It's yours, and it's free: FMS Question and Answer...

[NorthernSky]

In the thread about descending 'now', it seemed to be clear that many ATCOs are not as aware of the manner in which FMS-equipped jets are operated, as they perhaps might like to be. I offered to try to help by setting up a Q and A thread, and this is it.

Here, you'll find the opportunity to post questions and to receive answers. I'll try to help with Boeing stuff, but hope someone else might pitch in on the 'bus as my experience wth this technology is not current.

I hope this thread will run in a friendly, open, manner, and that we'll all remember we're professionals.

So, who'll start the ball rolling?

[OscarTango]

Very nice initiative...congrats... I am an ATCO and as you say I don't know as much as I'd like to know about FMS-systems. But what does it have to do with descending now or later ? Is it the 'to-be-level-at...' restriction that brings in the FMS... the descend is initiated by adjusting the FL on the auto-pilot, right ?

I've been on one fam-fligth so far... don't know when the next one will be :(

[GroundBound]

Not so much a question as an example of the problem, perhaps.

I was riding "up front' some wee while ago, and we received a clearance, descend to.. level by … The F/O duly tapped it into the FMS. A short while later ATC asked for confirmation, were we going to make the level (obviously due crossing traffic or some-such)? "Affirm", with the confidence its in the FMS. Then another request, for confirmation, because of other traffic. "Affirm, its in the FMS".

With about 1 minute to the restriction, the FMS pulled the plug, the aeroplane dropped like a stone, and we sailed through the restriction point spot-on.

However, I don't think that was what the controller wanted, and I don’t think the pilot perhaps appreciated the reason why the controller was becoming agitated.

[BOAC]

Congrats, NS.
My two euros worth for OT and GB.

OT: The Boeing 737 FMS warns of the computed top of descent point and prompts for 'reset altitude' at a variable point BEFORE computed TOD which depends on speed and forecast descent winds plus actual cruise wind. It is around 8-12 miles, I think, from memory, so around 1-2 minutes, for the 737. NORMAL operating practice (for economy) is to use the FMC for descent profile. The crew will probably ask for descent at the 'prompt', and reset the altitude window on the control panel (which 'allows' the autopilot to start the descent at the computed point). This is where the problem can arise, as they MAY then wait for the system to start descent, which is probably not what ATC want. Hence our suggestion to use the words 'NOW' in the clearance if that is what you want. I know it is not correct, but it WILL prompt the 'computer operators' to start descent either by reverting to 'old-fashioned' autopilot techniques, or by asking the FMS to begin descent early, which in the 737 is at 1000fpm until it reaches the pre-planned descent profile.

The 'level by' restriction (if known to the crew - see previous comments about the 'secret' ones!) will be pre-programmed. If another is given by ATC, it will be entered at that point, and the Boeing FMS takes about 15-20 seconds to re-compute its descent profile. Again, 'Descend NOW to be level at (new restriction)' alleviates the problem of the delay.

GB: This problem normally arises when the crew have a higher than normal (hence steeper profile) programmed descent speed OR strong head-winds which will both delay the ECONOMICAL descent point (and in the case of higher speeds, make it NOT where you have seen the others start). The FMS WILL achieve the target, but if it gives you a missed heart-beat or two, back to the use of the word 'NOW' might help to alleviate the worry? At least you will see some descent begin. Alternatively give a desired rate of descent which is a common feature in the Frankfurt approach sectors and elsewhere.

[ferris]

My question relates to speed control.
An aircraft like the A330 begins descent very early, then generally flies much slower (groundspeed) through the mid levels FL200- 10,000'. But every now and then you'll get one that behaves like a 'normal' plane. Who or what decides this profile? Obviously the crew can alter what the FMS is doing, but how is the descent profile calculation done? Is it active, different for each flight (considering wx, weight etc), or simply passive- a design thing? For a given miles-to run, the A330 is usually about 10,000' lower than other types (B777, A310, B747), making sequencing harder because of the groundspeed differential.

[Spearing Britney]

Not on the 330(am on 320/1) but presume its the same...

Basically prior to a flight you enter a cost index(CI), a number assigned by the company based on fuel cost at dep/dest time/speed requirement etc. CI 0 equals really fuel efficient, CI99 equals goes like rocket and BP smile.

With a typical cost index in the 30's/40's you cruise at .77/.78 :(and descend at 280-300 knots. If youre late or need to make closing time at the local you can alter any/all of climb,cruise,descent speeds. Altering the cost index does them all. Typically in the 320/1 you can cruise at selected speeds of up to .81 :cool: and descend if you want at 340 kts (350 if youre brave/silly:eek: ).

Some companies/skippers dont like to mess with the FMGS speeds, some never use them and fly selected speeds everywhere.

Hope that helps...

Oh one other thing, and you may well know it but I'll mention it anyway. The A320/1 has an incredibly clean wing compared to the 737. This means it glides really well. If we are doing 340 knots in level flight below FL100 it takes around 10 miles to slow up to flap extending speeds (230kts) - in level flght. If were descending we cant slow down in any real sense, below 250kts there is gear etc but its messy. Slow down or go down - not both.

The 330 I hear is even better, it glides like a dream (canada to funchal I believe;) ). No seriously I believe it really is clean so might explain why it gets lower quicker, so it can then slow up...

[NorthernSky]

Thanks everyone for chipping in! A thread that runs itself!

Just one thought so far, to add:

Remember that we don't always fly the aircraft through the FMS. There are several ways of actually 'flying' the aircraft, so let's review them, starting with the most basic:

(i) Hand-flown, raw data. Here, the aircraft is flown in the same manner as a Cherokee or Cessna single. The pilot moves the flying controls and throttles, scans the instruments, and adjusts pitch, power, and so on, in order to achieve his desired flight path. Workload is very high, and this is usually reserved for visual approaches or when something has gone very badly wrong!

(ii) Hand-flown, with Flight Director. Here, the pilot flies by hand, but the 'Non-flying Pilot' programmes the Flight Director to give appropriate modes of flight. Workload is lower, accuracy is improved.

(iii) Autopilot, and basic Flight Director modes. Here, the aircraft s flown automatically, and instead of one pilot flying and the other monitoring, both are effectively monitoring most of the time. The 'Flying Pilot' flies the aircraft using modes like Heading Select (HDG SEL) and Level Change (LVL CHG), to achieve the desired profile. Workload is lower, accuracy is better. Almost always, the authothrottle will be engaged too. Regarding modes available, these are some common ones, as Boeing call them. The Airbus ones are similar:

HDG SEL: the aircraft follows the selected heading;
VOR/LOC: the aircraft tracks the selected radial or LLZ course;
LVL CHG: the aircraft climbs with full thrust or descends with idle thrust;
V/S: The aircraft controls speed with thrust and rate of climb or descent with pitch attitude. Can be used in climb or descent, but not commonly employed in climb;
G/S: the aircraft follows the glideslope;
TO/GA: the aircraft climbs at a pre-determined rate of combined climb and acceleration, with go-around thrust, reduced go-around thrust, or take-off thrust as appropriate;

(iv) FMS navigation in lateral and/or vertical sense: The aircraft follows the pre-programmed FMS path in terms of altitude and tracking. Very little pilot intervention is required. The modes are:

LNAV: lateral tracking, following the 'magenta line';
VNAV: vertical profile, following the most economic, or fastest, or longest range, path. this is inluenced by many things including wind data inputs, sensed wind, and input Cost Index (a comparative metric relating to the relative costs of having the aircraft airborne and the cost of fuel).

Don't forget that you can mix and match. We often fly in LNAV, with the aircraft following tracks itself, but in more basic pitch modes than VNAV. Thgis is particularly so when making ATC restrictions or following instructions like 'Descend now'.

Purists will note that I've left out VNAV SPD, as it's efectively the same as LVL CHG, and Control Wheel Steering (CWS), as this is irrelevant to normal line operations on big jets.

Hope this is still of help.

[TrafficTraffic]

On the Airbus A33 and A34 series could somebody help me out here. Today I had cause to reclear an acft "..after ABC track direct XYZ"...

About 8 miles before ABC the acft commences a left turn and (by radar QDM) misses ABC by 4nm to the NE, then quite happily intecepts the dct ABC - XYZ track.

I undersand the FMC does conduct a certain amount of, lets call it track smoothing, but when the crew were quieried the reply came that we should tell them we want them to "Fly over" the point.

Is there a different input to made into the FMC to make a waypoint a "fly over" point because todays flight was routed via ABC to XYZ to take them around the Mil area they ended up very* near.

I thought I knew a fair bit about the way acft and their Nav systems operate - but this was a new one.

On the other side how many crews when cleared to a point think that means they dont actually have to overfly that point, 1 maybe 2 miles I can understand but 4 miles with RNP5 ?

[BOAC]

<Is there a different input to made into the FMC to make a waypoint a "fly over" point?>

Not on the 3/4/500 737. It is on the airlines' wish list to help with SIDs and the like. Pilot intervention is required for what you cleared them for, and from what you say, was, wrongly, not made.

[Spearing Britney]

Traffic Traffic

Again assume big bus is the same but on the baby buses there is an overfly key that you can use to make it overfly a point exactly but noramally it will as you say 'track smooth' whilst always staying within the airway... It will go further from the point the tighter the turn is too...

[terrain safe]

If you're flying direct a waypoint eg LAM and given a descent restriction to be level 250 abeam LOGAN, would this be easy to do or would you prefer a DME range from LAM instead? I think Boeing and Airbus might fly these differently but I am happy to be corrected.

Thanks for the info so far very useful.

[BOAC]

737 3/4/5 - you can always 'create' the abeam point + height restriction in the FMS so no problem, and the DME should be on the arrival chart anyway. OTOH, DME ranges always useful........

[Cough]

Quite curious this one. If you are sent direct LAM from a long way out, you would probably do the following. Execute direct LAM, with the 'abeam points' active. You end up with straight line to LAM with logan, braso, sabre abeams. Enter the restrictions as appropriate and descend with the FMC. What happens is you end up about 600 to (max) 1000' high at sabre (if you take the abeam to be an equivalent DME range) This is due to the way the FMC computes the abeam point. So I end up re-constructing it with range arcs to make the restriction.

Q. Have you noticed and does it matter?

[SPRATLEVEL300]

Some excellent stuff here, but the bottom line is that us ATCOs need to "ride up front" again:(

[crj-jockey]

@traffictraffic

If you want us to fly over a fix, please tell us. There are certain ways to do that (either fmc-programming or basic skills)! Otherwise, the fmc would take it as a fly-by-waypoint.
Please consider that some aircraft´s autopilot revert to a 1/2bank mode above a certain altitude, which leeds to a considarable larger angle of lead.

Greetings:)

[Scott Voigt]

It's a problem that we have noted in the US with the Airbus avionics. The smoothing sometimes has rather large lead turns which takes the aircraft outside of protected airspace, both for conventional separation, as well as terrain separation... Not a good thing. It also sometimes raises a surprise if you are expecting someone to go over the navaid for spacing and then they turn 9 or 10 miles early (big turn) and instead of going in trail with a sequence go side by side... Has caused more than one error here...

regards

[Spearing Britney]

Scott - not meaning to be confrontational and I stand to be corrected but my understanding is that the airbus avionics will not lead the aircraft outside the airway, it has predicted accuracy etc and will use all the airway it can be sure of but not more. Are you sure a Bus has been more than the 5 miles out? Just cos it aint boeings way doesnt mean its the wrong way. I was suspiious at first but now with plenty of bus hours it aint never been wrong...

Terrain safe - for a vast majority of 320/1's out there has to be range from a point we are going to (next/one after etc), if it isnt then its time to select in a VS with some dead reckoning by the pilot, abm points not (practically) possible. New 320/1's are available with a better box (cost option) that does abm points a la boeing (330 has always had the ability).

[BOAC]

I received this Q by P Message:

<Abeam points
How long does it take to get the abeam point into the FMS?>

A: On 737 3/4/5, around 40-60 seconds before the descent profile is 're-calculated'.

If abeam waypoint needs spelling to crew, longer!

To expand on Cough's post - not all 737's have this 'abeam points' function, depending on the mod state of the FMS. Therefore it will be nearer to 60 seconds on these a/c as the point has to be 'constructed'. As he says, of course, 'abeam SABRE' is no longer based on a DME range in his scenario, but 90 degrees abeam SABRE on the track to LAM, somewhat more difficult to calulate by DR!

It is 'seductive' for crews in this scenario to wait for computed descent point to appear: hence a re-inforcement of the trial of 'descend now' if that is what you want? If you cannot bring yourselves to make this non-CAP call, how about giving that minimum descent rate or "Descend Fl xxx to be level by YYY, leave now"? Another R/T call we get elsewhere. Then you would trigger those 'seduced' into actually starting down while the great computer machine does its bits with bytes.

Let's not forget those colleagues of ours (me not so long ago!), stumbling about Europe using antiquated navigation equipment, non-FMS, where all descent planning had to be based on DR, and 'abeam' points most certainly need a DME reference if available.

[TrafficTraffic]

If you want us to fly over a fix, please tell us. There are certain ways to do that (either fmc-programming or basic skills)! Otherwise, the fmc would take it as a fly-by-waypoint

Do I understand you correctly, if I reclear you "..from Pam dct Arnem" that the aircraft will make Pam a fly-by-waypoint?? and therefore no need for the acft to track VIA that point?

I find this hard to believe.

I know that if there are 4 waypoint between Pam and Arnem that there will become fly-by, (probably made abeam) but my example was a simple reclearance from A - D missing B and C, but the aircraft turned 8 miles before A and missed A by at least 4 miles...

[karaoke]

Ahh, all is more clear now.

This sounds as though the crew made (a not uncommon) error of assuming the clearance was 'direct' Arnem. Rather than Arnem Via PAM.

Someone wasn't listening ;)

[crj-jockey]

@karaoke
negativ

@traffic
correct. The FMC will force the autopilot to smoothly intercept the direct track between waypoint a and b, if not programmed to fly-over a, and then intercept the applicable track. This is not a big deal, if the track differs by only for example 20 degrees. However, if for instance the inbound to a is 90 and the track from a to b ist 180, the aircraft will normally turn quite early.
There are a lot of SIDs with fly-over wpts, they are underlined in our charts, and the fmc is programmed accordingly.
But once again! If you want us to fly over a specific fix, please tell us. Otherwise we will pass by!
To make it clear: If you clear us via pam to arnem, the a/c will track to pam initially, and then (before pam) initiate a turn to intercept the direct pam-arnem track.
In the 90 degree turn example the a/c will start the turn about 6.3 NM before the fix with 25 bank and a gs of 450kts. With half-bank mode about 12NM before the first fix.

[TrafficTraffic]

To make it clear: If you clear us via pam to arnem, the a/c will track to pam initially, and then (before pam) initiate a turn to intercept the direct pam-arnem track.

Thanks for the info, I have of course in the meantime figured out how to stop the problem in the future -

NO TRACK SHORTENING

;)

Only kidding

[Scott Voigt]

Spearing;

Yes we have seen the busses go outside of the airway...

Oh, for any other responses. I will be at our convention for the next week in Cleveland... Back in a week...

regards

[NorthernSky]

A quick post to (a) bring this back into circulation (because I'm sure we haven't finished discussions just yet); and (b) answer the 'fly-over'/'fly-by' point.

It is a basic 'rule' of instrument flying - given in CAP54 Apendix M if memory serves, if not, please correct me - that a pilot may 'anticipate' the turn overhead a beacon or waypoint, provided that the aircraft remains within the protection of controlled airspace. For this reason, controllers should not be surprised that the aircraft turns early.

In the FMC database, a few points on some routes (especially SIDs and STARs) are defined as 'fly over', others, as 'fly-by'. In the Boeing, there is no facility to make a waypoint a 'fly-over' one.

I must say, that if separation depends upon it (miss distance predicted at 10nm or less, say), I would hope to be on a heading or series of headings, rather than trusting my safety to the aircraft's tracking.

One infuriating thing about the Spanish is their unwillingness to use headings, and just a couple of days ago, I found myself 6nm from opposite direction traffic, both of us on our 'own nav', in the upper air!

One final point for the high-paid help: There is no phraseology for instructing a pilot to 'fly-over' a waypoint, the phraseology laid down does not make a distinction between fly-by and fly-over.

[pom]

Traffic Traffic

"Thanks for the info, I have of course in the meantime figured out how to stop the problem in the future -

"NO TRACK SHORTENING"

I know your tongue was in your cheek, but even without track shortening, the Airbus will turn early in order to intercept the next track - we never fly over any waypoints unless the outbound track is the same as the inbound, or we tell it to. Of course, then we turn "late" and probably approach the other side of the airway.


http://www.webdck.com/Angela/images/fun/felixanm.gif

[West Coast]

Northern sky
To make it a flyover point rather then a flyby, I develop a WP over the point. If ABC is a flyby, try ABC/-0 in the scratchpad and drag it up to the original ABC. Your mileage may vary.

[SLT]

Going back to the question re. the A330 descent profile - as already pointed out by Spearing Britney, the profile is worked out in advance by the FMGEC using various factors. The difference is between A320/1 and A330 - the A330 will, if you let it, compute an incredibly efficient (but slow) descent profile, despite calculating a perfectly normal climb and cruise. This is not a problem on A320/1 because the speeds are faster, but can be on the A330. We tend to modify the descent speeds to something more realistic - usually about .81/310knots. If you don't do this - the machine will use approx .79/280 knots - again dependent on cost index etc. She does come down slow, and she sure don't like slowing down!!! 160 kt to 4 miles is the absolute maximum you can do, if you want to be stabilised at 800-900 feet.

Great areoplane though!!

[NorthernSky]

West Coast,

Do you claim this works on the 737? -EFIS or -NG?

I've never seen it done or tried it. I'll give it a go next week.

I guess that the theory goes that in attempting to track from ABC/-0 to ABC, the aircraft flies over the waypoints. It's a nice idea.