[jhb_airlines] Re: FPI

  • From: "Bones" <bones@xxxxxxx>
  • To: <jhb_airlines@xxxxxxxxxxxxx>
  • Date: Thu, 28 Jul 2005 16:52:56 +0100

The fascinating thing about FPI or any other real time ATC simulation is
that when it gets busy it becomes very much like real world traffic. Allow
me to explain.

Life isn't all rosy in real ATC either. In FPI we get foreign pilots and we
have many newcomers who are still developing their skills. The former may
have the language barrier to overcome and the latter may just be slow
because they are still trying to keep ahead of the aircraft, never mind the
ATC on top of this. In real life we also get foreign pilots with slow uptake
and, especially in mixed VFR/IFR airports, some real dipsticks whose flying
skills would make even yourselves cringe. The trick is for ATC to assess how
good, or bad, pilots are and to adjust instructions accordingly. That's why
bad pilots get lots of orbits - it keeps them out of the way of other
traffic until ATC feel it is safe to bring them in.

Real life ATC isn't as difficult as FPI (believe it or not) because the
professionalism is greater and ATC expect pilots to follow the rules and to
respond immediately to instructions. This allows tighter vectoring, more
aircraft in a smaller area and snappier landing and take off rates. We can't
do this in FPI because the pilot skills are not quite up to the mark and so
we use a bigger chunk of airspace, keep aircraft a little bit more apart and
are generally more cautious about crowding everyone in.

I'm not saying that all pilots are slow. A good number of aircraft last
night were as good as I see in real life and I know I could rely on them to
do exactly as I asked. However there were also one or two foreign pilots
who, although competent, had to think about what they had been told and so
were not as crisp as most of you. That however is the rub - it just needs
one slow pilot in a stream and it blows it apart. In a busy environment
EVERYONE has to be good.  I'd love to get five or six of you into a stream
landing at Heathrow all exactly four miles apart - but I know I can't yet
until skills are just a bit better.

As in real life, a controller very much bases his instructions to aircraft
on a gut instinct - primarily based on how the pilot sounds when he talks to
you. Those with confident RT tend to get trusted and vectored in more
tightly. Those with less verbal confidence are given more sea room. For
example I trust most of you enough to offer an eight mile final but some
aircraft I will give 10nm or even 15nm to because I feel that they need it.
Subjective I know but that's the way it works.

The other problem is also true to real life - as it gets busier pilots have
to become more disciplined or else ATC can't keep the flow tight. This is
essential in a multi aircraft environment otherwise traffic flow grinds to a
halt. I trust almost all of you to fly the heading and heights I throw at
you but the one factor that really makes life difficult (and makes FPI stand
apart from real ATC) is speed management. Speed matters. There are a few
good reasons for this but I will only mention a few.

If you are the only aircraft on the tube then it might be thought that speed
control isn't essential. Actually it is - because speed control is not just
used for aircraft separation. Speed control also allows ATC to judge the
right points for vectoring onto final - the turn onto base leg and the turn
to intercept an ILS. If speed is too high the radius of turn increases
significantly and it becomes almost impossible to judge the correct points
for an ILS approach.

The critical turn is the one putting you on a 30 degree closing heading to
the ILS. Think about this and you will see that there isn't much leeway in
the timing of this turn. If your speed is too high and the radius of turn
too great then you will probably go through the ILS (and I'll bet everyone
has experience of this). If the opposite happens and I turn you too early
then you will intercept the ILS, not at 10nm but at 9 or even 8nm - which
means you may find yourself descending on the glideslope before hitting the
localiser (something that real pilots do not like at all). If the RT is busy
just at the point I need to turn you then things will still go wrong but
this happens in real life too <vbg>.

In a busy TMA speed control is even more essential because of the other
traffic around. It's no fun having an aircraft at FL80 doing the correct
250kts if a white van driver is thundering up behind him at 350kts. The
latter is the one who mucks it up for everyone because he has turn rates the
size of Wales. He will also have other problems - a high cockpit workload
because higher speeds mean less time (far less than you might realise) to
run through cockpit procedures coupled with almost continuous ATC
instructions because he is whistling through the traffic pattern at a much
higher rate. High speed does you no favours at all.

ATC very rarely manage speed control in real life because they are fixed in
the procedures. It's up to you to follow these but I'm sure not everyone
scrutinises the STAR's and SID's in fine detail or knows what a speed gate
is. I would therefore like to make a suggestion to improve the situation -
both to help ATC and yourselves and get FPI looking just that little bit
more professional.

Pilots using the STAR procedures should ignore what I say because I will
assume you are obeying the speed gates on the charts. For example the BNN
STAR to Heathrow has a speed gate (correctly a SLP or Speed limitation
Point) at WCO NDB of 250kts and this applies REGARDLESS OF LEVEL i.e. it
overrules the mandatory 250kts below FL100 rule. Not so clear on the STAR is
the 220kt limitation on reaching BNN.

Now it is impossible to suggest you all read the STAR's for every airport,
nor am I going to suggest (as I used to) speed reduction based on DME
distance from an airfield because you are not really going to keep an eye on
the DME when you are busy (and some airfields don't have this anyway). A
much simpler rule is one based on height because you are always aware of
this. Here is my suggestion.


1. I see some aircraft descending from high altitudes at 450kts or more but
this isn't real - and this is where an approach in FS actually starts to go
completely wrong.

Depending on aircraft type the descent should be between 250kts and 300kts
IAS - which means you start descent quite fast (probably 420kts TAS) but
your speed gradually reduces as you get into lower altitudes (because TAS
and IAS begin to merge) to about 320kts TAS by FL100. Yep, I know some FS
aircraft don't obey the speed selector - the default B737 will stay at
275kts regardless of what you dial in - but this means the ROD is too high
for the aircraft and you need to reduce it. Work out a ROD that keeps your
aircraft following the dialled speed and I will guarantee your life will be
a lot easier. Once you are happy with a ROD you know works update the FSNav
data so it gives to correct TOD points and you may also need to amend the
aircraft.cfg file so that the autopilot automatically enters your new ROD
rather than the default 1800fpm. The line is:


2. 250kts at FL100. We all know this is mandatory but I want to enforce it.
To get your speed back by FL100 means dialling in speed reduction beforehand
and I suggest the speed really should be on the way down as you are passing
FL120 - at 1800fpm you are only talking about 66 seconds here. This is where
Point 1 comes into play too. You can't really expect to reduce from 450kts
to 250kts in just over a minute and trying to do so means that even this far
back on the approach you are already fighting to keep on top of things.
Manage Point 1 and the FL100 restriction is a doddle as you will only be
reducing from about 320kts to 250kts. You shouldn't have to be working hard
at this point in the flight but here FS may make things a lot harder.

3. Not a real world procedure but I suggest that you reduce speed to 220kts
at FL70. It's not unrealistic because this is the stack height and speed
gate for many STAR's. You will not believe how much more thinking time you
will get once you get the speed back at this point.

4. 180kts at 4000ft. I say 4000ft rather than 3000ft because on some
approaches you will not be cleared down to 3000ft - but the logic is still
the same. 180kts is the best speed to be closing on the ILS and I'd be happy
to see even 165/170kts. If I see anyone approaching the ILS above 200kts in
future I shall get very cross.

The above is easy to remember and you don't need to cross check against
anything other than the altimeter. If you can do this I will guarantee that
things will get a lot slicker in FPI - ATC can manage you better and you
will also (probably) get shorter turns onto finals because the turning radii
are lower (so reducing speed may get you in quicker because your track
distances will be reduced).

The real point of all the above is flight management. If you feel that a
situation is running away from you there is usually a reason for this
further back in the flight. Good speed control is simply good flight
management and gets you flying a nice arrival procedure with hardly any fuss
at all.


-----Original Message-----
From: jhb_airlines-bounce@xxxxxxxxxxxxx
[mailto:jhb_airlines-bounce@xxxxxxxxxxxxx] On Behalf Of Kev Townsend
Sent: 28 July 2005 10:42
To: jhb_airlines@xxxxxxxxxxxxx
Subject: [jhb_airlines] Re: FPI

Thanks All for the plaudits, It's fantastic that all you guys turn up to
enable an experience like this!

I did have problems communicating with Mike Brook, although I received him
perfectly, I was unreadable at his end, but with a little ingenuity and
some help from Alistair and Mike he made a standard departure and we got
him over to Bones at London Control without incident. He gave me a shout on
approach to Ronaldsway, and all was working perfectly so NFF as they say!

The nature of the SID / STAR procedures at Stansted, particularly when most
movements are to the North West rely heavily on the BKY VOR. Outbounds
follow the SID south of the track of Incoming flights via CPT, and I felt
sure that I gave all clearances Initial 6000', particularly to avoid
conflict at BKY. Bones ensured that Inbounds were not cleared to below
FL080 so we should have maintained separation. I remember that EFP046
reported "heavy" on arrival and that he requested the hold for burn off
purposes, this remained (I believe) at FL080. It just goes to show that
there's no let up when responsible for Tower and Ground. My hats off to the

In quiet times Bones and I were able to discuss the problems of aircraft
"reaction", particularly when turning onto the glideslope where the
controllers judgement is critical. Real life pilots when receiving a course
or altitude change would immediately dial the new figure into the autopilot
and respond to ATC as a sort of confirmation of what they have input, this
speeds up the reaction of the aircraft to our instruction and is very much
easier to manage, those who respond to the instruction immediately and then
go back and make the changes may have travelled a mile or two before the
aircraft changes course, this then requires a "fine tune" or if busy a
greater angle to establish at a closer point to touchdown, even a decision
to call for go-around or an orbit, which when busy would really screw
things up!

Just off to do the paperwork I may be some time!

Thanks again for providing the experience



Weblog from one of the MSFS Team <http://blogs.msdn.com/tdragger/> may be
of interest - particularly the links.

At 09:47 7/28/2005, you wrote:
>A good evening in general.  I didn't suffer any of the voice problems
>which others clearly did - I found myself acting as R/T go-between for
>MikeB a couple of times.  Kev did brilliantly - I counted six inbound and
>3 outbound at EGSS as I departed.  I didn't have the heart to mention my
>airprox with EFP046 (2 miles at same altitude; blue warning on TCAS) under
>the circumstances, particularly since he said nothing about my running out
>of runway on landing at EGSS a little earlier. (Reverse thrust and
>autobraking failed on my CRJ - and again at EGNS). Kev was kind enough to
>send AA Roadside Rescue to find me (and provide a clean pair of trousers
>Mike L
>Bones wrote:
>>Grrr indeed - and you have my sympathy. Despite any announcement on
>>FPI it was a very busy evening. Lost count of the total number of
>>flights but I think we had 12 pilots and many flew two or more legs.
>>My hat goes off to Kevin who plugged in at Stansted, did little for
>>maybe an hour and then got a right gaggle of inbounds from all
>>directions. Very nicely handled too. Oddly enough, reminiscent of a
>>few weeks ago when Mike had total engine failure, I got another one
>>tonight. I felt sorry for the pilot having flogged right across from
>>Hamburg in a Cessna 411 as he never got anywhere near his destination
>>at EGNS. He lost an engine over the North Sea and I had to take him
>>into Coningsby. We hit slight voice problems during the session - it
>>started to break up about halfway through the evening - but this
>>cleared towards the end of the session. Can't really be server load as
>>there was zilch traffic outside the UK and we've seen lots more
>>aircraft online in the past. I'd also like to take my hat off to
>>everyone who flew into Stansted with its lousy weather. 3300m vis and
>>OVC400 wasn't nice at all! bones

Other related posts: