was to do it using the Mk1 grey matter all the way down and vary the speed a few knots here and there as necessary to keep bracketing the desired gradient profile .. and a lot of us flew the entire descent hand flown to maintain the scan rate skills .. FD either off or ignored. Much less conservative overall due to the higher data sampling rate than programming the box in cruise with comparatively significant conservatism.

Sometimes, just for fun, we might plug in some forecast winds. That just confused the issue further and we would still just ignore the box's prognostications.

The SOPs are built with a margin, and you never get to learn where the hard limit is because you never get to experience it.

And that's fine, if necessarily conservative and, to some extent, expensive.

couldn't make it to stable approach gate at 1000 and went around

Three observations -

(a) that sort of experience is indicative of stuffing the nose down and going along for the ride. The pilot has to be the boss and monitor all the way down, correcting deviations from plan as appropriate. If it isn't going to work, it becomes pretty obvious a long ways earlier than the final gate.

(b) alternatively, discuss with other pilots, especially those more experienced, try some tests in sim playtime, and revise one's own knowledge. Should the situation arise again, one should be able to perform much better than first time around.

(c) a miss is not a source of embarrassment .. it merely acknowledges that, for whatever reason, the history didn't match the plan to an extent that prudence dictated an escape option. It's only a concern if a pilot continually can't get things right .. the odd miss is a tick in the pass box.

-> will never start an approach with that much energy again

I believe the technical term for that is "learning" ?

made it to to stable approach gate at 1000 sweatting all over -> could do it again, but nothing more

.. but VERY satisfying. The difficult bit was carrying off the "doesn't everyone do it that way" nonchalance ... maybe, next time, one gives oneself an extra mile's worth of fat to reduce the workload ?

[Denti]

To be honest the forecast winds, especially if downloaded right at TOD are usually very accurate these days. Nevertheless being high on approach still happens for a number of reasons.

Personally how I cope with the situation depends how we got into it in the first place. The other day the approach controller into Orly set us deliberately up by clearing us only to 5000' and expecting us to fly the ILS from 10NM being 2000ft high. In a case like that the answer to the ILS clearance is of course an "unable". In other cases, especially if its my own fault, it is easiest to switch it all off and fly the rest at the edge of the envelope and stabilized approach criteria manually which is indeed much easier without the distraction of the auto flight system and FD. And if it doesn't work out a go-around is simply a nice change of pace since they are so extremely rare in normal line flying.

[BARKINGMAD]

Perhaps, as Denti says, when we all start saying loud and clear on VHF 1 to the guy/gal on the ground "UNABLE DUE ENERGY MANAGEMENT!", then this speculation on how to get rid of excess energy AT OUR EXPENSE will be consigned to history.

I note that most if not all the contributors to the maths exam answers have excluded the energy due to the mass, so these fancy figures may work one day in the average landing mass aircraft, but will turn round and bite you the day you arrive with tanking fuel or an unusually large payload?

Are the airlines as a block talking to the ATC authorities as a block, about the fundamentals of energy management? I doubt it very much, and as for the ORY case, they will only probably change their ways when one after another inbound refuses their crazy plan and they finally get the message that PHYSICS is involved, and not interpretation, presentation and style.

The odd fool will accept such a proposal as +2,000 ft at 10 miles and demonstrate their skill, panache and daring by getting in with a lot of sweat and possibly excited passengers.

I hope I'm not sitting beside or behind that character, life is too short...........:uhoh:

[HazelNuts39]

the contributors to the maths exam answers have excluded the energy due to the massAt Vref drag is proportional to mass, and since thrust is close to zero, the NM/1000' does not change appreciably with mass.

Kinetic and potential energy are both proportional to mass, so the exchange rate between them is independent of mass.

[BARKINGMAD]

Hazelnuts; I find your post baffling.

I have recently handled the 738NG at max landing weight, luckily our planes are mostly full and we were "tanking" fuel.

I can assure you the 'frame badly needed extra track miles to slow down when compared with the same 'frame 5 tonnes lighter.

Why does Mr Boeing provide me with distance tables showing increasing track miles needed with increasing weight e g 40/50/60/70 tons from identical heights?

And this ToD manoeuvre is only flown close to Vref in the last few miles so I don't get the connection?

Energy (momentum) is the product of the mass times velocity, so why do you think it is not a relevant and important factor? :confused:

[HazelNuts39]

Energy (momentum) is the product of the mass times velocity
Energy is the sum of kinetic and potential energy:
½*m*V² + m*g*H
where m = mass, V = velocity, g = acceleration of gravity, and H = height.

So for two points 1 and 2 of equal energy, we get: H2 - H1 = ½*(V2² - V1²) / g
Why does Mr Boeing provide me with distance tables showing increasing track miles needed with increasing weight The track miles are not proportional to weight, are they? I suspect that Mr Boeing's distance tables are based on a fixed speed schedule. At a fixed speed the L/D changes with weight, and therefore the glide angle changes with weight. If the descent speed is 300 kt, for example, that speed will be closer to the minimum drag speed at 60 tons than it is at 40 tons. Hence the L/D at 60 tons will be higher than at 40 tons. Also, the thrust may not be exactly zero, but we're discussing rules-of-thumb, not looking for exact answers.

Vref, on the other hand, varies with weight so that the L/D and glide angle are constant.

[HazelNuts39]

As an illustration to my post #49, I thought the following graph might be interesting. It shows the variation of lift-to-drag ratio (L/D) with airspeed.

For example at 1.4 times the minimum drag speed Vmd, the L/D is about 0.8 times the maximum L/D.

The relation is based on a 'parabolic' drag polar and is independent of configuration, weight or aircraft type.

http://i.imgur.com/bycqZ3S.gif?1

[mamad]

Sorry for the late response. I thought the thread deleted when it was moved from tech log to questions.

Thank you so much guys for these answers.

These discussions and the infos shared enhanced my confidence greatly at the time. It took me from trying

To be slighly low (because scared to be high and end up unstable ) to being a normal pilot (not bold). Yes in todays strict SOP environment we tend to know less and less of our aircraft real capabilities eventhough we might need it when **** hits the fence !